Volume
34, Number 2, October 2005
ISSN 1457-9723
Editor: Pirkko Oittinen, Pirkko.Oittinen@tkk.fi
Assistant Editor: Tiina Hartikainen, Tiina.Hartikainen@tkk.fi
|
|
Volume
34, Number 2, October 2005
|
|
|
|
|
Printed
Codes Patent Survey
Saarelma,
H. ABSTRACTInventions
related to printed codes were examined on the basis of the patent
literature. There are slightly over ten thousand patent families and
about thirty thousand individual patents on the topic. The leading
country in this field is Japan, followed by the United States. The most
researched sub-fields at present are printed watermarks and printed,
merged codes. 1 INTRODUCTIONWhen the bar code system was introduced many years ago, it was quickly adopted by retailers. The interest in using printed codes also in other applications than retail sales has increased strongly. Primary application fields include the integration of mass media technologies (print on paper, digital television and internet) and goods handling logistics. There has been considerable research and development activity in this field in Finland, so it was considered useful to examine this matter in the light of patents and patent applications. The objective was to find out who is doing what, where and how much. On the one hand, this information will help in predicting what kind of applications that are being developed in the world at present. On the other hand, it will reveal the protected areas where commercial productisation requires the application developer to obtain IPR licensing from the patent holder. 2 MethodsThe aim of the study was to determine the total number of patents covering the making and use of printed codes, their distribution over time and geographically, and the focus areas of current development efforts. The desired information was retrieved with the Espacenet patent search engine. The total number of patents was tracked with the aid of suitable keywords. The distribution of the most recent patents between countries was determined based on a sample. The distribution between countries of patents awarded and applied for during the ten past years was also determined. The field was divided into ten sub-fields, and within each of these sub-fields about ten of the most recent patents or patent applications were examined. There age reveals which fields that are now in focus and which are becoming obsolete. 3 General Findings There are slightly more than 10 000 families of patents related to printed codes. If the other members of these patent families are taken into account, the total number of patents can be estimated at about 30 000. In this study, only the mother patent of the family, i.e. the initially submitted application, was examined. In cases were supplementary patents have been assembled around the basic patent to provide flank protection (which is what large corporations tend to do), only the basic patent has been examined. In some cases, two separate patents constitute an entity, in which case both of them are mentioned. The first patent in this field was awarded in 1916 /19/. This was concerned with a control unit for a teleprinter. The patents of the early 1980s were primarily concerned with equipment that is now outdated. Starting in the 1980s, coding principles, code-making methods, system components and entire systems have been patented. Most of this technology is still, at least in principle, operational. About half of the patent applications have been filed before the year 2000 and half during the period of slightly less than six years of the current millennium. In other words, the number of applications has grown exponentially /Figure 1/.
This growth has been considerably faster that the growth of the number of patents on average. Table 1 shows how the number of US patents has increased during the period 1935-2004.
Obviously, exponential growth according to Figure 1 cannot continue indefinitely. A clearer picture is obtained by examining patents by sub-fields. Patents in the field of "printed watermarks" are here used as an example /Table 2/.
Table 2a shows the realised exponential growth of the number of patents as a function of time. According to Table 2b, the growth has been saturated in the first years of the new millennium. Similar (even declining) trends are visible in all the sub-fields of printed codes examined here. However, there are no signs that the number of patents would begin to decline, but growth rates similar to those recorded in the past five years are unlikely. Table 3 shows the distribution between countries of patent applications submitted during the past ten years.
In other words, the markets are dominated by Japan and the United States, whereas China and Korea are gaining ground. The percentage of Europe is high enough to provide a sufficient basis for growth. However, a clear know-how cluster is lacking in Europe. Patents have been filed here and there, often on the initiative of local inventors. This can be concluded from the fact that the inventor and applicant are often the same person, and a private person. A private person can usually finance one (domestic) patent. When 18 months have passed from the filing of the patent, the patent is freely available for use, except in the country where it was filed. A private person is usually completely unable to defend his/her patent rights. To get an idea of which way the trend is going, all patent applications on printed codes submitted in 2005 were reviewed. The findings are shown in Table 4.
Table 2 leads to the conclusion that China is becoming a major player in this arena. Korea appears to have lost its grip and India has for some reason failed to submit a single application. However, the conclusion cannot be drawn that Japan would be losing market share. Japan's patent agency is known to be slow, and as the search engine sorts applications according to the date of submission, recent Japanese applications are left out. The data also reveal that Russia and the other former east-block countries have adopted the patenting culture. The numbers are still so small that they do not turn up in rough statistics of this kind, but these countries are definitely emerging here. Europe's share is the same as in the statistics for the past ten years. However, regrettably, the know-how cluster in France, which was clearly visible in the ten-year data, appears to have dried up; this year not a single application has been submitted in France. To get an idea of what is going on, in each of ten subfields, about ten recent patents, all of them relevant to the research being done here, were identified. These were examined in more detail. The sample represents about one per cent of the total number of patents. Often a patent is difficult to allocate to a specific field; in a couple of cases, a patent has in fact been allocated to two fields. The results are the following:
Individualising codes are used both in security print products and in personalised printed products. Patented solutions include both code creation methods and methods and systems for reading and interpreting codes. An example is a device which reads bank note serial numbers and identifies notes reported stolen. Reportedly, this has not yet advanced to the application stage. Development in this field is active at present.
Printed watermarks have long been used for identification of security print products. There are about ten patents concerning innovative improvements of analogue printed watermarks, submitted from 1919 to the 1970s. Different features of an originally digitally generated printed watermark have been patented in a couple of hundred patents and applications. There is a lot of activity in this field at present.
Patents whose core content is concerned with the structure or some other feature of a bar code, the method for creating or the technique for reading a barcode total slightly over 19,000. It is difficult to estimate the relative weight of printing in a specific bar code patent. It is assumed here that one-third of the 10,000 patents examined for this study are bar code patents. Though the bar code is standardised as such and in extensive use, research and development activity is still lively.
A code can be merged into an image or text so that it is invisible to the human eye. The code can also be printed in colours which are invisible to the human eye but which are machine-readable. According to several experts, merged codes will replace conventional bar codes in many applications. Research and development activity is very brisk.
Exceptionally few coding principles have been patented. This is probably partly due to the fact that for a long time only a machine or piece of equipment or a component were considered patentable. Another likely reason is that coding principles have largely been developed by mathematicians, who tend to shun the patenting culture and prefer their discoveries to be publicised. The scientific community still prefers an article published in a leading professional magazine over a patent. Another, clearly separate, sub-field under this heading is the code-making method and the materials used for this. Numerous code-making methods and materials have been patented, especially for codes that are invisible to the human eye.
Even before the dawn of the "official" digital age, devices which convert printed code into digital impulses were patented. These devices were in fact used at an astonishingly early stage, notably in various media applications, though in image transfer, for example, analogue technologies dominated until the early 1980s. The hardware has not been a problem in analogue/digital/analogue conversions for about twenty years, so innovations in this field have waned.
The integration of communications media has been on the agenda for more than ten years. This involves some kind of interface between the printed product, the computer and digital television. It is obvious that one solution is a code in the printed product which is interpreted by the computer, which then operates in the desired way after reading the code. The ideas and implementations are in many cases trivial and have occurred to numerous (thousands of) people at the same time. However, someone has always decided to file a patent application and part of them have even been successful. This situation is somewhat problematic. Almost any kind of interface based on the code used in a printed product or computer enjoys patent protection somewhere in the world, even if it is a trivial technique. This can be compared with the situation in the field of printing technology in the early 1980s. Someone decided to patent the principle of digital halftoning and managed to get it approved. However, as far as it is known today, no licensing fees were paid and nobody was sued. Another dispute at the time was concerned with text font copyrights. The fonts had been designed by hand and scanned by computer, when someone got the idea to generate fonts digitally. This was the subject of extensive disputes, but the makers of digital fonts won in the end and the price of a font to the user dropped to fraction of its original price. It is obvious that the trivial feature of the interface between printed product and computer will be available to everybody in due course, but there is still potential for a few patent disputes. Therefore, in the present situation, it will be important to update and save the documents concerning these interfaces. In the future, it may be important to be able to prove that a certain method has been known and in use before someone else has submitted a patent application for the same method.
The technical solutions used in payment devices are related to the printed product-computer interface and to the individualised code. However, it has been taken as a separate sub-field in this context, because the patents are close to new applications. Reading these patents will help to make an educated guess about what the world will look like 5 to 10 years from now.
Code writing devices and systems are naturally indispensable. However, these devices are reaching maturity, because research and development work is still being made but not necessarily with the same intensity as before.
Code reading devices were the focus of intensive research and development activity at the turn of the millennium, though active research and development is still in progress in this field. The lists of references gives the complete name of each patent, which gives the reader some idea of its contents. With the aid of the patent number, the complete patent or its abstract can be retrieved with the patent search engine. An abstract in English is available, as a minimum, for all patents. 4 Conclusions The results show that the printed watermark and printed merged code are currently subjects of active research and development. In contrast, code writing and reading devices appear to have reached maturity; new patent applications have not been submitted during the past two or three years. Product development focusing on printed codes and applications based on these has attracted major inputs and is characterised by intense competition. However, just by reading patents it is difficult to conclude whether a specific patent in fact provides protection and whether one's own application infringes on the patent in question. In the event of a dispute, this is decided in a court of law. In pursuing commercial applications, great care must be taken. There is such a massive number of patents that, presumably, any solution infringes on some existing patent. From the viewpoint of the research and development being done in Finland at present, it would be vitally important to find openings in this barrage of patents and plug them with our own patents. Commercial success of course requires ideas of appealing applications, as well as application development and deployment. References
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
PrintAccess
Hautala, T., Kallenbach, J., Nuutinen, M., Salo, L., Venho,
T. Bäck, A., Bäckström, C.,
Hakola L., Järvinen, T.
ABSTRACTThe focus of PrintAccess project was on the hybrid solutions of the print and electronic media by using printed codes. It was funded by Tekes and the industry. The research organizations in the project were the TKK Media Technology and VTT Information Technology. In the project different code types and their potentiality as PrintAccess codes were investigated. Pilot applications for mobile devices and digital television were designed and realized. These applications were tested by usability test and the first smart magazine in Finland was published. In addition, workflows of the PrintAccess applications were investigated. The results showed that the code based integration between the camera phone and printed media are technologically mature and there are opportunities for commercial applications. Also usability evaluation gave positive results. On the other hand technologies around the digital television not yet mature enough for execution of the PrintAccess applications. 1 INTRODUCTION 1.1 Research Objectives and the Fields of Research The research in PrintAccess project was focused into hybrid solutions of print and electronic media. These so called hybridmedia solutions enable print-based interactive communication. The overall objective of the project was to enhance synergy between print and electronic media, and, as a result of this, to promote the competitiveness of print products and the diversity of the ways of using them. In pursuit of the overall objective new concepts and pilot applications were created. The pilot applications offer a mechanism for accessing digital content though print media. Some examples of the digital content in question are electronic marketing, ordering and purchasing, as well as published material and digital television programs. The development of the applications utilized the existing and developing infrastructures of electronic communication, i.e. the Internet protocol, mobile communications and the MHP-platform of digital television. The technical functionality and the usability of the pilot applications were evaluated experimentally. The research covered also the use of different types of code information and alternative printing methods in producing the print products, the reading and identification of the code information with camera-equipped devices, and the utilization of the code detection as an access mechanism. The fields of research in the PrintAccess project are depicted in Figure 1.
1.2 Background of the Research All around the world a lot of research is focused to the investigation and realization of solutions for integrating print and digital media. For example, mobile phone readable QR-codes are already in use in many solutions in Japan. All new mobile phone models offered by operator NTT DoCoMo include a QR code reader software /1/. QR codes can be found i.e. in magazines, on posters and on business cards. In magazines and on posters the codes offer quick one-touch access to specific Internet sites. Figure 2 shows an example from a sub-way station in Tokyo where mobile phone readable QR code is attached on a poster. In addition, some companies offer their own mobile phone readable code solutions. For example, a Swedish company OP3 offers so-called two-dimensional ShotCode technique /2/. The ShotCode includes a URL-address, and the main focus area is mobile commerce applications. Japanese ColorZip offers a code technology named ColorCode TM /3/. The ColorCodesTM are comprised of a matrix of cells rendered in various colors. According to the company the ColorCodeTM is the only technology in the world using three-dimensional color combination in camera phone readable codes. American company Digimarc has announced that they have developed a method to hide data in printed images that mobile phone camera can see and respond to /4/. According to the company the hidden data is invisible to the human eye.
1.3 Project Partners The research organizations in the project were the Laboratory of Media Technology of Helsinki University of Technology and VTT Information Technology. The company partners were Edita Prima Oy, M-real Oyj, Stora Enso Oyj and UPC Consulting Oy. The project participants are listed in Table 1.
1.4 The Structure of the Document This report is organized as follow. Chapter 2 gives an overview of the different code types and code manufacturing techniques that have been used in project and presents the results of the code detection study. Chapter 3 presents the matrix code reader programs, PrintAccess pilot applications and results of the usability evaluation. The parts that compose the business area and the potential end users, technology adapters and profit makers are discussed in chapter 4 and chapter 5 concludes the report. 2 CODE TYPES, PRINTING AND DETECTION 2.1 Two-Dimensional Code Types 2.1.1 ISO Standardized Two-Dimensional Symbologies The term code type can mean the way that is used to encode and decode information. In this case the term code type equals the term symbology. A symbology defines the grammar of a specific code type. ISO (the International Organization for Standardization) has standardized four two-dimensional symbologies. These symbologies are QR code (Quick Response Code), Data Matrix, PDF417 (Portable Data File 417) and Maxi Code /5/. In addition to the twodimensional code types standardized by ISO, there are also other two-dimensional code types: Aztec Code, Veri Code and Code One. Example images and the maximum data capacity of the ISO standardized two-dimensional symbologies are shown in Table 2. The principles of QR code and Data Matrix are introduced more specifically below.
QR code is a square-shaped two-dimensional symbology developed by Denso Wave /5/. QR code is the most used two-dimensional code type in Japan. There are three QR code types: Model 1, Model 2 and MicroQR. Model 1 and Model 2 include three position detection patterns whereas MicroQR includes one position detection pattern. QR code Model 1 is the original definition and Model 2 has been developed from it. The matrix size of Model 2 can be between 21×21 and 177×177 cells. The amount of encoded data defines the matrix size. A cell refers to the white and black squares that make up the QR code. A single cell is the smallest element of QR code and is used to encode one bit of data. Bit value 1 is marked by a dark cell and bit value 0 is marked by a light cell or vice versa. QR code employs Reed-Solomon error correction algorithm. One notable feature of QR code is the possibility to encode directly Japanese Kanji-characters. The structure of QR code is shown in left side of Figure 3. /18/ Data Matrix is a two-dimensional symbology developed by RVSI Acuity CiMatrix /7/. There are two Data Matrix code models: ECC 000 140 and ECC 200. Model ECC 000 140 is recommended to be used only in closed applications. Only the properties of the Data Matrix code model ECC 200 are described below. /15/ Data Matrix code consists of data regions, a finding pattern and possible alignment patterns. Data region consists of dark and light squared cells. Alignment patterns are only used in the larger Data Matrix codes. The larger Data Matrix codes consist of a number of data regions which are separated by alignment patterns. The finding pattern is composed of two slides forming an L pattern and the two opposite slides of alternating dark and light elements. As the QR codes, the Data Matrix codes employ Reed-Solomon error correction algorithm. Structure of a Data Matrix code with a single data region is shown in the right side of Figure 3. /15/
Data Matrix codes have an even number of rows and columns. Some Data Matrix codes are squared and some Data Matrix codes are rectangular. The matrix size of the Data Matrix code can be between 10×10 and 144×144 cells and eight cells constitute one code word. Figure 4 shows code word placements for a Data Matrix code of size 10×10 according to the Data Matrix standard. The first code word is 1.x, second code word is 2.x and so on. /15/
2.1.2 Techniques for Manufacturing Codes In Section 2.1.1 the term code type equals the term symbology. The term code type can also mean the way that is used to manufacture printed codes. Total of three different ways of manufacturing printed codes were investigated in the PrintAccess project. The basic way is to set a dark symbol on a light background /17/. For example, codes are printed on a white paper using black ink. These kinds of codes are clearly visible for human eye but they also can be esthetically disturbing. In addition, there can be restrictions for the size and alignment of the codes. These problems could be removed or reduced by using invisible codes. A second way of manufacturing printed codes that was investigated is printing codes using inks that are invisible without ultraviolet radiation /19/. The third investigated code type was multi-color codes /20/. With different colors it is possible to add a third dimension to a two-dimensional code. The third dimension increases the maximum data capacity and it is possible to embed more information compared to the two-dimensional codes of equal size. Multi-color codes could also be esthetically less disturbing than pure black and white codes. Technology in the manufacturing of multi-colored codes is based on additive color process. A multi-color code includes two or three sub-codes. The color of a single sub-code can be red, green or blue. Figure 5 shows a multi-color code that is combined of three sub-codes. Multicolor codes can be read by a normal digital camera or a mobile device with an integrated camera. Multi-color codes can be decoded based on the color channel information of the captured image.
2.2 Printing Codes 2.2.1 Introduction Several printing trials with different presses and printers were made during the Print Access project. The objective was to find the limitations that printing processes bring to the code size and quality. The objective was also to investigate how much information can be stored into publications. Another objective was to produce test prints for the trials of code detection and for the usability studies. When talking about the two-dimensional codes in this chapter the black areas are considered the code and white areas the background although the white areas are actually a relevant part of the code. An example is in Figure 6.
In a total of five printing trials were conducted: three at Edita Prima, one at UPC Print and one at HP Indigo. The test images for all printing trials included two-dimensional bar codes with different cell sizes, information capacities and colors as well as micro text with different fonts and font sizes. Also test fields for usability studies were included. Purpose of the micro text was mainly grading of print quality. The test pages from the printing trial at UPC Print are presented in Figure 7.
The cell sizes ranged from 100 µm to 600 µm with 50 µm intervals. The numbers of cells were 16×16, 24×24, 32×32, 40×40 and 48×48. The information capacities of these cell amounts for Data Matrix codes are presented in Table 3. Micro text fonts were Arial, Times New Roman and Courier New with font sizes 1, 1,5, 2, 2,5, 3, 4, 5, 6, 8, 10 and 12 pt. The text was written with small letters and capital letters and the distance of the letters relative to each other was varied. When printing two-dimensional codes that contained other colours than black and white only the code size 24×24 was used.
The purpose of the printing trial at HP Indigo was to get information for designing the test images for the other printing trials as well as to get information on the print quality with liquid based toner electrophotography. The press was HP Indigo s2000 with 800 dpi resolution and with speed of 1000 A3 images per hour (Figure 8). The press had the ability to print security features such as invisible ink, micro text down to 0.9 pt font size and digital watermarks. The test image was printed one-sidedly on glossy and matt paper.
At Edita Prima two presses and two printers were used. For printing two-dimensional bar codes (Data Matrix, Aztec Code and PDF 417) and invisible codes (Data Matrix) Concept NT 9-unit web-fed offset press with 130 m/min speed was used. Printing was done two-sidedly, but the different sides were also printed one-sidedly and analysis was made on these samples. Variable data i.e. two-dimensional bar code (Data Matrix) was printed online with Scitex continuous inkjet printer and offline with Océ dry toner electrophotography printer all on top-side. Two printing trials were done with Concept NT press. The purpose of the second printing trial was to improve print quality of invisible codes and colour registration and to print test pages for usability studies. For printing micro text 5-unit sheet-fed offset press with the speed of 8000 sheets per hour was used and printing was done one-sidedly. The paper in all printing trials was two-side single coated G-Print (115 g/m2).
At UPC Print 6-unit web-fed Rotoman D heatset offset press was used with 30 000 prints per hour speed. Two-dimensional bar codes (Data Matrix), micro text and test pages for usability studies were printed two-sidedly in order to investigate the differences between print qualities on different papers. Five paper grades were used: newsprint (52 g/m2), uncoated paper (52 g/m2), SC paper (50 g/m2), LWC paper 1 (54 g/m2) and LWC paper 2 (80 g/m2). 2.2.3 Differences in Code Quality Between Digital and Offset Printing In Figure 10 differences between digital and conventional printing can be seen. The print quality with offset and electrophotography was quite similar but with inkjet printing the ink spreading degraded the code quality especially with small cell sizes. The inkjet printer had a resolution of 240 dpi and such a low resolution affects ink spreading. The ink spreading, however, didn't affect code decoding. The cells also curved into print direction but this also was to so small extent that it didn't have an effect on code decoding.
It was concluded that digital printing methods are suitable for printing two-dimensional bar codes especially when variable data is needed. Inkjet printing can be used for online printing and electrophotography for offline printing. It was also found out that all the variable data has to be printed with the same printer in order to make sure that the different fields with variable data are congruent with each other. 2.2.4 Code Quality on Different Papers Cell width, cell area and hole percentage were analysed from the two-dimensional bar codes with a microscope, a CCD camera and a tailored image analysis software. All analyses were made on codes printed with black ink on white background. In Figure 11 two-dimensional bar codes on different papers are presented. On all papers the cells were square-like even at cell sizes as small as 200 µm. On coated LWC and SC papers the cells whose size was as small as 100 µm were square-like but on newsprint and on uncoated paper cells of as small size were indefinite due to ink spreading. Ink spreading was also seen in cell area measurements since areas on newsprint and uncoated paper were larger than on other papers.
In Figure 12 the hole percentage on different papers as a function of cell size is presented. It can be seen from the chart that on LWC and SC papers there were only a few holes in the cells but on newsprint and especially on uncoated paper there were a lot of holes. The hole percentage, however, decreased with increasing cell size. Large hole percentage on uncoated and newsprint is probably due to high roughness of paper surface and its tendency to strongly absorb ink.
The optical density, the area of the letter e and the area of the hole of the letter e were measured from the printed samples with a microscope, a CCD camera and a tailored image analysis program. In Figure 13 Arial 6 pt letters on different papers are presented. Even font size 1 pt was readable with microscope but no analysis could be made with them due to their small size.
Ink spreading can be estimated from the area measurements. The bigger the letter and the smaller the hole, the more the ink has spread. In Figure 14 the area of the hole of the letter e on different papers is presented. The biggest letter areas and smallest hole areas were found on newsprint and the smallest areas and biggest hole areas were found on LWC paper 2 and on coated paper. The papers behaved in this regard as expected: ink spread more on uncoated paper grades and less on high quality coated papers.
The differences between different papers were relatively small when printing micro text. Coated paper, however, was clearly better for printing micro text than other papers. This is probably because coated paper was printed with a different press, sheet-fed offset press, than the other papers that were printed with a web-fed heatset offset press. The sheet-fed offset press might produce overall better print quality than the web-fed heatset offset press. 2.2.5 Summary of the Results Concerning the Printing Trials All printing methods and papers were found suitable for printing two-dimensional bar codes and micro text. On high quality papers such as coated and LWC papers even two-dimensional bar code cell size 100 µm was square-like. Newsprint and uncoated papers are also suitable for printing two-dimensional bar codes but they are more suitable for larger cell size due to ink spreading and absorption that degrades the print quality with small cell sizes. When decoding two-dimensional bar codes with camera phones it is important that black cells are not spread too much on white cells. With very small cells spreading can cause white cells to appear black and thus hinder decoding. It is also important that there aren't much satellite drops that also can make white cells appear black. Digital printing methods are suitable for printing variable data either online with inkjet printing or offline with electrophotography. With online inkjet printing it is not possible to use as small cell size as with offline electrophotography printing. However, with online inkjet printing (resolution of 240 dpi) even cell size of 350 µm was occasionally found good enough for decoding. If an inkjet printer with higher resolution would be used even smaller inkjet printed cell sizes could certainly be detected. The papers and offset presses used in this study were found to be suitable for printing micro text. On high quality papers even font size 1 pt was found to be readable although no analysis could be made due to the small size. High quality papers such as coated and LWC papers are the best papers for printing micro text due to low ink spreading. Also lower quality papers are suitable for printing micro text but it is not possible to use as small font size as with high quality papers. 2.3 Code Detection 2.3.1 Introduction The objective of researching code detection was to investigate what kind of codes of which size and of which information capacity can be detected with camera phones. In this section, when talking about two-dimensional codes, the black areas are considered to be the code and the white areas the background although the white areas are actually a relevant part of the code. 2.3.2 The Effect of Code Type on Code Detection Three different two-dimensional bar code types were imaged with a Nokia 3650 camera phone with an add-on macro lens. The resolution of the camera was VGA resolution (640×480 pixels). The imaging distance was chosen so that the code appeared as sharp as possible on the camera phone screen. The imaging distance was typically between 4 and 8 cm depending on the cell size and information capacity. The Data Matrix codes were decoded with software manufactured by Intelcom installed into the camera phone /8/. Images of Aztec Code and PDF 417 codes were transferred to a PC and decoded with software manufactured by Intelcom. Different code types imaged with the camera phone are presented in Figure 15.
All of the decoded codes were printed with web-fed offset press at Edita on coated paper (G-Print 115 g/m2) with black ink. It was found out that the code type had no effect on decoding. The smallest decoded cell size was 200 µm with all the three code types. 2.3.3 The Effect of Paper on Code Detection Data Matrix codes printed with black and white on different paper grades were imaged with a Nokia 3650 camera phone with an add-on macro lens. The codes were decoded with software manufactured by Intelcom which was installed into the camera phone. The information capacity i.e. the amount of cells and the cell size varied. The results are presented in Figure 16. The smallest cell size decoded with the camera phone used was 200 µm on the high quality papers such as LWC and coated papers. When the amount of cells increased the smallest cell size decoded increased at least 50 µm. This is probably because when imaging larger cell amounts the imaging distance has to be longer in order to be able to fit the code into the focus area. If the distance is too high, the images with codes with smaller cell size are not sharp enough for decoding.
Although the minimum cell size for the camera phone used was found to be 200 µm the print quality is good enough for detecting smaller cell size on certain high quality papers. This was tested by imaging black and white codes with 24×24 cells with a Canon Power Shot G6 digital camera that has a resolution of 7 mega pixels and by decoding the codes with software installed in PC. The comparison between camera phone and digital camera is presented in Figure 17. It was found out that on LWC and coated papers the cell sizes as small as 100 µm were decodable and on SC paper 150 µm. On newspaper and uncoated paper the smallest decoded cell size was the same as with the camera phone i.e. 250 µm due to poor print quality of small codes on these paper grades. As a result, improving the camera performance even smaller cell sizes can be decoded on coated papers, but on uncoated papers the print quality sets limits to decoding.
Micro text on different papers was imaged with a Nokia 3650 camera phone with an add-on macro lens. The imaging distance was chosen so that the text appeared as sharp as possible on the screen of the camera phone. The text quality of the images was analyzed visually on a PC screen. Examples of the images are in Figure 18. The smallest font size that was readable based on visual examination was 2 pt when a dark and angular font such as Arial was used. With a squiggly font such as Times New Roman only font size 3 pt was readable. The micro text was more readable on high quality papers than on newspaper or uncoated paper due to ink spreading. If the distance of the letters relative to each other was increased or if only capital letters were used, micro text was more readable. With Arial even 1 pt was to some extent readable when using capital letters.
In camera phones suitable applications for micro text are e.g. storing text as an image file or using the camera phone as a magnifying device to be able to read a text stored in a small area. If micro text would be used for encoding URL addresses like two-dimensional bar codes, text recognition software for camera phone would be needed. Because the image quality with the camera phones is not very good yet a lot of image processing would be needed in the software. This might make the software too complicated and large for the memory capacities of the current camera phones. In addition, dots often present in URL addresses are difficult to detect in micro text even with naked eye. 2.3.4 Other Print-Related Effects on Code Detection All Data Matrix codes presented in this chapter were imaged with a Nokia 3650 camera phone with an add-on macro lens and decoded with software manufactured by Intelcom installed into the camera phone. The effect of using other than black and white codes was investigated. Codes were printed with different colors (100% C, M, Y, R, G or B) on white background and with black on different colored backgrounds (100% C, M, Y, R, G or B). On white background all codes were decoded down to a cell size of 200 µm or 250 µm except yellow codes. Yellow doesn't have high enough contrast on white background thus affecting the detection process. On colored background all other codes were decoded down to a cell size of 200 µm, 250 µm or 300 µm except codes printed on blue background. Black doesn't have high contrast on blue background thus affecting the detection process. The contrast between code and its background was investigated by printing white codes on 25%, 50%, 75%, 90% and 100% black backgrounds and by printing 25%, 40%, 50%, 60%, 75%, 80% and 90% black codes on white background. The codes were printed on coated paper. When the background was white even the codes whose color was 25% black were decodable down to the cell size of 350 µm. With 80% black the code was decodable down to 200 µm of cell size i.e. the same as with 100% black codes on white background. When the background was grey its color had to be at least 75% black, but the smallest decoded cell size was only 350 µm. With 90% or 100% black background the smallest decoded cell size was 300 µm i.e. not as good as when the colors were the other way around. The effect of contrast can be seen in Figure 19.
The effect of the quiet zone i.e. blank area around the code was also investigated by printing codes whose distance from the surrounding image or text varied from a size of one cell to ten cells. Different paper grades were used in the study. The Data Matrix standard defines that the quiet zone has to be at least the size of one cell. It was found out that also when imaging codes with camera phones a quiet zone of the same size is enough. The paper grade had no effect on the results. Example images are presented in Figure 20.
It was also found out that color registration has an effect on decoding when using black codes printed on colored background. The registration error from printing on coated paper can be seen in Figure 21. If the registration error is too high, the camera phone interprets the white areas around the cells as white cells or the white areas cause noise into the images. It was also found out that if the registration error is higher than half of the cell width, the camera phone can't detect the code. Registration error can be avoided by printing a colored background without leaving blank areas for black cells thus eliminating the white areas behind the codes. In that case the black cells would be printed on top of color.
The effect of printings on the reverse side of the page that has two-dimensional bar codes was also investigated with different paper grades. The printings on the reverse side were 100% black area, 75% black area, cyan area, text, images and an area that was composed of small black and white squares. It was found out that the printing on the reverse side had no effect on the smallest decoded cell size with the papers used that had grammage of 50-52 g/m2 on uncoated papers and 54-80 g/m2 on coated papers. 2.3.5 The Effect of Optics on Code Detection Data Matrix codes with different cell sizes were imaged with Nokia 3650 camera phone with and without the add-on macro lens. The add-on macro lens was used in both 50 mm and 100 mm modes. Data Matrix codes were decoded with software manufactured by Intelcom installed into the camera phone. The imaging distance varied between 1-20 centimeters. The cell size of the codes varied between 110-710 µm. Results are presented in Figure 22. The smallest cell size decoded with the camera phone without the add-on macro-lens was 570 µm. The smallest cell size decoded with the 50 mm add-on macro lens was 210 µm and with mode 100 mm it was 320 µm. The most suitable reading distance without the add-on macro lens was 8 cm. The most suitable reading distance with the 100 mm add-on macro lens was 5 cm and with mode 50 mm it was 3 cm.
It was found that the add-on macro lens has a huge effect on the readability of Data Matrix codes with small cell sizes. The basic idea of the add-on macro lens is that it decreases the focus distance of the camera phone. A consequence of this is the possibility to take sharp images near the object. Close-up images give higher spatial sampling rate for the individual cells and the result is a higher probability of detection for codes with small cell size. 2.3.6 The Effect of Reading Angle on Code Detection Data Matrix codes were imaged using different reading angle with a Nokia 3650 camera phone without the add-on macro lens. Codes were decoded with software manufactured by Intelcom installed into the camera phone. The imaging distance was 8 cm and eight different cell sizes between 570-2494 µm were used. The degree of success was calculated as a mean of the number of successful detections of codes of all different cell sizes in particular reading angle. Results are presented in Figure 23.
It was found out that lateral tilt of the reading angle has slightly more decreasing effect on the code detection rate than vertical tilt. When reading angle was over 45 degrees the degree of success decreased very rapidly. Anyway, when reading angle was below 25 degrees, the degree of success was constantly over 80 %. 2.3.7 The Effect of Illumination on Code Detection The effect of illumination on detecting two-dimensional bar codes on different papers was investigated. The bar codes were decoded in different illuminations with decoding software installed in a Nokia 3650 camera phone with an add-on macro lens. A total of three light sources were investigated: day light (160 cd/m2, x=0.33, y=0.36), LED light (139 cd/m2, x=0.25, y=0.21) and light bulb (129 cd/m2, x=0.47, y=0.42). The spectral images of these lights on coated paper are presented in Figure 24.
It was found out that the illumination had an effect that was dependent on the paper. Light sources with high luminance, such as day light, caused light reflection with glossy papers such as LWC and coated papers. This made the detection difficult. Light sources that had low luminance, such as light bulb, didn't produce enough illumination for newspaper and uncoated paper thus hindering detection. It was concluded that illumination that is good enough for reading text from a publication is also good enough for detecting and decoding two-dimensional bar codes with a camera phone. 2.3.8 Detecting Invisible Codes 2.3.8.1 Introduction Invisible codes were printed both with web-fed offset press at Edita and Epson Stylus Photo 870 ink-jet printer. Results of detecting invisible codes are presented in chapter 2.3.8.2 and 2.3.8.3. Invisible two-dimensional bar codes were printed with an ink that is visible as yellow under UV light. The codes were printed with a web-fed offset press at Edita on coated paper (G-Print 115 g/m2). The codes were printed with different cell sizes on white, on 100 % black, on 75 % black, on text, on a photograph and on an area that is composed of small black and white squares. The codes were decoded with decoding software installed in a Nokia 3650 camera phone with an add-on macro lens. The codes were illuminated with UV light from two 9W UV lamps. Pictures of the invisible codes under UV light taken with the camera phone are presented in Figure 25.
The codes were decodable only when printed on 100 % black background from the cell size of 250 µm upwards. The other backgrounds didn't provide high enough contrast or the texture disturbed the code detection. During the second printing trial at Edita more invisible ink was used for printing the invisible codes in order to make the codes appear darker under UV light. This was found out to have no effect on decoding. Invisible codes were also imaged with a Canon Power Shot G6 digital camera that has a resolution of 7 mega pixels. The codes were decoded with software installed in PC. With this better camera cell sizes down to 150 µm were decodable on black background, as also were cell sizes 150 µm on white background. The halftone structure of 75% black area disturbed detection because the detection software interpreted the halftone dots as cells. The images of the invisible codes taken with the digital camera are presented in Figure 26.
The yellow color of the invisible ink used was a difficult color since it is in general difficult to produce high contrast with yellow ink on almost any other background except very dark colors. If another ink would have been used such as red or blue under UV light, invisible codes would probably have been more easily decoded. The codes, however, require a uniform color surface behind them in order to avoid the background texture from disturbing the decoding. 2.3.8.3 Self-Made Ink Printed with Ink Jet Invisible two-dimensional codes were printed with ink jet printer using self-made fluorescence ink that was visible under UV light. The codes were printed on different white papers using cell sizes between 400-1200 µm. The codes were imaged with a Nokia 3650 camera phone with an add-on macro lens after which they were transferred to a PC and decoded with software manufactured by Intelcom. The imaging distance was between 1-30 cm. Suitable reading distances of the invisible codes that are printed with ink jet on different paper grades are presented in Figure 27. In addition, contrast between the invisible codes and different paper grades under UV-light were measured.
The codes were not decodable when printed on high quality papers such as LWC (80 g/m2) and coated papers (115 g/m2). Besides, the codes were decodable when printed on the other paper grades from the cell size 600 m upwards. These suitable paper grades were newspaper (52 g/m2), uncoated paper (52 g/m2), SC paper (50 g/m2) and LWC paper (54 g/m2). The LWC (80 g/m2) and coated (115 g/m2) papers did not provide high enough contrast with the ink used which is clearly visible in Figure 28. Optical brighteners are probably one of the reasons for the low contrast value between the high quality papers and the invisible ink.
The effect of illumination on code detection was investigated using visible light sources such as a light bulb and a fluorescent lamp alongside with the UV light. The illuminance values were set between 100-1000 lux. Figure 29 shows the invisible codes that are illuminated in addition to the UV light with a light bulb (100 lux) and a fluorescent lamp (100 and 200 lux).
It was found out that the illumination has a huge effect for the decoding of the invisible codes. The codes were decodable only when illuminance was not more than 100 lux regardless of the used light sources. In addition, the codes were decodable only on LWC (54 g/m2) or uncoated (52 g/m2) papers. Illuminance level of 100 lux is probably too low for the environments where the PrintAccess applications are thought to be used. 2.3.9 Detecting Multi-Color Codes A total of four difference multi-color Data Matrix code types were imaged with a Nokia 3650 camera phone with an add-on macro lens. In addition, black and white Data Matrix codes were imaged and used as a point of performance comparison. The imaging distance was 5 cm and the physical sizes of the codes were 0.1-2.0 cm. Images of the multi-color codes were input to the computer based color channel separation. After that red, green and blue color channel information were decoded with software manufactured by Intelcom installed into the PC. The matrix size of the multi-color codes was 20×20 cells. According to the Data Matrix standard, maximal data capacity of a regular code of this size is 20 bytes. Based on this, a multi-color code with two sub-codes has maximum data capacity of 40 bytes and a multi-color code with three sub-codes has maximum data capacity of 60 bytes. The objective was to investigate the possibility to embed the same amount of information on a smaller area using multi-color codes compared to a situation where normal codes are used. The maximum data capacity of a normal code with 26×26 cells equals the maximum data capacity of the 20×20 cells multi-color code with two sub-codes. Respectively, the maximum data capacity of a normal code with 32×32 cells equals maximum data capacity of the 20×20 cells multi-color code with three sub-codes. When the matrix size of the code is fixed, the cell size is relative to the physical size of the code. The physical size of the code equals the square area of the code. A consequence is that with the same maximum data capacity the cell sizes are smaller in normal codes than in multi-color codes. When cell sizes decrease, code detecting becomes more difficult. The performance results of multi-color codes are presented in Figure 30. Characters RG and GB denotes the colors of the sub-codes which are embedded in the multi-color codes. For example, the character R denotes red sub-code and G denotes green sub-code.
Nokia 3650 camera phone generates low quality color images. In particular the reproduction quality of the blue color is insufficient. This phenomenon is clearly visible in Figure 31. Because of this a multi-color code enhancement program was generated in Matlab. Details of this Matlab code can be found from the reference /20/. It was found that the performance of the RG- and GB-enhanced multi-color codes were better than the performance of the normal codes. It was also found that the performance of the RG-non-enhanced multi-color code was better than the performance of normal codes.
2.3.10 Summary of the Results Concerning Code Detection The smallest cell size decoded with the current camera phones is 200 µm depending on the paper grade used. Since at least offset printing is capable of producing smaller cells than 200 µm on certain papers, it is expected that also smaller cells could be decoded in the near future. The cameras and lenses on the camera phones improve constantly so the opportunities to achieve this improve all the time. But even with the current camera phones as much as 172 ASCII characters or 348 numbers can be stored in an area of 9,6 mm × 9,6 mm and this code can still be decoded with a camera phone. It is possible to read codes that are printed with fluorescence ink, but it requires careful application specific planning. Reading performance reduces significantly if lighting also includes visible light. Also, the emission of the ink should be high enough compared to the emission of the paper. 3.1 The Matrix Code Reader Programs The Scanning and decoding appliance that is developed as a prototype during the Printaccess project is a self contained unit that consists of a camera cell phone and a 2D barcode scanning /image analysis/decoding program which uses advanced image analysis algorithms. It requires a Symbian Series 60 phone in order to be able to use the 2D code interpretation program. The program is organized in four independent parts: the user interface, the image grabbing, the image analysis method for interpreting the 2D barcode and the decoding part. The camera is constantly scanning the view in low resolution mode using pattern recognition for the decision making. When a 2D barcode is found a high resolution image is snapped and classified and analysed by pattern recognition methods. The results is then decoded and sent to a server which performs the request that is contained in the barcode. This is very versatile method with numerous applications. The server can send back any information to the cell phone for displaying or the server can perform any predefined task. There are also some commercial matrix code reader programs on the market today and the number is growing. NPC Intelcom Ltd. has developed a Data Matrix reader application for the Symbian platform. The user points the camera to the printed Data Matrix code and reads the code by pressing a button. The decoded message is show on the phone's display. Intelcom has also developed a SMS version of the software that reads the receivers phone number and a message from the Data Matrix code. A new text message is opened automatically with the message and the phone number in place. /8/ Gavitec AG has developed a program called Lavasphere for decoding traditional barcodes, QR-codes and Data Matrix codes. The application shows the view from the camera on the phone's display and when user targets the camera over a code, the software reads the code automatically as soon as it decodes it from the picture stream. The user can configure the application to read only certain types of codes (barcode, QR & Data Matrix) or the software can decode all the types automatically. Lavasphere is available for Symbian Series 60 smartphones. /9/ Semacode is a Data Matrix reader from the Semacode Corporation. Semacode reader works similarly to the Intelcom's application - the user targets the code with the cameras viewfinder and reads the code by pressing a button. The software shows a read square in the viewfinder that helps the user to align the code correctly. Semacodes are primarily designed to contain URL's and when the software decodes a URL it is automatically opened in the phones browser. Semacode is available for the Symbian platform and the J2ME capable handsets. /10/ 3.2 Pilot Applications for Mobile Phones and Digital Television 3.2.1 Introduction Pilot applications were designed and developed in two phases in the PrintAccess project. First, ideas for several applications were documented as user stories /13/. Second, some applications were selected for implementation /14/. These applications were tested by usability tests. One application was tested in a larger trial, when first smart magazines in Finland were published in February 2005. 3.2.2 Documented User Stories for Mobile and Digital Television Applications The process of creating ideas for pilot applications produced nineteen candidates. These candidates were documented as so-called user stories. A user story tells in a simple way how an application works from the user's point of view. The term user story was borrowed from Extreme Programming, which is a discipline of software development. The following user stories were written:
These user stories were categorized in various ways for various purposes. One example of the classifications is illustrated in Figure 32. It classifies the applications according to two characteristics or axes. The first dimension is how the service is used - is it only based on the mobile phone that is used to read the code, or is additional equipment needed. The second dimension deals with the relation between the content in the printed article or advertisement and what can be found using the PrintAccess functionality.
Another way of classifying the PrintAccess applications is based on the content and the functionality of the applications. Two broad classes can be found: advertisement related applications and content related applications. Advertisement related applications either give more information of the product or they may even lead to an actual transaction where the service or product is bought. The purpose of content related applications is to add value to the print product. Advertising has been and will be an important source of revenue to print media publishers, and therefore, it is crucial that print media can be maintained and developed as an efficient advertising channel. The PrintAccess codes are in the position to provide reliable feedback for advertisers, and also make it possible to offer access to immediate purchasing. It also gives the opportunity to tailor the electronic message according to the user information (in those cases where there is user information available, and this information is utilised to vary the message content). Targeting the message and interactivity are very important criteria to advertisers. Interactivity also adds value to content-based services, and gives valuable feedback of the readers' interests and reactions. 3.2.3 PrintAccess Applications in the Mobile Environment 3.2.3.1 Introduction After writing user stories, some pilot applications where selected to be implemented. Most of the selected applications were implemented as prototypes with dynamic functionality with a connection to a database. Some applications were implemented as static XHTML page mock-ups. All of the selected applications are used by reading a two-dimensional matrix code which is printed on a newspaper or a magazine. In each code a URL-address is encoded. The reading of the code is performed with a matrix code reader program. After reading the code, an XHTML page is displayed in the browser of the phone. The flow of information is illustrated in Figure 33.
The pilot applications are described in more detail in chapters 3.2.3.2 - 3.2.3.8. 3.2.3.2 Discount by Using Code A code is a part of a printed advertisement. After the detection of the code, the browser displays information about the discount and a verification number. Additional information such as the location and the open hours of the shop are displayed as well. Based on the information, the user can buy the product. She goes to the shop, takes the product, goes to cash desk and shows the discount information to the sales person. The sales person checks the verification number (each number can be used only once). Finally, the user pays the product. 3.2.3.3 House Info / House Buyer's List In a newspaper there are advertisements of houses and apartments. Every advertisement consists of a picture and a textual description of the house. Next to the description there is a code. After reading the code, the browser displays the price, the size, the type of the apartment, the address, the driving instructions and other information about the house. The user can save the target in her favourites list. 3.2.3.4 Game Downloading In a newspaper or a magazine there is an advertisement about mobile phone games. A code is placed next to the game information. After reading the code, the browser displays a page from which the game can be downloaded. 3.2.3.5 Drawing In a newspaper or a magazine there is a page explaining how readers can rate the articles or ads in the magazine (for instance best article, best ad or the most interesting product advertised). The participation includes taking part in a drawing, which is actually the main incentive for the users to participate. A code is located next to the article or the ad where the instructions for the evaluation are given. After reading the code the browser displays a notice of whether the user has won in the drawing or not. If she wins something, she must provide her address to the system. 3.2.3.6 Poll In a newspaper or a magazine there is poll that has two or more answer options. There is a code next to the poll. After reading the code, a page that contains the question and the answer options is displayed in the browser. The user can select the answer option of her choice. The user could also write a comment to be displayed on the poll results page. After pressing a submit button, the results of the poll are displayed to the user. 3.2.3.7 Clip Book of Interesting Articles Next to an article there is a code. After reading the code, article content is displayed in the browser. The user can send the article to any email address. As a default browser feature, the user can also add the article to the bookmark list. Next to an event advertisement there is a code. After the code is read, the browser displays a page with detailed information about the event and a link to a page where tickets can be bought. The buying of tickets has three phases: 1. Select seat, 2. Confirm purchase and 3. Get barcode that is used as a ticket at the gate. 3.2.4 PrintAccess Applications in the MHP-environment 3.2.4.1 Introduction In addition to implementing mobile pilot applications also the possibility of using PrintAccess applications in MHP-environment was studied in the project /24/. MHP-platform (Multimedia Home Platform) is a software platform that provides a possibility to run interactive applications in a digital television set. The use of the applications is started by reading a code which is printed in a magazine or a newspaper with a mobile device including a camera. After the detection the information in the code is transferred to the MHP-platform where the information is used as a part of an interactive application. Examples of possible applications are polls and enquiries that are related to the programs or to some other kind of content in digital television. Another example is the pay-per-view concept which means that a certain kind of digital television content, for instance a movie, is ordered and paid by reading a code in a magazine. The movie may be offered only for the subscribers of the particular magazine. During the research many different means of transmitting the information of the codes to the MHP-platform were found. More thorough research was conducted considering the use of Bluetooth protocol in the communication between the devices. The flow of information between a camera phone, the MHP-terminal and the J2EE-server is depicted in Figure 34.
3.2.4.2 The PrintAccess Framework for Application Development in MHP-environment For the purpose of PrintAccess application development a PrintAccess framework was constructed. The framework offers the PrintAccess applications a service that is composed of the common features of the applications. The use of the framework makes application development faster because there is no need to implement the same functionalities separately for every application. The common features include the handling of the connections, the interpretation of the code information, the handling of the application's content and the construction of the user interfaces. The PrintAccess framework was used to implement some test applications, and at the same time the applicability of the PrintAccess framework was verified. 3.2.4.3 A Pilot Application Using Bluetooth-Connection One example of the applications that were implemented is an ice-hockey information service called Score. With the Score-application ice-hockey enthusiasts are able to follow the situation and the progress of a certain game even if the game isn't transmitted through the video broadcast. The application is used by detecting a code that is printed on a magazine, after which the information about the game is showed on a television screen. With the application it is possible to show personalized digital content for the readers of a printed media. This is a clear difference and improvement compared to the traditional analogue teletext service where the same information is shown to all spectators regardless of the choices they have made. 3.2.4.4 Pilot Applications Using GPRS-connection In addition to the Score application which utilizes Bluetooth protocol in forming the connection between a camera phone and an MHP-terminal, also three other applications using GPRS connection were implemented. The information flow is depicted in Figure 35.
The implemented applications were a news application, a movie information service and an information service for the customers of a bank. The idea of the news service is to offer current news. A code can be printed next to an article in a newspaper if it is expected that something important and interesting regarding the same topic will be happening after the printing of the newspaper. By reading the code next to the article the readers will get the newest information about the topic with the PrintAccess application. The movie information service offers information, for instance a short review of the plot or trailers about movies or other programs. It is used by detecting the codes that are printed in the television program section in a magazine. The information service of a bank offers the customers a possibility to read more information of a loyal customer program of the bank and also to join the program. The code is included in a letter that a bank sends to its customers. The realization of the pilot applications is not fully completed. The layout and the functionality of the applications are not finalized. However, the connection between the camera phone and the MHP-terminal is working, and the PrintAccess applications can be opened and part of the information read on the digital television screen. 3.2.4.5 Production and Distribution of PrintAccess Service to DVB-MHP environment To publish PrintAccess service content through digital television to consumer, the content has to go through broadcaster's system. Broadcasters have to offer flexible, efficient and easy to use methods to external content producer for delivering service content from their production system to consumer through broadcaster's system, that the concept can be popular. Test environment enabling efficient and flexible content distribution from producer to consumer through broadcaster's system by using technologies based on Web Services, SOAP-protocol and ICE syndication model was developed in the Laboratory of Media Technology /11/. 3.2.5 PrintAccess pilot application using RFID /16/ 3.2.5.1 RFID-codes The PrintAccess project allows print-based interactive communication by using mobile devices that read PrintAccess codes. These codes can be either be ink-based codes or utilize the RFID technology (Radio Frequency IDentification) in form of RFID tags. RFID tags, or smart tags, consist usually of an active, semi-passive, or passive RFID chip and an antenna. Active tags have a built-in battery from where they can draw off the power to actively send information via the antenna into their environment. Semi-passive tags do have a battery as well. However, they use this only to supply the chip itself with power. Passive tags do not have a battery at all and are therefore dependent on external power supply. To read data from a tag's chip a reader device is required. This device contacts a tag by sending a request and gets the data back as part of a response. Depending on the application the communication process is standardized in several ISO standards, i.e. ISO 10374 for container identification, e.g. in harbors, or ISO 10536 for contactless close coupling smart cards. In the PrintAccess project the idea was to integrate such a reader device into today's available mobile devices in order to allow easy access to information on printed media. 3.2.5.2 PrintAccess RFID Application Scenario and Implementation An example scenario of realizing access to print products by using RFID technology is the print-broadcast communication process in which information stored in a printed product (inside a sticker RFID tag) is read by mobile devices in order to enable the consumption of interactive broadcast services. This general use case envisions the following two major steps:
In the first step a print media company, such as a TV guide magazine, distributes information about broadcasted programs and services along with the magazine copy inside a RFID sticker tag. The information in the tag is then read by a mobile device, such as a mobile phone containing a built-in RFID reader module and sent further to the broadcaster, for example, as an SMS. This action can elicit several things. For example, the broadcaster can take care of remotely programming the video recorder at home via a return channel to the MHP box. It is possible to grant access to pay-content or to additional contents and services offered only during a limited time. This can take place even on the mobile device itself if it is able to receive and play back TV programs or downloaded movies. The above described scenario has been implemented on a mobile phone (Nokia 9210i Communicator). The example application is a mobile electronic program guide (MEPG), on the one hand, exploring the usability of receiving TV on a mobile phone with a sufficient display and user interface and, on the other hand, demonstrating the functionality of print-based interactive services in a broadcasting environment. As on a normal TV, the user can select and switch between channels as well as display program information such as starting time and duration. Furthermore, to realize true EPG functionality, recording a program, watching pay content and accessing additional interactive services the software utilizes print-based interactive communication via an RFID reader module (SkyeTek M1) and a sticker tag (Scotch/Texas Instruments). For example, to watch the pay content the user has to enter a password. This can be done either manually by typing or more comfortably by automatically reading it from an RFID tag supplied in a TV magazine. 3.3 Smart Magazines Marketing communications company Soprano published the world's first smart magazines (in Finnish "älylehti") in February 20051. A two-dimensional matrix code, so-called smart box (in Finnish "älyruutu"), is printed on a magazine. A camera phone with VTT's code reader software is used to read the code (see Figure 36). The code may provide many kinds of services, such as more information about a product in an advertisement or an article in its original language. Readers of the smart magazines can download the code reader software with their mobile phone's browser. The smart magazines are printed by printing house UPC Print.
1 Web sites of the smart magazines: http://www.uusitoimisto.net/ and http://www.kampanja.net/ 3.4 Usability Evaluation of the Pilot Applications and the Smart Magazines 3.4.1 Usability Study of the Mobile Pilot Applications 3.4.1.1 The Objectives of the Usability Evaluation Usability evaluation of the mobile pilot applications was conducted in two phases. In the first phase in autumn 2004 the study included usability testing of four pilot applications (poll, drawing, game downloading and discount coupon, and later in the end of the year 2004 all seven pilot applications were tested. The pilot applications are presented in chapter 3.2.3. Both tests included also usability evaluation of the code detection software (ProdInfo, chapter 3.1). After the first evaluation phase a new version of the code detection software was launched. In total 12 people took part in the usability tests. All test subjects, aged 22-29 years, were either students of the University of Technology or employees of the Laboratory of Media Technology. None of the test subjects had prior knowledge of the pilot applications or the code detection software but the technical background of the subjects presumably has an effect on the results. The goal of the usability testing was to assess the ease of use of the code detection software and the pilot applications and also to find out how interesting the applications are in the opinion of users belonging to the target group (young adults). The most important usability attributes were defined to be the pleasantness of the use of the pilot applications and the accuracy and the ease of learning of the code detection task. Commonly used measures for evaluating the usability attributes are the time needed to complete the tasks, the number of errors which occur during the task completion, the amount of help needed during the tasks, and the subjective opinions of the test subjects. 3.4.1.2 The Preparation of the Evaluation and the Course of the Usability Tests After defining the usability attributes and corresponding metrics the planning of the testing was started. At first the code detection software (ProdInfo) was installed to a Nokia 3650 camera phone with a 50 mm add-on macro lens. Secondly the descriptions of the test tasks were written. The descriptions included scenarios that depicted a customary use situation of each of the pilot applications. Also required test material was prepared. The material consisted of self-made printed magazines, advertisements and posters, each of which included a Data Matrix-code. A URL-address composed of 30-40 characters was coded to every Data Matrix-code. In the first phase of the usability evaluation the size of the codes was 1 cm × 1 cm and the codes were printed with a regular laser printer. Later in the second phase of the usability evaluation, the codes were taken from the materials that were printed at Edita in the fall 2004 (chapter 2.2.2). The size of the codes was about 0.8 cm × 0.8 cm. Usability tests were started with a short introduction after which the test subjects completed a set of pre-defined tasks using thinking-aloud method. In the end of the test the participants were asked to fill out a short questionnaire and they were also interviewed. The questionnaire and the interview included questions about for instance problems in the detection task and the use of the applications, opinions concerning the interestingness of the applications used and other possible application areas, and suggestions for improvements. The test subjects were also asked which way of using mobile applications did they prefer, code detection or for example SMS messages, and how well did the DataMatrix codes fit to the layout of the printed materials. In the first phase there were four tasks, three of which were conducted inside in the laboratory room and one outside the building. The intention was to test the performance of the code detection software in different illuminations and reading angles. During the first two tasks the test subjects were sitting on a couch and the materials were placed on a table in front of them. In the third task the test subjects sat on a chair and the test material was attached to a wall in front of them. The intention was to imitate a situation where a person is sitting on a bus and there is an advertisement glued to the front seat. The fourth task was performed outside the building. The test subjects were standing by the wall to which the test material was attached. In the second phase of the usability evaluation all tasks were carried out inside under normal illumination conditions. In the first phase of the usability evaluation the subjects performed the code detection in all tasks (4 times) whereas in the second phase the code detection was performed only three times when there were in total seven tasks. The intention was to shorten the time required for the evaluation. In four test tasks the subjects started the pilot applications by selecting a bookmark from the listing in the browser application. 3.4.1.3 Results of the Usability Tests Only the results of the second phase of the usability evaluation are covered below. Results are presented in more detail in references /23/ and /21/. The time required for code detection was from 5 seconds to 1 minute 55 seconds. On average it took 52 seconds to detect the code. The time required for the task completion was on average 3 minutes. Compared to the first phase the code detection time was somewhat shorter but still too long in some cases. In one case from 18 cases in total the completion of the task was unsuccessful because of problems in the code detection, and 10 times from 42 cases the subjects didn't succeed in opening some of the web pages. The code detection software gave error messages six times which was a couple of times less than in the first phase. Two times the code detection seemed as it had succeeded because the browser of the phone got started, but the forming of the GPRS connection didn't occur automatically as in a normal situation. The test subjects needed help only twice during the tests. In the first phase majority of the test subjects didn't receive any instructions in the use of the code detection software in advance. In the second phase the subjects were shortly guided how to use the software. Therefore there are differences in the completion times, but of course also the improvements in the code detection software itself affected the results. Some problems in the code detection resulted from incorrect angle and distance of reading. The subjects also moved the camera phone too much in both vertical and horizontal directions. The test subjects sometimes pressed the keys of the phone in vain which might have had some effect on the operation of the software. Regarding the pilot applications, some problems were due to the inexperience of the subjects in the use of the phone model which had Symbian operating system (series 60) and a round-shaped keypad. The terms that were used in the menu labels caused some uncertainty, as was also the case with the sequence of numbers below the alignment box which, however, didn't cause any problems. In the interview the subjects commented that it took too much time, up to 30 seconds, for the code detection software to start up. Also the subjects commented that the refresh rate of the screen was too slow, and because of the jerkiness of the image on the display it was difficult to find the correct reading angle and distance. If there were many codes close to each other the subjects were sometimes unaware which code was detected. According to the subjects the operation of the code detection software was understandable due to the automatic functioning. In the first phase of the usability evaluation the subjects said that there weren't big differences in the performance of the code detection in different use situations. Some subjects said that it was difficult to see the screen when the display reflected the light from the lamps in the ceiling. The problem was present when the subjects had to hold the phone in a horizontal position. Some of the subjects said that it was more difficult to detect the code when the phone had to be kept in a vertical position compared to other situations. In general the tasks were thought to be easy. The ordering of the discount coupon was the easiest task because it only required the opening of the page. Only some problems in the applications were listed: there wasn't always feedback about which code had been detected or which selection hade been made, there were some wrong letters (no Scandinavian letters) and there were also some difficulties in navigation. The slowness of the GPRS connection was said to lower the willingness to use the applications. In the questionnaire the event info application was rated as the most interesting application although the answers were varied. The home info service and sending an article to email were the second most interesting applications and the discount coupon was the fourth most interesting application. Game downloading and drawing were thought to be the least interesting applications and poll was only slightly more interesting. The results of the questionnaire are presented in Table 4.
The interestingness of the game downloading application was not very high because the test subjects werent interested in playing mobile games in general. The subjects thought that they would rather use a normal web browser in a PC to answer to polls and drawings because it is cheaper to use broadband connections. When the subjects were asked what kind of applications they might be willing to use in the future they listed news services, extra information regarding news articles or advertisements also in other formats than text e.g. audio or video, lottery or other games, and ordering of files, timetables, tickets or mobile phones logos. The subjects said they would use code detection rather than SMS messages if detection was faster and easier. It is also required that the connections would be faster. Also the prices should be competitive. However one test subject said in favor of the SMS messages that they don't require printed material or anything else than the number that can be stored to the memory of the phone. The subjects also evaluated the appearance of the codes related to the layout of the printed materials. Some subjects said that the codes are somewhat disturbing or even ugly, but others said that it is important that the codes can be easily distinguished. The subjects thought that colorful codes, better graphical design or locating the codes on top of images would make the appearance nicer looking and the codes more unnoticeable. 3.4.1.4 Suggestions for Improvements Based on the results of how much time it took to detect the code it was recommended that the code detection should be further developed to be faster. The start up of the software should also be faster. The operation of the software didn't prove to be very intuitive so it was recommended that operating instructions should be written and possibly included in the menu of the code detection software. It is also a possibility to place the instructions next to the printed codes. Because only one of the functionalities in the menu of the software is necessary (start reading barcode), all other functions should be removed from the menu. Also it was said that the sequence of numbers below the alignment box should be removed. The error messages should be written so that users can understand them more easily. It was recommended that the code detection software would give feedback of the success of the code detection and possibly instructions how to read the code. 3.4.2 Usability Study of the Smart Magazines 3.4.2.1 The Objectives of the Usability Evaluation Relating to the smart magazines and the code detection software, a usability study was conducted in May 2005. The goal of the usability study was to find out how easy it is to download the code detection software to a mobile phone, how easy and fast it is to detect codes and how interesting the applications in the smart magazines are. In total eight test subjects, aged 22-39 years, took part in the usability tests. Five of the test subjects didn't have prior knowledge of the use of the Nokia 3650 camera phone and none of the subjects knew the smart magazines beforehand. The inexperienced users were briefly introduced the use of the mobile phone. The subjects had to accomplish four tasks and also to take part in an interview and a questionnaire. The first task included downloading the code detection software to a mobile phone according to the instructions given in a smart magazine (Uusi toimisto 1/2005), detection of one code and using an application (i.e. reading a wap page). The following tasks included only the detection of codes and the use of the applications. The material consisted of two articles and two advertisements in two smart magazines. Information about the used material is listed below:
3.4.2.2 Results of the Usability Test and Suggestions for Improvements The results of the usability study are presented in more detail in reference /22/. The averages of the completion times of the tasks are presented in Table 5. The completion times include the times needed for the start up of the code detection software and the establishment of the GPRS-connection which took about 10-15 seconds. The experienced users didn't cut the GPRS-connection in between the tasks so only the first task includes the time needed for the connecting.
The test subjects had some problems with the instructions that were written in the smart magazine. Some words in the instructions were a bit difficult to understand and the subjects gave suggestions of better wording. Test subjects had some difficulties in finding a correct distance and angle of code detection although it was said in the instructions that the distance of reading is about 20 cm. However, all users learned quite quickly how to detect the code. Most of the problems were due to inexperience of the use of the mobile phone and therefore the experienced users completed the tasks faster and more easily. If there were problems in the use of the software, for instance if the screen appeared like frozen, especially the experienced subjects managed to solve the problems quite smoothly by closing the software and starting it again or by using the functionality "start reading barcode" which can be found from the menu of the software. According to the questionnaire results that are presented in Table 6 the code detection didn't take too much time and it was easy to conduct. The operation of the software was also easy to understand as were the instructions in the magazine (although some test subjects seemed to have considerable problems in following the instructions). The fastness of the GPRS connection and the interestingness of the applications got more negative ratings.
The application offering a possibility to read an article in original language was rated to be the most interesting application. The rest of the applications got neutral grades. One reason for the low ratings of the interestingness of the applications may be due to the fact that none of the participants were subscribers of the smart magazines nor did they know the magazines in advance. Instead of being targeted to a broad audience the magazines are for meant for a special group of readers that are interested in office equipment and furniture or marketing communications. Therefore the themes of the articles and advertisements and also the applications relating to them weren't targeted for the participants of the usability test. 3.5 Prepress Processes to Produce and Canage Codes The use of PrintAccess applications sets challenges and requirements on the prepress processes. In the prepress process co-ordination of several parties is needed. The process of producing content-related services is simpler than the process of creating advertising-related services. Both processes are described in the following chapters. The prepress processes and relating XML vocabularies are discussed in more detail in reference /12/. 3.5.1 The Process of Creating Advertisement Applications Many partners participate in producing advertising related applications. If PrintAccess applications are included in advertisements, the complexity of the process becomes even higher. If a print advertisement offers access to some online service or content, it must be made sure that the electronic service is functioning and accessible with the code given in the advertisement. An important question is to consider what kind of information is needed to define the code. If there is only one code for the whole print run it can be directly included in the art work. In that case the organization that makes the advertisement should either be able to generate the code itself or be able to get hold of it and include it as a component into the ad. If there are more codes that are unique and produced individually using digital printing, space needs to be left for the codes. To create and include a fixed code for an advertisement, the advertiser must
This information must also be communicated to the code generating party. The advertiser must also make sure that the PrintAccess service is created which may require utilising a content creator. If the codes are to be produced as variable data printing applications (with digital printing) the data must be produced to meet the requirements of the application to be used in printing. Personalized Print Markup Language, PPML, is applicable here. The PrintAccess codes may bring two new players into the advertising process for newspapers: PrintAccess service provider and digital content provider (the party that creates the digital information that can be accessed via the PrintAccess code). Advertiser's or publisher's existing web services can be utilised, so the number of participants does not necessarily increase. Additional issues that need to be decided are, at first, how long the offered service is available, and secondly, what kind of notice is shown to the users when the service has ceased to exist. Most probably this can be managed at the PrintAccess server where some default functionality can be offered for those trying to access content or service that has been removed. Some of the information that can be transmitted via the PrintAccess codes could be described as a "classified ad" or a database entry. Event information, estate information, retail location information (address, opening times) can well be formalized, and utilised to produce the code. This formalization also supports the use of this information in all electronic applications. If the code is produced as a component, AdsML includes messages that support managing component production. There are two levels of vocabularies relating to the adverting process management:
Considering the PrintAccess project some conclusions about the XML vocabularies can be made. AdsML is expected to support the management of complex advertising processes, so there will be support to handle additional components in an efficient way. Vocabularies that define the data content of different types of advertisements, such as housing, are very useful, because this formalised content can be utilised to offer services, such as the different kinds of shopping list and event information applications. The vocabularies are still under development, and their relevance to the Finnish circumstances is unclear. If a widely accepted vocabulary does not exist, the quickest way to pilot is to develop a temporary vocabulary for that purpose; later, this can be offered for general use, or there must be capability to adopt a public vocabulary, if such a vocabulary emerges later. 3.5.2 The Process of Creating News Services If PrintAccess applications are used to add value to content such as news articles and feature stories the following process steps are needed.
Utilising PrintAccess codes in news applications is mainly a management issue. The article with all its content including the PrintAccess code and content that is accessed using it should be managed as a whole. The code is very similar in nature to web links and it could be managed in a similar manner. Two most important news related XML vocabularies are The News Markup Language (NewsML) and News Industry Text Format (NITF). NewsML is similar to AdsML in that it is a way to package news components into a package. There is, however, an important difference in the vocabularies: in NewsML the components included in one NewsML package relate to one and only piece of news whereas in AdsML, the components within the package don't have to be related to the same advertisement. Another article related vocabulary is the Publishing Requirements for Industry Standard Metadata (PRISM), which has been developed more from the magazine industry's point of view. The goal of the PrintAccess value chains study was to define costs and profits when using Print Access applications. The objective was to specify the concepts, outline parties involved, payers, profit makers and investments as well as define the overall system based on the afore-mentioned factors. 4.1 Specifying the Concepts The applications chosen for the business research were getting additional information, ordering a ticket, sending an article to e-mail, downloading information from housing notice and downloading a game. The applications were classified into two categories:
The first category includes all the other afore-mentioned applications except sending an article to e-mail. 4.1.1 Getting Additional Information, Buying a Product In applications related to getting additional information or buying a product the purpose is to:
Figure 37 presents the parties involved in the applications. Advertiser requests the publisher to print two-dimensional bar codes into the advertisement. Software developer provides the licenses to use the detection software. The decoding software both decodes the bar codes and gives automatically access to Internet if needed. Advertiser or publisher prepares the electronic service that requires maintenance possibly with help from a sub-contractor. Maintenance can be done in two ways:
The publication is printed in a print house that has the systems for printing two-dimensional bar codes. The publication is delivered to the consumer that downloads the decoding software with the help of the instructions in the advertisement. This download happens only one time. The consumer reads the two-dimensional bar code with his camera phone and gets access to the electronic service through data connection provided by the operator. The consumer gets the supplementary service. If all the additional information is encoded into the bar code, such as contact information, electronic service or data connection isn't needed. The decoding software can also be installed into consumer's camera phone by phone manufacturer or operator before consumer purchases the phone. This way the publisher or advertiser doesn't need a license to use the software, but instead the license price is included in phone or connection price and the consumer doesn't have to do software downloading.
Figure 38 presents the money streams between the parties. Advertiser buys advertisement space from the publisher and the publisher buys the printing and delivery service. The publisher and/or the advertiser pay for the software developer to develop the decoding software and to get a licence for the software. The consumer pays for the publisher by buying the publication and potentially for the publisher and/or advertiser by downloading the decoding software or using the electronic service. If the application is for buying a product, the advertiser gets the product price from the consumer. The consumer pays for the operator for using the data connection.
The publisher can benefit from the applications by:
The advertiser can benefit from the applications by:
4.1.2 Ordering an Article to e-mail When ordering an article to e-mail the publisher maintains the electronic service and offers the decoding software. Consumer downloads the software when using the service for the first time. After decoding a request to send the particular article to a given e-mail address is sent to the electronic service. Figure 39 presents the parties involved.
Figure 40 presents the money streams in the application. The consumer pays for the publisher for buying the publication and potentially for using and/or downloading the software as well as for the operator for using the data connection.
The publisher can benefit from the application by:
4.2 Investments In order the publishers and the advertisers to be able to offer services based on two-dimensional bar codes, investments have to be made in order to create an overall system:
4.3 Summary The parties involved in Print Access applications are:
Publisher and/or advertiser pay for the building and maintenance of systems needed in applications. The investments relate to:
From the consumer profits can be made by:
The first two options require that the consumer finds the services useful enough to be willing to pay for them. Instead of charging the consumer, the publisher can also make profit from advertisers by:
In general the publisher is a better option for maintaining the system than the advertiser. In this way all the publications from one publisher use the same system thus forcing the consumer to download only single software for all services. If the advertiser is the administrator, every advertiser might use its own decoding software. In this case the consumer is forced to download different software for every advertisement thus making the situation confusing. The simplest solution seems to be that the operator or the phone manufacturer installs the decoding software into the phone before consumer purchases the phone. When planning of the PrintAccess project was started, it was argued that to maintain competitive in the long run, printed media need to be somehow integrated with digital media. The argument was consistent with timely paradigms such as for instance media convergence and ubiquitous computing, or "ubiquitous photography". Printed codes studied in the project offer a way of integration. They can be produced without investments in new manufacturing technology by printers. The proliferation of camera phones on the other hand guarantees that consumers have devices for reading the codes and need not make investments either. Previous efforts to introduce applications based on printed codes were dwarfed by the need for separate reading devices. The project focused on 2D code technologies and readability, reading software, code based applications (mobile and digi-TV), usability and workflows. In code readability study it was found out that the smallest possible cell size in 2D-code was 200 µm when reading device was the Nokia 3650 camera phone with an add-on macro lens. That gives possibility to decode 172 ASCII characters or 348 numbers from Data Matrix code when physical size of the code is 9,6 mm × 9,6 mm. The add-on macro lens is an accessory for the Nokia 3650 mobile phone. The smallest readable cell size with same camera phone without an add-on macro lens was only about 600 µm. That gives possibility to decode only 10 ASCII characters or 24 numbers. Based on this, it is obvious that optics of the camera phone is remarkable effect for decoding performance. Anyway, since Nokia 3650 model was launched, pixels count of camera phones is constantly increasing and the same time the performance of optics is improving. Resolution of Nokia 3650 is 640 × 480 pixels. For example, Sony Ericsson has launched camera phone that has 2,0 megapixels image sensor and auto-focus lens with macro mode option. Especially, by force of auto-focus lens with macro mode it is possibility to get better decode performance than Nokia 3650 camera phone with tricky add-on macro lens. Nokia has also announced that they will launch phone model with integrated camera module of same performance level than in the Sony Ericsson phone. Anyway, couple of the next years most of the consumers in Finland will still have the mobile phones equipped with low performance camera module. In consequence of this, the cell sizes of the codes used with PrintAccess applications should be big enough. Also, the consumer should download the decoding software into mobile phone. Another alternative is that the decoding software is installed into consumer's camera phone by the phone manufacturer or operator before consumer purchases the phone. The idea is that decoding software is installed only into phone models with the high enough performance camera module. The former alternative was used in Smart Magazines published by Sopranos within the PrintAccess project. The latter alternative is current practice in Japan where decoding software is installed into camera phones by operators. Data Matrix was mainly used code type in the PrintAccess project. Meanwhile in Japan the QR-code type is widely used in commercial applications. In this project there was not carried out in-depth performance evaluation between Data Matrix and QR codes. That could be one valuable research topic in future. One solution in the future could be to use decoding software which recognizes both the Data Matrix and QR codes. Someone can think that black and white codes are disturbing. Someone else can think that it is important that codes can easily distinguish from content. Multi-color codes that were investigated in the PrintAccess project are easy distinguished and less esthetically disturbing than black and white codes. Next generation code technologies, based on information hiding into printed image, will offer invisible codes for human eyes. That kind of code technologies will be focus on the incipient PrintInteract project. RFID tags could include same information than printed codes and reading device can be used as non-orientation way. Anyway, RFID tags need a specific reading device and in contrary of printed codes, the purchase of it brings extra investment to the consumers. Also, in the PrintAccess applications area, the manufacturing costs of tags have to decrease and the possibility to print tags directly to the paper surface must be possible. One scenario in future could be RFID-tags in parallel with printed codes. The choice between technologies would be application specific. In addition to the 2D-code reader program developed in the PrintAccess project, there can be found plenty of commercial reader programs for mobile devices. These programs are developed i.e. in Germany, Russian and USA. So it is certain that in addition of Japan and Finland there is found much of willingness to make business with 2D-codes and mobile phones. In the PrintAccess project seven pilot applications in the mobile environment were implemented and theirs usability were evaluated. Also four pilot applications for the digital television and one application using RFID-tag were investigated. Based on the opinions of the test persons the code detection is more agreeable alternative for way to order the service than SMS-messages if the detection would be faster and easier. Anyway, the new version of the ProdInfo reader software is faster than version that was used in usability study. Subjects' opinions of interestingness of different pilot applications were dispersed. The average feelings were positive and in the future there will probable come out different applications for different user segments. Probable younger people are more interesting applications like game downloading and older people are more interested applications like sending article to Email or home info. Summarizing, it is evident that code based integration of print and mobile device with integrated camera has reached sufficient technological maturity for commercial applications. It is also evident that more researching is needed to development business models and application specific workflows between different partners. Commercial applications in the digital television environment will be workable in the future when technology of the required devices is more mature. REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
