2
Counterfeiting Technology Trends

Current digital technology has evolved over the past 20 years in a number of ways. Some of these were predicted, but many—especially the low cost for high quality—were difficult to foresee in the early days of photoreproduction. The digital technology revolution has had a profound effect on counterfeiting technology. Each of the three major steps involved in digital imaging technology is used in counterfeiting: (1) acquiring the image, (2) processing the image, and (3) printing the image.

Each of these three steps may be carried out in either binary or analog mode, there may sometimes be hybrid processes, and steps may be done in various combinations. Both binary and analog images have advantages and disadvantages. Producing images in binary systems requires patterning the image in such a way as to make it appear to have gradations; upon close inspection, these are dense on-and-off spatial patterns. Each pixel must be printed with either ink or no ink. Analog printing produces images with an array of elements that have different optical densities. Hybrid systems can draw from both paradigms, but the results tend toward the artistic rather than producing multiple images that are all closely identical.

IMAGE ACQUISITION

To replicate the Federal Reserve note (FRN) using digital technology, a counterfeiter must first transform an analog copy—a physical note, printed by the Bureau of Engraving and Printing (BEP)—into a digital format. This is also called image capture. A variety of low-cost, high-quality consumer-grade devices for image capture are available. Capture devices can also be custom-built by enterprising counterfeiters. Table 2-1 provides a summary of image-capture devices, their capabilities, and relative costs.

Scanners

U.S. currency is approximately 2.5 by 6.0 inches, which means that the scanned image of a U.S. banknote from a 2,500 pixel per inch (ppi) scanner would be about 6,250 by 15,000 pixels. This would result in an RGB (red, green, blue) file size of about 94 million pixels, which is well beyond the performance capabilities of a digital camera. A counterfeiter would have to piece together several images taken with a digital camera to use this method. However, this image size is well within the capabilities of even lower-priced home copiers and amateur flatbed scanners.

An artifact of the image-scanning system revolves around the fact that the majority of the printed features on FRNs are binary, meaning that at every pixel, ink is either deposited on the substrate or it is



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes 2 Counterfeiting Technology Trends Current digital technology has evolved over the past 20 years in a number of ways. Some of these were predicted, but many—especially the low cost for high quality—were difficult to foresee in the early days of photoreproduction. The digital technology revolution has had a profound effect on counterfeiting technology. Each of the three major steps involved in digital imaging technology is used in counterfeiting: (1) acquiring the image, (2) processing the image, and (3) printing the image. Each of these three steps may be carried out in either binary or analog mode, there may sometimes be hybrid processes, and steps may be done in various combinations. Both binary and analog images have advantages and disadvantages. Producing images in binary systems requires patterning the image in such a way as to make it appear to have gradations; upon close inspection, these are dense on-and-off spatial patterns. Each pixel must be printed with either ink or no ink. Analog printing produces images with an array of elements that have different optical densities. Hybrid systems can draw from both paradigms, but the results tend toward the artistic rather than producing multiple images that are all closely identical. IMAGE ACQUISITION To replicate the Federal Reserve note (FRN) using digital technology, a counterfeiter must first transform an analog copy—a physical note, printed by the Bureau of Engraving and Printing (BEP)—into a digital format. This is also called image capture. A variety of low-cost, high-quality consumer-grade devices for image capture are available. Capture devices can also be custom-built by enterprising counterfeiters. Table 2-1 provides a summary of image-capture devices, their capabilities, and relative costs. Scanners U.S. currency is approximately 2.5 by 6.0 inches, which means that the scanned image of a U.S. banknote from a 2,500 pixel per inch (ppi) scanner would be about 6,250 by 15,000 pixels. This would result in an RGB (red, green, blue) file size of about 94 million pixels, which is well beyond the performance capabilities of a digital camera. A counterfeiter would have to piece together several images taken with a digital camera to use this method. However, this image size is well within the capabilities of even lower-priced home copiers and amateur flatbed scanners. An artifact of the image-scanning system revolves around the fact that the majority of the printed features on FRNs are binary, meaning that at every pixel, ink is either deposited on the substrate or it is

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes TABLE 2-1 Capabilities of Digital Image Capture Devices Capture Device Resolution (pixels per inch) Cost Comments Digital camera Varies Low to High Limited number of total pixels available Simple consumer-grade reflective flatbed scanner 2,500 Low Ideal for counterfeiting owing to high quality and low cost High-quality reflective flatbed scanner 2,500-4,000 Moderate High quality with only slightly higher costs Professional drum scanner 4,000-5,000 High High cost presents barrier to entry Artwork software No limit Free to Moderate Time consuming and limited to use by experienced artists not. This is characteristic of virtually all of today’s high-quality printing processes.1 An image sourced from an input scanner is usually retained, not in binary form, but in analog. This means that each pixel of the image is retained with more than 1 bit of information, usually at least 8 bits, or 256 levels of intensity. These levels are then sent to the image-processing software. When the image is prepared for printing, however, the bit depth is reduced again to 1 bit because most electronic printers are binary devices. Most often, because halftoning or similar techniques are used to generate the appearance of gray in the image that is printed, the sharp binary image is lost in the transition to analog and back again. A scanning resolution of up to 4,000 to 5,000 pixels per inch is currently available on the best graphic arts scanners. These scanners are often referred to as copy-dot scanners because they attempt to copy all of the halftone dots in the original. Printing on currency paper, however, is difficult above 2,500 pixels per inch because of surface roughness. Thus, advances in image capture will not improve image acquisition for counterfeiting since that is at a practical limit already. An exception to the binary nature of printing on the current U.S. banknote is the watermark. This is not strictly a binary image, and it is likely that analog image capture would do a better job than binary could in reproducing a watermark through printing. Artwork Software The process of visually inspecting currency and then re-creating it line by line using software illustration programs is a painfully slow process. However, this approach is also one in which great detail can be properly executed. Re-creating a design feature by feature may be more efficient, and programs such as Adobe Illustrator, Adobe Photoshop, Adobe Pagemaker, and Quark Xpress provide a rich set of tools with which to generate exceptionally high-quality images. In addition, these programs can leave the created or re-created data in a form that can allow “resolution-independent” printing options. This means that the quality of the output will be materially dependent on the quality of the printer. This is generally not true of scanning or other pixel-by-pixel methods. Bitmap editing, or pixel-by-pixel editing, can be done using very inexpensive software such as Microsoft Paint, but it would take a tremendous investment in time and a good dose of eyestrain to generate a useful image this way. While illustration programs and software will improve over the next several years, the hard work of faithfully capturing a “squiggly” line bit by bit will be difficult to turn over to the machine. Using an illustration program in this manner is not unlike engraving the false duplicate by hand. Thus, while a software rendering of currency in a pixel-by-pixel fashion is a viable approach, it is likely to be uncommon. The labor content of such an approach is not expected to abate in the next 5 to 10 years. 1   Photographic film and thermal dye printers are two exceptions in that they are analog rather than binary technology.

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes Image-Acquisition Implications Image capture for the average consumer or user is at a high state of development. Flatbed input scanners costing a few hundred dollars or less can now scan at high quality, at reasonable speed, and with spatial pixel densities up to 3,000 pixels per inch or higher. This resolution is more than adequate to capture most printed material very well, especially with some post processing of the image to enhance fine features. While there may be improvements in capture technology, such as increased bit depth of the image and larger tonal ranges, increasing the scanning density beyond an effective 4,000 pixels per inch is both impractical and unnecessary. Current flatbed scanners usually use a single silicon sensor chip that has sufficient sensors to cover the entire page image. This means that if one is scanning at 3,000 pixels per inch, the chip would need about 25,000 sensors to cover an image 8½ inches wide. Furthermore, the imaging lens must image the platen of the scanner at sufficient quality to permit the sensor density to be adequately utilized. Image scanner design always presents a trade-off between speed and spatial resolution. The light sources used in most flatbed scanners are either special fluorescent lamps or light-emitting diodes, and the lens f-number, or ratio of focal length to optical aperture, must be such that it can adequately generate the required flux on the silicon sensors in the time required while maintaining the required image quality. This is often no easy task, but significant progress has been made in the past few years. It thus appears that the quality available today for image acquisition may not improve greatly. Consider that improved scanning pixel density will not be needed for most applications and that increasing this density would require costly improvements in optics and storage. The profitability of flatbed scanners is not high enough to warrant large investments for specialty or niche applications. In addition, increasing the bit depth to 12 or 14 bits per pixel would significantly improve the quality of scanned continuous-tone material such as film or other analog images. However, because authentic currency is printed using a binary process, this would not improve the scanned image quality of banknotes. Digital cameras will certainly continue to improve their quality levels, and performance-for-cost will significantly increase in the next few years. Cell phone cameras are becoming so ubiquitous and functional that an increase in quality will make them a serious threat. Today, digital cameras cannot provide the image quality of a flatbed scanner owing to many optical effects as well as to limited pixel density. Capturing the approximate appearance of security strips and watermarks, however, can be done quite well with either digital cameras or flatbed scanners. The use of masking software, automated subroutines, and other such software shortcuts can greatly reduce the amount of work required of a determined counterfeiter. Less with hardware than with artistic design software, innovations are made and integrated into each generation of upgrades, making it easier for the skilled user to reproduce complex effects. In addition, software that seeks to reproduce the look and feel of traditional art materials has advanced by leaps and bounds in the past few years and continues to improve—intaglio being a traditional medium that can now be simulated more easily than ever. In summary, innovations in flatbed scanners, artistic design software, and other methods for capturing images are not likely to be an area in which technology will further aid the counterfeiter in the next 5 or 10 years. Scan quality along with the capture of color features and other image characteristics is already adequate for many counterfeiting activities. IMAGE PROCESSING Once an image has been captured, it must undergo a number of processing steps before a good reproduction can be made. Some of these steps, listed below, would likely be used in preparing an image for high-quality reproduction. Of course, poorer-quality reproductions are possible without this additional effort.

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes Removing scanning artifacts such as dust specks; Adjusting the brightness and contrast of the image; Performing color adjustments; Applying software filters to enhance edges and sharpen the image, often called unsharp masking; and Rotating of the image if it has been scanned at a small angle.2 Image-Processing Capabilities One of the significant improvements in image processing has come about through generally available software. Packages such as Adobe Photoshop, Microsoft’s Digital Image Pro, and others provide powerful tools for improving and enhancing digital images. Personal computer (PC) operating systems also are sold with image tools that can be accessed by the most casual users. The capabilities of the very low cost tools are expected to increase with new operating system releases. The growing demand for accessible digital photography tools makes image processing a very rich area of innovation that will be available to the counterfeiter. Additional tools available through online photo-sharing Web sites can easily—and anonymously—improve the reproduction of color images. While image scanning is approaching practical cost limits, image processing has just begun its advance toward impressive automation and application capabilities. Scanning of sharp lines such as those found in intaglio printing is a particular challenge. Often, the images of lines scanned at angles or lines that are wavy or that possess other artistic features will not necessarily reproduce with uniform thickness in an image. When printed, the lines can vary by a pixel or so, which can be quite obvious in certain circumstances. Often such lines need to be retouched to remove problematic artifacts. Software to perform this task automatically is under development; it would greatly reduce the labor cost of making high-quality images and eliminate the need for special skills. To prepare an image for most digital printing, the scanned image needs to be converted to a binary format. The scanned image can be halftoned, as in conventional printing, or it can be made without halftone patterns using an on-off process often referred to as thresholding. While technologies are in development, it seems that the practicalities of binary printing processes will be limiting for some time. Image-Processing Implications Image processing will no doubt be an area of significant development in the next 5 or 10 years. The modern personal computer can now readily handle gigabyte-size images. A $20 bill captured at 4,000 pixels per inch, in color at 8 bits per image plane, is not a particularly large data file. A 2.5 by 6.0 inch FRN, at 4,000 pixels per inch, represents only 720 megabytes of disk storage space. This amount of data would have been considered immense only a few years ago; today, however, storing such an image costs only about 50 cents! Because capture quality is already at a high level of fidelity, image processing can be used to enhance the scanned image and can enable high-quality printing. However, most of the software available today, while having somewhat sophisticated automated tools, delivers the best results when it is in the hands of experienced users. In the next 5 years, much of what occurs today in application software may be incorporated in the scanner or operating system. The growth of the digital imaging market will require that such capabilities be close to automatic. A few years ago, one could do a very good job of processing color film and prints at home. To do so required a darkroom, chemistry, some basic photoinstrumentation, and a lot of experience if it was to be done well. As is clear from any number of trends, the darkroom is now digital, 2   This can be quite important in banknotes with features that rely on front-to-back registration. However, today’s U.S. banknotes do not utilize such a feature.

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes and today’s experienced digital imaging enthusiasts are constantly improving their skills. However, like their predecessors in chemical photography, most do not “mix their own chemicals”; they use what is commercially available. This is becoming easier; several products introduced in 2005 will automatically produce multiple versions of a color-corrected image, requiring the user simply to select the best image from a group. Most users want to take a picture with their digital camera or scan an image with their scanner and then be able to print out a faithful reproduction of the original. While good images can be readily obtained today, faithful reproductions are not so easy or repeatable. Color-management tools are available in image-processing software and are intended to deliver good color reproduction. A skilled user will be able to match colors easily, and these tools are becoming accessible even to casual users. Technology development will result in increasing automation in image processing in the next 5 to 10 years. Techniques such as automatic contrast adjustment, unsharp masking, line width control, and feature smoothing are all under development as automatic features. The driver for these changes is to make the user’s experience as pleasant and simple as possible. Automatic comparison of the input and output image is currently available in some software packages. Significant improvement is expected in image reproduction because the market will demand it. In 1981, SciTex provided a stunning image-processing system for graphic artists that cost approximately $1 million. Today, systems that dwarf SciTex system capabilities can be purchased along with their host PC for less than $2,000, a 500 times performance-for-cost improvement. There is no indication that we are nearing the limit of software or computing technology. The ability to use image-processing software to customize individual notes by changing the serial numbers, plate numbers, or other identifiers may enhance the ability to pass a counterfeit note. IMAGE PRINTING Image printing is the final step in digital reproduction except for finishing operations, which would include cutting, trimming, and the addition of simulated security elements. Electronic printing began more than 35 years ago with the invention of the laser printer. When the first laser printers were built, no one could have envisioned the quality levels that would be attainable today. The same can be said of early ink-jet printers which were interesting, but around the time of their invention they were never seen as being capable of what can currently be purchased today for less than $100. Where will these technologies go in the future? What might limit quality improvements from both a technological and operating performance for cost standpoint? The basic categories of printer technology are described in the subsections below, where both innovations and variations on the processes are considered. The electronic printing technology area is in a constant state of flux, and other fields of development including printed displays and electronics may spur the development of printing well beyond the paper-based markets of today. Often, one field drives technology that provides for another field significant opportunity that would not have otherwise occurred. Printer features and their capabilities are also summarized in Table 2-2. Note that “printers” also include copiers, or devices that only scan and print an image. This encompasses commercial copiers, but also stand-alone multifunction devices that may act as printers, scanners, and fax machines on the home desktop. While ink-jet technology has the highest number of printers in the market, electrophotography—used in laser printers and most large copiers—provides the maximum overall revenue to the industry. Both of these printer technologies are lucrative investments for corporations, and for this reason their capabilities are expected to continue to advance. It is interesting to note that printer business lines are highly profitable, whereas scanner business lines are not. Currently, the revenue generated by consumables such as toner, ink, and paper makes the difference.

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes TABLE 2-2 Some Printer Technologies and Capabilities Technology Resolution (µm) Cost Comments Electrophotographic (laser) 7-10 Low Potential primarily for low-volume counterfeiting Ink jet 7 Low Attractive for all classes of counterfeiters owing to low cost and high resolution Thermal 2 Low Limited by low-image-quality substrate chemistry requirements Chemical NA High High cost, variable quality, and limited availability of adequate substrate materials Photographic NA High Produces quality similar to that obtainable with ink-jet printers but at a higher cost Electrophotography The electrophotographic process is the basis of the most widely used document-copying machines.3 It begins with a photoconducting surface that is uniformly statically charged. In many copiers, this is a metal plate with a selenium-based coating. The charged surface is then exposed to a pattern of focused light, usually from a laser or light-emitting diode. This pattern is the image to be printed. Where light falls, the charge dissipates and a “charge image” of the light pattern remains on the photoconductor surface. The image is developed by dusting the charged surface with a pigmented powder, called toner, which is attracted to the charged areas of the pattern. The powder is then transferred electrostatically to paper and, finally, the toner is fused to the paper with heat. A continuously rotating metal drum moves the plate and paper through the various steps—charging, exposing, developing, and transferring—in a seamless manner. The quality of electrophotographic images has improved continually since its introduction in the mid-1900s. The critical elements of electrophotography technology for counterfeiting are the development, transfer, and fusing steps. These control the image quality and are also where advancements have resulted in higher quality and lower cost. In the development of the image, in which the toner material forms the text and images on the printed paper, the size of the toner particles controls the print resolution. As first implemented, particle size in toner averaged about 12 micrometers (µm). In order to improve image resolution to 600 dots per inch, the particle dimension was reduced to 8 µm and the uniformity of particles was improved. Further reductions in size will be necessary for improvements in resolution. Toners that are substantially smaller than this are not only hard to manufacture properly but also are difficult to keep confined to the printer itself; that is, they can be become airborne. Toners with particles as small as 1 or 2 µm can be made, but they can be hazardous if they become airborne and are inhaled by users. Advances in the materials properties of toner can also help improve resolution during fusing and transfer. By varying the chemistry of the polymer in the toner, printed images can be made to look more like ink on paper rather than having the look of electronic printing. Fusing can have serious implications for durability. Lower-cost printers often do not use fusing techniques that are as durable as those of the faster, more-expensive machines. Folding or creasing of the print can cause poorly adhered toner to fall off at the crease, thus exposing the white substrate and producing an obvious defect. As fusing technology improves, home printers could have the potential to do one-pass, two-sided printing. This means that 3   Xerox Corporation was the principal investor in electrophotography in its early days. It became more commonly known as “xerography,” apparently a catchier name than electrophotography.

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes registration of the images on each side of the substrate can be significantly better than that achieved by removing the printed sheet, flipping it over, and reinserting it for printing of the second image. While the quality of electrophotographic products is expected to improve in the next several years, significant engineering challenges remain that will limit the performance of these systems. Improvements in ink-jet printing, the major market competitor for lower-cost electrophotographic printers, will continue to drive print quality. Ink-Jet Printing While ink-jet printers predominate in the home printer market, they are also capable of producing some of the highest-quality images of any type of printer. An ink-jet printer is any printer that shoots extremely small droplets of ink on to paper to create an image. The droplets form small dots on the printed page; these dots can range from 10 to 30 dots per millimeter. The placement of the dots can be very precise. Ink-jet technology is particularly useful because registration in the system is an electromechanical issue that can be solved by straightforward engineering methods and techniques. For this reason, flatbed ink-jet printing has the greatest potential for future use in home printing of high-registration, simultaneously printed front and back images. One disadvantage of this method is ink bleeding into the paper, resulting in a blurry appearance on some types of paper. These effects may be pronounced with certain types of paper, especially when the ink is water-soluble. The introduction of ultraviolet-curable, water-impervious ink-jet inks would solve the problems of stability, lightfastness, water solubility, dot gain, and spread. Ink-jet printers fall into three categories: continuous, thermal, and piezoelectric ink jet. Their features and applicability to counterfeiting are described next. Continuous Ink Jet Continuous ink jet is one of the oldest ink-jet technologies in use; it is fairly mature and arguably produces the best-quality image obtainable on any type of printer. Continuous ink jet is actually better than photography in many cases in reducing image noise. Continuous ink-jet printers are not, in general, printers for home use. They are often large, and a fair level of maintenance and operator skill is required to get the most out of the device. In continuous ink-jet technology, a high-pressure pump directs liquid ink from a reservoir through a microscopic nozzle, creating a continuous stream of ink droplets. One of the advantages of continuous ink-jet printing is the very high velocity of the ink droplets, which allows the ink drops to be thrown a long distance to the target. Another advantage is freedom from nozzle clogging, as the jet is always in use. Another key advantage of continuous ink-jet printers has been their ability to produce small drops of ink that can be repeatedly printed at the same or carefully adjacent spot to vary the effective droplet size. This capability produces stunning quality. It is not expected that continuous ink-jet printer technology will move significantly forward in use in the next 5 to 10 years for two reasons. First, they are inherently complex to maintain and operate. Second, newer ink-jet printers or other types are now approaching the same quality levels attainable using continuous ink-jet technology. Thermal Ink Jet Thermal ink-jet printers, also known as bubble-jet printers, are widely available for home use. These printers operate by rapidly heating a small volume of liquid ink and forcing a steam bubble to form that ejects the ink from an orifice. As the bubble cools, the vacuum created draws fresh ink back into the nozzle.

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes A large investment in this technology has produced a remarkably reliable process. The print cartridge is highly engineered, with 64 or 128 tiny nozzles that are produced using photolithography. To produce an image, the printer runs a pulse of current through the heating elements. Each of the tiny chambers can fire a droplet simultaneously to produce the image. A disadvantage of this technology for counterfeiters is that the ink must be water-based. Such inks are inexpensive to manufacture, but they may perform best on specially coated media that do not have the feel of currency paper. However, because the print head may be produced at less cost than that for other ink-jet technologies, it is a relatively low cost process. Research and engineering are currently aimed at producing multiple drop sizes on demand. However, it is not an easy task to control the bubble and its associated heat cycle to produce different-size bubbles reliably. Possible advances in the next 5 or 10 years may also lead to increased process speed. It is not clear at this stage if either of these advances is likely to be realized for home users or only for higher-end users. Piezoelectric Ink Jet Most higher-end ink-jet printers use a piezoelectric crystal in each nozzle instead of a thermal heating element. “Piezo-jet” printers utilize a piezoelectric actuator to produce pressure in the liquid-ink chamber. The pressure and drop size generated by the piezoelectric actuator can be controlled better than in bubble-jet printers and can also produce smaller droplets of ink. These variable-size drops and smaller droplets can generate impressive images. Today, the smallest commercial droplet size available is in the range of 1 picoliter. This corresponds to a drop diameter of about 7.1 µm, or about 3,500 or so drops per inch. Such small droplet sizes are generally not attainable in the bubble-jet schema. However, this very small drop size can be lost on typical paper used for home printing, so it is unknown whether the investment in such improvements would be profitable in the home market over the next 5 to 10 years. An advantage of piezoelectric ink-jet printing is that it allows a wider variety of inks than does thermal or continuous ink-jet printing. The emerging ink-jet material deposition market uses ink-jet technologies, typically piezoelectric ink-jet, to deposit materials on substrates. Printers that can deposit glues, resins, or waxes onto a variety of substrates have been available since the early 1990s; they are called solid ink-jet or wax-jet printers. While it has advantages in color intensity, solid ink-jet technology has limited droplet size compared with that of bubble-jet or piezo-jet liquid-ink printers because the viscosity of the melted material complicates the jetting of small drops. The piezoelectric ink-jet technology is currently available on commercial flatbed digital presses, and while they are currently very expensive, they are easily and quickly configured for counterfeiting. This is a major departure from the setup time required for commercial printing presses; it could allow counterfeiters who might work in printing shops to use the digital equipment without alerting their management. In addition, advances in this technology in the next 5 to 10 years may mean that this level of quality may be available for consumer units. Advances may also mean that metallic or plastic particles could be “printed” and someday may be able to simulate some non-image features. For these reasons, piezoelectric ink-jet printers may prove to be the most useful technology to counterfeiters now and in the future. Thermal Printing Thermal printing commonly is done in one of three forms: thermal transfer, thermal dye transfer printing, and custom-substrate thermal printing. Thermal transfer, one of the early printing technologies, has come and gone as a technology of choice. Its main advantage has been its reliability and high edge acutance. However, achieving printer pixel density higher than 1,000 pixels per inch is problematic owing

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes to the requirements for the heater heads. It is not expected that thermal transfer will be able to compete with bubble-jet, piezo-jet, or electrophotographic images as time goes on. The second form of thermal printing, thermal dye printing, has also been referred to as dye sublimation printing. However, in most cases, sublimation is not the physical process that causes image formation, and hence “thermal dye printing” is the preferred nomenclature. Thermal dye is much like thermal transfer, except that instead of transferring material from a ribbon surface to a substrate, dye is migrated from a ribbon and absorbed into a special substrate under pressure. The main advantage of thermal dye is that it can produce true gray-scale images without any of the usual halftone patterning or other geometrical structures that trick the eye into seeing gray-scale. Such printers produce images very close to those achieved in photography. The main disadvantages of thermal dye are limitations in spatial resolution and speed. Currently, about 600 pixels per inch is the maximum practical density, which is generally low for quality currency generation. The third form of thermal printing uses custom chemically-treated paper and prints with heater heads not unlike those used in thermal transfer. It generally cannot produce good colors, and like the other forms of thermal printing, this technology is generally low quality and is unlikely to improve in the future to any extent. The specialty substrate, in which the chemicals used for image generation are in the paper, would significantly complicate its use for currency counterfeiting. All of these technologies are slow, expensive, and require special substrates to achieve the promised quality. Because none of them is a focus of current industry innovation, concerns about future thermal printing developments with respect to counterfeiting are likely unwarranted. Chemical Printing No imaging process has seen its world grow smaller more quickly than that of conventional chemical photography. Photography can produce outstanding images, but such images are no longer unique to this process. Electrophotographic and ink-jet processes can produce images in many ways as good as those of photography. Because genuine currency is produced in general with commercial processes—offset, intaglio, and letterpress printing—the images are binary, meaning that each pixel on the banknote either has ink or it has no ink. Hence, photography offers little value over simpler digital processes that are much more reliable and lower cost. Photography is still of use in the generation of printing plates, but this area is also quickly becoming digital. There is little current activity in either using or improving the quality of nonphotographic chemical printing, and it is not expected that this technology will benefit counterfeiters in the foreseeable future. Problems of chemistry control, quality, and maintenance were the factors that killed the early chemical copiers when electrophotographic processes came on the market. These factors are still a barrier to this technology today. Image-Printing Implications Modern digital printers available to ordinary users can provide stunning quality at or above that available from photochemistry just a few years ago. The various electronic and other printing processes described above have varied advantages in image quality. The quality of digital presses and advanced electrophotographic printing has risen to the point that it is expected to replace offset printing in the near future. This means that the value of using offset printing as a tool to deter the counterfeiting of the images on banknotes is limited. As in image capture, image printing is now at pixel density levels—2,400 to 3,600 pixels per inch—that are at the limits of human vision. However, improvements in image quality could come from printing images other than in binary mode. Current ink-on-paper printing such as offset and intaglio are binary processes. To simulate the look of gray—as in a watermark—commercial printing devices and most electronic printers print black ink in a patterned fashion. This patterning results in “aliasing” artifacts,

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes because lines are not single strokes but a collection of dots that look at normal viewing distance like single strokes. The most useful feature of electrophotographic laser printers is that the laser beam can be turned on and left on while it is writing horizontal image structures such as a horizontal rule or line. Vertical rules are created by “stacking” pixels on top of each other. The ability to keep the beam on in the horizontal (often called the fast) direction yields higher-quality lines in electrophotographic as compared with ink-jet printers. These timing issues that affect beam placement are well understood and well used. New approaches in both ink-jet and electrophotographic printers can produce some levels of gray, and multiple gray-level printing could produce dramatic improvements in image quality. In electrophotography, toner particulate size control and toner management as well as fusing technology (which depends on toner chemistry) will continue to set the limits for electrophotography laser printers. Current ink-jet printers must produce only one drop of ink at a time, and each drop must be placed adjacent to the previous drop or drops to produce a line. Such lines often look a bit ragged and do not reproduce intaglio-printed structures well. The size of the ink-jet orifice and the ability to print reliably using small orifices will set the limits for ink-jet pixel density. Inks with multiple densities of the same hue may achieve different tones in the image but with much-reduced patterning. This approach has the advantage of using existing transport and mechanical systems and changing only the supply packages for printers in place. Thermal printers can produce excellent vertical or horizontal lines, but their limited pixel density compared with that of electrophotographic and ink-jet printers is expected to render them inferior both now and in the future. Market drivers are a final consideration for high-end printing equipment. One result of gains in desktop technologies and their capability to produce professional-looking results is that small professional printing houses across the country are going out of business and often selling their equipment for as little as 10 percent of its original cost. This may bring a professional setup well within the price range of the petty criminal. With a good computer, a pirated copy of Adobe Photoshop, a digital direct-to-plate machine, and a used Heidelberg press, a petty criminal with considerable skill might start an offset money-printing workshop in his garage. Straight-to-plate technology, in which a computer design can be digitally inked onto printing plates, can turn older equipment into very modern presses. This technology in particular may be a threat because the ease of setup is a significant advantage. SUBSTRATES AND ADDITIONAL ELEMENTS Each type of digital printing method discussed above has its own special substrate requirements for optimal image printing. While new substrate capabilities are always emerging, the nature of the various electronic printing technologies restricts the range and types of substrates that work well. Most types of paper work well in electrophotographic printers. The two principal restrictions in these printers are the ability of the paper-feeding system to handle the substrate material and the substrate interactions with the electrostatic subsystems such as image transfer and fusing. Many electrophotographic printers can easily handle light- to heavyweight papers, well within the range of currency substrates. Some newer printers can print on both sides automatically, but back-to-front registration of printed features would in general not be nearly as good as that achieved in offset printing. The look and feel of the substrate—including the paper surface, stiffness, and intaglio texture—are difficult for most counterfeiters and technologies to reproduce using current desktop image-reproduction technology. Ink-jet printing can be either more or less tolerant of substrate differences, whereas electrophotographic printers are less sensitive to substrate quality. Paper-feeding requirements are still important, but ink jets can generally feed a wider array of substrate thicknesses. Card stock, shiny or matte finishes, and vellums can be used, because the paper path is usually very short and throughput speeds are low. The only additional requirement of substrates used with ink-jet printers is that they permit

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes the ink to dry and adhere to their surface properly. Liquid ink must to some extent wick into the substrate surface for adhesion. Solid ink is almost substrate-independent, although the quality of the image might vary. Features that affect light transmitted through the substrate—the watermark and security strip—also pose challenges, because most image reproduction is based on reflected light. Banknote features used in machine authentication (e.g., special inks) are difficult to simulate with desktop printing technology. While thermal printers are unlikely to be used for counterfeiting, they may be used as part of a larger system. For example, laser thermal printing could be used to apply special symbols or thin foil features over a previously ink-jet-printed page. Electrophotographic printing can also be used to affix foil or holographic features. SUMMARY OF DIGITAL IMAGING TRENDS The counterfeiting threats from the digital imaging system components are summarized here to contribute to an understanding of trends as they relate to planning for new currency features. Table 2-3 summarizes the usefulness of security features in deterring counterfeiting that uses digital age tools. A number of features for digital imaging tools expected to be introduced in the next 5 years would allow the casual counterfeiter to achieve what only a specialist can do today. Image-Acquisition Technology Digital images that appear on FRNs can be acquired using art software, digital cameras (including cell phone cameras), and a variety of digital scanning equipment. Current technology encompasses a variety of devices offering very cost-effective capture of images with adequate quality. In the future, improvements in consumer-grade scanners are possible, but they will not overcome the limitations on counterfeiting presented by substrate quality and the printing processes used to produce FRNs. Expected significant improvements in digital photography will enable the ordinary user to obtain results of the quality that professional counterfeiters can produce today. Image-Processing Technology Digital image processing of FRN images can be done using a wide variety of software tools, ranging in cost from free to inexpensive to very expensive. Current capabilities with such software are highly dependent on the skill of the user. In the future, substantial improvements in automation are expected to help the ordinary user process images like an expert. These improvements may include automatic contrast and brightness enhancement, optimal unsharp masking, and color balancing, and they could extend to many other areas. Significant increases in processing speed and automation capabilities may enable a user with little or no training to optimize images with high bit depth. Automated capabilities such as line-width control, uniform image appearance, and color balance would enable an ordinary user to easily obtain an optical image that is very faithful to the original. Image-Printing Technology The capability needed to print captured and processed images is the limiting step in the counterfeiting process at the present time. Current printers for home use, primarily thermal ink-jet printers, offer very-low-cost image printing. Other types of ink-jet printers and electrophotographic printer technologies produce counterfeits that can be passed, even though no counterfeits produced with such equipment are currently able to withstand minimal scrutiny by a trained money handler. Thermal

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes TABLE 2-3 Usefulness of Security Features in Deterring Digital Age Counterfeiting Features Ink-Jet Printer All-in-One Device Color Copier Flatbed Ink-Jet Printer Digital Press High-Quality Scanner Imaging Software Overt   Substrate ++++ ++++ +++ ++ ++ NA NA Tactility (or feel) +++ ++ +++ ++ ++ NA NA Watermark +++ ++++ +++ + + +++ + Plastic strip +++ +++ +++ ++ + + — Intaglio printing ++++ ++++ +++ + — NA NA Offset color blending ++ ++ + — — — — Optically variable ink +++ +++ +++ + ++ + — Intaglio microprinting ++ ++ + — — — — Offset microprinting ++ ++ ++ + — — — Colored threads ++ ++ ++ + + — — Machine-readable   Paper fluorescence +++ +++ ++ — — NA NA Magnetic ink ++++ ++++ ++ — ++ NA + Magnetic ink pattern ++++ ++++ ++++ + + NA NA Color-shifting inks ++++ ++++ ++++ + + NA — Digital CDS ++++ — — + — + +++ Digital BDS — — ++ — — — — Fluorescent thread ++++ ++++ +++ — + NA NA NOTE: ++++, high deterrence value; +++, good deterrence value; ++, some deterrence value; +, low deterrence value;—, does not use this technology; NA, not applicable; CDS, counterfeit deterrence system; BDS, banknote detection system. printing and digital photography are available but are not specifically useful for printing counterfeit notes; however, they are useful for simulating specific features. In the future, the capabilities of electrophotographic printers will continue to advance in terms of image quality, image maintenance, and color quality. While fine lines and small details may be possible to reproduce through the use of smaller toner particles than are available today, the introduction of particles smaller than 4 to 5 micrometers is unlikely because of environmental and performance factors. A more useful improvement in electrographic printers for counterfeiters would be an improvement in gray-scale printing. Having even four levels of gray compared with the option for on-or-off imaging of most current products could provide significant advantages in image appearance. Ink-jet printers will continue to improve, but it is unlikely that droplet volumes will fall below the current level of 1 picoliter. Improvement in the number and design of ink nozzles is also expected to increase the print speed, but more importantly, would also allow for the use of inks with the same color but differing density, thus improving both color and gray-scale image printing. Variable droplet size and placement, combined with new ink formulations and innovative curing cycles, may also help impart texture to the note or could enable printing of optically variable features.

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes An important class of printers consists of copiers, or devices that only scan and print an image. This includes commercial copiers, but also stand-alone multifunction devices that may act as printers, scanners, and fax machines on the home desktop. When these are used only to scan and copy, they may bypass the image-acquisition and processing steps. EMERGING TECHNOLOGIES The growing availability of low-cost, high-performance hardware and software for scanning, processing, and printing images remains the primary threat to secure currency. However, the emergence of some technologies that go beyond image printing may soon overtake those recognized methods. The Convergence of Printing, Manufacturing, and Biology A number of emerging tools use technology invented for printing on paper, although the tools themselves are designed for other processes. Examples include the high-volume manufacture of microelectronic circuits,4 smaller-volume prototyping processes,5,6 and even printing of biological material to aid analysis.7 Printing-based manufacturing systems offer levels of resolution and registration that significantly exceed the current or projected capabilities of conventional ink-on-paper printers. The target applications for this technology—such as printing flexible electronic circuits—demand very low-cost high-throughput operation, and large-area printing. Goals include rates of 100 square meters per hour, with the ability to process substrates larger than 5,000 square meters with accurate registration. The range of “inks” that can be printed by these systems include not only the dyes and toners used in conventional printers, but also color-shifting and other optically active inks. Functional materials for the semiconductor elements of thin-film transistors or the electroluminescent layers of emissive displays have also been printed. New ink-jet printing processes can also print “inks” made of a variety of metals, glasses, and plastics, which implies that simple printing processes may be able to simulate non-image features, such as the metallic print on the security strip. Although access to these manufacturing systems is not expected to be widespread, it will increase if their use broadens to more distributed manufacturing systems. It is also possible that some of the features of these advanced systems could migrate into low-cost desktop printers. Some specific technologies in this area include the following: Flexography, or flexographic printing, is a dominant printing technology for labels, tags, boxes, packaging, and objects with rough or uneven surfaces. This method uses flexible (often elastomeric) printing plates mounted on cylindrical supports and inked with rollers. The resolution and registration that can be achieved with standard systems is approximately 30 microns. It may be possible to integrate high-resolution printing plates, formed using soft lithography technology, with commercial flexographic printing systems to improve the resolution and registration by a factor of 10—to approximately 3 microns, or better. These systems are being investigated for their potential to print metallic interconnects in large-area circuits and antennas. High-speed ink-jet printing in manufacturing systems uses piezoelectric or thermal print heads for high-speed, large-area patterning—up to 100 square meters per hour. These printers involve hundreds of active nozzles, each operating independently, at frequencies of up to tens of kilohertz. Recent work focuses on applications in the manufacturing of display systems and 4   Proceedings of the IEEE. 2005. Special issue on Flexible Electronics Technology. Vol. 93. 5   A.V. Kumar and A. Dutta. 2003. Investigation of an electrophotography based rapid prototyping technology. Rapid Prototyping Journal 9:95-103. 6   See http://www-me.mit.edu/people/research/sachs.htm. Accessed April 2006. 7   See http://www.shimadzu-biotech.net/pages/news/1/press_releases/2004_07_23_chip.php. Accessed April 2006.

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes certain components of packages for microelectronics. Large-area, high-resolution displays that use organic light-emitting diodes or circuits patterned by ink jet are possible. Sophisticated control devices enable excellent repeatability in the printing. Machine vision provides the ability to perform registration in real time during printing. Experimental printers are capable of printing droplets with diameters of 5 microns or less, with registration at even finer scales, particularly when patterns of wetting and nonwetting regions on the substrate are used to confine the printed droplets. The range of inks that can be printed is broad. Screen printing uses a squeegee-type blade to push viscous inks through patterned openings in a screen mesh. Existing applications include the low-resolution patterning of decals, signs, and textiles. Screen printing is also commonly used to define some features on printed circuit boards. Recent developments indicate that improved printers and screens can achieve resolution near 10 to 20 µm with good registration. Laser-induced thermal transfer printing uses a focused laser beam to selectively transfer layers of solid-material “inks” from a “donor” sheet to a target substrate. The basic printing mechanism is similar to that of a conventional thermal transfer printer. The use of lasers in place of resistive heaters, however, can improve the resolution significantly, to levels that are comparable to the spot size of the laser, near 2 µm. The overlay registration is as good as the resolution. Large areas and high patterning speeds are possible with rigid and flexible substrates. Inks range from electroluminescent organics, to metals, to colored and black dyes, to semiconducting polymers, to carbon nanotube composites, to biological tissues. The most well developed potential applications of this printing technique are for the production of organic light-emitting display devices, and color filters for liquid-crystal displays. In the latter case, the printed material consists of stacks of charge transport and emissive layers. In the former, dye-doped polymers are used. Ink-jet printers can print a wide range of materials in addition to those commonly used for printing images. The “inks” include liquid suspensions of nanomaterials, such as carbon nanotubes and buckyballs, colloidal particles, nanoparticles, and nanowires—metals and semiconductors, with magnetic and dielectric properties. Biomaterials, including deoxyribonucleic acid, or DNA, have also been printed. Active materials—semiconducting or light-emitting materials—as well as passive dielectric and photocurable polymers are also printable. Many of these unconventional inks might be used to simulate features, such as optically variable images, that appear on currency. Ink-jet technology can also be used to pattern classes of materials that are themselves not directly printable. In these cases, printed polymers or waxes may be used as sacrificial masking layers for patterning other materials. After use, these layers can be removed. In this way, both positive and negative patterns can be produced by ink-jet techniques. Improvised Counterfeiting Devices Probably the most significant change in digital imaging trends in the past decade is the increasing availability of sophisticated software toolsets and modular hardware. While clever hobbyists have always been able to use innovative methods to produce counterfeits, it is important to remember that the dedicated counterfeiter may also be a skilled technologist. Improvised printed devices, made from parts available from various printer vendors, may enable smart operators to build hybrid devices for improved counterfeit image quality. No hardware or software is immune to tampering by smart users, and some users will always be able to evade the software countermeasures intended to prevent counterfeiting by either writing their own code or detecting the applicable subroutines. Enterprising counterfeiters may be able to produce or appropriate their own toner materials or inks and thus may be able to further improve existing printing technology. Modern currency is produced using three different printing processes—intaglio, offset, and letterpress—and special substrates that include watermarks and inserted identification strips. A growing

OCR for page 15
Is That Real? Identification and Assessment of the Counterfeiting Threat for U.S. Banknotes question is whether modern digital reproduction an d its developments over the next few years can simulate all features found in currency that is authentically produced. The use of anticounterfeiting techniques to protect FRNs is by its nature primarily a defensive approach. Advancing imaging and printing technology may be making better weapons available to potential counterfeiters faster than authentic currency is gaining deterrents; thus, the judicious exploration of the ways that a counterfeiter might exploit weaknesses in the currency system using available and improving equipment is an important part of any comprehensive defensive plan. CONCLUSIONS A range of excellent, reliable, and cost-effective digital printers for consumer use are available today at very affordable prices. Innovation and skilled engineering have resulted in this progress, and while innovation will continue, some physical limits may dominate the possible improvements in image quality. Image-capture, processing, and reproduction technologies, both current and predicted, pose a significant threat to the security of Federal Reserve notes—particularly because the security of FRNs depends on the casual viewing of two-dimensional printed features in reflected light. Emerging technologies are targeted at dramatic improvements in desktop capabilities. These improvements will continue to limit the ability of any two-dimensional printed image to deter widespread counterfeiting successfully. The committee concludes that images involving other classes of features—images viewed in transmitted light, light-reflecting features, or other complex optical features—offer a substantial challenge to primitive and hobbyist counterfeiters and a costly barrier to petty and professional criminal counterfeiters. An obvious consideration for the future is the goal of incorporating in image-processing software the ability to disable the processing of image patterns that are unique to currency in all digital tools. Because digital printing devices depend on software, the potential to disable the devices from reproducing these identified patterns is also a pertinent issue. Whereas simple copy protection may deter an opportunistic counterfeiter, the growing availability of online “hacks” means that a criminal with intent will not be deterred. A more sophisticated approach would be to add features to banknotes that intentionally frustrate image-capture capabilities or that generate unwanted patterns when scanned and processed images are printed. Some successful features on currency today are optical features that cannot be directly captured by present-day input scanners but can be seen by the human eye.