Q&A: The Future of Color
Q & A: The Future of ColorIt is common knowledge that the latest development in the document management industry is the emergence of color. The editorial staff at isitDOCIMAGING.com had the opportunity to invite six industry authorities to enlighten them with their views on color's future. To find out what these experts think about the outlook of color as it penetrates the industry, read on. |
1. What can color do in the future better than black and white and grayscale can do now? |
|
|
TMSSequoia Vaughn Henson, Senior Product Development Manager Capturing in color can simplify a system, making it less expensive to develop as well as more flexible. For example, using color images can virtually eliminate key from paper, because color images more closely reflect the original document, especially with colored inks and handwritten content. Exception handling can generally be implemented without rescanning, and can generally be handed closer to the point of exception without going back multiple steps in the process. Again, this is because color images don't lose vital information the way bitonal images do. In addition, the information in a color image can increase the accuracy of recognition which can reduce the need to perform manual operations. With the speeds of computers and networks increasing faster than the speed of scanners we believe color has a definate ROI potential now. |
|
Eastman Kodak Company Stephen Welk, Worldwide Marketing Manager, Mid-Volume Scanners, Document Imaging Color scanning takes document capture to the next level by providing images as the human eye would see them. Grayscale, although providing additional bit-depth, and hence, more information than bitonal, does not offer all the information that color can and in a way we are accustomed to. It also increases file sizes. Color maximizes image information capture by "seeing" everything on the page and conveys critical information lost in black and white capture. It also adds an intrinsic value to the way we interact with information. And color is the benchmark for Web-enabled processes, allowing for better use of Internet-related applications. It streamlines the capture process by reducing time and labor because there is less need to presort incoming documents by color types or backgrounds. |
|
2. What compression technology will become the standard for color? Why? |
|
|
Kofax Image Products J.D. Moons, Product Marketing Manager, Image Processing Business Unit JPEG is the standard for color capture. If the Internet embraces the JPEG 2000 standard, the document imaging world will follow. However, JPEG 2000 comes in many flavors. The multilayer JPEG 2000 format, specifically designed for document imaging, is the one that everyone is waiting for. It will provide both image quality and small file sizes. The first JPEG 2000 SDK's are emerging but only produce the single-layer wavelet-compressed JPEG 2000. These are already an improvement over traditional JPEG, producing good quality document images at about 60% of the file size of traditional JPEG. Kofax is committed to endorse single-layer and multilayer JPEG 2000 and add support of these formats in our products. |
|
Pixel Translations Mike Cohn, Director of Marketing There are 3 main contenders: JPEG2000, DjVu and TIFF-FX. I believe that JPEG2000 will win out in the future because it looks to be efficient and independent of any monolithic giant in the industry. |
|
3. In what amount of time will color become the majority requirement for Document Management/Imaging installations? |
|
|
Microsystems Technology, Inc. Martin A. Greif, Vice President of Marketing Color scanning is still very new in document imaging. Since it has not been available in mass, few are considering it to be a requirement at this time. The other consideration is how many documents exist where color is a factor? The vast majority of forms, invoices, documents, white mail, etc. are still black and white. In some specialized organizations where the forms are already standardized by color, color is a requirement now. Unless the types of documents dictate color, it will be a nice feature to have, but may never qualify for a must-have feature. |
|
Fujitsu Computer Products Victor Kan, Director, Product, Management, Imaging Products Group. Color will become the majority requirement for DIM installations as new color standards are implemented (i.e. JPEG 2000) as well as when the price for color scanning hardware decreases. The recent decrease in the cost of storage combined with the increase in network bandwidth has made color scanning more prevalent today. We estimate that color scanners will become the majority requirement in the next 18 months to three years. |
|
|
Stephen Welk, Eastman Kodak Company Prior to 1998, our research showed that 30 percent of the market would shift to color reasonably quickly if affordable products were available. On the positive side, we have seen a steady adoption of color since we introduced affordable production scanners. And now, other vendors have followed suite. However, growth expectations have not been as rapid as we would have liked due to a number of factors. It is hard to assess the speed of adoption although we know what factors need to be in place. End-users must be thoroughly educated to the cost benefits of color capture and color image utility. Experience has showed that a majority of end users exposed to a color alternative will seriously evaluate implementing color within their document imaging applications. |
4. What is the expected minimum DPI needed for color scanning? What is the desired requirement?J.D. Moons, Kofax Image ProductsFor good quality originals with 10-point text and standard 1D bar codes, 100 DPI is good enough. For degraded originals with high density, degraded 1D bar codes, or good PDF417 bar codes, 150 dpi is required. For documents containing 6-point text or smaller, or containing degraded 200 dpi PDF417 bar codes, or being used for OCR recognition, 200 dpi is required. Higher than 200 dpi resolution is only recommended for special applications containing small pictures that have to be scaled up, finger prints, extremely fine print, etc. |
|
Mike Cohn, Pixel Translations The minimum is a function of what the color content one is scanning and what amount of fidelity is needed later on. If one needs to merely view than 75 dpi would work. If one wants to print than 240-300 dpi or more may be required and if it is a photograph than even higher may be desirable. Fortunately high performance CPUs and compression will tend to control the file size increases that occur as resolution increases. |
5. Is Multi-stream output (simultaneous output of color, black and white and/or grayscale) a requirement for color scanning?Mike Cohn, Pixel TranslationsMultiStream(tm) is something Pixel created (and there are other functionality implementations) to allow a "best of both worlds" scenario, color data streams for viewing and bitonal for OCR/ICR so that existing applications can continue to work. I believe MultiStream-type capabilities will eventually be supplanted by "native" color processing something that we and others were showing at AIIM this year. In fact, the previewing of color image processing capabilities was one of the most asked for technologies in our booth at AIIM. We are committed to providing color image processing and are surveying our extensive customer base to get insights into EXACTLY what capabilities ISVs and end-users are looking for. So, MultiStream is a technology/capability that can be used NOW! based on specific application needs. Most of the newer scanners are producing multiple streams from the scan and our toolkits are able to accommodate this capability. |
|
Martin A. Greif, Microsystems Technology, Inc. It's not a requirement at this time. Our information capture software, OCR for Forms, can convert the color image to black and white for data processing. The color image is retained and passed along through the system for viewing and then for the final storage. Multi-stream output will eliminate a step on the front end but won't be a requirement. |
|
Victor Kan, Fujitsu Computer Products Multi-stream output is not a requirement for current capture applications today. Many ISVs will threshold a color image to black and white within the image since true color OCR processing is not available. Black and white images are required for OCR applications. The concept of multi-stream output was made for color scanners that could not output to black and white. Today's OCR engines can threshold a color image to black and white without having to get both from the scanner. |