|1||Please provide a short description of the state-of-the-art and/or current trends in the field? How does the result fit into it?|
|In many cases nowadays, big amounts of paper documents (financial, medical, etc.) have to be stored for years. This requires the existence of special stores, where certain temperature and humidity have to be maintained. The transformation of paper documents into electronic files offers many advantages concerning the specific requirements for their storage and facilitates their use. Most of the documents, which have to be archived, are text documents (newspapers, magazines, books), photos, medical images, etc. Together with the conveniences, which the electronic archiving offers when the requirements for the storage of documents are concerned, numerous problems exist as well. The electronic documents are usually stored in one of the widely used formats, like doc (for text documents), tiff, or bmp (for scanned documents), etc. As it is known, the doc files do not contain reliable identifying information or content protection. This is why it is suitable the documents to be stored as images and to be transformed into one of the widely used file formats: bmp, jpg, tiff, pdf, etc. The efficient storage and access of the archived files depend on their size (usually large) and this requires their compression. The main methods for image compression consist of two large groups: statistical and psycho-visual. The statistical methods comprise: Run-Length Coding (RLC), Lempel-Ziv-Welch (LZW) coding, based on dictionaries, Huffman coding, arithmetic coding, linear prediction coding. The psycho-visual methods are based on linear orthogonal transforms (for example, the Karhunen-Loeve Transform (KLT), the Discrete Cosine Transform (DCT), the Walsh-Hadamard Transform (WHT), wavelet transforms, etc.), vector quantization (VQ), pyramid decomposition, fractal transforms, etc. The traditional compression techniques are usually based on the standards for image compression, JPEG or JPEG2000, which offer very efficient lossy compression for natural images of any kind, but for same compression ratio the quality of the text or graphic information is significantly deteriorated. An example for such distortions is shown in Figure1, where the JPEG image is obtained after lossy compression with compression ratio 15:1. The same compression applied on natural pictures retains their visual quality to a high degreeOne of the very efficient methods for processing of compound images is DjVu, which decomposes the documents into three components: foreground, mask and background. Specific for DjVu is that the components are processed with different resolution, but this does not ensure lossless compression for texts and graphics in the image.Another significant problem for documents archiving is that the contents of the archived documents could be easily edited and their authenticity – violated. For some documents (medical, financial, etc.) the problems concerning the reliable content protection are of highest importance. One of the possible solutions is to use a “digital watermark” as a way to integrate some hidden identifying information in the still images. There are two main kinds of digital image watermarking: with “fragile” and with “resistant” watermark. The fragile watermark is used for image content protection against unauthorized access and editing. In the first case (unauthorized access) the fragile watermark could be applied as a “mask”, i.e. it overlaps the protected part of the image, and the watermark removal could be performed by authorized users only. The second application (protection against unauthorized editing) is based on the watermark “fragility”, because after editing of any kind the inserted watermark is destroyed (partially or in total). Another possible application is the data hiding (in this case the watermark carries the “hidden” information). The watermark extraction is permitted for authorized users only. Most of the methods for resistant digital watermarking are based on processing in the image spatial or frequency domain. The analysis of these methods shows that one of the most efficient approaches for still image watermarking is based on the modification of selected coefficients from the image spectrum, obtained using one of the famous discrete transforms, like these of Fourier (DFT), WHT, Fourier-Mellin (FMT), Radon (RT), etc. The basic advantages of the image watermarking in the spectrum domain are the practical watermark invisibility and the high resistance against various effects or manipulations, such as editing, lossy JPEG compression, affine transforms, cropping, contrast enhancement, linear and non-linear filtration, etc.The use of the fragile watermark usually needs lower computational complexity than that of the resistant one, but its’ efficiency is high enough to prove the image authenticity.|
|2||What is the problem/need/knowledge gap that the research result is responding to? How was it addressed before?|
|The renowned compression techniques, based on the JPEG and JPEG2000 standards are very efficient and widely known. The main problem is that they are not adaptive to image contents, when compound images are processed. In the proposed approach is suggested that the image content is evaluated and the parts, containing pictures, are compressed with lossy compression, while the part(s), containing texts (graphics) are compressed with lossless compression. For this will be used the two versions of the IPD method (for lossy and lossless compression). One more new technique is the ability to insert resistant and fragile watermarks in the processed images, which to be used to prove the image authenticity (with the resistant watermark) and possible editing (with the fragile watermark). The fragile watermark is used also to hide predefined regions of interest, which to be removed by authorized users only.The layered representation of the processed images (based on the IPD) permits the development of tools for hierarchical access in large image databases|
|3||What is the potential for further research?|
|The further research will be aimed at the development of more flexible compression techniques, based on the adaptive Karhunen-Loeve transform and branched pyramid decomposition. These new investigations will ensure significant impact for processing of multispectral and multi-view images; sequences of medical images, satellite images, etc|
|4||What is the potential of the research result for synergy with other research areas either in the same or in a different discipline?|
|The potential is in the areas of development of adaptive tools for efficient compression for scanners, faxes, photo cameras|
|5||What is the proposed method of IPR-protection? (patent, license, trademark etc.)|
The team already has 3 patents:
- European patent on the Inverse pyramid decomposition (R. Kountchev. Method and Device for Pyramidal Image Coding. Patent PCT/BG99/00027, Publ. No WO 01/10130 A1, Intern. Bureau of WIPO, Geneva, Switzerland.)
- USA patents on Image watermarking and on Lossless data compression:
1. Vl. Todorov, R. Kountchev, M. Milanova, R. Kountcheva, C. Ford. Method and Apparatus for Contents-Based Run-Length Data Encoding, Patent Application Publication, Publ. No: US 2007/0279261 A1, Publ. date: Dec, 6, 2007, United States.
2. R. Kountchev, Vl. Todorov, R. Kountcheva, M. Milanova, C. Ford. Method and System for Digital Watermarking of Multimedia Signals. Pat. Application Publ. Pub. No. US Patent 2007/0014428 A1, Pub. Date: Jan. 18, 2007
The fragile watermarking and the adaptive processing of compound images are not patented.
After development of the new features, mentioned in the previous parts, and the corresponding interfaces, the new solutions could be protected by patent, license or trademark.
|6||What are the steps that need to be taken in order to secure the IPR-protection? What is the cost of IPR-protection?|
|The European patent is abandoned; the USA patents are active for USA only. In our opinion for the future development of the product, the know-how and the participation of the team should be paid.|
|7||What is the expected impact of the research result? (industry, society, administration etc. and target groups of beneficiaries)|
|The impact of the research result could be in various areas: in industry – for example for efficient lossless compression of graphics, texts, formulas; for the society – for efficient archiving and content protection in medicine; for everyday use in faxes, scanners, photo cameras; for administration: creation of image databases with reliable access control; for efficient data hiding, etc.|
|8||What is you overall assessment of the scientific maturity of the research result?|
|The research result is already seriously investigated concerning the possible application areas; It needs some more investigations concerning the adaptive processing based on contents analysis and the resistant watermarking.The format for single images compression, archiving and watermarking is already created. Additional efforts will need the creation of the new format for archiving in image databases with hierarchical access control.|
KEYWORDS QUANTITATIVE ASSESSMENT (0-5).