The Research On Digital Watermark Technology Of Binary Text Images

With the rapid development of information technology, the copyright protectionand integrity of the identification of digital resources has become an increasinglyimportant. The development of digital watermarking technology is leaps and bounds,and the relatively slow progress for binary text image watermarking technologyresearch. Most of the watermarking algorithm is based on changes in the text of blackpixels to achieve watermark embedding, To expand the embedding capacity, andsometimes the need for a large number of effective black pix els changes, thisapproach can easily lead to the distortion of the text is too large and cause people toperceive. The same time, print, fax and photocopier growing popularity of theseelectronic devices has become a part of people’s work and life, a larg e number ofdigital resources need to be repeated after the digital-to-analog, analog-to-digitalconversion, Watermarking efforts on the carrier in this process, most of the watermarkprint scans is difficult to resist such attacks.About these shortcomings, this paper, basing on the characteristics of humanHVS and making use of imperceptible yellow dots describe the watermarkinginformation embedding in the blank of the blocked text, under the premise of theblack pixels of the binary image of text changes, study design two digitalwatermarking algorithm:1. Black pixel information accordance ranks of the statistical text watermark, thewatermark is encoded with error control code to bind, and then text blocks, make useof imperceptible yellow dots characterize the coding sequence watermark hidden totext in the blank space, using the generated yellow point information to monitorwhether text pixels change from vertical and horizontal two directions so that thealgorithm has the performance of tamper localization2. Imperceptible yellow dot has stronger robustness, it also can restore the yellowdot through digital image processing, according to this characteristic, combined withthe size of the text and estimated maximum embedding capacity of the text, u singimperceptible yellow dot to characterize binary sequence which generated by themeaningful watermark to hide in the text, in order to achieve text copyrightprotection.This thesis demonstrates the two algorithms.Meanwhile through experimentswere carried out to achieve, in the experiment test the feasibility algorithm,Experiments show that the images which have been embedded watermark have fine image quality, watermark capacity is also relatively large. But algorithm is still somedefects, the content text have certain requirements.