Design And Evaluation Of Information Hiding Algorithms For Binary Image

The binary image is also can be called black and white images. Because of its easy storage and unique compact structure, it was widely used. With the global digital information growing up, much important text information, such as personal files, medical records, academic certificates, handwritten signature, design drawings, collection of books are all scanned into digital documents and stored in binary images. It is very valuable to study the binary image. So, it is very important to hide the protect information in binary image. Therefore, it is meaningful to study the binary image information hiding technology in information security and property rights protection.This paper studies the information hiding technology based on binary image, and the main results are as follows:1. A new authentication algorithm used for binary image is presented. According to the visual characteristics of pixels, we present the standard of selecting turning point and select the revisable degrees of the pixels as the embed element. According to the score, the positions of turning points can be found in sequences. Then we can use the MME (Modified Matrix Encoding) algorithm to choose the best way to embed the authentication information. This method can reduce the changes of the original image and the distortion of the carrier images. At the same time, a novel measurement of the binary image's distortion is proposed to measure the distortion caused by the information hiding algorithm.2. A new information hiding algorithm used for binary image is proposed. Aiming at the feature that the changing trace is obvious in binary image, the set of pixels suitable for turning are denoted. Considering the visual features, the method for computing the scores of the turning pixels is presented, which can distinguish the level of distortion pixels. Then the STC (Syndrome-trellis Codes) algorithm is used to choose the optimal path in the sequence of cover blocks for hiding message. The information hiding algorithm reduces the changes of the host image, avoid the large distortion of the local point emergence and ensure the invisibility of the stego images.3. A lossless data hiding algorithm for binary image is presented. Considering the binary image pixel and its neighborhood connectivity and smoothness, the selection criteria of turning point is presented. Using the selection criteria we can construct a suitable modification of the vector sequence for being modified. Then we use the arithmetic coding to compress the vector sequence, so the redundant hidden information space can be obtained. Bit control method is selected to reduce the original image changes of information embedded. Experimental results show that: the algorithm can not only extract the hidden information but also restore the original host image. 4. A method of evaluating the information hiding algorithms based on binary image is proposed. According to their characteristics, an index set for evaluating the information hiding algorithms is established. The index set includes hiding capacity, visual effects, image distortion, embedding efficiency, statistical characteristics and time complexity. We evaluate the algorithms from the performances of six aspects. A comprehensive assessment of the algorithms is made after setting up the appropriate index weight according to the actual application needs. The new assessment program can not only evaluate the performance of the information hiding algorithms based on binary image, but also help to choose the best one from all the known algorithms.