Based On Cluster Information Hiding Strategy And Mutual Hiding Method

This paper is about information hiding technique on digital image. First it analyze the existing information hiding techniques. With the reference to the literature read, it gives a brief statement on the advantage and disadvantage of each hiding technique, from airspace technique to transform-space technique, from DCT to DWT, followed by the statement of sameness and difference between the two main branches——steganography and watermarking. To the common requirement of invisibility of the two, the paper presents a information hiding strategy which first uses cluster analysis methods (this paper uses K-means algorithm) to classify the image to get the nature of the image, and then uses one embedding algorithm, accounting the improvement on the conceal effect using image's character. Using cluster analysis to classify the image can overcome the subjective threshold value affecting the result of classification, which often appears in the existing image classifying methods. Experiments also proved that, using bright-factor, grads-factor, texture-factor, contrast-factor and entropy-factor mentioned in the paper to cluster the pixels of an image can cluster the corresponding character of the image well. Based on the classification of images, this paper gives out four digital image steganography techniques which have good conceal effect, including low-three-bits hiding strategy, odd-even of hypo-low bit hiding strategy, deviation of adjacent-field hiding strategy and self-adaptive deviation of adjacent-field hiding strategy. On digital watermarking aspect, this paper gives two blind watermarking strategies which have good conceal effect and high robustness, they are single- watermarking strategy based on minimal-value exchanging and multi-watermarking strategy which hides the same watermark copy on the different parts of an image. Each strategy is depicted in detail and followed with experiment result. Finally the paper gives out a mutual hiding strategy, which is based on interested region.