Research On Digital Image Information Hiding Algorithms Based On Hiding Capacity

With the rapid development of computer technology and the rapid popularization ofnetwork, people’s life and work are more convenient and digital information disseminationhas taken on a new appearance. More and more businesses, institutions and individualscommunicate with each other via network, and digital files become the main carrier ofcommunications. A lot of digital information transmitted and shared on the network, but at thesame time a series of safety problems appear, and people have higher requirements ofcommunication security. Traditional secure communication encrypted the content, whichtends to attract the listener’s attention. Information hiding technology appeared in this case,which hid secret information in the host information unlikely to attract the listener’s attentionand then achieved secret transmission. Due to the large redundancy space, the host image isthe most frequently used carrier at present. Depending on the application purpose ofinformation hiding technology, different characteristics are required such as imperceptibility,security, undetectability, robustness and high embedding capacity. For general informationhiding system, the three indicators imperceptibility, robustness and embedding capacity areusually mutually dependent. In this paper, a systematic study was made on the digital imageinformation hiding algorithm, a solution was proposed for the key issue of hiding capacityincrease in information hiding algorithm, and simulation experiments and analysis werecarried out for the proposed algorithms. The main research work of this paper is as follows:In the host image analysis, the characteristics of human visual system’s masking ofimages were analyzed. Based on the insensitive characteristics of human eyes to brightnessand complicated texture area, two kinds of insensitive areas in the host image wererespectively screened, and the integrated area screening method was proposed. These twokinds of masking characteristic were weighted selectively, and the threshold was used todetermine the number of masked areas and the size of embedding capacity. The higher bitplane of the image contain partial image information, and its change may cause a largervariation of the image and attract human eyes. In the embedding stage, the characteristics ofthe masked area were utilized. First the secret information was processed with pseudo randomordering of Arnold transform to eliminate the correlation between adjacent pixels. Then themulti-bit planes replacement embedding algorithm was used for modifying the spatial domainvalue of the multi-bit planes to embed secret information to the lower bit plane of image andthe higher bit plane in the masked area. The algorithm had very high embedding capacity aswell as high imperceptibility.For purpose of removing redundancy, efficiently utilizing host image space and embedding more useful information, the predictive coding technology is used to eliminate theredundancy between pixels before embedding, so as to achieve the purpose of informationcompression. The prediction coding coefficients are embedded into the original image insteadof the secret information, the capacity of coefficients is much less than the original secretinformation.The characteristics of the reversible information hiding algorithm based on differenceexpansion were analyzed. According to the principle of prediction error expansion algorithm,the more accurate the prediction, the higher the embedding capacity, the different expansionalgorithm based on the gradient prediction was proposed. The existing embedding rule wasimproved in the hiding scheme, and based on the value of the secret information, the originalhost image value was modified selectively. As the secret information was0, no modificationwas made to reduce the damage to the host image and increase the image’s imperceptibility.Results of the algorithm simulation experiments indicated that this algorithm increased theembedding capacity and ensured the imperceptibility of images.