An Adaptive Grayscale Watermarking Method in Spatial Domain

摘要

In the past a few years, many watermarking approaches have been proposed for solving the copyright protection problems. This paper presents a new grayscale watermarking method in spatial domain. In the host image analysis, the characteristics of human visual system's masking of images were analyzed. Based on the insensitive characteristics of human eyes to complicated texture area, the insensitive areas in the host image were screened. The threshold was used to determine the number of masked areas and the size of embedding capacity. The higher bit plane of the image contain partial image information, and its change may cause a larger variation of the image and attract human eyes. In the embedding stage, the characteristics of the masked area were utilized. First the secret information was processed with pseudo random ordering to eliminate the correlation between adjacent pixels. Then the multi-bit planes replacement embedding algorithm was used for modifying the spatial domain value of the multi-bit planes to embed secret information to the lower bit plane of image and the higher bit plane in the masked area. The algorithm had very high embedding capacity as well as high imperceptibility.