Universal embedding strategy for batch adaptive steganography in both spatial and JPEG domain

Abstract

In this paper, we present a universal embedding strategy for batch adaptive steganography in both spatial and JPEG domain. This strategy can make up for the problem when applying batch steganography to adaptive steganography in JPEG domain meanwhile improve the secure performance in spatial domain. When embedding, the strategy are employed to determine the sub-batch of cover images to carry the total message. To evaluate the feasibility of images selected as sub-batch, we calculate the histogram equilibrium for each image and the images with larger “size” and more equilibrated histogram are set a high priority when determining the sub-batch. The histogram equilibrium of images is the combination of entropy and standard deviation for the probability of histogram coefficients. We keep the repetitive information of using entropy and standard deviation to measure the histogram equilibrium. Experimental results show that histogram equilibriums vary according to images and employing images with larger “size” and more equilibrated histogram to carry the total message is a universal embedding strategy for both spatial and JPEG domain. Batch adaptive steganography with this embedding strategy can improve the secure performance effectively.