Steganalysis of JPEG images by block texture based segmentation

The current JPEG steganalysis systems have attained outstanding achievements. A considerable variety of strategies for feature extraction are developed. However, a common shortcoming in the traditional image steganalysis techniques is that they are conducted on the entire image and do not take advantage of the content diversity. In this paper, a new steganalysis algorithm based on image segmentation is proposed to enable us to utilize the content characteristics of JPEG images. The images are segmented into several sub-images according to the texture complexity. The steganalysis features of each type of sub-images with the same or close texture complexity are extracted separately to build a classifier. The steganalysis results of the entire image are determined through a weighted fusing process. Experimental results demonstrate that the proposed method exhibits excellent performance and significantly improves the detection accuracy.