Research On Steganalysis For JPEG Images

As the main sub-discipline of information hiding, Steganography is the art and science of communicating in a way which hides the existence of the communication. Steganalysis is the attacks to steganography, and its important task is the detection of hidden messages. This paper focuses almost exclusively on Steganalysis technology based on JPEG image. The research includes the following sections:1 .This paper introduced the concepts of information hiding,steganography,steganalysis and the JPEG images compression standard. Based on the review of the usual steganography method and steganalysis algorithms for JPEG images, we simulated Jsteg,F4 and MB steganography,and compared the capacity,embedding efficiency and using rate. A test database of cover-images and stego-images is constituted.2. A modified detecting method of breaking MB steganography is proposed.One of the first order statistics of individual DCT histograms is bigger after cropping the stego-image by 4 pixels,but this feature of the cover-image is almost same as the cropped cover-image.So we can class the cover and stego images using the above-mentioned difference.Experimental results show a better detection reliability for stego-images,and the detection rates of stego-images with 50% capacity are more than 94% .3.A LSSVM-based blind steganalysis method for JPEG images is constructed. Each JPEG image is characterized using 18 calibrated features calculated from the DCT and spatial domain.The LSSVM classifier trained on the feature vectors corresponding to cover and different stego images can apply to blind steganalysis .The experiments show that it is effective to classify the cover and stego images using the binary classifier, and it is also reliable to classify high-embedding-rate stego images to their embedding techniques using the multi-classifier.