A Novel Image Authentication and Recovery Method for Applications of Image Sharing

Image authentication techniques should be able to distinguish legitimate images from forged ones. In practical applications, a shared image may be verified in compressed format such as the popular JPEG file. In this paper, an image authentication method is proposed to meet this requirement. The quantized DCT coefficients are used as the image feature to distinguish JPEG compression from malicious manipulations, and further used to predict the first five AC coefficients of the tampered blocks and recover the damaged images if malicious manipulations are detected. The concept of the regions of interest (ROI) is also included in the proposed method to improve the visual quality of the recovered image. Experimental results show that the proposed method has better performance in both detection and recovery of the tampered image than our previous work.