Active tamper detection of digital images using VQ compression

Abstract

This paper presents an active tamper detection and content recovery scheme for digital gray images. It mainly adopts vector quantization compression and treats the index codes as the authentication and recovery data of an image. In order to achieve the purpose of detecting and locating whether an image has been tampered with, this scheme embeds the essential information of authentication and recovery into each pixel value's LSB bit. In addition, this paper also utilizes Chan's LSB matching function to improve the quality of embedded image. Experimental results showed that the authentication data of index codes are able to detect those possible modified regions effectively and clearly. Moreover, the embedded image quality also improves nearly to 1.66dB.