In this paper, we focus on image tampering detection and tampered region localization. We find that the probability distributions of the DCT coefficients of a JEPG image will be influenced by tampering operation. Hence, we model the distributions of AC DCT coefficients of JPEG image and detect the tampered region from the unchanged region by using their different distributions. Based on an assumption of Laplacian distribution of unquantized AC DCT coefficients, Laplacian Mixture Model (LMM) is employed to model the quantized AC DCT coefficient distribution of a suspicious JPEG image. With the help of Expectation Maximization (EM) algorithm, the probability of an 8 x 8 block being tampered can be estimated; and then, a sophisticated image segmentation method, graph cut, is applied to determine the tampered region. Extensive experimental results on large scale databases prove the effectiveness of our proposed method which is suitable for different tampered region sizes at all levels including pixel, region and image level. © 2011 IEEE.