With the rapid development of image processing technology, it is becoming increasingly easy to tamper with digital images without leaving any obvious visual trace. Today those who see no longer believe[1]. Image falsification, like any other illegal and harmful activity, could cause serious harm to society. Forensic analysis of digital images is a new emerging field. The purpose of image forensic analysis is to detect whether an image has been tampered with. With the widespread use of high resolution digital cameras and highly advanced photo editing software, image tampering has become more common and it may often be necessary to resample (resize/rotate/stretch) the image to make it look natural. Therefore, verifying the authenticity of digital images has become a very important issue. JPEG is one of the most commonly used compression schemes in many practical applications. Therefore, forensic analysis of JPEG images has attracted more and more attention recently. Typically, there are two properties that are significant for forensic analysis. The first and most obvious property is the artifacts that block the spatial domain. Due to block-based processing in lossy JPEG compression, discontinuous pixels are usually found in the boundary between two adjacent 8×8 blocks. Such a block signature can serve as proof of JPEG compression [2] and some tampering operations [3]. Another important property is the quantization artifacts in the DCT frequency domain. During JPEG compression, each DCT frequency component in the 8×8 block is quantized using a quantization step. This will lead to a specific shape of the corresponding DCT histogram. That is, those dequantized coefficients will simply cluster into multiples of the quantization step. In combination with Laplac...... at the center of the article ...... the disparity map was based on belief propagation and mean shift segmentation [19]. The disparity map and reference image (JI_L) are segmented into some objects. The objects and the average disparity of these objects are denoted by O_(JI_L)^i and d_(JI_L)^i, i = 1,2,…,m, respectively. If d_(JI_L)^i is in [D_b,D_f ], O_(JI_L)^i is considered to be the main content, O_(JI_L)^i∈O_maipart. If d_(JI_L)^i is not in [D_b,D_f ], O_(JI_L)^i is considered to be the background, O_(JI_L)^i∈O_background. That is,O_(JI_L)^i∈{█(O_mainpart d_(JI_L)^i∈[D_b,D_f ] @O_ background d_(JI_L)^i∉[D_b,D_f ] )→(1)┤A seam is a optimal path of eight pixels connected on a single JPEG image from top to bottom (vertical) and consisted of one and only one pixel in each row, which ensures that the JPEG image maintains a rectangle when stitching is removed. In [20], an energy function defines the