关键词: 图像伪造,合成图像,归一化割,共生矩阵,支持矢量机 中图法分类号TP309．7文献标识码A

Abstract: The paper proposed an algorithm to expose spliced photographic merging two or more parts from different photos into one composite．The proposal algorithm firstly segments photo into several parts under perceptual grouping criterion,minimizing the disassociation between parts and maximizing combination within part with normalized cut algorithm,then predicts each pixel’s value according to its eight neighbors by minimal least square methods,conducts statistical features of inharmonic points．Since these features are influenced by texture characteristics,features of co-occurrence matrix,which display the image’s texture,are also calculated．Finally,all features are feeded to a support vector machine (SVM) classifier．The test experiments show that the proposal method is effective in exposing splicing image.

Key words: Image forgery,Spliced photographic,Normalized cut,Co-occurrence matrix,Support vector machine