基于SIFT的图像盲取证方法

摘要

This paper describes an effective method to detect copy-move forgery in digital images. This method works by extracting Scale Invariant Feature Transform(SIFT) descriptors of an image and by seeking for approximate nearest neighbor based on Product Quantization(PQ).The method of approximate nearest neighbor search is to decompose the space into a cartesian product of low dimensional subspaces and to quantize each subspace separately. Asymmetric Distance Computation(ADC) computes the euclidean distance between two vectors. Experimental results show that the approach can correctly detect the copy-move forgery which is preprocessed by different methods and decrease the memory usage and the complexity of learning the quantizer, at the same time, reduce the search time.%采用一种基于尺度不变特征变换的算法提取图像特征,使用乘积量化的近似最近邻搜索方法对子空间分别进行量化,运用非对称距离算法计算特征向量之间的欧氏距离,提出一种新的数字图像复制粘贴被动盲取证方法.实验结果表明,该方法能够准确地对复制区域经过预处理的伪造进行检测,减少内存的使用量和空间复杂度,缩短搜索时间.