基于交叉划分的灰度图像篡改检测与恢复算法
Algorithm for Gray Image Temper Localization and Recovery Based on Intersection Division
-
摘要: 针对单像素篡改检测算法和分块篡改检测算法的缺点,提出了基于交叉划分的篡改检测与恢复算法.根据图像内容生成的2×2分块水印被两次嵌入到不同的对应块中,两次分块方式采用交叉划分方式.在水印嵌入时采用混沌映射序列对水印进行了加密.根据不同的分块位置进行两次篡改检测,并根据两次检测结果交叉后逐步求精,得到最终检测结果.图像恢复采用像素级恢复方式,先确定水印嵌入区域是否被篡改然后再根据两次嵌入的信息进行恢复.实验结果表明,该算法篡改检测定位精度可以达到像素级,恢复图像质量好,能够抵抗矢量攻击和噪声攻击.Abstract: Aiming at disadvantages of pixel-level temper detection algorithm and block-level temper detection algorithm,a temper detection algorithm based on intersection division is been proposed in this paper.Watermarking is produced according to image for blocks of 2×2pixels and embedded in various blocks two times.These blocks of the first division are intersected with these blocks of the second division.Watermarking is been encrypted by chaotic sequence.Twice temper detections are processed according to various block location.The last detection result is obtained using the twice detection results by stepwise refinement method.Image recovery is pixel-level.In order to enhance recovery image quality,temper area is twice recovered using embedded information in watermarking block after the watermarking block is judged to be reality.The proposed algorithm can reach pixel-level tamper localization precision.Experiment results show that the algorithm has good recovery performance and may resist noise and vector attack.
下载:
