摘要
研究图像的篡改识别问题,由于数字图像能够被轻易的篡改并且很难发现改动痕迹,对篡改像素不融合现象识别不清,导致图像篡改很难被肉眼识别。为解决上述问题,提出了一种根据小波的篡改图像鉴定识别技术,通过提取图像块特征得到图像块的特征向量矩阵,并对特征向量矩阵进行小波分解,从而提取篡改图像特征点灰度融合程度,利用定义的判别阈值实现对篡改区域的准确定位与检验。实验表明,改进方法通过检测可以准确识别出篡改后的图像,为篡改图像的识别提供了依据。
Research identification of image tamper.Digital images can be easily tampered and it is difficult for human eyes to find the changes.This paper puts forward a method based on wavelet image identification technology.First,image block features were extracted to get the image block matrix.Then,the wavelet decomposition was done with the feature vector matrix to extract the gray-scale fusion feature points of the tampered image.Finally,the defined discriminant threshold value was used to accurately position and inspect the tamper area.Experiments show that the proposed method is can achieve good effects.
出处
《计算机仿真》
CSCD
北大核心
2011年第11期238-241,共4页
Computer Simulation
