摘要
目前图像处理和编辑软件的迅速发展使得图像伪造和篡改变的更加简单,同时也使通过视觉检测来区分图像真伪变的更加困难。为此,从数码相机成像特性出发,深入研究其成像原理及相机响应过程,将传感器曝光度转换为图像亮度的非线性映射过程描述为最大后验概率模型。根据不同相机响应的物理特征不一致原理,提出了一种基于相机响应常态性的图像真伪检测方法。实验结果表明,本文方法可以获取较高的正确识别率。
Currently,the rapid development of image processing and editing software makes not only the image forgery and tampering become easier,but also the distinction of the authenticity of the image by visual inspection become tougher.For this reason this paper makes a deep study on its imaging principle and camera response with the digital camera imaging feature as a starting point,and converses the nonlinear distribution of the process of transforming the sensor visibility into image brightness as the maximum posteriori probability model.In accordance with the inconsistent physical characteristics of the response of the different cameras,a method of image authenticity detection based on the normality of the camera response is proposed in this paper.Experimental results show that this method can obtain better recognition results.
出处
《西安理工大学学报》
CAS
北大核心
2011年第1期88-91,共4页
Journal of Xi'an University of Technology
关键词
图像真伪鉴别
相机响应常态性
相机响应逆函数
image authentication
the normality of camera response
inverse camera response function
