摘要
针对自然图像与计算机生成图像在统计特征上存在的差异,提出一种基于分形维数和小波域特征的计算机生成图像盲鉴别算法,该算法基于统计特征对图像进行真伪识别.在使用支持向量机作为分类器的情况下,对800张标准图像进行实验的结果表明,该算法对计算机生成图像检测准确率达96.5%,明显提高了计算机生成图像的识别精度,为数字图像的真实性提供了保证.
In view of the perceptual differences between photographic and computer generated images primarily existed in statistical features,an identification algorithm for photographic and computer generated images was proposed,which is based on fractal dimension and wavelet domain features.The algorithm that is based on the statistical characteristics was used to recognize the authenticity of images.In the process of our experiment,800 standard images were used as the image database,the support vector machine was used as the classifier,and the detection accuracy rate of the computer generated images could reach 96.5% in the algorithm.It has significantly improved the accuracy of identifying computer generated images,and provides a guarantee of the authenticity of the digital image.
出处
《吉林大学学报（理学版）》
CAS
CSCD
北大核心
2013年第4期653-658,共6页
Journal of Jilin University:Science Edition
基金
吉林省自然科学基金(批准号:201215165)
符号计算与知识工程教育部重点实验室开放基金(批准号:93K172010K05)
关键词
自然图像
计算机生成图像
小波域
分形维数
支持向量机
photographic image
computer generated image
wavelet domain
fractal dimension
support vector machine
