摘要
通过对图像阴影二值化后的分析,提出一种基于阴影不一致的简易人像篡改鉴别算法。该算法先手工选取鼻翼区和下巴区的图像,经过Ostu法二值化处理后统计分析阴影面积,对比鼻翼区和下巴区阴影的一致性来判断图像是否经过篡改。大量实验结果表明,该方法简单易行、精确度高、可靠性强,在图像盲取证中具有一定的实用价值。
A simple tampered portrait identification algorithm based on inconsistent shadow was proposed by analyzing of binarization of the image shadow.The algorithm determines whether the image is tampered by first manually selecting the local image of the nose area and chin area,then statistically analyzing shaded area after Ostu binarization process,and finally comparing the shadow consistency between the nose area and chin area.A lot of experimental results show that the method has simplicity,high accuracy,strong reliability and has some practical value in image forensics.
出处
《计算机科学》
CSCD
北大核心
2014年第S1期129-131,共3页
Computer Science
基金
国家自然科学基金:船载航行数据记录仪(VDR)图象篡改鉴别技术研究(61250006)
基于调制解调的图象多分辨率分解理论与方法(61273262)
广义测不准原理及其应用研究(61002052)资助
关键词
阴影
二值化
盲取证
Shadow,Binarization,Blind forensics
