计算机科学与应用

Vol.5 No.7 (July 2015)

一种基于关键点的复制粘贴盲检测算法
Copy-Move Forgeries Detection Based on SIFT Algorithm

 

作者:

王珺斌 , 杨正洪 :中国农业大学理学院,北京

牛少彰 :北京邮电大学计算机学院,北京

 

关键词:

图像取证复制粘贴篡改SIFT算法向量匹配Image Forensics Copy-Move Detection Scale-Invariant Feature Transform (SIFT) Vector Matching

 

摘要:

鉴于传统的复制粘贴篡改检测时间复杂度较高,而且对于图像的旋转缩放等后续润饰操作的鲁棒性不好,误报率高。提出了一种基于SIFT关键点的复制粘贴检测算法。首先利用SIFT算法对整幅图像提取SIFT关键点,再对提取到的关键点进行匹配。将匹配的关键点连接起来,来定位篡改区域。实验效果表明,相比较于传统的复制粘贴检测算法,该算法可以有效地检测出经过旋转缩放等几何变换的后续篡改。

In view of the higher time complexity of traditional copy-move forgery detection, bad robustness for the image rotation zoom and other follow-up retouching operations, a copy-move forgeries detection algorithm based on SIFT key points is studied. First, extract SIFT key points on the image by using SIFT algorithm, and then match the key point of extraction. Connect the key points of the match to locate the tampered area. Experimental results show that the algorithm can effectively detect the subsequent tampering of the geometric transformation such as rotation and zoom.

文章引用:

王珺斌 , 杨正洪 , 牛少彰 (2015) 一种基于关键点的复制粘贴盲检测算法。 计算机科学与应用, 5, 255-263. doi: 10.12677/CSA.2015.57033

 

参考文献

分享
Top