电子测量与仪器学报

吴鹏,陈北京,郑雨鑫,高野.基于双流 Faster R-CNN 的像素级图像拼接篡改定位算法[J].电子测量与仪器学报,2021,35(4):154-160

基于双流 Faster R-CNN 的像素级图像拼接篡改定位算法

Pixel-level image splicing localization algorithm basedon dual-stream Faster R-CNN

DOI：

中文关键词: 图像拼接 Faster R-CNN 错误等级分析特征金字塔全卷积神经网络

英文关键词:image splicing Faster R-CNN error level analysis feature pyramid fully convolutional network

基金项目:国家自然科学基金(62072251)、江苏省大学生创新创业训练计划(201910300022Z)、江苏高校优势学科建设工程项目（PAPD）资助

作者	单位
吴鹏	1. 南京信息工程大学计算机与软件学院
陈北京	1. 南京信息工程大学计算机与软件学院，2. 江苏省计算机网络技术重点实验室，3. 南京信息工程大学江苏省大气环境与装备技术协同创新中心
郑雨鑫	1. 南京信息工程大学计算机与软件学院
高野	1. 南京信息工程大学计算机与软件学院

Author	Institution
Wu Peng	1. School of Computer and Software,Nanjing University of Information Science and Technology
Chen Beijing	1. School of Computer and Software,Nanjing University of Information Science and Technology,Nanjing，2. Key Laboratory of Computer Network Technology of Jiangsu Province,3. Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology, Nanjing University of Information Science and Technology
Zheng Yuxin	1. School of Computer and Software,Nanjing University of Information Science and Technology
Gao Ye	1. School of Computer and Software,Nanjing University of Information Science and Technology

摘要点击次数: 867

全文下载次数: 6

中文摘要:

基于双流 Faster R-CNN 的图像拼接篡改定位算法因综合考虑彩色图像及其噪声图像作为输入而获得良好性能。但是,它仍存在两个不足,定位精度只是块级且经过隐写分析富模型产生的噪声图易夹杂大量冗余非篡改区域信息。为此,提出一种基于双流 Faster R-CNN 的像素级拼接篡改定位模型。针对第一个缺陷,增加一个全卷积网络分支实现像素级定位。针对第二个缺陷, 采用错误等级分析噪声模型替代隐写分析富模型用于提取噪声图。实验结果表明提出算法较现有算法提高了近 10%的准确率。

英文摘要:

The image splicing localization algorithm based on dual-stream Faster R-CNN achieves a good performance because it considers both the color image and its corresponding noise image as inputs. However, it still has the following two drawbacks, it only achieves block-level precision and the noise images generated by SRM filter are likely to contain a lot of redundant non-forged semantic features. Therefore, this paper proposes a pixel-level image splicing localization model based on dual-stream Faster R-CNN. Regarding the first drawback, a fully convolutional neural network branch is added for pixel-level localization. Regarding the second one, the steganalysis rich model is replaced by error level analysis noise model for noise map extraction. Experimental results show that the proposed algorithm improves the accuracy by nearly 10% compared with some existing algorithms.

查看全文查看/发表评论下载PDF阅读器