An improved method for SIFT-based copy–move forgery detection using non-maximum value suppression and optimized J-Linkage

作者：

Highlights：

•

• A method to improve the detection performance in smooth tampered regions is proposed.

• A new solution to acquire uniformly distributed keypoints in a test image is proposed.

• Discriminative power of SIFT keypoints is enhanced using OpponentSIFT descriptors.

• A matched pairs grouping algorithm utilizing the correspondence between superpixels is proposed.

• J-Linkage clustering is optimized by using image segmentation and the matched pairs grouping algorithm.

摘要

•A method to improve the detection performance in smooth tampered regions is proposed.•A new solution to acquire uniformly distributed keypoints in a test image is proposed.•Discriminative power of SIFT keypoints is enhanced using OpponentSIFT descriptors.•A matched pairs grouping algorithm utilizing the correspondence between superpixels is proposed.•J-Linkage clustering is optimized by using image segmentation and the matched pairs grouping algorithm.

论文关键词：Copy–move forgery detection,Non-maximum value suppression,OpponentSIFT,Simple linear iterative clustering,J-Linkage

论文评审过程：Received 13 January 2017, Revised 15 May 2017, Accepted 15 May 2017, Available online 26 May 2017, Version of Record 4 June 2017.

论文官网地址：https://doi.org/10.1016/j.image.2017.05.010