Region duplication detection based on hybrid feature and evaluative clustering

2019 ◽  
Vol 78 (15) ◽  
pp. 20739-20763 ◽  
Author(s):  
Cong Lin ◽  
Wei Lu ◽  
Xinchao Huang ◽  
Ke Liu ◽  
Wei Sun ◽  
...  
Keyword(s):  
Region Duplication ◽  
Region Duplication Detection ◽  
Duplication Detection

Region duplication detection based on image segmentation and keypoint contexts

2017 ◽  
Vol 77 (11) ◽  
pp. 14241-14258 ◽  
Author(s):  
Cong Lin ◽  
Wei Lu ◽  
Wei Sun ◽  
Jinhua Zeng ◽  
Tianhua Xu ◽  
...  
Keyword(s):  
Image Segmentation ◽  
Region Duplication ◽  
Region Duplication Detection ◽  
Duplication Detection

Copy-Move Forgery Detection Using DyWT

2017 ◽  
Vol 8 (2) ◽  
pp. 1-9
Author(s):  
Choudhary Shyam Prakash ◽  
Sushila Maheshkar
Keyword(s):  
Wavelet Transform ◽  
Input Image ◽  
Discrete Wavelet ◽  
Forgery Detection ◽  
High Similarity ◽  
Dyadic Wavelet ◽  
Region Duplication ◽  
Copy Move Forgery Detection ◽  
Duplication Detection ◽  
Better Than

In this paper, we proposed a passive method for copy-move region duplication detection using dyadic wavelet transform (DyWT). DyWT is better than discrete wavelet transform (DWT) for data analysis as it is shift invariant. Initially we decompose the input image into approximation (LL1) and detail (HH1) sub-bands. Then LL1 and HH1 sub-bands are divided into overlapping sub blocks and find the similarity between the blocks. In LL1 sub-band the copied and moved blocks have high similarity rate than the HH1 sub-band, this is just because, there is noise inconsistency in the moved blocks. Then we sort the LL1 sub-band blocks pair based on high similarity and in HH1 blocks are sorted based on high dissimilarity. Then we apply threshold to get the copied moved blocks. Here we also applied some post processing operations to check the robustness of our method and we get the satisfactory results to validate the copy move forgery detection.

Sign in / Sign up

Export Citation Format

Share Document