Abstract:
There are usually some issues such as high-dimension and high computational complexity that need to be solved in digital image copy-move forgery detection. In this paper, we propose a new method for digital image copy-move tampering detection based on slicing transform (SLT). The original image is divided via Slicing Transform, then image slices is grouped before overlapping blocking. The local slice density feature vector of each group slicing block is extracted for copy-move forgery detection. Experimental results demonstrate that the extracted feature vector which has lower dimension compared with other methods representing the image block. The proposed method can detect copy-move forgery regions faster and more accurately, and it is robust to post-processing such as rotation and scaling.