Image Copy-Move Forgery Detecting Based on Local Invariant Feature

Now digital images are widely used in many fields. Making image forgeries with digital media editing tools is very easy, and these image forgeries are undetectable by human eyes. Copy-move forgery is common image tampering where a part of the image is copied and pasted on another parts. Up to now the useful way to detect copy-move forgeries is block matching technique. This paper firstly analyzes and summarizes block matching technique, then introduces a copy-move forgery detecting method based on local invariant feature matching. It locates copied and pasted regions by matching feature points. It detects feature points and extracts local feature using Scale Invariant Transform algorithm. Matching local features is based on k-d tree and Best-Bin-First method. Through analysis we learn computational complexity of the proposed method is similar to existing block-matching methods, but has better locating accuracy. Experiments show that this method can detect copied and pasted regions successively, even when these regions are operated by some process, such as JPEG compression, Gaussian blurring, rotation and scale.