Radial search for accelerating block matching to detect copy-move forgery in digital photographs
- Copy-move image forgery, block based, radial search, radix sort, DCT.
One of the exciting areas of image processing research right now is image forensics. The most common kind of image forgery is copy-move (CM) forgery. In this study, a method for quickly detecting CM counterfeiting is proposed. The suggested approach speeds up the block matching process by employing the proposed radial search method. DCT features are extracted from each block and only the most "R' significant features are considered to create the feature vector which in turn shortens the processing time for feature matching. Then radix sorting is used to sort the blocks based on the DCT coefficients, and the proximity of blocks is used to indicate how similar they are. Each pair of adjacent blocks has their difference measured, and if the difference between two blocks is less than a predetermined threshold, the blocks are deemed comparable. Once a match is found, the nearby eight blocks are located using the radial search strategy, which eliminates the need for the traditional sequential block comparison. As opposed to exhaustive block matching, the radial search approach is experimented with which gives commendable improvement in copy move forgery detection.