Feature Point Extraction by Adaptive Over-Segmentation and Feature Point Matching for Effective Digital Image Forgery Detection

The invention of the web has introduced the incredible growth and developments within the noted analysis fields like medication, satellite imaging, processing of image, security, biometrics, and biological science. The algorithms applied in the twenty 21st century has created the human life easier and secure, but the security is a major concern in the digital image processing domain. A new study is presented in this paper to detect forgery using the adaptive over-segmentation and feature point matching. Both block based and key point based forgery detection methods is integrated here. Adaptive over segmentation algorithm is used to segment the host image into non -overlapping and irregular blocks adaptively. Feature points are extracted from every block as block features and the block features are matched with one another to locate suspected forgery region. The feature points are gradually replaced by super pixels in the proposed Forgery Region Extraction algorithm and then neighboring blocks that have similar local color features are merged into the feature blocks to form merged regions; finally, morphological operation is applied to the regions which shows the detected forgery regions. The proposed forgery detection algorithm achieves much better detection results even under various challenging conditions than earlier methods in all aspects.