A Robust and Fast Technique to Detect Copy Move Forgery in Digital Images Using SLIC Segmentation and SURF Keypoints

摘要

Digital image forgery is the alteration of an image in any form. Increasing advancements in the quality of image capturing devices and photo editing softwares have made the process of image forgery easy. Copy move forgery is the most frequent type among the various types of forgery. This paper presents a novel technique to detect copy move forgery that uses Speeded Up Robust Features (SURF) keypoints. First, the image is sectioned into nonoverlying blocks using Simple Linear Iterative Clustering (SLIC) algorithm. After this, the feature points are selected using SURF. The matching of the SURF points of different segmented regions is done using cosine similarity. The matched regions represent the forged areas. This method has a low computational complexity and shows good results even if the forged area is rotated or scaled.