Identification of Image Forgery based on various Corner Detection methods

Digital image indubitably has lost virginity in both its source and surroundings due to the encroachment of high-resolution camera, state-of-the-art image handling tools and contemporary personal computers. As a result, authenticating the integrity of digital image became very imperative and discovering any indications of falsification in the image has turned into a sizzling turf to carry out research for last few years. In this paper, image forgery has been identified using Min Eigen feature extraction based on Shi-Tomasi Corner Detection method which detects interest points. Initially, various corner detection and feature extraction methods have been studied and analysed to extract features from the input images. From the gray-scale image, distinctive localized features are extracted based on Harris feature extraction, surf features and fast features. Subsequently, Euclidean distance is calculated between the feature vectors of the images. Then the resultant feature values are further implemented using classifiers to obtain accurate result analysis.