Detection and Prediction of Modifications in Medical Images using SIFT Feature Extraction Algorithm

Abstract—This paper proposes a method of digital blind forensics in medical image field. The proposed system detects the global modification of medical images and also finds the nature of the modifications done on the victim medical image. Scale Invariant Feature Transform algorithm used as the medical image feature extraction technique. Scale Invariant Feature Transform descriptors are invariant to illumination differences and it affine present the local features of an image. Keypoints and keypoint descriptor are extracted features from the image, which is input to a vector classifier built in order to detect modifications as well as to identify the nature of the modification. In performance evaluation different types of medical images are analyzed by feature extraction in the proposed system. It shows above 80 percentage modifications detection and Identification rate