Fusion of SIFT and hue moments features for cloning tamper detection

Image tamper detection is a new branch of digital forensics. Image tamper detection is having two approaches that are active and passive approach. In active approach we have pre-embedded knowledge of an image but passive approach we do not have that. Hence passive approach is more challenging as compared to active approach. The passive approach is categorized into image splicing, image re-touching and cloning. Again cloning tampered detection is divided into key-point based and block-based depends on pixel level and block level feature extraction respectively. But, there are some drawbacks in these techniques. We can elevate these drawbacks by combining global and local features of block-based and key-point based respectively. This paper shows that fusion of these approaches gives effective results in less time as compared to traditional approach.