Efficient Keypoint based Copy Move Forgery Detection Method using Hybrid Feature Extraction

Abstract

Digital images are considered as one of most sorted mediums as an evidence or proof in various areas. For example, in forensic analysis, military intelligence, medical imaging, journalism, crime scene investigation and so on. With increased use of smart phones and cheap availability of internet has led to increased circulation of digital image. Along with, availability of wide range of digital image editing tools has made tampering image much easier. Such tampered image generates an improper message and could be used to influence events especially when the society based important decisions are carried out using these tampered image. Copy move tampering method has been widely applied passive image tampering method. Extensive work has been carried out in recent time for detecting such copy move tampering attack using key-point and block based method. However, these methods does not extract enough feature points considering small and smoothed region. For overcoming research challenges, this paper presents an efficient copy move forgery detection(CMFD) technique using key-points employing hybrid feature extraction, detection and hierarchical clustering method. Experiment result shows the proposed method attain significant performance when compared with other forgery detection methods.