| Summary: | In recent years, digital image forgery detection has become an active research area due to the advancement of photo editing software. In general, image forgery detection can be classified into two types, namely active and passive detection. Active forgery detection relies on embedded authentication code in the image while passive forgery detection relies solely on the images for authentication. The forgery detection techniques are used to identify images tampered with common techniques such as copy move, slicing, contrast alteration and sharpening/blurring. This project focuses on passive forgery detection on images tampered using copy move technique, better known as Copy Move Forgery Detection (CMFD). A CMFD technique consisting of oriented Features from Accelerated Segment Test and rotated Binary Robust Independent Elementary Features (Oriented FAST and rotated BRIEF) as the feature extraction method and 2 Nearest Neighbour (2NN) with Hierachical Agglomerative Clustering (HAC) as the feature matching method is proposed. The ORB parameters, namely the number of features to retain and patch size are optimized using Particle Swarm Optimization (PSO). The optimization is essential in obtaining a balance between performance and runtime. Evaluation of the proposed CMFD technique is performed on images which underwent various geometrical attacks. With the proposed technique, an overall accuracy rate of 84.33% and 82.79% is obtained for evaluation carried out with images from the MICC-F600 and MICC-F2000 databases. Forgery detection is performed accurately, with True Positive Rate of 91% and above, for tampered images with object translation, different degree of rotation and enlargement. However, the performance degraded for tampered images with reduced copied object size and asymmetrical scaling, with True Positive Rate of 73.68% and 38.15% respectively.
|
|---|
