A comparison between Speeded Up Robust Features (SURF) and Discrete Wavelet Transform (DWT) as feature extraction in Copy-Move Forgery Detection
Keywords:digital image, copy-move forgery detection, SURF, DWT
The advancement today technology have leads to various new technological revolution including in image editing software. Modern and easy to use editing software have allowed the content of the digital images easier being tampered. Therefore, the validity, credibility and authenticity of such digital images has become an essential concern. There are many types of digital but copy-move forgery is the hardest to detect because of their consistency with the rest of the image with the noise variable, colour palette, dynamic range and most other essential properties, and will therefore not be observable with methods that search for anomalies in statistical measurement in different parts. There are two categories for copy-move forgery detection (CMFD) which are (a) Keypoint based and (b) Block based. The block based methods are useful for the precise identification of forged areas but are incredibly complex in computer technology. The alternate ways to solve the problem by using keypoint based which is involving others feature vectors to reduce the computational complexity. In this paper we review the differences keypoint approach using SURF (speeded up robust features) technique and block-based approach using DWT (discrete wavelet transform) technique using MATLAB platform. The performance of these two techniques are applying using dataset MICC-F220.