Perceptual image hashing using SVD based Noise Resistant Local Binary Pattern

Abstract

Image hashing has become a major research area due to rapid growth of image alteration techniques that can tamper digital images. The major concern of all image hashing schemes is the selection of robust features. Local Binary Pattern (LBP) is a technique that selects robust features for different image applications. This paper presents a perceptual image hashing scheme by the utilization of Noise Resistant Local Binary Pattern (NRLBP), a modified form of the LBP. The features of NRLBP are extracted from non-overlapping blocks of a gray scale image. The NRLBP is combined with Singular Value Decomposition (SVD) to provide good robustness characteristics against a number of non-malicious distortions. Another major advantage of the proposed scheme is to detect localized tampered regions. Experimental results exhibit that the proposed scheme has the capability to detect tampering as small as 3% of the image size and at the same time offers good robustness properties.