Content Authentication and Tampered Localization Using Ring Partition and CSLBP-Based Image Hashing

In the present world, innovation, and creative thinking were pivotal in bringing about the digital revolution. As a result of consistent development in multimedia processing algorithms, altering the contents of digital images has become more accessible. Image hashing has been found to be one of the most suitable approaches for content authentication applications. A ring partition and center-symmetric local binary patterns (CSLBP) approach has been introduced in the work for generating image hash. Our ring-based statistical characteristics remain unchanged when an image is rotated at any angle. As a result, it provides rotation invariant property. CS-LBP considers pairs of symmetrically opposite pixels around the central pixel. This symmetric encoding captures texture information that is invariant to certain rotations. The tampered detection and localization are done by computing the hash correlation between the original and tampered images. The experimental findings demonstrate that the suggested image hashing method can provide desirable robustness and discrimination in all content preserving operations (CPO). In comparison to specific current schemes, the proposed method offers shorter hash length, improved classification performance, and reduced hash generation time. Geometric rectification used in the suggested method can identify tampering even when geometric adjustments occur instantly. The receiver operating characteristics (ROC) curve demonstrates that the suggested model exhibits superior performance compared to other cutting-edge techniques.