A Novel Approach of Fragile Watermarking for Authentication and Tamper Detection Exploiting Local Binary Pattern (LBP)

Data communication and transmission through internet has been rapidly increased now a days specially during and after COVID-19 pandemic situation. So, it is essential to preserve multimedia data from outlawed access. Here, a novel approach of fragile watermarking for authentication and tamper detection has been suggested. Firstly, an input image is split up into non-overlapping blocks of sized \(3\times 3\) . A reference block has been chosen randomly to get the watermark which is hidden at the LSBs (Least Significant Bits) of the non-reference blocks of input image. The proposed scheme takes the advantage of Local Binary Pattern (LBP) feature of the image which is used to generate the watermark. The proposed scheme maintains high quality watermarked images with 51.68 dB average PSNR and 0.9955 average SSIM. The proposed scheme has been analysed against different attacks like cropping and copy-move forgery which shows 100% Tamper Detection Rate (TDR) for cropping attack and 99.39% for copy-move forgery attack. The presented scheme also exceeds other methods with respect to PSNR and SSIM value as depicted in comparative analysis.