An effective watermarking technique using BTC and SVD for image authentication and quality recovery

Abstract

Digital image protection has become a major concern in digital communication. Watermarking digital images has long been thought to be a good way to ensure the image quality. This paper proposes a watermarking scheme for image authentication and recovery of tampered images based on block truncation coding (BTC) and singular value decomposition (SVD). Here the image is divided into 4x4 non-overlapping blocks. Then the watermark information for authentication and recovery is embedded into it. The authentication watermark for each block will be generated using SVD and XOR operation, then it gets embedded into it. Simultaneously, the recovery watermark of each block will be generated using BTC and gets embedded in a mapped block. The watermark information gets embedded in two least significant bits (LSBs) of each pixel in a block. The proposed scheme got a higher degree of watermarking with better recovery against various kinds of attacks.