Abstract
The paper proposes an efficient watermarking scheme by incorporating singular value decomposition (SVD), discrete cosine transform (DCT) and homogeneity information in the carrier image. The algorithm decomposes the carrier and watermark images into non-overlapping blocks and analyzes the homogeneity of pixels in them to identify the possible blocks of carrier image where the suitable watermark block can be attached so that the watermarking process maintain better visual image fidelity in the watermarked image. The algorithm avoids the decomposed high frequency carrier image blocks while attaching watermark blocks to reduce the visual distortion in the watermarked images. Experimental results show that the proposed algorithm produces good quality watermarked images and are less immune to potential attacks.