Evaluation of Singular Value Decomposition based Digital Image Watermarking with Dither Quantization and Edge Detection

The use of singular value decomposition (SVD) in watermarking area has been developed by many researchers. In this paper, we describe our evaluation result against an SVD based watermarking scheme with Dither quantization and edge detection proposed by Mohan et al. As an evaluation benchmark, we use StirMark to produce modified images by various image attacks. The extracted watermarks from these images were compared with the original watermark by calculating its bit correlation ratio (BCR). The result shows that the scheme is robust against JPEG compression, rotation, scaling, and PSNR attack. The scheme is also quite robust to some extents against removeline attack and cropping.