A Digital Watermarking Algorithm Based on Digital Wavelet and Singular Value Division

Abstract

Discrete wavelet has excellent characteristics such as multi-scale analysis, fast operation speed, noise elimination, and singular value decomposition has stable characteristics, so the blind watermarking algorithm combined with the two has high practical value. At the same time, the watermark is scrambled by Arnold transform to improve its secrecy. After using JPEG compression and shearing to attack the image after embedding the watermark, the clearer watermark can still be extracted from it. The obtained PSNR value is greater than 30. At the same time, the NC value is very close to 1. Therefore, the algorithm has good robustness and can provide effective protection for the copyright of various digital works.