Based on SVD Intelligent Image Watermarking

• Main Authors: Yu-Jie Jhuang, 莊御捷
• Other Authors: Hung-Hsu Tsai
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/hepeeh

碩士 === 國立虎尾科技大學 === 資訊管理研究所 === 98 === This thesis presents two image watermarking for image copyright protection. The first method employs the Particle Swarm Optimization (PSO) algorithm to select nearly optimal solution parameters for embedded Singular Value Decomposition (SVD) coefficients and watermark strengths. It is called SVD-based Image Watermarking using PSO (SIWP). The second method which is called SVD-based Image Watermarking in Wavelet Domain Using SVR (SIWS)， is proposed to improve the robustness of SIWP method. The SIWS method is developed in Discrete Wavelet Transform (DWT) and SVD domains. It utilizes the SVR to estimate original coefficients of SVD domain during watermark extract. Experimental results show the two methods possess significant transparency， and the robustness of the SIWS method is better than the SIWP method. Additionally， the SIWS method is superior to the existings method on both transparency and robustness under consideration here.