A New JPEG Tolerance Detector System for Image Authentication

• Main Authors: Cheng-Shian Lin, 林承賢
• Other Authors: Chien-Chang Chen
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/91716288967815017475

碩士 === 玄奘大學 === 資訊管理學系碩士班 === 93 === In this thesis, we first propose a DCT-based image authentication approach which can survive JPEG lossy compression. In our approach, the image feature is calculated and then embedded into DCT middle frequency domain by our proposed quantization properties. The experimental results demonstrate that the proposed image authentication approach can effectively resist the JPEG compression. Then we introduce how Genetic Algorithm (GA) can be utilized to find nearly optimal authentication embedded positions, so as to achieve high quality protected image. In the evolution process, the embedded positions are simulated as chromosomes. Then we can obtain the nearly optimal embedded positions by natural selection and GA operators, such as reproduction, crossover, and mutation. Finally, the experimental results demonstrate that GA can improve the protected image effectively.