Digital Image Watermarking Scheme Based on DCT and DWT Algorithms

This graduation thesis presents a digital image watermarking scheme utilizing Discrete Cosine Transform (DCT) and Discrete Wavelet Transform (DWT) algorithms, accompanied by comprehensive program documentation. The paper elaborates on the fundamental principles of digital image watermarking, implementation procedures, and examines the advantages and limitations of both algorithms. The implementation typically involves dividing images into frequency domains using block-based DCT processing or multi-resolution analysis through DWT decomposition. Key functions include embedding watermarks in mid-frequency DCT coefficients to balance visibility and robustness, or inserting watermark data in specific DWT subbands (LL, LH, HL, HH) to leverage spatial-frequency characteristics. The thesis also provides a survey of related research and proposes future research directions with potential enhancements. Experimental results and analytical data are included to validate the effectiveness and feasibility of the proposed scheme, demonstrating performance metrics like PSNR (Peak Signal-to-Noise Ratio) and NCC (Normalized Cross-Correlation) for robustness testing against various attacks. In summary, this work serves as a comprehensive and detailed guide for research in the field of digital image watermarking, offering practical code examples for coefficient modification, inverse transformation, and watermark extraction procedures.