A Hybrid Approach to Digital Image Watermarking Integrating DWT, DFT, and Genetic Algorithm
摘要
Digital multimedia security demands innovative solutions to address the dual challenge of preserving image quality while ensuring robust authentication. This paper introduces a hybrid digital image watermarking method that integrates Discrete Wavelet Transform (DWT), Discrete Fourier Transform (DFT), and Genetic Algorithm (GA) to meet these demands effectively. DWT decomposes the original image into frequency sub-bands, creating a stable and efficient foundation for embedding the watermark. DFT enhances robustness against spatial and geometric attacks by embedding the watermark into carefully selected DWT coefficients. GA optimizes the embedding process through iterative evolution, ensuring minimal image quality degradation while achieving a balance between invisibility and robustness. Experimental validation using multiple 512x512 color host images and 32 × 32 watermark images resulted in an impressive average Peak Signal-to-Noise Ratio (PSNR) of 41.67 dB. The results demonstrate the method's effectiveness in providing reliable copyright protection and image authentication with minimal visual distortion, making it a powerful solution for digital multimedia security.