MATLAB Audio Watermarking Implementation

This MATLAB implementation provides comprehensive audio watermarking solutions with two primary modules: wavelet domain audio watermarking incorporating Modified Discrete Cosine Transform (MDCT) for frequency-domain embedding, and advanced detection algorithms utilizing HRT (Harmonic Rejection Technique) with linear regression for improved watermark extraction accuracy. The codebase supports additional functionalities including audio preprocessing (such as normalization and filtering) and postprocessing operations to enhance watermark robustness against common audio manipulations. The implementation employs matrix operations for efficient MDCT computation and uses statistical regression models for detection threshold optimization. Developers can extend the system by integrating more sophisticated algorithms like spread spectrum techniques or psychoacoustic models to increase embedding capacity and improve detection rates. The modular design allows separate testing of embedding functions (watermark_encode.m) and detection functions (watermark_detect.m) with configurable parameters for wavelet decomposition levels and regression coefficients. Overall, this MATLAB framework offers a complete and effective methodology for implementing robust audio watermarking systems with customizable feature sets.