Dual Tree Complex Wavelet Transform and Spread Spectrum Technique for Robust Medical Image Watermarking
Ganta Catur Paramitha, Ledya Novamizanti, Gelar Budiman · 2023
The rapid evolution of technology and computer networks has brought about a paradigm shift in data transmission within the medical field, facilitating the utilization of wireless networks for seamless communication. However, this convenience also introduces various challenges, including copyright infringement, data theft, and the imperative need for robust ownership identification during data transmission across open networks. Addressing these concerns, watermarking emerges as a practical solution by discreetly embedding confidential personal information within other data. This paper introduces a watermarking technique designed to enhance the security of medical images, utilizing a fusion of the Dual-Tree Complex Wavelet Transform and spread spectrum (DTCWT-SS) methods. The watermark image undergoes a pre-processing stage involving a spread spectrum modulation technique employing a random pseudo-noise code. Subsequently, the watermark series is meticulously embedded into the DTCWT-based host image, resulting in the creation of a watermarked image. At the receiving end, the watermarked image undergoes processing by DTCWT, and the optimal subband is carefully selected to extract the original watermark with high fidelity. The proposed scheme achieves an impressive peak signal-to-noise ratio of over 50 dB for each medical image modality, underscoring the preservation of image quality despite watermarking. Moreover, the watermark retrieval process demonstrates exceptional performance, achieving a normalized correlation of 1 and a bit error rate of 0 under no-attack conditions, ensuring accurate and reliable watermark extraction. Notably, the proposed model exhibits robustness against signal processing attacks and noise addition, confirming its efficacy in various scenarios and resilience against potential threats.