Hybrid Architecture For Reversible Watermarking Of Medical Images Using Wavelet Transforms

Abstract

ct
To overcome the difficulties in the performance analysis of the real time implementation of
image processing algorithms; an attempt is made by proposing a hybrid architecture combining CPU
and FPGA. FPGA is used as hardware accelerator. The proposed architecture is tested by
implementing a real time reversible watermarking system using wavelet transform. Here, Two level
Haar wavelet transform based watermarking approach is used. Results have shown that the proposed
architecture provides robustness with low computational complexity. Peak Signal to Noise Ratio
(PSNR) is used to compare simulation results (MATLAB) with the real time implementation (DE2-
115) with respect to signal to PSNR. The experimental analysis shows the improvement in the PSNR
when compare to its simulation result.