Reduced reference image quality assessment metric using optimized parameterized wavelet watermarking

We propose a novel reduced reference quality assessment metric for image transmission rooted in an optimization approach toward parameterized wavelet-based data hiding. The approximation coefficients of one level parameterized wavelet transform of the original image at the transmitter are embedded into that of the horizontal and vertical detail coefficients in a robust and invisible manner to be used as a feature of the original image for comparisons at the receiver side. The best wavelet type used for decomposition is obtained by a curve fitting process, resulting in the most optimum parameters in terms of which the wavelet transform is formulated using genetic algorithm. Simulation results demonstrate the efficiency of our proposed image quality metric in that it has strong correlation with already established image quality metrics, e.g., Structural Similarity Index (SSIM).