Improved PCNN Polarization Image Denoising Method Based on Grey Wolf Algorithm and Non-Subsampled Contourlet Transform

Abstract

In this article, an adaptive pulse-coupled neural network (PCNN) polarization image denoising method based on Grey Wolf Optimization (GWO) and Non-Subsampled Contourlet Transform(NSCT) is proposed. Different from the traditional PCNN denoising method, the captured polarization image was firstly devised by the NSCT and enforced band-decomposition to denoised by PCNN. The evaluable index of the image was used for quantitative analysis. Then, GWO is used to update PCNN inherent voltage constant and attenuation time constant and neurons connected intensity factor three model parameters, after looking for multiple optimal solutions, and then to polarized image denoising to achieve the best effect. This method not only avoids the image edge blurring caused by the traditional image denoising method, but also solves the problem that the parameters of the PCNN are difficult to accurately estimate. Hence, it is more suitable for polarization images containing noise. The experiment and the quantitative analysis of image evaluation indices showed that NSCT-GWO-PCNN effectively suppresses the noise in polarization image by reducing salt-and-pepper noise while protecting edges.

Enhancing Positron Emission Tomography Image Reconstruction: A Bayesian Approach Incorporating Total Variation and Median Root Prior

“…The Normalized Root Mean Square Error (NRMSE) and Signal-to-Noise Ratio (SNR) values of the images were calculated to provide a quantitative measure of the algorithm's effectiveness [20]. Figure 5 illustrates the plot of NRMSE values over the course of iterations for the Shepp-Logan phantoms reconstructed by the four algorithms.…”