Removal of salt and pepper noise from SAR images using optimized APCNN in Shearlet transform domain

Abstract

The simultaneous elimination of salt and pepper noise and the conservation of the edge still present an unresolved challenge. In this paper, the optimized adaptive pulse-coupled neural network (APCNN) is proposed in Shearlet transform domain to reduce the salt and pepper noise present in the Synthetic Aperture Radar (SAR) images. The APCNN is optimized by Grey Wolf optimization technique. In Shearlet transform (ST) domain, PCNN upgrades the subtleties of images in the low-recurrence and high-recurrence sub-bands. Then the improved low-recurrence and high-recurrence sub-bands are utilized for inverse ST to acquire the improved images. The outcomes demonstrate that the optimized APCNN in Shearlet transform technique yields better quality of filtered SAR images. Also, the salt and pepper noise are smoother and show superior to the existing methodologies which are based on the parameter analysis of structural similarity index measure (SSIM) and peak signal-to-noise-ratio (PSNR). This method improves the PSNR value of 3.78%, and SSIM of 22.4% when compared with Adaptive PCNN denoising technique.