A SONAR images conditioning system in the wavelets domain

During their acquisition, SONAR images may face two issues: the presence of speckle noise and the appearance of areas with missing pixels. Image denoising is the process of reducing or removing the noise from images while the inpainting procedure involves the reconstruction of missing pixels in the images. This paper presents a SONAR images conditioning system composed by a wavelet shrinkage based denoising method used to reduce the speckle noise and an exemplar-based inpainting method used to reconstruct the lost parts of the SONAR images. The results are analyzed in terms of the Peak Signal to Noise Ratio (PSNR) and the Structural SIMilarity (SSIM) index.