Contoulet denoising of natural images
Atif Bin Mansoor, Shoab Ahmad Khan · 2008
The paper investigates the image denoising utilizing a new discrete transform, contourlet transform through thresholding technique for natural images. We investigate the new transform with varied amount of noise for three sigma thresholding on a block of four natural images. Signal to noise ratio and visual judgement are made to assess the denoised images. The method displays improved signal to noise ratio for noisy images, but with observable artifacts.