Directional Total Variation Filtering Based Image Denoising Method

In this paper, we study signal denoising technique based on total variation (TV) which was reported by Ivan W. Selesnick, Ilker Bayram [1]. Here, we present a directional total variation algorithm for image denoising. In most of the image denoising methods, the total variation denoising is directly performed on the noisy images. In this work, we apply a 1D TV denoising algorithm in sequential manner on the pixel sequence obtained in different orientations including zig-zag, horizontal and vertical. The performance of the proposed method is evaluated using the standard test images and the quality of the denoised images is assessed using various objective metrics such as peak signal to noise ratio (PSNR), structural similarity index metrics (SSIM), visual signal to noise ratio (VSNR). Various experiments show that the proposed method provides promising results with less computational load.