A Novel Fast NLM Denoising Algorithm Based on Improved Kernel Function Parameters: Applied to RFID Labels

This paper proposes a novel fast NLM algorithm. The integral diagram is introduced to speed up algorithmic processing and the kernel function coefficient is changed through the function diagram to improve its performance. The denoising effects of three RFID label images collected by RFID system which added the gaussian white noise is studied. This paper collects and analyzes the running time, PSNR and SSIM of various NLM denoising algorithms under different RFID label images added the gaussian white noise of the same standard deviations. The results verify that the proposed algorithm has both high speed and high efficiency for the applications to RFID labels.