Video denoising using spatio-temporal filtering

Abstract

Video denoising is very important to be used as a pre-processing stage, which should be considered both spatial and temporal dimensions. In this paper, video denoising scheme based on spatio-temporal filtering is researched, which contains spatially filtering in the two dimensional wavelet domain and temporal filtering in the time domain. Spatial filtering is firstly performed on individual video frames by wavelet shrinkage with an adaptive threshold selection. We use motion detection and recursive time weighted averaging to perform temporal filtering. It should be noted that both previous and following frames are considered to get the final filtered frame during the temporal filtering. Experimental results show that the proposed scheme yields better results in terms of both objective and subjective quality of the reconstructed video sequence.