Naslov
Video Denoising Based on Adaptive Temporal Averaging

Autori
Bartovčak, David ; Vrankić, Miroslav

Izvornik
Engineering review (Technical Faculty University of Rijeka) (1330-9587) 32 (2012), 2; 64-69

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni

Ključne riječi
signal processing ; video denoising ; temporal averaging ; averaging interval ; pixel-domain method ; additive white Gaussian noise

Sažetak
This paper proposes a video denoising algorithm based on adaptive, pixel-wise, temporal averaging. The algorithm decomposes videos into a set of 1-D time dependent signals and then removes the noise by establishing temporal averaging intervals throughout each signal from the set. Temporal averaging intervals are established by simple, yet effective comparison processes which include two- way thresholding. The proposed algorithm is tested on several types of 1-D signals and benchmark videos. Experiments suggest that the proposed algorithm, despite its simplicity, produces high- quality denoising results and even outperforms some state-of-the-art competitors.

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika

POVEZANOST RADA