Izvorni znanstveni članak
https://doi.org/10.17559/TV-20130918123509
Adaptive Methods for Video Denoising Based on the ICI, FICI, and RICI Algorithms
Jonatan Lerga
orcid.org/0000-0002-4058-8449
; Faculty of Engineering, University of Rijeka, Vukovarska 58, HR-51000 Rijeka, Croatia
Edi Grbac
; Faculty of Engineering, University of Rijeka, Vukovarska 58, HR-51000 Rijeka, Croatia
Victor Sucic
; Faculty of Engineering, University of Rijeka, Vukovarska 58, HR-51000 Rijeka, Croatia
Nicoletta Saulig
; Department of Engineering, Juraj Dobrila University of Pula, Zagrebačka 30, 52100 Pula, Croatia
Sažetak
In various applications, reducing noise from video sequence is of crucial importance. In this paper, we have presented performance analysis of the novel video denoising method based on the relative intersection of confidence intervals (RICI) rule, and compared it to the methods based on the intersection of confidence intervals (ICI) rule and the fast ICI (FICI) rule. Detailed comparisons, based on two test video signals, are provided for a range of noise levels and different noise types. The RICI video denoising method has shown to outperform the original ICI based method, both in the algorithms execution time, reducing it by up to 11 %, and in the level of noise suppression, improving it by up to 10 dB. It also outperforms the FICI based video denoising method by up to 12.7 dB for the two test videos.
Ključne riječi
intersection of confidence intervals (ICI) rule; locally adaptive filters; video denoising
Hrčak ID:
200590
URI
Datum izdavanja:
26.5.2018.
Posjeta: 1.599 *