Open AccessEnhancement of LLLIs with improved BCP and matrix completion
Author(s) -
Yang Jie
Publication year - 2017
Publication title -
electronics letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.375
H-Index - 146
eISSN - 1350-911X
pISSN - 0013-5194
DOI - 10.1049/el.2016.4686
Subject(s) - image enhancement , contrast enhancement , matrix (chemical analysis) , brightness , computer science , outlier , channel (broadcasting) , artificial intelligence , process (computing) , point (geometry) , matrix completion , contrast (vision) , computer vision , image (mathematics) , algorithm , mathematics , optics , materials science , physics , telecommunications , medicine , geometry , composite material , quantum mechanics , gaussian , magnetic resonance imaging , radiology , operating system
The problem of enhancement of low light level images (LLLIs) with the characteristics of extremely low brightness and low contrast is addressed. A coarse‐to‐fine framework to settle the problem is proposed. First, a point‐wise linear relationship is constructed between an LLLI and its enhancement result, and a prior named bright channel prior (BCP) is introduced to guide the enhancement process. The BCP to avoid over enhancement is improved, and the idea of matrix completion to remove the outlier noises in the enhanced image is used. Experimental results on real images demonstrate that the method is effective for the enhancement of LLLIs.