Font Size: a A A

Research And Implementation On Image Clarity Algorithms

Posted on:2017-05-01Degree:MasterType:Thesis
Country:ChinaCandidate:Y LiuFull Text:PDF
GTID:2348330485456578Subject:Control engineering
Abstract/Summary:PDF Full Text Request
Image clarity algorithms are widely used in transportation, security, medical and military fields, which has important applications significance. Image clarity technology has been a hot research topic in computer vision and digital image processing.In this thesis, this paper focus on the problem of image clarity, focusing on the image clarity algorithms of fog, night, underwater and dust conditions. The main research contents are as follows:(1) based on dark priori algorithm for single image defogging and based on semi-inverse defogging algorithm of layers existing the color distortion and time-consuming disadvantages, this paper presents a improved semi-inverse defogging algorithm. The algorithm can detect the fog area, then calculate the overall atmospheric optical value, according to the atmospheric model,, the image can be restored, and the restored image ultimately maintain a good color and detail information.(2) Retinex algorithm is improved using RGB color space conversion to HSV color space, only enhanced luminance component, to solve the color distortion of the underwater and dust degraded image; at the same time, the equivalence ratio of low illumination image enhancement algorithm has been improved, the enhanced proportion coefficient is replaced by Retinex algorithm processing to deal with the problem of the low illumination image dark area noise suppression and highlight areas over enhancement; combined with the advantages of the improved color space conversion retinex algorithm in the processing of underwater and dust images the improved the low illumination enhancement algorithm in the processing of low illumination images. the paper weighted fusion of the two kinds of enhancement algorithms, proposes an improved fusion retinex algorithm.(3) respectively, from the subjective and objective indicators to evaluate the restored image visual quality by processed of image clarity algorithm, after experiments showed that the improved semi-inverse algorithm has strong ability to dehaze, calculation speed. Improved fusion retinex algorithm for night, underwater and dust have good image detail enhancement and color restored.
Keywords/Search Tags:image clarity, semi-inverse, image enhancement, image restored, fusion
PDF Full Text Request
Related items