计算机技术与发展2017，Vol.27Issue(8)：73-78,6.DOI:10.3969/j.issn.1673-629X.2017.08.015
基于图像融合策略的Retinex背光图像增强算法
Backlight Image Enhancement Algorithm of Retinex Based on Image Fusion Strategy
摘要
Abstract
Retinex is a model of brightness and color perception based on human vision,which has wide applications in the field of image enhancement.As a common scene,backlight could make captured images blur in their dark areas,loss of information and others,which could impact further analysis and detection on the image.In order to improve the enhancement effect of such images,backlight image enhancement algorithm using Retinex based on image fusion strategy is proposed.It gets both the white balance and enhanced images by original image to carry on color correction and contrast enhancement.Then two images are achieved weight maps to implement Laplacian pyramid fusion for enhanced image.Weights maps highlights the details of the backlight region,which can effectively raise the enhancement effect of backlight area combined with fusion technology and acquire identified images with high contrast and ample hues.Compared with other existing methods,experimental results show that the proposed algorithm has retained the details and promoted the contrast and definition of backlight image.关键词
Retinex/背光图像增强/权重图/融合策略Key words
Retinex/backlight image enhancement/weight maps/fusion strategy分类
信息技术与安全科学引用本文复制引用
马康,崔子冠,干宗良,唐贵进,刘峰..基于图像融合策略的Retinex背光图像增强算法[J].计算机技术与发展,2017,27(8):73-78,6.基金项目
国家自然科学基金资助项目(61501260,61471201,61471203) （61501260,61471201,61471203）
江苏省自然科学青年基金(BK20130867) （BK20130867）
江苏省高校自然科学重大项目(13KJA510004) （13KJA510004）
江苏省六大人才高峰(2014-DZXX-008) （2014-DZXX-008）
南京邮电大学校科研基金(NY214031) （NY214031）
南京邮电大学"1311人才计划"资助课题 （）
