欢迎来到天天文库
浏览记录
ID:54369099
大小:625.82 KB
页数:8页
时间:2020-04-30
《保持图像亮度的局部直方图均衡算法.pdf》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库。
1、第5期电子学报Vol.34No.52006年5月ACTAELECTRONICASINICAMay2006保持图像亮度的局部直方图均衡算法1,2111江巨浪,张佑生,薛峰,胡敏(1.合肥工业大学计算机与信息学院,安徽合肥230009;2.安庆师范学院物理与电气工程学院,安徽安庆246011)摘要:现有的局部直方图均衡算法用于图像的对比度增强时,输出图像的亮度与输入图像无关.由于图像增强的很多实际应用需要保持原始图像的亮度,为此提出一种新的局部直方图均衡算法.以子块部分重叠的局部直方图均衡算法(POSHE)为基础,改进其对子块图像的均衡方式:根据亮度均值对子块图像进行递归分解,得到一系列
2、不同灰度范围的子图像,然后对每一个子图像在其相应的灰度范围内进行直方图均衡,最后合并这些子图像的均衡结果.实验结果表明,这种局部直方图均衡新算法既能充分增强图像局部细节,又能保持输入图像的亮度,且亮度的保持程度可由用户进行控制.本文算法使所增强的图像具有更加自然的视觉效果,能更好地满足消费类电子工业对图像对比度增强的应用要求.关键词:对比度增强;直方图均衡;局部直方图均衡;亮度保持中图分类号:TP391.41文献标识码:A文章编号:0372-2112(2006)05-0861-06LocalHistogramEgualizationwithBrightnessPreservation
3、1,2111JIANGJu-lang,ZHANGYou-sheng,XUEFeng,HUMin(1.SchoolofComputerScienceandInformation,hefeiUniversityofTechnology.hefei,Anhui230009,China;2.SchoolofPhysicsandElectricalEngineering,AngingTeacher'sCollege.Anging,Anhui246011,China)Abstract:Asthetraditionalalgorithmsforlocalhistogramegualizationi
4、susedtoenhancethecontrastofanim-age,theoutputbrightnessisalwaysregardlessoftheinputimage.However,thisisnotthedesirablepropertyinmanyap-plicationswherepreservingtheoriginalbrightnessisessential,soanewalgorithmoflocalhistogramegualizationispro-posed.Basedonthealgorithmofpartiallyoverlappedsub-blo
5、ckhistogramegualization(POSHE),wechangethewayofegualizingsub-block:basedonthemean,eachsub-blockisseparatedrecursivelyintoaseriesofsub-imageswithdifferentgrayranges,thenthehistogramsareegualizedovertheirrespectivegrayranges,andfinallytheoutputofthesub-blockiscomposedwiththeresultingegualizedsub-
6、images.Thesimulationresultsshowthat,withthenewalgorithm,thelocalde-tailsontheimagecanbeenhancedeffectively,andtheoriginalbrightnesscanalsobepreservedinthemeantime.Besides,scalabledegreeofbrightnesspreservationmaybeachievedbychoosingthenumberofrecursiveseparation.Theproposedalgorithmcanenhanceth
7、eimageinamorenaturalmannerandhence,isveryusefulformanycommercialpurposessuchasconsumerelectronicswherebrightnesspreservationisreguired.Keywords:contrastenhancement;histogramegualization;localhistogramegualization;brightnesspreserv
此文档下载收益归作者所有