用于遥感影像亮度均衡的亮度补偿方法  

The Brightness Compensation of Remote Sensing Image for Brightness Equalization

作　　者：刘小帆 鲁金金 邢鹏威 张炎 LIU Xiaofan;LU Jinjin;XING Pengwei;ZHANG Yan(Nanning College of Technology,Guilin 541006,China;Beibu Gulf University,Qinzhou 535037,China;Guilin University of Technology,Guilin 541006,China)

机构地区：[1]南宁理工学院,桂林541006 [2]北部湾大学,钦州535037 [3]桂林理工大学,桂林541006

出　　处：《航天返回与遥感》2022年第4期97-110,共14页Spacecraft Recovery & Remote Sensing

基　　金：南宁理工学院校级重大科研项目(KY202101、KYZ202201);广西高校中青年教师科研基础能力提升项目(2021KY1670、2020KY58008)。

摘　　要：影像亮度均衡是从事航天遥感分析工作的重要前提。由于Retinex法与亮度拉伸法等传统亮度均衡方法具有依赖固定模型、易造成影像失真等缺点,文章提出了用于遥感影像亮度均衡的亮度补偿模型。该模型首先根据遥感影像的亮度特征值计算亮度补偿值,再根据亮度补偿值建立亮度补偿模型,最后将亮度补偿模型用于影像来实现遥感影像的亮度均衡。为验证该方法的效果,文章使用ARGON KH-5卫星的格陵兰岛影像与Google Earth中采集的航空影像与CORONA卫星拍摄的中国云南的影像进行亮度均衡实验,结果表明:相较于Retinex与亮度拉伸模型,亮度补偿模型既兼顾了两者的优势,又降低了影像的标准差(影像亮度更加均衡),提高了影像的平均梯度(地物信息更加清晰)。Achieving brightness equalization is an important work before RS-image analysis.However,the current methods,such as Retinex and brightness stretch have some disadvantages like demand reference histogram,rely on the fixed function,and produce color distortion,etc.According to above disadvantages,this paper proposed the brightness compensation method to achieve RS-image color equalization.Firstly,calculating the value according to the brightness characteristics of the RS-image.Secondly,establishing the brightness compensation model.Finally,the brightness compensation model is used to realize the brightness balance of the RS-image.In order to verify accuracy of this method,the experiments were using the Greenland image taken by the ARGON KH-5 in the 1960s and the aerial image taken by Google Earth.The results shows that compared with brightness stretch and Retinex,the brightness compensation method has the following advantages:Compared with the Retinex,the standard deviation of the result is reduced obviously,and the average gradient of the result is increased compared with the brightness stretch.

关 键 词：航天遥感 亮度均衡 亮度补偿值 标准差 平均梯度 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]