遥感图像质量分析与评价方法研究  

Remote Sensing Image Quality Analysis and Assessment Methods Study

作　　者：吕赛[1,2,3] 李绪志[3] 张九星[3] 

机构地区：[1]中国科学院光电研究院,北京100094 [2]中国科学院研究生院,北京100190 [3]中国科学院空间应用工程与技术中心(筹),北京100094

出　　处：《计算机仿真》2012年第9期266-270,共5页Computer Simulation

基　　金：国家自然科学基金项目资助(90718010)

摘　　要：评价遥感图像质量,由于观测器在轨运行,造成环境噪声干扰,图像配准质量差,直接反映了空间遥感器在轨成像性能,而且对遥感数据的后期融合也有影响。针对传统遥感图像质量评价方法精度不高和通用性差的特点,提出了一种阈值法结合多波段遥感影像相关性信息的云含量检测方法,并改进了刃边的MTF参数估计方法,从而提高了评价结果的计算精度,通用性好,便于自动化实现。利用遥感图像进行仿真,结果表明所采用方法能较好的满足应用需求,具有较高实用价值。The cloud ratio and MTF estimation are two important indicators for the quality of remote sensing ima-ges. They not only represent the performance of remote sensing equipments, but also affect subsequent applications. In view of the inaccuracy and poor universality characteristic of the traditional remote sensing image quality evaluation methods, a cloud detection method with the correlation information of multi-band remote sensing image was pro- posed, and the MTF parameter estimation method was improved. These methods have improved the precision and can be used universally and automaticly. The simulation experiments were carried out utilizing remote sensing images, and the results indicate that the simulations are adequate for comprehensive assessment of remote sensing images and of great value in practice.

关 键 词：图像质量 评价 遥感图像 云检测 调制传输函数 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论]