论文部分内容阅读
针对卫星遥感图像可用度分类问题,提出基于图像信息的云覆盖率、云厚度、云破碎度的云结构特征描述概念.研究了它们对于卫星遥感图像可用信息量的影响程度,构建了由厚度相关的云覆盖率及云破碎度构成正交可用度评估空间.并进一步分析了云结构特征与可用度之间的近似线性及单调特点,证实了所提描述方法对于可用度表征的合理性和有效性.通过与500幅不同可用度等级的专业人工判读结果比较,证实综合这3个变量的可用度判读结果与人工判读结果的准确度比较达到95%.该方法为建立卫星遥感图像可用度评估模型,自动分类得到用户需要的图像数据提供理论及技术支持.
Aiming at the availability classification of satellite remote sensing images, the concepts of cloud structure feature description based on cloud information coverage, cloud thickness and cloud fragmentation are put forward, and their influence on the amount of information available for satellite remote sensing images is studied. Cloud coverage and cloud fragmentation constitute an orthogonal availability evaluation space, and further analyzes the approximate linearity and monotonous characteristics between the cloud structure features and availability, and proves the rationality and effectiveness of the proposed method for the representation of availability By comparing with the results of 500 manual interpretations of different degrees of usability, it is proved that the accuracy of the interpretation of the availability of these three variables is 95% compared with the accuracy of the manual interpretation.The method is to establish the evaluation of the availability of satellite remote sensing images Model, automatic classification to get the user needs image data to provide theoretical and technical support.