月表图斑数字量化分类方法研究  

Research on Digital Quantitative Classification Method of the Lunar Surface Spots

作　　者：陈垦[1] 廖平[1] 王玉川[1] 肖鹏飞[1] 王勇[1] 苗放[2,3] 

机构地区：[1]四川省国土勘测规划研究院,四川成都610045 [2]北京大学数字中国研究院政策与战略研究中心,北京100871 [3]成都理工大学空间信息技术研究所,四川成都610059

出　　处：《新技术新工艺》2015年第5期106-109,共4页New Technology & New Process

基　　金：863项目(2010AA122202)

摘　　要：图斑是月球表面的重要特征之一,对月球科学研究及月表制图具有重要意义。现有的图斑分类技术主要基于手动的经验判断为主,存在效率较低和主观性较强等问题,以月表图斑影像几何特征分类研究为导向,利用面向图像信息提取的需求进行综合分析,提出一种基于大小、边缘清晰度、曲率、宽窄(坑唇)、中央复杂度和深度等六因子的"六位十级"月表图斑影像几何特征数字量化方法。为实现快速、智能、自动判识提供理论基础;为开展月表形貌和构造的综合研究提供依据;为开展月表的形成、演化以及制图研究提供技术支持。The impact craters have the important meaning to the moon with scientific research and drawings. The existing classification technology of impact craters is based on manual experience judgment problem such as low efficiency and strong subjectivity. As it being a guide, that comprehensive analysis with image information extraction, putting forward a kind of six-ten-level geometrical characteristic methods based on the moon impact craters which combined with size, edge of clarity, curvature, pit lip, central complexity and depth. The paper wants to achieve rapid, intelligent and automatic theoretical foundation knowledge, to develop the moon morphology and structure of the comprehensive research provide the basis, and for the development of the formation and evolution of the moon and drawing research provides the technical support.

关 键 词：六位十级 判识因子 几何特征 嫦娥一号数据 月表图斑 

分 类 号：TP753[自动化与计算机技术—检测技术与自动化装置]