项目名称: InSAR连接点自动稳健提取理论与方法研究
项目编号: No.41474010
项目类型: 面上项目
立项/批准年度: 2015
项目学科: 天文学、地球科学
项目作者: 靳国旺
作者单位: 中国人民解放军战略支援部队信息工程大学
项目金额: 80万元
中文摘要: InSAR技术具有快速、高精度、全天时、全天候等突出优势,在我国西部地区和常年阴雨地区的快速地形测绘与更新中具有重要应用价值。但是,在当前的InSAR测绘方案中,区域网平差与控制点加密所需的连接点仍依赖于人工选取,严重影响了生产效率。因此本项目拟研究InSAR连接点的自动稳健提取理论与方法,重点针对相同侧视方向近似平行轨道的InSAR数据,研究其特征点的自动提取和匹配方法,并初探相反侧视方向InSAR数据的特点及匹配方法。具体研究内容包括:SAR图像广义点特征提取、基于线特征增强的连接点自动匹配、匹配点的粗差探测与剔除、连接点的分布方案等。本项目的研究将解决InSAR连接点自动提取难题,突破对人工选点的依赖,奠定InSAR自动化选点的理论和技术基础。研究成果对于提高我国InSAR技术的理论水平和工程化水平、促进InSAR技术在地形图测绘与更新中的应用具有重要意义。
中文关键词: 雷达干涉测量;连接点;自动;提取;匹配
英文摘要: It has the advantages of fast, accurate, all time and all weather capabilities for InSAR. It can be used in rapid terrain surveying and mapping as well as updating of maps for our country's west regions and rainy areas. However, The tie points needed in block adjustment and densification of ground control points still depend on manual selection in current InSAR Processing. It has affected the producing efficiency seriously. Therefore, The theories and methods for robust and automatic extraction methods of tie points will be researched in this project. The InSAR Images with the same side-looking direction and approximate parallel tracks will be laid on emphasis to research the extraction of feature points and their matching algorithms.The characteristics and matching algorithms of the opposite side direction images will be studied preliminary. Topics include: generalized point feature extraction of SAR images, automatic matching based on line feature enhancement, gross error detection and removing for matched points, the distribution of tie points,etc. The difficulty in automatic extraction of tie points for InSAR will be resolved, as can break the dependence on artificial selection of tie points,and can lay the theoretical and technical foundations for automatic selection of tie points for InSAR. It is significant for improving the theoretical level and the technical level of InSAR,promoting the application of InSAR in topographic mapping and updating.
英文关键词: InSAR;Tie Point;Automatic;Extraction;Matching