一种面向对象的机场跑道变化检测方法  

An Object-oriented Method for Airport Runway Change Detection

作　　者：张艺明 肖文[1] ZHANG Yiming;XIAO Wen(School of Instrumentation Science and Opto-electronics Engineering,Beihang University,Beijing 100191,China)

机构地区：[1]北京航空航天大学仪器科学与光电工程学院,北京100191

出　　处：《航天返回与遥感》2019年第1期102-111,共10页Spacecraft Recovery & Remote Sensing

摘　　要：利用遥感影像变化检测技术获取机场跑道变化信息,可以为机场跑道打击效果评估等多种军事应用提供决策支撑。为了快速、准确的检测出机场跑道的变化区域,并定量、定性的获得变化属性,文章以"资源三号"卫星影像为例,提出了一种新的面向对象的机场跑道变化检测方法。首先,将经过配准处理的前后时相遥感影像进行多尺度分割,分割尺度利用尺度参数估计(Estimation of Scale Parameters,ESP)算法确定;之后,利用影像分割结果对不同时相的对象进行切割,形成前后时相上位置、大小一致的对象单元,再利用不同时相对象间的变化向量大小确定变化区域;最后,利用特征空间优化之后的特征集合对已确定的变化区域对象进行监督分类,获得机场跑道内部各对象的变化属性。结果表明:在变化区域的检测上,该方法避免了单张影像的分类过程,可有效提高检测效率;在变化类别的检测上,该方法在检测出的变化区域基础上进行分类,可大幅提高变化类别的检测精度,并且能够获得更为丰富的变化属性和满足快速准确检测机场跑道相关变化的信息。The remote sensing image change detection technology can be used to obtain the information of airport runway change,thus providing decision support for various military applications such as airport runway strike effect evaluation.In order to detect the change area of airport runway quickly and accurately,and to obtain the change attribute quantitatively and qualitatively,a new object-oriented method of airport runway change detection is put forward in this paper by taking ZY-3 satellite image as an example.Firstly,the pre-and post-time phase remote sensing images after registration processing are segmented,and the segmentation scale is determined by ESP algorithm.After that,the objects of different time are cut using the result of image segmentation to form the object unit with the same position and size on the different time image,and then the change region is determined using the change vector size between different time objects.Finally,the objects in the determined change area are performed Supervised Classification using the feature set optimized by feature space,and the change attributes of the objects in the airport runway are then obtained.The results show that in detecting changing area the proposed method can effectively improve the detection efficiency by avoiding the classification process of single image,and in detecting changing categories,the method can greatly improve the detection accuracy by classifying on the basis of the detected variation area.In addition,more abundant change attribute information can be obtained by the method.It should be noted that the method can meet the military needs in rapid and accurate detection of airport runway related changes.

关 键 词：面向对象 影像分割 监督分类 变化检测 航天遥感应用 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]