建筑物轮廓边缘检测方法对比分析  

Comparative Analysis of Building Outline Edge Detection Methods

作　　者：张宏伟[1,2] 赵宝强[3] 陈小卫 

机构地区：[1]信息工程大学 [2]空军航空大学 [3]61363部队

出　　处：《影像技术》2017年第5期76-78,共3页Image Technology

摘　　要：遥感影像中建筑物的提取一直以来都是研究的一个难点问题,而建筑物轮廓边缘的检测是建筑物准确提取的基础。本文对Edison算子、OFMM亚像素边缘检测方法和利用极值梯度的通用亚像素边缘检测方法进行了分析与实验,结果表明Edison算子在弱边缘的检测以及检测结果的连续性上要优于另外两种方法。Building extraction in remote sensing images has always been a difficult problem of research. And building outline edge detection is the foundation of building extraction accurately. In this paper, the Edison operator, OFMM subpixel edge detection method, and universal subpixel edge detection algorithm based on extremal gradient were analyzed with the experiment. The results show that Edison operator is better than the other two methods in weak edge detection and continuity.

关 键 词：遥感影像 建筑物 边缘检测 分割 

分 类 号：P237[天文地球—摄影测量与遥感]