摘要
针对高分辨率图像上部分规则建筑物阴影轮廓的特点,基于曲率尺度空间角点探测法(CSS,Curvature Scale Space),提出了一种阴影长度半自动量算方法。该方法首先计算角点最近距离,然后进行长度和角度筛选,最后进行统计平均。在以北京师范大学校园为例的实验中,较准确地得到建筑物高度信息,从而显示该方法不仅可以弥补传统手工量算阴影宽度工作繁琐和效率低下的缺陷,而且可以有效地提高计算精度。
In the light of characteristics of some building shadow outlines in high resolution images, the authors put forward a new method for measuring the shadow width based on the Curvature Scale Space (CSS) method. This semiautomatic method first computes the shortest distance between corners, then performs length and angle filtering, and finally calculates statistical averages. The campus of Beijing Normal University was chosen as the test field. The result demonstrates that this method can make up such drawbacks in the manual work of shadow width measurement as the cumbersome process and low efficiency and at the same time effectively improve the calculation precision.
出处
《国土资源遥感》
CSCD
2008年第3期32-36,共5页
Remote Sensing for Land & Resources
基金
国家自然科学基金(40671127)
"111"计划(B06004)
长江学者和创新团队发展计划资助
关键词
角点
统计平均
建筑物
阴影长度
Corner
Statistical average
Building
Shadow length