首页 | 本学科首页   官方微博 | 高级检索  
     

基于高分遥感影像的建筑物侧面信息提取及其高度计算
引用本文:宋旭东,曹文峰,冯德俊,张星星. 基于高分遥感影像的建筑物侧面信息提取及其高度计算[J]. 北京测绘, 2020, 0(3): 296-300
作者姓名:宋旭东  曹文峰  冯德俊  张星星
作者单位:成都市国土规划地籍事务中心;西南交通大学地球科学与环境工程学院;成都市土地整治中心
摘    要:本文选取成都市某一区域建筑物A、B为研究对象,采用分辨率为0.61 m的Quick bird影像,运用图像分割法和LVQ神经网络算法,提取建筑物侧面信息,根据假设法原理,构建高度计算物理模型,求取建筑物高度。对比实测数据,结合可能影响实验结果的实地因素、遥感影像因素进行精度分析与评价,探讨基于高分遥感影像的建筑物侧面信息提取和高度计算的方法。结果表明,LVQ神经网络算法在建筑物侧面提取和高度计算中有更好的应用价值,精度高达94%。

关 键 词:高分遥感影像  侧面信息提取  图像分割法  LVQ神经网络  高度物理模型

Building Side Information Extraction and Height Calculation Based on High-Resolution Remote Sensing Imagery
SONG Xudong,CAO Wenfeng,FENG Dejun,ZHANG Xingxing. Building Side Information Extraction and Height Calculation Based on High-Resolution Remote Sensing Imagery[J]. Beijing Surveying and Mapping, 2020, 0(3): 296-300
Authors:SONG Xudong  CAO Wenfeng  FENG Dejun  ZHANG Xingxing
Affiliation:(Chengdu Land Planning and Cadastre Center, Chengdu Sichuan 610072, China;Faculty of Geosciences and Environmental Engineering, Southwest Jiaotong University, Chengdu Sichuan 610097, China;Chengdu Land Management Center, Chengdu Sichuan 610072, China)
Abstract:This paper chooses buildings A and B in a certain area of Chengdu as the research objects.Quick bird image with resolution of 0.61m is used.Image segmentation and LVQ neural network algorithm are used to extract side information of building.According to the principle of hypothesis method,a physical model of height calculation is constructed to calculate the height of building.By comparing the measured data,combining with the factors that may affect the experimental results and remote sensing image,the accuracy analysis and evaluation are carried out,and the methods of building side information extraction and height calculation based on high-resolution remote sensing image are discussed.The results show that LVQ neural network algorithm has better application value in building side extraction and height calculation,and the accuracy is as high as 94%.
Keywords:high resolution image  side information extraction  image segmentation  Learning Vector Quantization(LVQ)neural network  altitude physical model
本文献已被 CNKI 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号