地球信息科学学报 ›› 2015, Vol. 17 ›› Issue (8): 995-1000.doi: 10.3724/SP.J.1047.2015.00995

• • 上一篇    下一篇

无参数高分辨率遥感影像的建筑高度快速提取方法

乔伟峰1,2,3(), 刘彦随1**(), 项灵志2, 王亚华2   

  1. 1. 中国科学院地理科学与资源研究所,北京 100101
    2. 南京师范大学地理科学学院,南京 210023
    3. 江苏省地理信息资源开发与利用协同创新中心,南京 210023
  • 收稿日期:2014-10-22 修回日期:2014-11-18 出版日期:2015-08-10 发布日期:2015-08-05
  • 作者简介:

    作者简介:乔伟峰(1975-),男,博士后,研究方向为土地遥感和GIS应用。E-mail: qwf@263.net

  • 基金资助:
    国家自然科学基金项目(41371172、41130748);中国博士后科学基金项目(2014M561040);江苏高校优势学科建设工程资助项目(164320H116)

Research on Extracting Building Height Rapidly Based on High-resolution Remote Sensing Images Without Parameters

QIAO Weifeng1,2,3(), LIU Yansui1,*(), XIANG Lingzhi2, WANG Yahua2   

  1. 1. Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China
    2. School of Geography Science, Nanjing Normal University, Nanjing 210023, China
    3. Jiangsu Center for Collaborative Innovation in Geographical Information Resource Development and Application, Nanjing 210023, China
  • Received:2014-10-22 Revised:2014-11-18 Online:2015-08-10 Published:2015-08-05
  • Contact: LIU Yansui E-mail:qwf@263.net;liuys@igsnrr.ac.cn
  • About author:

    *The author: SHEN Jingwei, E-mail:jingweigis@163.com

摘要:

无参数高分辨率遥感影像快速提取建筑高度,在城市建设和土地管理中有重要的现实意义。当前的研究多以已知参数的遥感影像获取,但其提取方法受限制条件多。本文提出以无参数高分辨率遥感影像,综合利用单张影像上的特征点所构成的4类特征线换算建筑高度的方法。4类特征线包括屋顶位移点与其阴影点的连线、建筑高差引起的屋顶像点位移、阴影全长和建筑遮挡后的阴影长。通过已知的少量建筑的实际高度和推导出的4类特征线换算建筑高度的计算模型,可对大量建筑进行快速、精确地高度量算。结合南京市的Google Earth影像进行了验证,结果表明,该方法采用的影像易于获得,综合量算方法大大增加了单张影像上提取建筑高度的可操作性,并解决了量算建筑高度时无相关参数的问题。该方法精度较高,可大面积、快速提取建筑高度,在生产实践中有较大的实用价值。

关键词: 无参数影像, 特征线, 换算模型, 建筑高度, Google Earth

Abstract:

There is a pressing need to extract building height rapidly based on high-resolution remote sensing images without parameters in the field of urban construction and land management. Current studies are mostly based on remote sensing images with parameters, however the images used for extraction are difficult to get, and current extraction methods have a lot of restrictions. In this paper, a new method is proposed with the use of four types of characteristic lines. The characteristic lines are comprehensively formed by the characteristic points on a single image, which is used to convert the building height based on high-resolution images without parameters. The four types of characteristic lines include: the connection line of the roof displacement point and roof shadow point, the displacement of the roof image point caused by the building height, the full-length shadow, the remaining length of the shadow excluding the part occluded by building. Four types of calculation model for acquiring building height based on the corresponding characteristic lines are deduced. According to the known heights of a small amount of constructions and using the four calculation models, the relevant parameters of remote sensing images can be derived conversely. Then, we can select the characteristic lines that are extracted most accurately on each building , and use the corresponding model to convert the building height. With the application of this method, a large number of building heights can be calculated quickly and accurately. The method is verified based on Google Earth image in Nanjing city and the results show: the images used in this approach are easy to acquire; the method of comprehensive measurement and calculation does not merely rely on the use of shadow lengths to calculate building height, so it significantly increases the practicality of extracting building height on a single image; it solves the issue that there is no related angle parameters of the sun and the satellite position when calculating the building height. Case study indicated that the precision of the proposed method is high, and it can extract the building height quickly in a large area. Generally, the method proposed in this paper has significant practical value in production applications.

Key words: image without parameters, characteristic line, calculation model, building height, Google Earth