地理学报

地理学报 ›› 2005, Vol. 60 ›› Issue (1): 158-164.doi: 10.11821/xb200501018

• 区域环境 • 上一篇    下一篇

基于遥感影像的城市人口密度模型

吕安民1,2, 李成名3, 林宗坚3, 王兴奎1   

  1. 1. 清华大学水沙科学教育部重点实验室,北京 100084;
    2. 郑州市国土资源局,郑州 450006;
    3. 中国测绘科学研究院,北京100039
  • 收稿日期:2004-01-21 修回日期:2004-05-29 出版日期:2005-01-25 发布日期:2005-01-25
  • 作者简介:吕安民, 男, 河南人, 博士, 教授, 1986年毕业于武汉测绘科技大学, 2002年获武汉大学博士学位, 现为清华大学博士后, 从事GIS应用研究,发表论文20余篇。E-mail: anminlu@sina.com
  • 基金资助:

    国家计委2000年高科技应用项目 (中国人口地理信息系统建设技术支撑体系)

Modeling Urban Population Density with Remote Sensing Imagery

LV Anmin1,2, LI Chengming3, LIN Zongjian3, WANG Xingkui1   

  1. 1. School of Civil Engineering, Tsinghua University, Beijing 100084, China;
    2. Zhengzhou Land Resource Bureau, Zhengzhou 450006, China;
    3. Chinese Academy of Surveying and Mapping, Beijing 100039, China
  • Received:2004-01-21 Revised:2004-05-29 Online:2005-01-25 Published:2005-01-25
  • Supported by:

    SDPC Hi-tech Application Project in 2000 (Technology Support System of China PGIS Construction)

PDF (PC)

赞

被引次数

116

摘要：

分析了常用的城市人口遥感估算方法,指出了其特点和不足,在土地利用密度法的基础上作了改进。土地利用密度法的关键是抽样区的代表性,抽样区选择的不当,就会给人口计算带来较大的误差。实际上,抽样区的选择有较大的难度,而且实地区域和抽样区相比总会有一定的差异,所以人口估计误差是不可避免的。作者试图避开抽样区的选择而估计出各种居住类型的人口密度,提出的方法不需要实地抽查人口密度,减少了估算的工作量,建立的数学模型没有抽样的随机性误差影响,结果从整体上达到了最优。

关键词: 遥感影像, 人口密度, 土地利用密度法, GIS, 城市, 郑州

Abstract:

The population in the study region is the sum of the area of each habitation type multiplying its respective sampled population density. The population density of each habitation type is estimated. In some experimental region, there is mathematical relation among the area, the population density and its total population of each habitation type. The population density of each habitation type can be estimated via the mathematical relation. A new land use density method is proposed based on least square principle. The main idea is: First, habitation type is defined in the study region; and the boundaries of all the habitation types are lined out based on remote sensing imagery and the area of each habitation type are calculated; then mathematical models according to the population data of every sub-region are established; the best population density estimation of each habitation type is calculated with the least square principle when the number of sub-regions is more than the number of habitation types. The population estimation of any region can be calculated since the population density of each habitation type is known as well. The method need not sample the population density of each habitation type. The estimation workload of population density of each habitation type is low. The mathematical models are not influenced by random error of samples.

Key words: remote sensing imagery, population density, land use density method, GIS, city