湖北农业科学 ›› 2018, Vol. 57 ›› Issue (20): 132-137.doi: 10.14088/j.cnki.issn0439-8114.2018.20.031

• 信息工程 • 上一篇    下一篇

基于高分二号遥感影像的东北地区村庄建设用地提取

梁书维a, 王建国a, 温馨燃a, 陈圣波b   

  1. 吉林大学,a.地球科学学院; b.地球探测科学与技术学院,长春 130026
  • 收稿日期:2018-07-30 发布日期:2018-10-25
  • 通讯作者: 王建国(1976-),男,副教授,主要从事土地资源遥感技术应用研究,(电子信箱)wang_jg@jlu.edu.cn。
  • 作者简介:梁书维(1993-),男,吉林吉林人,硕士研究生,主要从事土地资源遥感技术应用研究,(电话)18844194035(电子信箱)liangswty@163.com
  • 基金资助:
    国家自然科学基金项目(41201158)

Methods to Extract Rural Construction Land in the Northeast ChinaBased on GF-2 Remote Sensing Data

LIANG Shu-weia, WANG Jian-guoa, WEN Xin-rana, CHEN Sheng-bob   

  1. a.College of Earth Sciences; b.College of GeoExploration Science and Technology, Jilin University,Changchun 130026,China
  • Received:2018-07-30 Published:2018-10-25

摘要: 准确高效地识别和测算村庄建设用地的类型及数量,可以为村镇规划、村庄整治等提供依据。为研究应用高分二号遥感影像提取村庄建设用地的技术方法,选取吉林省长春市和松原市的两个村庄作为典型研究区,针对遥感影像的不同时相特征,分别采用直接提取法和间接提取法进行村庄建设用地提取试验。结果表明,高分二号遥感影像可以应用于村庄建设用地的精确识别。直接提取法以支持向量机的监督分类法效果最优,可作为精确提取地类的方法;基于植被指数并辅以归一化蓝色屋顶指数的间接提取法,适合村庄建设用地的快速估算。

关键词: 村庄建设用地, 遥感影像, 监督分类, 植被指数, 归一化蓝屋顶指数

Abstract: The rural planning and land consolidation rely on the accurate information of rural construction land. To accurately measure the existing rural construct land area, two villages which in Changchun area and Songyuan area of Jilin province with different phases were selected as typical study areas. GF-2 remote sensing images were used as data source to study on the method of extracting rural construction land by direct extraction and indirect extraction respectively. The experimental results showed that GF-2 high-resolution remote sensing images can be used to identify the construction land inside of village accurately. SVM supervised classification, the best way of the direct extraction methods, is used in accurate extraction. The indirect extraction based on NDVI combined with NDBBI is a good fit for rapid estimation of village construction land area.

Key words: rural construction land, remote sensing image, supervised classification, NDVI, NDBBI

中图分类号: