测绘通报 ›› 2019, Vol. 0 ›› Issue (11): 79-84.doi: 10.13474/j.cnki.11-2246.2019.0356

• 学术研究 • 上一篇    下一篇

机载多光谱LiDAR的随机森林地物分类

曹爽1, 潘锁艳2, 管海燕1   

  1. 1. 南京信息工程大学遥感与测绘工程学院, 江苏 南京 210044;
    2. 南京信息工程大学地理科学学院, 江苏 南京 210044
  • 收稿日期:2019-07-19 修回日期:2019-09-14 发布日期:2019-12-02
  • 作者简介:曹爽(1977-),女,博士,讲师,主要研究方向为摄影测量数据处理。E-mail:sh_cao2004@aliyun.com
  • 基金资助:
    国家自然科学基金(41671454)

Random forest-based land-use classification using multispectral LiDAR data

CAO Shuang1, PAN Suoyan2, GUAN Haiyan1   

  1. 1. School of Remote Sensing & Geomatics Engineering, NUIST, Nanjing 210044, China;
    2. School of Geographic Sciences, NUIST, Nanjing 210044, China
  • Received:2019-07-19 Revised:2019-09-14 Published:2019-12-02

摘要: 机载多光谱LiDAR技术利用激光进行探测和测距,不仅可以快速获取地面物体的三维坐标,还可以获得多个波段的地物光谱信息,可广泛用于地形测绘、土地覆盖分类、环境建模、森林资源调查等。本文提出了多光谱LiDAR的随机森林地物分类方法。该方法通过对LiDAR强度数据和高程数据提取分类特征,完成多光谱LiDAR的随机森林地物分类;并分析随机森林的特征贡献度特性,采用后向特征选择方法实现分类特征选择。通过对加拿大Optech Titan多光谱LiDAR数据的试验表明:随机森林方法可以获得较好的地物分类精度,而且可以适当地去除部分冗余和相关的特征,从而有效提高分类精度。

关键词: 多光谱LiDAR, 随机森林, 地物分类, 变量重要性, 特征选择

Abstract: Airborne LiDAR systems can quickly obtain three-dimensional coordinates of ground objects, which has been widely used in topographic mapping, engineering construction, environmental monitoring, and land-cover and land-use classification, and so on. This paper, by means of random forest algorithm, performs land-cover classification using airborne multispectral LiDAR data. The proposed method extracts features from elevation and multispectral images combined by three individual intensity images, performs a backward feature selection according to the variables importance calculated by RF, and finally applies RF to the multispectral images. All experiments are conducted on the Optech Titan multispectral LiDAR data.The experimental results show that RF can achieve a good performance in land-cover classification, and the proposed RF-based backward feature selection method contributes to the improvement of classification by iteratively removing redundancy and related features.

Key words: multispectral LiDAR, random forest, land-cover classification, variable importance, backward feature selection

中图分类号: