欢迎访问林业资源管理

林业资源管理 ›› 2020, Vol. 0 ›› Issue (3): 118-121.doi: 10.13466/j.cnki.lyzygl.2020.03.022

• 技术应用 • 上一篇    下一篇

基于林业遥感的树种分类应用分析与展望

马鸿伟(), 刘海(), 姚顺彬, 周蔚   

  1. 国家林业和草原局华东调查规划设计院,杭州 340019
  • 收稿日期:2020-03-08 修回日期:2020-04-16 出版日期:2020-06-28 发布日期:2020-07-30
  • 通讯作者: 刘海 E-mail:hdymhw@126.com;liuhai.hdy@qq.com
  • 作者简介:马鸿伟(1968-),男,山东禹城人,工程师,农业推广硕士,主要从事森林一体化监测等工作。Email: hdymhw@126.com

Analysis and Prospect on the Application of Tree Species Classification Based on Forestry Remote Sensing

MA Hongwei(), LIU Hai(), YAO Shunbin, ZHOU Wei   

  1. East China Inventory and Planning Institute,National Forestry and Grassland Administration,Hangzhou,340019
  • Received:2020-03-08 Revised:2020-04-16 Online:2020-06-28 Published:2020-07-30
  • Contact: Hai LIU E-mail:hdymhw@126.com;liuhai.hdy@qq.com

摘要:

树种分类是林业遥感中重要的应用领域,在林业可持续管理、生物多样性研究和生态环境保护领域都有广发的应用场景。结合2000年后该领域的研究成果以及近些年林业生产过程中的应用实践,对多源数据在树种分类的应用加以汇总。从工作流程及相应算法两个角度,基于图像分类与数理统计层面对该问题进行了解析与比较。面对林业遥感在树种分类应用过程中遇到的问题与挑战,提出了基于语义分割及实例分割的不同工作思路,对未来多源遥感数据融合获取及硬件处理设备的发展前景进行了展望。

关键词: 林业遥感, 树种分类, 激光雷达数据, 高光谱数据, 理论应用

Abstract:

Tree species classification is an important application area in forestry remote sensing.It has widely used application scenarios in the areas of sustainable forestry management,biodiversity research,and ecological environmental protection.Based on the research results in this field after 2000 and the application practice in the forestry production process in recent years,the application of multi-source data in tree classification is summarized.From the perspectives of workflow and corresponding algorithms,the issue is analyzed and compared based on image classification and mathematical statistics.Facing the problems and challenges encountered in the application of forestry remote sensing in the classification of tree species,different working ideas based on semantic segmentation and instance segmentation were proposed,and the future development of multi-source remote sensing data fusion acquisition and hardware processing equipment was prospected.

Key words: forestry remote sensing, tree species classification, lidar data, hyperspectral data, theory application

中图分类号: