基于无人机航测技术的林草区植被快速分类方法研究
Research on the rapid classification method of vegetation in the forest and grassland areas based on UAV aerial survey technology
贡丹均 1张志明2
作者信息
- 1. 广东省地质测绘院,广东 广州 510800
- 2. 广东工业大学自动化学院,广东 广州 510800
- 折叠
摘要
林草区植被种类多,覆盖面积大,调查难度大,如何利用新兴技术手段快速实现植被识别与分类,成为当下重要研究内容.以广东省FY湿地公园A1测绘项目区为研究对象,研究了基于无人机航测技术,通过正射影像建模、色斑差异化分区、神经网络模型训练,对林草区植被快速分类方法.研究结果表明:随机抽取的20个不同类别植被色斑与人工现场复核的植被区域分类识别正确率达到90%.基于无人机航测技术的林草区植被快速分类方法能够快速有效地对林草区植被进行分类,且植被分类效果和精度满足要求,不仅显著提升了林草资源监测的效率和精度,还为生态保护、资源管理、灾害预警及生态修复规划提供了强有力的技术支持.
Abstract
There are many types of vegetation with a large coverage area in the forest and grassland areas and it's difficult to get them investigated.How to use emerging technologies to quickly achieve vegetation identification and classification has become an important research topic at present.The A1 surveying and mapping project area of FY Wetland Park in Guangdong Province is taken as the research object in this article,based on unmanned aerial vehicle surveying technology,through orthophoto modeling,color spot differentiation zoning and neural network model training,a rapid vegetation classification method for forest and grassland areas is studied.The research results show that the classification accuracy of 20 randomly selected vegetation patches of different categories and the vegetation area reviewed manually on site is up to 90%.The rapid classification method of vegetation in forest and grassland areas based on drone aerial survey technology can quickly and effectively classify vegetation in forest and grassland areas,and the vegetation classification effect and accuracy can basically meet the requirements.This method not only has significantly improved the efficiency and accuracy of forest and grass resource monitoring,but also provided strong technical support for ecological protection,resource management,disaster warning,and ecological restoration planning.
关键词
无人机航测/正射影像/神经网络模型/植被快速分类Key words
UAV aerial survey/orthophoto/neural network model/rapid vegetation classification引用本文复制引用
基金项目
广州市科技计划(202007040125)
出版年
2024
NETL
NSTL
万方数据