毛竹立竹度无人机遥感识别有效高度的研究  

Study of Effective Height of the Living Bamboo Density Identification based on Unmanned Aerial Vehicle(UAV)Remote Sensing

作　　者：张仓皓 杨樟平 谢巧雅 邓洋波[1,2] 余坤勇 刘健[1,2] Zhang Canghao;Yang Zhangping;Xie Qiaoya;Deng Yangbo;Yu Kunyong;Liu Jian(College of Forestry,Fujian Agriculture and Forestry University,Fuzhou 350002,China;Fujian Province Key Laboratory of 3S Technology and Optimal Utilization of Resources,Fuzhou 350002,China)

机构地区：[1]福建农林大学林学院,福建福州350002 [2]3S技术与资源优化利用福建省高校重点实验室,福建福州350002

出　　处：《遥感技术与应用》2020年第6期1436-1446,共11页Remote Sensing Technology and Application

基　　金：“十三五”国家重点研发计划“材用竹林高效培育与监测技术”(2018YFD0600103);福建农林大学科技创新计划专项;福建省教育厅省中青年教师教育科研项目(JT180132)。

摘　　要：确定最优的飞行高度,是无人机遥感快速、有效识别毛竹立竹度的重要基础。以2100株/hm^2密度的毛竹林为研究对象,利用无人机获取4个不同飞行高度(50、100、150和200 m)的毛竹林影像数据,通过分析各影像的波段光谱特征信息和纹理特征信息,选取识别毛竹立竹度的有效特征波段组合(蓝光波段、红光波段、近红外波段、纹理特征均值、归一化植被指数(NDVI)),采用面向对象多尺度分割,构建毛竹立竹度的识别单元。在此基础上,依据识别单元,通过K邻近值算法(K-Nearest Neighbor)实现特征波段组合影像的分类和毛竹立竹度信息的识别,比较不同飞行高度下毛竹立竹度的遥感识别效果,确定无人机遥感识别毛竹立竹度的有效高度。结果表明:4个不同高度水平分类总体精度分别是74.24%、85.32%、91.12%、83.84%,航高为150m时,2100株/hm^2密度的毛竹林的立竹度识别精度最佳。The optimal flying height will play a significant basis for the rapid and accurate identification of the Moso bamboo density based on UAV(Unmanned Aerial Vehicle)remote sensing.The Moso bamboo image data of 2100/hm^2 were obtained by the UAV at four different flying heights(50 m,100 m,150 m,200 m),the blue band,red band,the near-infrared band,texture feature mean,and Normalized Difference Vegetation Index(NDVI)were selected as feature band combination for Moso bamboo thmatic information based the spectral feature and texture feature information.Using image segmentation to construct the identification unit of the Moso bamboo density.According to the unit,the Moso bamboo density were measured through the"k"-Nearest Neighbor(KNN).Comparing to the remote sensing identification impacts of various flying heights,the effective height of the UAV remote sensing recognition of the Moso bamboo density was determined.The results present that the overall classification accuracy of the four various flying height levels was 74.24%,85.32%,91.12%,83.84% respectively.When the flight height is 150 m,the classification accuracy of Moso bamboo of 2100/hm^2 is the highest,which showed that the height can be used as the effective height for UAV remote sensing to identify the Moso bamboo density.

关 键 词：无人机 有效高度 毛竹 立竹度 遥感识别 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]