无人机影像红边波段K-L变换火烧迹地提取方法  

UAV Image Red-edge Band K-L Transform to Extract the Fire Burned Areas

作　　者：曹国军 赵爽 曾玉婷 陈兵 叶江霞[1] Cao Guojun;Zhao Shuang;Zeng Yuting;Chen Bing;Ye Jiangxia(Southwest Forestry University,Kunming 650224,China)

机构地区：[1]西南林业大学,云南昆明650224

出　　处：《山西林业科技》2022年第1期32-34,共3页Shanxi Forestry Science and Technology

基　　金：国家自然科学基金项目(3176-0212)

摘　　要：基于2020年5月9日云南省安宁市青云街道森林火灾多光谱无人机多波段高分辨率影像,分别采用最大似然法监督分类和第3主成分分析阈值法进行火烧迹地提取,并进行精度验证分析。结果表明,融合红边波段后的图像用2种方法提取火烧迹地的效果均优于普通Red图像,而第3主成分分析法的效果好于监督分类,融合红边波段且采用第3主成分分析的图像更适合于火烧迹地的提取。Based on the multi-spectral UAV multi-band high-resolution image of the forest fire in Qingyun Street of Anning City of Yunnan Province on May 9,2020,the maximum likelihood method supervised classification and the third principal component analysis threshold method were carried out to extract the fire burned areas.And the accuracy verification analysis was carried out.Results showed that after fusing the red edge band image,the effect of extracting the fire burned areas by the 2 methods was better than that of the ordinary red image.And the effect of the third principal component analysis was much better than the supervised classification.The image fused with the red edge band and using the third principal component analysis was more suitable for the extraction of fire burned areas.

关 键 词：红边波段 监督分类 K-L变换 火烧迹地 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] S762[自动化与计算机技术—计算机科学与技术]