无人机多光谱遥感技术监测松材线虫病疫木研究

Application of UAV multis-pectral remote sensing technology inmonitoring dead trees of pine wilt disease (PWD)

为探索无人机多光谱遥感技术监测大面积松材线虫病疫木的可能性和精度,该研究利用多旋翼无人机加载6通道多光谱相机,获取了典型松材线虫病疫区遥感图像,进而通过波段配准、拼接、辐射定标处理获得试验区的正射全景影像,并以此对试验区病死树进行识别和定位分析。结果显示:本次航拍试验区共有松材线虫病死树83株。通过对识别出的病死树分布位置和对应坐标的实地验证表明,本次航拍监测的准确率为95.2%,平均坐标精度为2.6 m,漏检率为9.8%,能够满足大面积松材线虫病监测工作应用需求。

To explore the application possibility and precision of UAV multi-spectral remote sensing technology in monitoring large-scale pine wood suffered in pine wilt disease(PWD),we used a multi-rotor UAV with a 6-channel multi-spectral camera to obtain the remote sensing image of the typical symptom areas of PWD,and the orthophoto panoramic image of this area was obtained through band registration,splicing and radiometric calibration.The diceased tree in the experimental area was identified and located by means of system automatic identification and manual correction.The results showed that 83 diceased trees suffered by PWD were identified in the aerial photography test area,and their distribution map and coordinate points also succeeded in obtaining.The field verification showed that the monitoring accuracy was 95.2%,the coordinate point accuracy was 2.6 m while the missed detection rate was 9.8%,which basically meets the needs of large-scale monitoring PWD.