基于无人机多光谱影像的松材线虫病早期动态监测

Early Dynamic Detection of Pine Wilt Disease Based on UAV Multispectral Imagery

研究利用搭载多光谱传感器的无人机航拍监测感染松材线虫病早期未表现症状的马尾松Pinus massoniana。试验方法是航拍后生成每株松树的NDVI值,然后观察NDVI值最低的松树(理论上代表最衰弱的松树)是否首先被清理。结果表明,2021年5月监测判定的衰弱松树(第一轮监测NDVI值最低为0.512 815和0.523 893;第二轮监测NDVI值最低为0.507 846和0.526 113)均不是最早清理的松树。根据这一结果调整了试验时间,改为2021年11月监测,得出的最衰弱松树(NDVI值最低为0.318 315和0.383 832)与最早清理的松树一致。这一结果为航拍监测病死松树的时机提出了新的思路,也从另一个角度证明了松材线虫病潜伏侵染的存在。

In this study, the UAV equipped with multispectral sensors was used to monitor the pines infected with pine wilt disease, but without symptoms. The test method is to generate the NDVI value of each pine after aerial photography, and then observe whether the pine with the lowest NDVI value(theoretically representing the weakest pine) is cleared first. The results showed that the pines(the lowest NDVI values in the first round of monitoring were 0.512 815, 0.523 893;the lowest NDVI values in the second round of monitoring were 0.507 846, 0.526 113)determined by monitoring in May 2021 were not the first to be cleared. According to this result, the monitoring time was adjusted to November 2021. The weakest pines(the lowest NDVI values were 0.318 315, 0.383 832) were consistent with the pines that were first cleared. This result provided a new idea for the time of monitoring dead pines by aerial photography, and also proved the existence of latent infection of PWD from another angle.