Cicada (Magicicada) Tree Damage Detection Based on UAV Spectral and 3D Data

The periodical cicadas appearin regions of the United States in intervals of 13 or 17 years. During theseintervals, deciduous trees are often impacted by the small cuts and eggs theylay in the outer branches which soon die off. Because this is such an infrequentoccurrence and it is so difficult to assess the damage across large forested areas, there islittle information about the extent of this impact. The use of remote sensingtechniques has been proven to be useful in forest health management to monitorlarge areas. In addition, the use of Unmanned Aerial Vehicles (UAVs) has becomea valuable tool for analysis. In this study, we evaluated the impact of the periodical cicada occurrence on amixed hardwood forest using UAV imagery. The goal was to evaluate the potential of this technology as a tool forforest health monitoring. We classified the cicada impact using two MaximumLikelihood classifications, one using only the high resolution spectral derivedfrom leaf-on imagery (MLC 1), and in the second we included the Canopy HeightModel (CHM)—derived from leaf-on Digital Surface Model (DSM) and leaf-offDigital Terrain Model (DTM)—information in the classification process (MLC 2).We evaluated the damage percentage in relation to the total forest area in 15circular plots and observed a range from 1.03% -22.23% for MLC 1, and 0.02% - 10.99% for MLC 2. The accuracy of the classification was 0.35 and 0.86,for MLC 1 and MLC 2, based on the kappa index. The results allow us tohighlight the importance of combining spectral and 3D information to evaluateforest health features. We believe this approach can be applied in many forestmonitoring objectives in order to detect disease or pest impacts.