Classification of pine wilt disease at different infection stages by diagnostic hyperspectral bands

Pine wilt disease (PWD) is a very destructive forest that causes the mortality of pine. The infected trees usually die within three months, and spreads fast with long-horned beetle as medium if are not removed from in time. Therefore, detecting at different infection stage, especially early infection, crucial for preventing PWD spread. This study aims to exhibit spectral differences pine needles between healthy pines stages reveal diagnostic bands classifying stage trees. We collected needle samples healthy, early-, middle-, late-stage Japanese (Pinus densiflora) Korean koraiensis) northern China explore biochemical properties these four classes, selected sensitive combining competitive adaptive reweighted sampling (CARS) successive projections algorithm (SPA). were used classification by linear discriminant analysis (LDA) algorithm. results show Chlorophyll a, chlorophyll b, carotenoids, moisture content decreases aggravation infection. green (510–530 nm), red-edge (680–760 short-wave infrared (1400–1420 nm 1925–1965 nm) bands, overall accuracy 77 % 78 respectively when using demonstrate physiological parameters including can be PWD, feasible stress symptoms