Abstract: 【Objective】A hue-saturation-value(HSV)threshold method was used for the recognition of red attack pine trees from unmanned aerial vehicle(UAV)imagery. To improve the investigation efficiency of red-attack damage and provide basic data for sanitation logging. 【Method】The extraction rules of the HSV threshold method were built according to the difference between red attack pine trees and other elements in the “H-V space”. The recognition results of red attack pine trees between the RGB threshold method and HSV threshold method were compared under different scenarios. Next, the evaluation accuracy of both methods was calculated. 【Result】① Samples of red attack pine trees plotted in the “H-V space” point cloud showed significant clustering and a banding distribution in the “G-R space” point cloud. ② A strong correlation among the components in the RGB color space was a main factor for the large omission errors of red attack pine tree extraction using the RGB threshold method. The overall accuracy of the HSV threshold method, based on decoupling of the hue and value from the HSV color space and its convenient threshold segmentation, outperformed that of the RGB threshold method for red attack pine tree recognition based on UAV imagery. ③ The HSV threshold method performed well for monitoring pines affected by pine wilt disease in a late stage, which achieved 60%-65% overall accuracy when the distribution of pine trees before infection from high-resolution imagery was used as a mask. 【Conclusion】 The HSV threshold method shows some advantages for monitoring red attack pine trees based on UAV imagery compared to manual interpretation. This method provides theoretical and methodological support for the recognition of red attack pine trees from UAV imagery.