Abstract: In studies on pollination ecology at the community level, the use of traditional survey methods for quantitative analysis of large-scale sample plots has certain technical limitations. Here, we explored the feasibility of applying unmanned aerial vehicle (UAV) image data to study pollination biology at the herbaceous plant community level in a subalpine wetland of Shennongjia Dajiuhu. The support vector machine (SVM) classification method was used to calculate flowering coverage of different-colored flowers based on UAV visible light images of four sites (1600~3000 m²) across different seasons, which were then combined using ContextCapture software. Combined with field surveys of flower-visiting insect activity within 16 quadrats (2 m×2 m each), the results showed that:(1) Flowering coverage of different-colored flowers in the UAV images was significantly correlated with number of pollinators, showing an exponential relationship. (2) The observed value of flowering coverage showed a decreasing trend with the increase in UAV flight altitude. (3) In the different sample plots, the number of pollinators within a unit flowering area showed no significant differences. We also explored the feasibility of calculating flowering coverage through UAV images to monitor flowering season dynamics in the study area and estimate the number of pollinators.