AUTHOR=Wu Wei , Liu Shengping , Zhong Xiaochun , Liu Xiaohui , Wang Dawei , Lin Kejian TITLE=Brandt’s vole hole detection and counting method based on deep learning and unmanned aircraft system JOURNAL=Frontiers in Plant Science VOLUME=Volume 15 - 2024 YEAR=2024 URL=https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2024.1290845 DOI=10.3389/fpls.2024.1290845 ISSN=1664-462X ABSTRACT=Rodents are essential to the balance of the grassland ecosystem, but their population outbreak can cause major economic and ecological damage. Rodent monitoring is the first stepcrucial for its scientific management, but traditional methods are highly heavily dependent on manual operation labor and are difficult to be carriedy out on a large scale. To address this, researchers have developed method to monitor rodent densities using unmanned aircraft system with deep learning. However, there is still a need for a more efficient and field-applicable method, as well as the verification of hole numbers between artificial cognition and DL translation. In this study, we used UAS to collect high-resolution RGB images of steppes in Inner Mongolia, China in the spring, and using used variousa multi-object detection algorithm algorithms to identify the holes of Brandt's vole (Lasiopodomys brandtii). The model was optimized by adjusting the regression strategy evaluation indexes. Optimizing the model by adjusting evaluation metrics, specifically, replacing classification strategy metrics such as precision, recall, and F1 -score with regression strategy-related metrics FPPI, MR, and MAPE to determine the optimal threshold parameters for IOU and confidence. Then, we mapped the distribution of vole holes in the study area, using position data derived from the optimized model. Results showed that the best resolution of UAS acquisition was 0.4 cm pixel -1 , and the improved labeling method improved the detection accuracy of the model. The FCOS model had the highest comprehensive evaluation, and an R 2 of 0.9106, RMSE of 5.5909, and MAPE of 8.27%.