“Ion adsorption rare earth minerals cause soil pollution due to leaching mining, resulting in poor vegetation growth and low survival rate in reclaimed areas. The use of drone imaging monitoring is helpful for scientific supervision. However, the complex environment in the mining area leads to significant differences in the overall characteristics of reclaimed vegetation, making automatic recognition of drone images difficult and resulting in low recognition accuracy. In order to improve the fast and accurate automatic recognition and localization of single plants in the reclaimed vegetation of mining areas in drone images, a mining area reclaimed vegetation detection method (YOLOv8-AS) that integrates global feature YOLOv8n network is proposed. This method has made the following improvements on the basis of YOLOv8n: (1) using the downsampling module ADown for feature convolution operation, reducing the feature loss caused by standard convolution during the process of deepening the model training depth; (2) The SPPF-GFP (Spatial Pyramid Pooling Fast Global Feature Pool) module is used for feature extraction to improve the model's ability to detect reclaimed vegetation with significant differences in overall features. The results indicate that YOLOv8-AS outperforms YOLOv8n in the self built reclaimed vegetation dataset mAP@0.5 and mAP@0.5 0.95 increased by 1.6% and 2.4% respectively; The model size, parameter count, and floating-point computation of YOLOv8-AS have decreased by 11%, 10%, and 9% respectively compared to YOLOv8n. The YOLOv8-AS algorithm mAP@0.5 and mAP@0.5 0.95 achieved 91.1% and 46.8% respectively, compared to SSD, Faster R-CNN, RT-DETR, YOLOv5, YOLOv7, and YOLOv7 tiny models mAP@0.5 They increased by 14.07%, 23.32%, 1.2%, 2.3%, 3.3%, 2.9%, and 1.2% respectively. In addition, YOLOv8-AS can quickly and accurately detect reclaimed vegetation for small targets, simple and complex scenes, while also improving its ability to identify and locate individual plants of reclaimed vegetation. This method can provide accurate and effective technical support for ecological restoration in mining areas.”