The Loess Plateau is one of the most severely affected regions by soil erosion in the world, with a fragile ecological environment. Vegetation plays a key role in the region's ecological restoration and protection. Most existing studies focus on the trends of vegetation cover change, while fewer studies investigate the driving factors or only conduct quantitative analyses. This study uses the Geodetector (Geographic Detector Model) to assess the impact of natural and human factors such as temperature, precipitation, soil type, and land use on vegetation growth, revealing the characteristics and driving mechanisms of vegetation cover changes on the Loess Plateau from 1982 to 2022. It also quantitatively calculates the influence of different factors. The findings indicate that vegetation cover on the Loess Plateau has generally risen from 1982 to 2022, but there was a noticeable difference around 2000. The annual mean slope of vegetation cover from 1982 to 2000 was 0.00129, which increased to 0.01075 from 2000 to 2022. The Geodetector indicates that the factors with the greatest impact on vegetation cover in the Loess Plateau are temperature, precipitation, soil type, and land use, and the interaction between these factors has a greater effect on vegetation cover than any single factor alone. This study identifies the ideal ranges or types for vegetation growth on the Loess Plateau over four time periods, offering a scientific basis for soil restoration and soil and water conservation efforts in the area.

Underwater visual detection technology plays a pivotal role in fields such as marine exploration. With the increasing demand for underwater monitoring, the quest for efficient and reliable methods for underwater target recognition has become particularly significant. To address this requirement, this study developed an innovative underwater object detection architecture based on YOLOv8, named YOLOv8-MU, aimed at significantly enhancing detection accuracy.By integrating the LarK module proposed in UniRepLKNet to optimize the backbone network, YOLOv8-MU aims to achieve a larger receptive field without increasing the model’s depth. Further, this research introduces C2fSTR, an innovative method that combines Swin Transformer with the C2f module. Additionally, we have incorporated the SPPFCSPC_EMA module, which combines Cross-Stage Partial Fast Spatial Pyramid Pooling (SPPFCSPC) with attention mechanisms, significantly improving the detection accuracy and robustness of various biological targets. Moreover, by introducing a fusion block based on DAMO-YOLO into the neck of the model, we further enhanced the model’s capability in multi-scale feature extraction. Finally, the adoption of the MPDIoU loss function, designed around vertex distance, effectively tackles the challenges of localization accuracy and boundary clarity in underwater organism detection. Experimental results on the URPC2019 dataset demonstrate that the YOLOv8-MU model achieved an mAP@0.5 of 78.4%, marking improvements of 5.6%, 1.1%, and 4.0% over YOLOv5s, YOLOv7, and YOLOv8n respectively, indicating the leading performance (SOTA) of this method. On the other hand, further evaluation on the URPC2020 dataset confirmed the generalization capability of the YOLOv8-MU architecture, with its mAP@0.5 reaching 80.4%, surpassing various models including YOLOv5x and YOLOv8n, showcasing the wide applicability of our proposed improved model architecture.