Microscopic Insect Pest Detection in Tea Plantations: Improved YOLOv8 Model Based on Deep Learning
茶园微观害虫检测：基于深度学习的改进YOLOv8模型
- 关键词：
- 来源：
- Agriculture-Basel 期刊
- 类型：
- 学术文献
- 语种：
- 英语
- 原文发布日期：
- 2024-10-02
- 摘要：
- Pest infestations in tea gardens are one of thecommon issues encountered during tea cultivation. This study introduces animproved YOLOv8 network model for the detection of tea pests to facilitate therapid and accurate identification of early-stage micro-pests, addressingchallenges such as small datasets and the difficulty of extracting phenotypicfeatures of target pests in tea pest detection. Based on the original YOLOv8network framework, this study adopts the SIoU optimized loss function toenhance the model’s learning ability for pest samples. AKConv is introduced toreplace certain network structures, enhancing feature extraction capabilitiesand reducing the number of model parameters. Vision Transformer with Bi-LevelRouting Attention is embedded to provide the model with a more flexiblecomputation allocation and improve its ability to capture target positioninformation. Experimental results show that the improved YOLOv8 networkachieves a detection accuracy of 98.16% for tea pest detection, which is a2.62% improvement over the original YOLOv8 network. Compared with the YOLOv10,YOLOv9, YOLOv7, Faster RCNN, and SSD models, the improved YOLOv8 network hasincreased the mAP value by 3.12%, 4.34%, 5.44%, 16.54%, and 11.29%,respectively, enabling fast and accurate identification of early-stage micropests in tea gardens. This study proposes an improved YOLOv8 network modelbased on deep learning for the detection of micro-pests in tea, providing aviable research method and significant reference for addressing theidentification of micro-pests in tea. It offers an effective pathway for thehigh-quality development of Yunnan’s ecological tea industry and ensures thehealthy growth of the tea industry.
- 所属专题：
- 60
