基于深度学习的智能胡蜂防治系统的设计  

Design of an Intelligent Wasp Control System Based on Deep Learning

作　　者：马中宇 周文静 汤舒涵 Ma Zhongyu;Zhou Wenjing;Tang Shuhan(Xinjiang Institute of Science and Technology,Korla 841000,China)

机构地区：[1]新疆科技学院,新疆库尔勒841000

出　　处：《黑龙江科学》2025年第6期68-71,共4页Heilongjiang Science

基　　金：2024年度自治区级大学生创新训练项目“基于深度学习的智能胡蜂防治系统”(S202413561038)。

摘　　要：设计了一种基于深度学习的智能胡蜂防治系统,旨在解决胡蜂对蜜蜂养殖业的严重威胁。采用YOLOv3-tiny目标检测算法,实现对胡蜂的实时、高效识别,结合自动化防护机制,通过舵机控制巢门的闭合状态与风机捕捉装置,将胡蜂隔离并捕获。设计了独特的单向滤门结构,确保胡蜂无法逃脱,以保护蜜蜂的安全。此外,采用模块化设计和太阳能供电,确保在不同环境下的稳定运行,通过远程监控功能,实现对养蜂过程的智能化管理。该系统为蜜蜂养殖业提供了一种高效、智能且环保的防治方案,具有广泛的应用前景。The study designs a deep learning-based intelligent wasp control system,which has been developed to address the significant threat that wasps pose to the beekeeping industry.The system employs the YOLOv3-tiny target detection algorithm to facilitate real-time and efficient identification of wasps,and integrates an automated protection mechanism to isolate and capture wasps by controlling the closing state of the hive door and the fan capture device through a servo.The system is designed with a distinctive one-way filter door configuration to guarantee that the wasps are unable to escape while safeguarding the bees.Furthermore,the system employs a modular design and a solar power supply to guarantee operational stability across diverse environmental conditions.It also facilitates intelligent beekeeping process management through a remote monitoring function.The system offers an efficient,intelligent,and environmentally friendly prevention and control solution for the beekeeping industry with a multitude of potential application.

关 键 词：深度学习 胡蜂防治 树莓派 智能农业 

分 类 号：S895[农业科学—特种经济动物饲养] TP18[农业科学—畜牧兽医]