基于机器视觉的大闸蟹自动分级分选设备研究与开发  

Research and development of automatic grading and sorting equipment for hairy crabs based on machine vision

作　　者：吴思 诸定莲 鲁梦瑶 万以磊 高亮亮 陈龙 汤卫荣 吴文彪 WU Si;ZHU Dinglian;LU Mengyao;WAN Yilei;GAO Liangliang;CHEN Long;TANG Weirong;WU Wenbiao(Nongxin(Nanjing)Smart Agriculture Research Institute,Nanjing 211800,China;Research Center of Information Technology,Beijing Academy of Agriculture and Forestry Sciences,Beijing 100097,China)

机构地区：[1]农芯(南京)智慧农业研究院,南京211800 [2]北京市农林科学院信息技术研究中心,北京100097

出　　处：《扬州大学学报（农业与生命科学版）》2024年第4期137-146,共10页Journal of Yangzhou University：Agricultural and Life Science Edition

基　　金：江苏省重点研发计划(现代农业)项目(BE2022366)。

摘　　要：为降低大闸蟹品质分级劳动成本,提高大闸蟹分级分选准确率及效率,开发一种基于机器视觉、自动称重技术的大闸蟹自动分级分选设备,利用深度学习技术获取大闸蟹公母、残缺等表型特征对大闸蟹进行品质分级。设备包含活蟹装载仓模块、自动输送模块、图像采集模块、动态称重模块、自动分拣模块,具备自动分料、传输、品质分级功能。经试验测试,大闸蟹公母识别准确率为98.45%,残缺识别准确率为93.28%,重量识别准确率为94.68%,识别准确率和分拣效率可满足实际应用要求,稳定性良好。结果表明该设备能自动、准确地完成大闸蟹的分级分选,可为农业智能化提供技术支撑。In order to reduce the labor cost of hairy crab quality classification and improve the accuracy and efficiency of hairy crab classification and sorting,an automatic classification and sorting equipment based on machine vision and automatic weighing technology was developed.Deep learning technology was used to obtain the phenotypic characteristics of hairy crab,such as male and female,and incomplete.The equipment included live crab loading bin module,automatic conveying module,image acquisition module,dynamic weighing module,automatic sorting module,with automatic material separation,transmission,quality classification functions.The accuracy of male and female identification was 98.45%,the accuracy of partial identification was 93.28%,and the accuracy of weight identification was 94.68%.The accuracy and sorting efficiency of hairy crab can meet the requirements of practical application,and the stability was good.The results show that the equipment can automatically and accurately complete the classification and sorting of hairy crabs,and can provide technical support for agricultural intelligence.

关 键 词：大闸蟹 机器视觉 自动称重 分级分选设备 

分 类 号：S226.5[农业科学—农业机械化工程]