摘要
针对糖果生产企业采用人工方法分选缺陷硬糖存在的漏检、成本高、效率低等问题,研究设计一种基于卷积神经网络的缺陷硬糖智能分选系统。通过工业相机采集硬糖图像,利用YOLOv5卷积神经网络模型进行缺陷硬糖的检测识别,使用喷阀剔除缺陷硬糖。测试结果表明,实时检测准确率高达98%,具有高度自动化和智能化水平,在食品生产工业中具有一定的应用和推广价值。
In order to solve the problems of missed inspection,high cost and low efficiency when sorting defective hard candies by workers in candies production enterprises,an intelligent sorting system of defective hard candies was developed based on convolutional neural network.This system uses industrial cameras to collect images of hard candies,uses YOLOv5 network for detection and classification of defective hard candies,and uses a spray valve to reject defective candies.The testing results show that the real-time detection accuracy rate reaches 98%.The proposed sorting system with a high level of automation and intelligence has certain application and promotion value in the food production industry.
作者
朱婷婷
程磊
王锦亚
居荣华
倪超
ZHU Tingting;CHENG Lei;WANG Jinya;JU RongHua;NI Chao(College of Mechanical and Electronic Engineering,Nanjing Forestry University,Nanjing 210037,China)
出处
《包装与食品机械》
CAS
北大核心
2022年第1期34-39,共6页
Packaging and Food Machinery
基金
国家重点研发计划项目(2017YFF0207200)。
