随着自助购物已成为零售业的发展趋势,为满足消费者日益增长的购物需求,本文设计并实现了一款自助售卖系统.该系统具有可视化界面,采用Jetson Nano硬件平台,通过训练深度学习模型实现果蔬分类.实验结果表明,MobileNetV2网络分类准确率达到 98.65%,相较于传统CNN卷积神经网络模型,准确率提高了 2.13%,模型大小减小了 20.54%,具有较为显著的优化效果.
To meet the growing shopping needs of consumers,self-service shopping has become a development trend in the retail industry.This article designs and implements a self-service sales system with a visual interface,using the Jetson Nano hardware platform to train deep learning models for fruit and vegetable classification.The experimental results show that the classification accuracy of MobileNetV2 network reaches 98.65%.Compared with traditional CNN convolutional neural network models,the accuracy is improved by 2.13%,and the model size is reduced by 20.54%,showing significant optimization effects.
杨易润;刘钰;李信源;杨天乐
金陵科技学院计算机工程学院 南京 211169
计算机与自动化
智慧超市;自助售卖;嵌入式系统;深度学习
Smart Supermarket;Self Service Sales;Embedded System;Deep Learning
《福建电脑》 2024 (001)
82-88 / 7
本文得到江苏省大学生创新训练计划重点项目(No.202313573024Z)资助,以及江苏省重点实验室数据科学与智慧软件平台支持.
10.16707/j.cnki.fjpc.2024.01.016
评论