摘要
零售智能化改革不仅有效地提升了销售效率,而且一定程度降低了人工成本,为零售业带来了便利和效益。文章基于深度学习算法,利用自建的商品数据集进行模型训练,在嵌入式系统上运行并进行了研究。采用树莓派4B开发板作为硬件核心,设计了一套集成称重、识别和交互功能的智能零售系统。这套系统充分利用了深度学习算法的优势,通过对商品数据的分析和处理,能够快速准确地识别商品。一系列的测试结果表明,该系统实现了预期的功能,为研究的正确性提供了有力的支持。
Retail intelligence reform has not only effectively improved sales efficiency,but also reduced labor costs to a certain extent,bringing convenience and benefits to the retail industry.Based on Deep Learning algorithm,this paper uses a selfbuilt commodity dataset for model training,which is ran on embedded systems and researched.A Raspberry Pi 4B development board is used as the hardware core,and an intelligent retail system integrating weighing,recognition and interaction functions is designed.This system fully utilizes the advantages of Deep Learning algorithm,and can quickly and accurately identify goods through the analysis and processing of commodity data.After a series of tests,the results show that the system has achieved the expected functions,providing strong support for the correctness of the research.
作者
朱镕佳
杨宇轩
李振东
陈硕
唐朝阳
唐晓雨
ZHU Rongjia;YANG Yuxuan;LI Zhendong;CHEN Shuo;TANG Chaoyang;TANG Xiaoyu(School of Communication and Artificial Intelligence,School of Integrated Circuits,Nanjing Institute of Technology,Nanjing211167,China)
出处
《现代信息科技》
2024年第19期189-192,198,共5页
Modern Information Technology
基金
2023年江苏省大学生创新培训项目(202311276078Y)
2021年度江苏省高校哲学社会科学研究一般项目(2021SJA0415)。
关键词
树莓派
深度学习
物品识别
人机交互
Raspberry Pi
Deep Learning
item recognition
human-computer interaction
