摘要
在突发性公共安全事件中,仓库内医疗物资的高效精准识别是保证分拣作业准确性和即时性的前提。利用爬虫技术获取医疗物资图片数据,建立基于卷积神经网络的医疗物资图片识别模型。通过三种不同算法在单通道和三通道中1000次模型训练的结果对比发现,基于卷积神经网络的三通道医疗物资智能识别模型的测试准确率可达89%,进一步拓展了卷积神经网络在物流仓储拣选活动中的应用范围。
In this paper,crawler technology is used to obtain the image data of medical supplies,and a medical supplies image recognition model based on convolution neural network is established.Through the comparison of three different algorithms in the single channel and the three channel model training results of 1000 times,it is found that the test accuracy of the three channel medical material intelligent identification model based on convolution neural network can reach 89%,which further expands the application scope of convolution neural network in logistics warehouse picking activities.
出处
《工业控制计算机》
2021年第4期89-91,共3页
Industrial Control Computer
基金
国家自然科学基金项目“仓储拣选系统拥堵的影响因素与联合控制策略研究”(71501015)
北京社会科学基金重点项目“基于大数据技术提升首都物流服务品质的策略研究”(18GLA009)。
关键词
仓储拣选
卷积神经网络
医疗物资识别
warehouse selection
convolutional neural network
medical material identification
intelligent handling
