Recognition of Material Status in Workshop and Logistics Automatic Scheduling Technology Based on RestNet

Material transfer and material status monitoring have always been an important part of operation and maintenance in the workshop. Too much reliance on manual operation will greatly reduce work efficiency and cause frequent errors. This article is based on Residual Neural Network (ResNet) transfer learning (TL) for model training. The status of material points in the workshop, namely empty frame, no frame, and full-frame, has been well recognized by using a small amount of surveillance video stream data for image analysis, which realizes the reverse optimization of the model. The accuracy of material point status recognition is as high as 99.7%. Based on the material point status recognition technology, the manufacturing operation management system interface can be called through the HTTP protocol to issue tasks, and the intelligent logistics system can be combined to realize the automatic circulation of materials in the workshop and improve production efficiency.