Research on Digital Instrument Detection Model Based on Machine Vision

In reference to the characteristics of the MobileNet network, a lightweight design method for the YOLOv4 model is proposed. A machine vision object detection model targeting 'dumb sensors' in digital workshops is constructed, and compared with other commonly used object detection models. The performance of the model on the dataset is verified, demonstrating that the model not only ensures detection accuracy but also reduces training and prediction time.