Object detection and recognition system based on computer vision analysis

Artificial intelligence based on deep learning enables the machine to have the ability of understanding and cognition, but the application of artificial intelligence technology in supermarket shopping scene is limited. In the post-epidemic era, the contactless self-checkout of unmanned supermarket is more in line with the development needs of modern society. We build Pytorch environment, first to collect pictures of a large number of commodities and labeling information, and training model is obtained by YOLO neural network algorithm, finally through a call to model to realize the recognition of goods. Neural network algorithm is used to improve the recognition rate of goods step by step and achieve the detection and recognition of objects. We have tested our model on the real supermarket commodity data set and the public data set ImageNet, and the results show that our model can achieve a certain practical effect.