Method and system for monitoring production of milk beverages based on ai vision and medium

Abstract

The application discloses a milk beverage production monitoring method and system based on AI vision and a medium, relates to the technical field of AI vision, and comprises the following steps: analyzing abnormal nodes of milk beverage component composition, and generating an abnormal chain; collecting production data in the whole process, and extracting instantaneous and time sequence image features; performing instantaneous and cumulative abnormality identification and prediction on the dual-state features based on the abnormal chain, and outputting instantaneous identification results and cumulative prediction results; positioning and tracking the production process according to the instantaneous identification results, and feeding back instantaneous monitoring results; and blocking and tracking the production process according to the cumulative prediction results, and feeding back cumulative compensation monitoring results. The application solves the technical problem that, in the traditional milk beverage production monitoring, the abnormality in the production process cannot be identified and corrected in time due to the dependence on manual experience and discrete sensor data, and the product quality is affected, achieves the technical effect that the whole-process real-time dynamic monitoring and abnormality accurate prediction are realized through AI vision, the monitoring effect of the production process is improved, and the stability of the product quality is ensured.

Images