Design of Pest Infestation Early Warning System Based on Deepsort Algorithm

This paper uses DeepSORT as a framework to design an efficient and convenient early warning system for insect infestation [1]. The system uses the multi-object detection algorithm DeepSORT to track, identify, and count the pests in the video, and uses STM32 single-chip microcomputer with GY30, DHT11, ATGM332D and other modules to collect and upload environmental data, which can comprehensively generate pest alarm levels according to the number of pests and environmental data [2]. This design has good identification ability for different types, different speeds, and different numbers of pests, and can accurately generate alarm levels based on environmental data. Therefore, this design has a good application ability of insect plague early warning, and provides reference and reference for the design and optimization of insect plague early warning system.