Insect pest evaluation and early warning method based on multi-terminal acquisition and AI image recognition algorithm

The invention relates to the technical field of pest control, in particular to a pest evaluation and early warning method based on multi-terminal acquisition and an AI image recognition algorithm, and the method comprises the steps: S1, placing a plurality of cameras at different crop positions of an agricultural factory, carrying out the real-time monitoring of crops, uploading a monitored video to an image processing model, and carrying out the recognition of the image processing model; or the mobile phone APP is adopted to take a picture of the crops, and the taken picture is uploaded to the interior of the image processing model to be processed; and S2, processing data uploaded by the camera through an edge computing technology, or correcting a picture shot by the mobile phone APP through an AI training mode. According to the method, the accuracy and the speed of pest recognition are improved through a deep learning technology optimization algorithm, the current pest type is matched with the pest types in