Mouse Recognition And Tracking Based On Infrared Thermal Imaging Technology

In the grain depot,mammals such as mice have caused great damage to agricultural production.Infrared monitoring has a strong application value as an important monitoring technology.In the process of infrared monitoring,the prospect of the extraction of moving targets such as mice is the key to infrared monitoring.This topic applies infrared thermal imaging technology to the identification and tracking of rodent movement targets,and has a strong reference value for rodent control in grain storage.In this paper,the infrared video camera is used to collect the mouse motion video image as the basic data of the experiment,followed by a series of preprocessing such as graying processing,noise filtering processing and digital morphology.In order to obtain more complete foreground target contour information,this paper uses the gradient direction histogram to classify the captured video images and extract the background model.Under the fixed temperature dynamic range,the background model is pixel-corrected with the maximum inter-class variance method as the threshold discrimination basis,and the foreground target is initially detected,and the threshold is deleted according to the detected result combined with the gray histogram to the foreground target.Perform secondary extraction and update.And the parameters such as mean and variance are combined with the test results to obtain better detection results.The experimental results also prove the effectiveness of the algorithm and adapt to the needs of agricultural monitoring.Finally,we use the fuzzy clustering algorithm to segment the image captured by the video.Combined with the segmentation result of the image,we use the mean shift algorithm to complete the tracking of the moving target.The tracking effect is good and the error is within the allowable range.