摘要：AbstractAgricultural pests and diseases seriously affect the yield and quality of crops. In order to study the incidence of pests and diseases, and improve the utilization rate of pesticides, an Android rapid processing system of Android-based insect pest thermal infrared image was designed and implemented. In this paper, maize tumor powdery mildew was used as the research object, infrared thermal imaging lens based on Android system was used to obtain corn disease images. Meanwhile, OpenCV technology was used for image processing. Moreover, in order to better achieve the thermal infrared segmentation effect, the image segmentation effect based on the traditional Otsu algorithm (Otsu) and the Otsu single-threshold segmentation method based on the bare bones fireworks algorithm (BBFWA) was analyzed. Importantly, the results showed that the method of Otsu single-threshold segmentation based on BBFWA had the best effect on thermal infrared image segmentation. Since thermal infrared images are greatly influenced by the ambient temperature, the shooting needs to be performed under cloudy or cloudy conditions. Finally, the ratio value between the total number of pixels in the white area and the total number of pixels in the divided binary image was calculated to determine the degree of pest infestation, which can provide data reference for the UAV spraying amount to some extent.
