Research on Identification Technology of Field Pests with Protective Color Characteristics

Accurate identification of field pests has crucial decision-making significance for integrated pest control. Most current research focuses on the sticky card or case great differences between target and background. There is little with protective color characteristics. Aiming at problem that it difficult to identify characteristics in complex environment, a method based near-infrared imaging technology YOLOv5 proposed this paper. Firstly, an appropriate infrared filter ring light source have been selected build image acquisition system according wavelength largest spectral reflectance difference curves (Pieris rapae) its host plants (cabbage), which are formed by specific Then, images collected construct data set, trained tested through YOLOv5. Experimental results demonstrate average time required detect one 0.56 s, mAP reaches 99.7%.