Research On Crop Pest Detection Technology Based On Image Recognition

Crop pests restricted the agricultural production.because the characteristics of many types of crop pests and high density,causing huge production of crops,and at the same time due to the imperfect in crop pests identification,lead to deal with aspects of insect pest,a large number of pesticides also caused the serious ecological imbalance.With the development of the concept of precision agriculture and intelligent agriculture,the detection and recognition methods of traditional agricultural diseases and insect pests have the disadvantages of slow recognition speed,low recognition accuracy and strong subjectivity.The information technology is used to assist agricultural production to realize the intelligent identification and detection of crop diseases and pests,plays a very important role in protecting the balance of the ecosystem,ensuring the safe production of the crops and improving the quality of the crops.This paper uses SpringBoot framework for development and microservice design patterns,and deploy in a distributed way on the VMware cloud computing framework.For image recognition,denoising is based on median filtering method,image preprocessing method based on Otsu and super green feature segmentation method.With the method of feature extraction based on multi parameter characteristics,and based on a selection strategy and fusion kernel function,M-SVM multi-class recognition algorithm to achieve recognition of six kinds of soybean leaf borer pest recognition,and verified that the image recognition method adopted in this paper is effective and efficient in the detection of insect pests of soybean leaf through experiment.