Experimental Study On Identification Of Cabbage Pests Based On Digital Image Processing

Plant diseases and insect pests is a major cause of agricultural production,the present stage mainly by pesticide management,but much of pesticide application not only pollute the envirornment but also affect the food safety of agricultural products.Precision applying pesticide is an effective way to solve a large number of application of pesticide,but first you need to type and make accurate judgment to the position of the plant diseases and insect pests.In this paper,through test box to collect pests and cabbage samples images,after the pretreatment of the sample image are extracted image features of insects.This paper to extract the insect pest of 13 kinds of characteristics,selection of which has the characteristics of independence and stability,area,perimeter,R-G,G-B,energy,entropy six characteristics to the pests pecies identification.Design formula of fuzzy recognition membership,by comparing the size of the membership degree of insect pest species to distinguish,establish a fuzzy classifier Through the MATAB graphical user function to establish a simple interface.Using fuzzy classifier on the field of insect pest of image recognition,get the sprouts three common pests identification accuracy is above eighty-five percent.Obtained by the experimental results have serious influence on identification accuracy of factors is the shape of the pests and the shadow cast by the sun.and the complexity of the background of pests.Provide theoretical help for the follow-up precision agriculture.