Method and system for performing plant lesion classification based on computer vision

Abstract

The application discloses a method and a system for performing plant lesion classification based on computer vision. The method comprises the steps of: selecting a sampling block from a field, and collecting images of the plant leaves from top to bottom; sorting each pixel value in the neighborhood filtering window according to an order of increasing and descending, and obtaining the pixel value of the pixel in the middle position of the arrangement as the pixel value of the neighborhood filtering window to process the images of the plant leaves; segmenting the processed images of the plant leaves to highlight the image area of the lesion through color space conversion; extracting the pixel features of the lesion image area obtained by image segmentation by pixel features in the images toobtain the pixel features of the lesion images of the plant leaves; and classifying the pixel features of the lesion images by the support vector machine classifier to obtain and save the classified pixel features of different lesions and a contrast model of the disease to which it belongs. The invention effectively improves the recognition and classification efficiency of the lesion, and reducesthe cost.

Description

technical field [0001] The present invention relates to the technical field of plant disease spot classification, and more specifically, to a method and system for classifying plant disease spots based on computer vision. Background technique [0002] Plant disease is one of the main limiting factors of agricultural production, and plant leaf disease spots are the main basis for judging the degree of disease occurrence. Classifying plant leaf disease spots is helpful for people to judge the degree of disease transmission that plants are currently suffering from more efficiently and quickly. [0003] At present, the classification of plant disease spots is mainly realized in the following two ways: (1), plant disease technicians directly go to the diseased plant site to investigate the disease situation of the plant, and directly give the cause of disease and disease status by the diseased state of the diseased plant. treatment method. This method is relatively common, but ...

AI Technical Summary

Problems solved by technology

[0006] In view of this, the present invention provides a method and system for classifying plant disease spots based on computer vision, which solves the cumbersome and delayed process of plant disease spot classification and plant disease spot classification in the prior art where professional and technical personnel are required to go to the field to sample and investigate. time technical issues

Images