Pest image classification method and pest image classification system based on morphological multi-feature fusion

Abstract

The invention relates to a pest image classification method based on morphological multi-feature fusion. The method comprises the following steps that: a training process is carried out: data in a training image set is subjected to image segmentation, the segmented training images are subjected to preprocessing, morphological features of the training images are extracted, and a training image feature matrix is obtained through the multi-morphological-feature fusion; a test process is carried out: test images to be recognized are input, the test images are subjected to image segmentation and preprocessing, morphological features of the test images are extracted, and a test image feature matrix is obtained through multi-morphological-feature fusion; a pest type is recognized: the similarity between the test image feature matrix and the training image feature matrix is calculated, the class with the highest similarity is found out, and the pest type and the control method are obtained according to the similarity. The invention also discloses a pest image classification system based on the morphological multi-feature fusion. The method and the system provided by the invention have the advantages that the pest recognition rate and the program robustness are improved; and the actual application value of pest recognition in agricultural production is improved.

Description

technical field [0001] The invention relates to the technical field of intelligent agriculture and pattern recognition, in particular to a pest image classification method and system based on morphological multi-feature fusion. Background technique [0002] Pests are the enemy of crop growth, they occur throughout the growth period of crops, and can cause a large amount of crop yield reduction. The current pest classification and identification work is mainly done by a small number of plant protection experts and agricultural technicians. However, there are many types of pests, and every plant protection expert can only identify some pests. There are more and more indications that the contradiction between the increasing demand for pest identification and the relatively few pest identification experts has been intensified. The automatic identification of pest images is of great significance, but the recognition rate of automatic pest image identification methods is low. The...

AI Technical Summary

Problems solved by technology

The current pest classification and identification work is mainly done by a small number of plant protection experts and agricultural technicians. However, there are many kinds of pests, and every plant protection expert can only identify some pests

There are more and more indications that the contradiction between the increasing demand for pest identification and the relatively few pest identification experts has been intensified. The automatic identification of pest images is of great significance, but the recognition rate of automatic pest image identification methods is low. The robustness is poor, and only exists in the experimental stage. It is of great significance to seek a pest identification method with high recognition rate and strong robustness.

Images