Lightweight optimization Yolo v4-based tea disease identification method and system

Abstract

The invention discloses a tea disease identification method and system based on lightweight optimization Yolo v4, and the method comprises the steps: collecting a tea disease picture, and carrying out the preprocessing of the tea disease picture, and taking the preprocessed tea disease picture as a data set for training a Yolo v4 model; performing lightweight optimization on a feature extraction trunk module and a feature extraction fusion module in the Yolo v4 model to obtain an optimized Yolo v4 model; the optimized Yolo v4 model is trained and verified through the data set for training the Yolo v4 model, and the optimal Yolo v4 model for recognizing the tea diseases is obtained; and identifying the tea disease image by using the obtained optimal Yolo v4 model. According to the invention, the huge parameter quantity and the model volume of the original Yolo v4 network model are effectively reduced, and the detection efficiency and the identification precision of the tea disease target are improved.

Description

technical field [0001] The invention relates to the technical field of crop disease target detection, and more specifically relates to a tea disease identification method and system based on lightweight optimized Yolo v4. Background technique [0002] The total tea output of China's four major tea-producing regions accounts for more than 40% of the global tea output, ranking first in the world. However, frequent and easily transmitted diseases of tea leaves (such as tea white spot, tea anthracnose, tea algae spot disease, etc.) have always been the key factors that seriously affect the yield of tea. Therefore, realizing rapid identification of tea diseases and reducing the use of pesticide doses as early as possible in the tea growth cycle is of great significance for improving the response speed of tea farmers to diseases, assisting tea farmers in controlling diseases, and promoting the intelligent development of tea production in my country. Common tea diseases mainly inc...

AI Technical Summary

Problems solved by technology

[0003] At present, in the study of tea disease classification based on deep learning, researchers often cut the collected pictures of tea diseases into only one picture of diseased tea, and then perform simple classification through classification networks such as AlexNet, VGG16, and lightweight MobileNet. Disease classification, which does not fully take into account the complexity of mutual occlusion of leaves in real tea garden scenes, unstable imaging quality, and real-time requirements for disease identification

In recent years, the end-to-end target detection algorithm represented by Yolo V4 has good recognition accuracy and speed in a variety of specific scenarios, but the disadvantage is that the corresponding network model parameters and calculation amount are too large, which is difficult to calculate on GPU. The deployment and operation of embedded devices or mobile terminals with limited human resources; and the distribution of tea diseases on tea has a high degree of randomness, and different disease types have certain similarities in shape, color, texture and other characteristics in their respective frequent periods The original Yolo V4 algorithm without improvement and optimization has a high false detection rate for tea diseases

Images