Citrus huanglongbing image recognition method based on attention mechanism

Abstract

The invention relates to the field of image recognition, and particularly discloses a citrus huanglongbing image recognition method based on an attention mechanism, and the method specifically comprises the steps: collecting the data of diseased leaves and diseased fruits of citrus huanglongbing as positive samples, collecting the data of other non-huanglongbing as negative samples, constructing an image classification network based on the attention mechanism, designing a loss function, inputting the training set into the image classification network based on the attention mechanism, carrying out supervised training by adopting the loss function, and inputting the verification set into the trained model for verification in the training process; and loading the trained model parameters to an image classification network based on an attention mechanism, and sequentially inputting thecitrus huanglongbing images of the test set into the network for reasoning to obtain a citrus huanglongbing image classification result. By introducing an attention mechanism module, key features of the citrus huanglongbing are autonomously selected, and the citrus huanglongbing detection method which is low in cost, high in efficiency and high in recognition rate is achieved based on image recognition.

Description

Technical field[0001]The present invention relates to the field of image recognition, and more particularly to a citrus grain drain image recognition method based on a focal mechanism.Background technique[0002]Citrus yellow dragon disease is the most devastating disease in the current global citrus production, caused by citrus yellow dragonfill, with citrus seedlings carrying citrus yellow dragon diseases, with citrus mahc monsoon, etc. for the media. The citrus planting area of ​​nearly 50 countries and regions in the world has been infected. It has reached the US Florida, Brazil, Sao Paulo and China Guangdong, Guangxi and other citrus important production areas. The agricultural departments of the world have attached great importance to the prevention and treatment of citrus yellow dragon disease. On September 15, 2020, citrus Huanglong disease was included in a class of crop pests in the Department of Agricultural Rural Areas.[0003]Early diagnosis early treatment, citrus Huanglon...

AI Technical Summary

Problems solved by technology

[0003] Early diagnosis and early treatment, the detection of citrus Huanglongbing is very important for the prevention and control work. At present, there are several detection methods: PCR detection, which uses instruments and equipment to detect whether the DNA sequence of the pathogen in citrus is consistent with the DNA sequence of Huanglongbing, which needs to be identified by a professional organization. The detection efficiency is low; using UAV hyperspectral remote sensing for large-scale detection is costly; computer vision-based solutions have been widely used in pest identification due to their low cost and high efficiency in recent years, but the disease characteristics of citrus Huanglongbing It is a fine-grained feature, and the general image recognition algorithm has a low recognition rate

Images