Crop disease detection algorithm based on deep learning

Abstract

The invention relates to a crop disease detection algorithm based on deep learning. The crop disease detection algorithm comprises the following steps: (1) obtaining a training set and a test set, andperforming image preprocessing on the training set; (2) inputting the training set into a model with adaptive multi-scale for training, connecting the model with two global pooling layers, using transfer learning with differential learning rates, and using Focalloses as a loss function at the same time; and (3) inputting the test set image into the trained detection model to predict the disease type of the image. Compared with other algorithms, the method provided by the invention has the advantages that the detection performance and the generalization capability are improved by using transfer learning of a difference learning rate; and meanwhile, the two global pooling layers are connected, so that the model keeps more detail information and has size adaptability, model convergence is accelerated through a progressive learning strategy, the problem of class imbalance is relieved through Focus loss, and the algorithm has better detection performance on diseases with similar pathologies.

Description

technical field [0001] The invention relates to the field of image classification, in particular to a crop disease detection algorithm based on deep learning. Background technique [0002] The previous image classification was mainly through the traditional machine learning method, which is usually divided into two parts: the method based on feature extraction and the method based on template matching. These methods require artificial selection of features, and the generalization is not strong. With the development of deep learning, especially the application of convolutional neural network in image classification, image detection and image segmentation, the effect achieved is unmatched by traditional algorithms in the past. The diagnosis of crop diseases is crucial to agricultural production. The detection of disease types through intelligent means can improve farmers' understanding of diseases and take corresponding measures to prevent and control them, improving agricultu...

AI Technical Summary

Problems solved by technology

[0003] Crop diseases have the problem of quantity imbalance between categories and the problem of too similarity between individual diseases. In order to solve these problems, the present invention proposes a crop disease detection algorithm based on deep learning

Images