Abstract:The agricultural production is greatly affected by various plant diseases. Classifying the severity of crop diseases is the requirement for formulating disease prevention and control strategies. However, the differences between different severity of the same crop disease are very tiny. It increases the difficulty of correct crop disease recognition. For example, at the early stage of the disease, the lesions on the leaves are not obvious. And it is very difficult to extract the features of the lesions. However, these very small color and texture differences of the lesions are the key patterns to distinguish different kinds of diseases of the same species. In order to achieve better performance in the fine-grained classification of the crop diseases, a modified light-weight convolution neural network was proposed. Multi-scale convolution kernel and coordinate attention mechanism are introduced in SqueezeNext to extract the features of the lesions accurately. The performance of the proposed model was evaluated using the AI challenger 2018 plant disease recognition dataset, and the recognition accuracy can reach 91.94%, which is 3.02% point higher than the original SqueezeNext model. In order to verify the effectiveness of the proposed model, comparative experiments were carried out using ReseNet50, Xception and mobilenetv2. The experimental results showed that the accuracy of the proposed method was slightly better than Xception, while the model size is only 2.83 MB, which is only 3.45% of Xception. The proposed method balances the performance and efficiency very well. Thus, it is suitable for deployment on mobile terminals and other embedded resource-constrained devices, which help to promote the popularization of smart agriculture application.

Lightweight Inception Networks for the Recognition and Detection of Rice Plant Diseases

Identification of Rice Plant Diseases Using Lightweight Attention Networks

Attention Embedded Lightweight Network for Maize Disease Recognition

An Attention-Based Lightweight Residual Network for Plant Disease Recognition

Efficient Real-Time Recognition Model of Plant Diseases for Low-Power Consumption Platform

Multi-Plant Disease Identification Based on Lightweight ResNet18 Model

An Ultra-Lightweight Efficient Network for Image-Based Plant Disease and Pest Infection Detection

Lightweight Isotropic Convolutional Neural Network for Plant Disease Identification

RIC-Net: A Plant Disease Classification Model Based on the Fusion of Inception and Residual Structure and Embedded Attention Mechanism

Using deep transfer learning for image-based plant disease identification

Identifying Plant Diseases Using Deep Transfer Learning and Enhanced Lightweight Network.

Deep Learning-Based Methods for Multi-Class Rice Disease Detection Using Plant Images

Detection of Rice Plant Diseases Based on Deep Transfer Learning.

Deep Learning Based Plant Disease Detection for Smart Agriculture

Crop Disease Recognition Based on Modified Light-Weight CNN With Attention Mechanism

MobileNetV2-Incep-M: a hybrid lightweight model for the classification of rice plant diseases

Convolutional Neural Network in Rice Disease Recognition: Accuracy, Speed and Lightweight

MS-DNet: A Mobile Neural Network for Plant Disease Identification

Mobile Convolution Neural Network for the Recognition of Potato Leaf Disease Images

Identification of Plant Disease Images Via a Squeeze-and-excitation MobileNet Model and Twice Transfer Learning

SK-MobileNet: A Lightweight Adaptive Network Based on Complex Deep Transfer Learning for Plant Disease Recognition