Analysis and Classification of Plant Diseases Based on Deep Learning

Abstract

AbstractPlant diseases identification is generally done by visual evaluation. The diagnosis quality depends strongly on professional knowledge. However, agricultural expertise is not easily learned. To overcome this issue, automatic analysis and classification of plant diseases through image processing and artificial intelligence is an encouraging solution and can reduce the lack of agricultural knowledge. This alliance between image processing and machine learning and/or deep learning approaches has produced very good results in medical imaging and has enabled the development of robust systems to assist in the diagnosis of several diseases. Deep learning (DL) and Machine learning approaches became the most promising tools for image analysis and classification. The objective of our work is to conduct a comparative study between different deep learning architectures for the analysis and classification of plant diseases. Four DL-based architectures have been implemented such as MobileNet, AlexNet, Inception V3, and VGG16. Simulation results have shown that the MobileNet architecture, even simple, has allowed better classification accuracy.KeywordsPlant diseases detectionPlant diseases classificationImage classificationDeep learningMachine learningConvolutional neuron network