Survey on Crop Disease Detection and Identification Based on Deep Learning

Abstract

Plant diseases lead to a reduction in both quality and quantity of agricultural production. 50% of agricultural losses are due to these diseases. Due to poverty and lack of infrastructures in some countries, their identification remains difficult. Plant pathologists use several techniques to identify these diseases. But these techniques are time-consuming and relatively expensive for farmers. Nowadays, several models based on image processing (IP) techniques, machine learning (ML) algorithms and deep learning (DL) algorithms have been proposed for automatic detection and identification of plant diseases. In this study, we divided these models into two groups: models based on IP and classical ML algorithms, and those based on DL. DL coupled with the transfer learning (TL) technique has become the most widely used method because of its impressive performance. The critical analysis of these models has allowed us to identify potential challenges in the field of automatic plant disease diagnosis.