Classification and Detection of Citrus Diseases Using Deep Learning

Abstract

In agriculture, plant diseases are primarily responsible for competitive devaluation that causes financial losses. Recognition of plant disease is a major problem in the world of agriculture. The quality and production of citrus fruits is badly affected by citrus-focused diseases. Thus, recognition of citrus disease is the most promising approach in agriculture and attracts considerable attention in both the farming and computer communities. Advances in deep learning provide a platform to expand and enhance the practice of precise plant protection. It also expands the market in precision farming for computer vision applications. We propose an advanced CNN technique in this chapter to care for farming by classifying and recognizing citrus disease in order to help grow healthy plants. The model proposed was trained using different training epochs, batch sizes, and dropouts. The dataset includes images of unhealthy and healthy citrus leaves and fruits that can be used to prevent plants disease using deep learning techniques. The main diseases in the datasets are canker, black spot, greening, scab, and melanose.