Identification and Classification of Plant Leaf Diseases using YOLOv4-tiny Algorithm

The Deep Learning (DL) and Machine Learning (ML) algorithms are applied for improvement of computer vision for right way of efficient for detection of plants leaf diseases. In this study, we have proposed an object detection approach with help of You Only Look Once (YOLO) detection methods to determine and diagnose multiple diseases. Our system is designed to identify and classify numerous plant leaf diseases accurately and reliably. The dataset of plant leaves was manually annotated to meet the model’s training requirements. The object detection model is trained to recognize the presence or absence of plant diseases using the annotated dataset. Our proposed system takes a full image of a plant leaf in a single instant and predicts bounding boxes and class probability. Farmers can use their mobile phones to capture a random picture of a plant leaf, and our model can detect and identify the disease present in the leaf, helping them take the necessary steps to prevent disease spread. Our system is especially useful for farmers who can identify plant diseases as soon as they appear and take measures to prevent their spread, resulting in higher crop yields and better crop quality. Overall mAP of YOLO v4-tiny is 77.0%. Our research proposes a reliable, safe, and accurate system for detecting and diagnosing plant leaf diseases, contributing to the advancement of precision agriculture.