Real-Time Detectionand Identification of Plant Leaf Diseases using YOLOv4-tiny

Abstract

Agriculture is one of the main sectors that contributes to our economy. It plays a vital role in raising the livestock in every country. Farmers and their families depend heavily on their crops for survival. When their crops are infected with diseases, it has a significant impact on their livelihood. This paper propose a system that can detect and identify the plant leaf diseases using Object Detection techniques in Image Processing. For the real time obj ect detection, we used Yolo v4 framework which is based on Convolutional Neural Network. In the concerned work, we have concentrated on different plant leaves of vegetables and fruits like Tomato, Mango, Strawberry, Beans and Potato. Bacterial, Fungal, and Early Blight are some of the most common plant leaf diseases. These diseases can be caused by living (biotic) agents or pathogens. This study aims to highlight the detection and identification of plant leaf diseases using a custom model called YOLOv4-tiny and thereby providing a remedial action to prevent the corresponding disease. Finally the system has been integrated with an application using android to provide an interface to the end users for the realtime detection of leaf disease.