Deep Learning approach for detecting diseases in rose plants

Abstract

Many places in Bangladesh grow roses for sale. People like to call them "the queen of flowers." It's very important to find diseases in rose plants early so they can grow well and make a lot of flowers. It takes a lot of work and time to find diseases the old-fashioned way, which makes it hard to act quickly. The main goal of this thesis is t find out how well the YOLOv8 model can find diseases in rose trees in all of its forms, including YOLOv8n, YOLOv8s, YOLOv8m, YOLOv8l, and YOLOv8x. The YOLOv8 deep learning method is used to find common rose diseases like black spot and downy mildew. The thesis looks at the different types of YOLOv8's precision, recall, and mAP scores to see how well they can find diseases after a lot of training and review. The results show how well the YOLOv8 model can find diseases on rose leaves, especially when the leaves are small or medium-sized. This means that the model could be used right now to help keep diseases from spreading in crops. This essay talks about how YOLOv8, a type of deep learning, can help keep track of the health of rose plants. It also sets the stage for future work on finding crop diseases.