Plant Disease Detection Using Advanced Deep Learning Algorithms: A Case Study of Papaya Ring Spot Disease
Prajwal Maski, Asokan Thondiyath
Abstract
Faster and accurate detection of plant diseases can enhance agricultural productivity and overall yield. Many methods are proposed in the literature for plant disease detection using optical sensors mounted on agricultural robots. There are many deep learning algorithms proposed to identify plant diseases. Many such algorithms require high computation power and take more time during training. Implementation of such algorithms on robot controllers for online detection is quite difficult. In this paper, we propose the use of lighter versions of YOLO which are more efficient and have high detection speed in plant disease detection. A case study of deadly viral disease that affects Papaya plants is considered, and a dataset consisting of images of both healthy and diseased leaves which are affected by the Papaya ringspot virus has been constructed. The proposed algorithms have been trained on the dataset built and tested. Test results show that lighter versions of YOLO are far more efficient than improved or original versions of YOLO, especially in the case of plant disease detection using mobile agricultural robots.