Abstract
Early detection of plant disease is a primary challenge in smart agriculture. Image processing can be used for detecting plant disease. When it comes to detecting plant disease, a variety of algorithms are built Hound these four stages. The performance of earlier designed algorithms is computed with regard to different parameters such as accuracy, recall, etc. In this paper, we propose a machine learning approach that will process images captured from an IoT camera-based approach that periodically sends :photos. The proposed approach uses a voting classifier for determining if a plan is healthy or not. The voting classifier was compared against the SVM and provided 26% better accuracy and precision and 27% better recall. (C) 2021 INT TRANS J ENG MANAG SCI TECH.