Development of plant disease detection for smart agriculture

Plant leaf diseases have become a serious concern for the agricultural industry, yet timely diagnosis and recognition are challenging in numerous regions of the globe owing to a shortage of automated crop disease identification methods. If plant diseases are not recognized in a prompt way, food insecurity will rise, affecting the country's income. Plant disease identification is critical for successful crop prevention and control of diseases, as well as farm production management and decision-making. Plant disease detection technologies aid in finding infected plants in their early phases and also help the user in cost-effectively expanding plant disease identification system to a variety of plants. This paper's major contribution is a stacked ensemble technique based on Machine learning and Deep learning techniques. This research article also elaborates on how plant disease detection framework will be realized using novel segmentation and feature extraction strategies for extracting significant features for classification. Once the features are extracted, they are transmitted to the cloud platform to implement web enabled automated monitoring system. The proposed stacked ensemble learning is evaluated by comparing different machine learning and Deep learning techniques models utilizing precision, recall, and F_score. When compared to traditional machine learning and deep learning techniques approaches, the findings show that the proposed technique achieves about 99% accuracy.