Analysis of Leaf Disease Detection in the Solanaceae family plants using Machine Learning Algorithms

The paper aims to classifying and predicting diseases for leaf images using random forest and convolutional neural networks and compare these two methods. The Indian economy is heavily dependent on agricultural productivity to meet the demands of citizens and support farmers. Unfortunately, crop diseases and pests lead to significant reductions in crop production and quality, which underscores the importance of early detection in precision agriculture. However, manually identifying these problems over large areas of the farm is both time-consuming and labor-intensive. The current work is implemented for Solanaceae families such as potatoes, tomatoes, and pepper plants. Initially, the Random Forest classification method was implemented, which achieved an accuracy of 67.90%. Further, the experiments with the CNN model extended resulted in a significant improvement with an accuracy level of 95.53% after training and testing.