Integrating Deep Learning and Ensemble Methods for Robust Tomato Disease Detection: A Hybrid CNN-RF Model Analysis

Abstract

The timely and accurate diagnosis of plant diseases is what determines the efficiency of farming and food security in the world. In the situation of tomato cultivation, diseases like tomato spot, wilt, blight, and rot are the ones that cause a lot of damage to the yield and quality. This paper presents a new hybrid model that combines convolutional neural networks (CNN) and random forest (RF) for the effective and accurate classification of tomato diseases. The model uses the CNN deep learning capabilities for feature extraction from photos of the tomato leaves and then utilizes RF robust classification for differentiating between four major disease categories and healthy ones. Our approach includes data collection and preprocessing, design and training of model architecture, and performance evaluation through various metrics. The hybrid model has been rigorously assessed on a dataset containing images of diseased and healthy tomato leaves, all in all yielding an impressive overall accuracy of 97.03%. This framework not only expresses the utility of this model as a trustworthy instrument for agricultural disease management, but it also clearly surpasses other models in the field. The results of the study emphasize the advantage of CNN and RF integration, which confirm higher performance in terms of disease classification accuracy, precision, recall, and F1 score. The success of the hybrid model leads to an exploration of the potential role of artificial intelligence for farmers in agriculture, as it demonstrates a promising way to tackle the existing challenges of plant disease detection and classification. By its high accuracy and stable performance, this model is an important part of creating smart and automated agricultural diagnostic systems, which directly leads to enhanced disease management, increased crop yields, and sustainable agriculture.

Topics & Concepts