TomatoLDP-Net: A Light-Weight Convolutional Neural Network for Revolutionizing Tomato Leaf Disease Diagnosis

In addressing the critical challenge of plant disease detection in agriculture, particularly within African and Asian regions, this study introduces TomatoLDP-Net, a lightweight convolutional neural network (CNN) optimized for the precise identification of tomato leaf diseases. With global tomato production exceeding 180 million metric tons, disease vulnerability poses a significant threat to both economic stability and food security. Traditional detection methods, often manual and error-prone, are increasingly inadequate. Utilizing artificial intelligence (AI), specifically deep learning (DL) techniques, TomatoLDP-Net leverages a multi-column CNN architecture with varied filter sizes (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$3\times 3,\ 5\times 5$</tex>, and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$7\times 7$</tex>) to effectively capture the diverse range of disease symptoms across scales. This approach, enhanced by dropout layers and softmax-averaging for prediction, showcases superior accuracy and reliability over conventional models, significantly improving disease classification with an impressive accuracy rate of 99.69%. TomatoLDP-Net not only marks a significant advancement in precision agriculture but also plays a pivotal role in ensuring the sustainability of tomato production and bolstering global food security, thereby addressing key agricultural challenges with cutting-edge AI technology.