Plant Leaf Disease Prediction using Deep Learning Algorithm with Fertilizer Suggestion

Abstract

Motivation: Early disease detection and accurate diagnosis of tomato leaf diseases are crucial for enhancing crop yield and ensuring food security in modern agriculture. Problem: One of the challenges faced by the agriculture sector is the spread of viruses, fungi and bacteria that cause plant diseases. Traditional diagnostic methods require manual inspection by farmers or experts, which is time-consuming and prone to inaccuracies. Approach: Proposed Convolutional Neural Network (CNN) system aims to classify ten classes of tomato leaf diseases using sophisticated deep learning techniques. The input images were pre-processed with a Wiener filter to enhance clarity and suppress noise without compromising important edge information. Computational complexity is decreased and precision is increased by selecting features through the Particle Swarm Optimisation (PSO) algorithm. Finally, the proposed CNN model can be utilised to effectively detect relevant plant leaf diseases based on the selected features. Result: It achieved promising results, with 96.5% precision, 94.7% recall, and 95.6% F1-score, to evidence its effectiveness. With the incorporation of pre-processing, PSO-based feature selection, and CNN-based classification this method offers a practical and easy-to-use tool that assists farmers with their decision-making, minimising crop loss.

Plant phenotyping relevance

トマト葉画像から病害状態をCNNで推定する手法を開発・評価しており、植物病害表現型の取得・分類が研究の中心である。

Code and data availability

公開本文の所在を確認できませんでした。非公開または購読が必要な可能性があります。

No evidence-backed public reproduction asset is currently recorded.