Monochromatic LeafAdaptNet (MLAN): an adaptive approach to spinach leaf disease detection using monochromatic imaging

Abstract

A country's economic growth heavily relies on agricultural productivity, specifically nutrition derived from vegetables and leafy greens. Spinach, abundant in iron, vitamins, and other essential nutrients, plays a vital role in maintaining the health of human tissues, cartilage, and hair. However, extreme summer heat and plant diseases can significantly reduce spinach yields, making it less nutritious and harder to obtain. Implementing improved detection and classification of bacterial and fungal diseases affecting spinach leaves is crucial for minimizing pesticide use and enhancing agricultural output. A cutting-edge approach was introduced for identifying diseases in spinach leaves through deep learning object detection. To tackle these issues, the DenseNet-121-DO model served as the basis for developing the Custom Monochromatic LeafAdaptNet (MLAN). Spinach leaves were classified as Half-Spinach, Curry Leaves, Drumstick Leaves, and Lettuce, with the aid of Google-Colaboratory. This model displayed impressive results, achieving an accuracy of 99.10% and a mean Average Precision (mAP) of 98.16%. Such outcomes promote higher agricultural productivity and reduced pesticide costs by showcasing the system’s effectiveness in accurately identifying and classifying spinach leaf diseases.

Plant phenotyping relevance

葉画像から病害を検出・分類する深層学習手法の開発が中心であり、植物の病徴状態を直接推定するため、植物フェノタイピング手法として採録する。

Code and data availability

公開状態または取得可能な本文経路を確認できませんでした。

No evidence-backed public reproduction asset is currently recorded.