Comprehensive smartphone image dataset for Aegle Marmelos, Hog plum, and lemon plant leaf disease and freshness assessment.

Abstract

Fruits, which are packed with nutrients, vitamins, and antioxidants, have been known for their numerous health benefits and curative powers, and are utilized in conventional medicine. Aegle Marmelos, Lemon, and Hog Plum are tangy fruits widely recognized in Asian countries for containing a plentiful supply of bioactive substances. They are also highly valuable in boosting metabolism, possessing tremendous therapeutic properties, and holding financial significance. The leaves of these fruit trees are as essential as their fruits, as they contain versatile medicinal and dietary benefits of immense value. However, these leaves are often affected by various fungal and other diseases, which reduce the ability for healthy growth and productivity of both fruits and leaves. Plants infected with various leaf diseases can produce fewer fruits, which are also of lower quality due to failure to reach maturity and lack of sufficient nutritional value. For these reasons, there is a risk of an outbreak in orchards, which can lead to significant financial losses for both producers and the agricultural sector. This signifies that the early identification of leaf diseases and the management of orchards are essential to minimize the impact of leaf diseases and mitigate these issues, ensuring the healthy production of valuable medicinal fruits. In this paper, various infected leaf images are collected from different regions of Rangpur, providing a comprehensive dataset comprising 3941 images. The dataset includes images of three different plant leaves, where 1513 images of Aegle Marmelos, 1232 images of Lemon, and 1196 of Hog plum, where each of the categories encompasses several classes of common leaf diseases. Through this dataset, an early and accurate digital detection system can be employed, allowing producers to clearly identify diseases instead of relying on traditional methods. The precise and timely identification of leaf diseases enables the control of these diseases by taking necessary actions, ensuring the sustainability of plants, and promoting the healthy growth of these invaluable medicinal fruits.

Plant phenotyping relevance

植物葉の病害状態を画像で評価する大規模データセットの構築が中心であり、病害表現型の画像ベース解析に該当する。

Code and data availability

The paper's own smartphone leaf-image dataset (3,941 raw + 12,295 augmented images of Aegle Marmelos, Hog plum, and lemon leaves) is publicly deposited on Mendeley Data with an explicit direct URL and DOI, matching the allowed URL list. No separate analysis code or trained model checkpoint is stated as available.