Solanaceae Family Leaf Disease Image Dataset

The Solanaceae Family Leaf Disease Dataset (SFLD Dataset) is a hierarchical leaf image dataset comprising healthy and diseased leaf samples from three Solanaceae crops: eggplant (Solanum melongena), potato (Solanum tuberosum), and tomato (Solanum lycopersicum). The dataset contains 2,593 original leaf images and 17,000 augmented leaf images, organized into two subsets: (i) Original Leaf Image Dataset and (ii) Augmented Leaf Image Dataset. The dataset includes healthy leaves and disease-affected leaves belonging to bacterial, fungal, and viral categories. The disease classes represented in the dataset are: Eggplant: Healthy, Downy Mildew, Septoria Leaf Spot, and Eggplant Mosaic Virus. Potato: Healthy, Bacterial Soft Rot, Late Blight, Potato Leaf Roll Virus, Potato Virus X (PVX), and Potato Virus Y (PVY). Tomato: Healthy, Bacterial Canker, Cercospora Leaf Spot, Late Blight, Tomato Leaf Curl Virus, Tomato Mosaic Virus, and Tomato Spotted Wilt Virus. The dataset is organized in a hierarchical directory structure based on crop species, disease category, and disease class, enabling efficient use in supervised learning tasks. The augmented subset was generated from the original images to increase sample diversity and improve the robustness and generalization capability of machine learning and deep learning models. The SFLD Dataset is intended for applications including plant disease classification, automated disease diagnosis, computer vision research, transfer learning, and precision agriculture systems for crop health monitoring and management.茄科植物叶片病害数据集（Solanaceae Family Leaf Disease Dataset, SFLD Dataset）是一个采用分层结构的叶片图像数据集，涵盖茄子（Solanum melongena）、马铃薯（Solanum tuberosum）、番茄（Solanum lycopersicum）3种茄科作物的健康与染病叶片样本。该数据集包含2593张原始叶片图像与17000张增强叶片图像，分为两个子集：(i) 原始叶片图像子集与(ii) 增强叶片图像子集。 本数据集包含健康叶片以及隶属于细菌、真菌、病毒类别的染病叶片，其涵盖的病害类别如下： 茄子类：健康叶片、霜霉病、壳针孢叶斑病、茄子花叶病毒； 马铃薯类：健康叶片、细菌性软腐病、晚疫病、马铃薯卷叶病毒、马铃薯X病毒（Potato Virus X, PVX）、马铃薯Y病毒（Potato Virus Y, PVY）； 番茄类：健康叶片、细菌性溃疡病、尾孢叶斑病、晚疫病、番茄卷叶病毒、番茄花叶病毒、番茄斑萎病毒。 该数据集采用基于作物种类、病害类别与病害类型的分层目录结构，可高效应用于监督学习任务。增强子集由原始图像生成，旨在提升样本多样性，增强机器学习与深度学习模型的鲁棒性与泛化能力。 SFLD数据集可应用于植物病害分类、自动化病害诊断、计算机视觉研究、迁移学习以及用于作物健康监测与管理的精准农业系统等场景。