Dataset for Detecting Diseases in Sweet Orange Leaves

Abstract:
This dataset is designed for the detection and classification of common diseases in Sweet Orange leaves, specifically foliage damage and mealybug infestations. The dataset was curated to support advancements in agricultural disease detection through machine learning, providing a comprehensive resource for researchers.
Data Summary:
1. Raw Data:
o Total Samples: 3,569
o Disease Dataset: 2,660 samples, including:
 Foliage Damage: 1,200
 Mealybug Infestation: 560
o Fresh Dataset: 909 samples, including:
 Healthy Leaves: 864
 Medium Healthy Leaves: 45
2. Augmented Data:
o Total Samples: 21,409
o Disease Dataset: 15,955 samples, including:
 Foliage Damage: 12,596
 Mealybug Infestation: 3,359
o Fresh Dataset: 5,454 samples, including:
 Healthy Leaves: 5,184
 Medium Healthy Leaves: 270
Purpose:
The Sweet Orange Leaf Disease Detection Dataset is a versatile resource for machine learning applications in agriculture. It is particularly suited for training image-based models, such as Convolutional Neural Networks (CNNs), to:
• Identify leaves affected by diseases like foliage damage or mealybug infestation.
• Assess varying levels of healthiness in fresh leaves.
This dataset facilitates:
• Automated Disease Detection: Streamlining monitoring processes in Sweet Orange cultivation.
• Early Intervention Strategies: Supporting timely responses to prevent crop losses.
• Enhanced Model Development: Integrating image features (e.g., shape, color, texture) with environmental parameters like humidity and temperature to improve predictive accuracy.
By fostering precision agriculture practices, this dataset contributes to sustainable citrus farming and demonstrates the potential of AI-driven solutions for crop health management.

摘要：
本数据集专为甜橙叶片常见病害的检测与分类任务设计，具体涵盖叶片损伤与粉蚧侵染两类典型病害。本数据集旨在通过机器学习技术推动农业病害检测领域的技术进步，为相关科研人员提供全面的研究资源。

数据概况：
1. 原始数据：
o 总样本量：3569份
o 病害数据集：2660份样本，包含：
 叶片损伤样本：1200份
 粉蚧侵染样本：560份
o 健康样本集：909份样本，包含：
 健康叶片：864份
 中等健康叶片：45份
2. 增强数据：
o 总样本量：21409份
o 病害数据集：15955份样本，包含：
 叶片损伤样本：12596份
 粉蚧侵染样本：3359份
o 健康样本集：5454份样本，包含：
 健康叶片：5184份
 中等健康叶片：270份

研究用途：
本甜橙叶片病害检测数据集是面向农业机器学习应用的通用型研究资源，尤其适用于训练基于图像的模型，例如卷积神经网络（Convolutional Neural Networks, CNNs），以实现以下目标：
• 识别受叶片损伤或粉蚧侵染等病害影响的叶片；
• 评估健康叶片的不同健康等级。
本数据集可支持以下应用方向：
• 自动化病害检测：优化甜橙种植中的监测流程；
• 早期干预策略：助力及时采取应对措施以避免作物减产；
• 进阶模型开发：将形状、颜色、纹理等图像特征与湿度、温度等环境参数相结合，提升预测精度。
本数据集通过推动精准农业实践，助力可持续柑橘种植业发展，并展现了人工智能驱动的作物健康管理方案的应用潜力。