Jute Disease Image Dataset

This meticulously curated collection of images, encompassing healthy jute plants and those affected by Dieback, Holed, Mosaic, and Stem Soft Rot, provides an invaluable foundational resource for training and evaluating machine learning models. The creation of this dataset directly addresses the crucial data void in jute disease imagery, positioning it as a foundational resource that can accelerate ML research and application in jute agriculture. By making this dataset publicly available, the aim is to empower researchers, developers, and agricultural stakeholders to build and deploy automated diagnostic tools. Such tools can enable timely interventions, significantly mitigating yield losses, enhancing food security, and bolstering the resilience and economic prosperity of the "Golden Fibre" industry in Bangladesh and beyond. The dataset comprises 1,390 unique images, meticulously collected and preprocessed, with a uniform resolution of 1024x1024 pixels. It is structured into raw and augmented/train-test split versions, catering to various research and development needs.

本经过精心遴选的图像数据集涵盖健康黄麻植株，以及感染顶枯病（Dieback）、穿孔病（Holed）、花叶病（Mosaic）与茎软腐病（Stem Soft Rot）的患病黄麻植株，可为机器学习（Machine Learning）模型的训练与评估提供极具价值的基础支撑资源。本数据集的构建直接填补了黄麻病害图像领域的关键数据空白，可有效加速黄麻农业场景下机器学习研究与应用的推进。通过公开共享本数据集，旨在赋能研究人员、开发者与农业利益相关者，使其能够开发并部署自动化病害诊断工具。此类工具可实现及时的病害干预，显著降低产量损失、提升粮食安全水平，并增强孟加拉国及全球范围内“金色纤维”产业的韧性与经济活力。本数据集共计包含1390张独特图像，所有图像均经过精心采集与预处理，统一分辨率为1024×1024像素。数据集分为原始版本与增强/训练测试分割版本两类，可满足多样化的研发与应用需求。