Thermal Image Dataset of Okra

The Okra, a fruit widely appreciated globally for its substantial seeds, holds significant importance at the international level. Its value transcends mere culinary appeal, positioning it as a valuable contributor to nutrition, agriculture, and environmental sustainability. The classification of harvested okra into over-matured and sufficiently matured pods post-harvesting brings forth numerous advantages, influencing both quality and culinary aspects. The field of quality inspection for edible items is rapidly evolving, and utilizing thermal images to assess okra maturity offers a non-invasive method for farmers and consumers to efficiently categorize it for various purposes. The FLIR E75 thermal imaging device was employed to capture thermal images of okra, with both the cameras and okra positioned at a distance greater than 0.5 meters. The image acquisition process maintained a controlled environment with a consistent room temperature. The data collection involved systematically selecting representative okra specimens, considering their maturity levels. Visual examination of the selected okra specimens was conducted, and samples were chosen for imaging to ensure accurate labeling and categorization. The collected dataset comprises 501 images, categorized into two main groups: over-matured okras and adequately matured okras. Since thermal images capture temperature distribution over the surface, FLIR-supplied software such as FLIR Thermal Studio and FLIR Research Studio can be utilized to determine the temperature of individual okra specimens. Various color palettes for representing thermal images are available, and the rainbow color palette is employed in this dataset. The FLIR software facilitates temperature measurement at specific points or selected regions, enabling temperature analysis. Moreover, computer vision techniques based on thermal imaging can be applied to these images, ensuring thorough scrutiny for accurate analysis, detection, and classification.秋葵（Okra）因其籽粒饱满而广受全球青睐，在国际层面具有重要地位。其价值绝非仅局限于烹饪食用价值，更是营养供给、农业生产与环境可持续发展领域的重要贡献者。对收获后的秋葵荚果进行过熟与充分成熟分级，可带来诸多利好，同时对秋葵的品质与烹饪应用均产生积极影响。食用农产品质量检测领域正快速发展，利用热成像图像评估秋葵成熟度，可为种植者与消费者提供一种非侵入式方法，使其能高效针对各类应用场景完成秋葵分级工作。本数据集采用FLIR E75热成像设备采集秋葵热成像图像，拍摄时相机与秋葵的拍摄距离均大于0.5米。图像采集过程在可控环境下开展，保持室温恒定。数据采集环节结合成熟度等级，系统选取具有代表性的秋葵样本。对选取的秋葵样本进行目视检查，再从中挑选样本进行图像采集，以确保标注与分级的准确性。本次采集的数据集共包含501张图像，分为两大类别：过熟秋葵与充分成熟秋葵。由于热成像图像可采集物体表面的温度分布，因此可使用FLIR官方提供的FLIR Thermal Studio与FLIR Research Studio等软件，对单份秋葵样本的温度进行测定。热成像图像可采用多种调色板进行可视化呈现，本数据集采用彩虹调色板。FLIR配套软件支持对指定点位或选定区域进行温度测量，便于开展温度分析工作。此外，基于热成像的计算机视觉技术可应用于本数据集的图像，助力开展全面细致的分析、检测与分级工作，确保结果准确可靠。