Cucumber Images Dataset – Original Collection

This repository contains the original collection of 1,037 images of cucumber plants (Cucumis sativus L.) captured in the field during different developmental stages of the crop. The images are organized into 10 folders that reflect the combination of phenological stage (seedling, early vegetative stage, vegetative growth, flowering, and fruit formation) and the type of device used for capture (digital camera or mobile phone). The distribution of images varies across folders according to real acquisition conditions. This set provides a direct visual record of the crop under heterogeneous capture conditions and can be used as a basis for image analysis tasks, computer vision experimentation, or studies on phenological variability. Related Repositories This dataset is part of a larger collection composed of three independent repositories: Original Dataset (this repository) Classical Augmentation Dataset: includes images generated through geometric and photometric transformations and soft effects (https://doi.org/10.5281/zenodo.17835867). StyleGAN2-ADA Generative Dataset: contains synthetic images produced from the original set. (https://doi.org/10.5281/zenodo.17820422)本存储库包含1037张田间拍摄的黄瓜（*Cucumis sativus L.*）植株原始图像，覆盖该作物的各个生育时期。图像被划分为10个文件夹，分类依据是物候期（幼苗期、营养生长早期、营养生长期、开花期及果实形成期）与拍摄设备类型（数码相机或移动电话）。受实际采集条件影响，各文件夹内的图像分布存在差异。本数据集提供了异质采集条件下作物的直接视觉记录，可作为图像分析任务、计算机视觉实验或物候变异性研究的基础数据集。 相关存储库 本数据集隶属于由三个独立存储库组成的更大数据集集合： 原始数据集（即本存储库） 经典增强数据集：包含通过几何变换、光度变换及柔化效果生成的图像（https://doi.org/10.5281/zenodo.17835867） StyleGAN2-ADA生成式数据集：包含基于原始数据集生成的合成图像（https://doi.org/10.5281/zenodo.17820422）