pallet-block-2696
收藏资源简介:
该数据集名为pallet-block-2696,由德国弗劳恩霍夫物流与材料流研究所和多特蒙德工业大学创建,包含2696张欧洲托盘图像。数据集图像在4个月内每隔一到两周进行拍摄,以记录托盘的自然老化和人为损坏过程。该数据集旨在解决托盘在物流过程中因老化而难以识别的问题,并为生成具有老化特征的人工托盘图像提供数据基础。数据集的创建过程包括收集30个EPAL托盘,并从三个角度(中央视角、右侧旋转和左侧旋转)拍摄RGB图像。图像在自然光照条件下进行拍摄,并记录了天气和光照条件的变化。最后一天拍摄时,还对部分托盘进行了人为损坏。数据集包含的图像已进行手动裁剪和缩放,并标记了ID、视角和拍摄日期等信息。
This dataset, named pallet-block-2696, was developed by the Fraunhofer Institute for Logistics and Material Flow (Germany) and Technische Universität Dortmund. It contains 2696 images of European pallets. The images were captured every one to two weeks over a four-month period to document the natural aging and human-induced damage processes of the pallets. This dataset is designed to address the challenge of difficult pallet identification caused by aging during logistics operations, and provides a foundational data resource for generating artificial pallet images with aging-related characteristics. The dataset creation workflow involved collecting 30 EPAL-compliant pallets, and capturing RGB images from three perspectives: central view, right-rotated view, and left-rotated view. All images were taken under natural lighting conditions, with variations in weather and lighting conditions recorded during the capture period. On the final day of image capture, artificial damage was also inflicted on a subset of the pallets. All images included in the dataset have undergone manual cropping and scaling, and are annotated with metadata such as unique ID, shooting perspective, and capture date.

- 1Enhancing Long-Term Re-Identification Robustness Using Synthetic Data: A Comparative Analysis德国弗劳恩霍夫物流与材料流研究所, 德国多特蒙德工业大学 · 2025年



