TIGPR: A Multi-View Ground Penetrating Radar Detection Data for Damage Assessment of Transportation Infrastructure
收藏资源简介：
The TIGPR dataset is a comprehensive collection of ground-penetrating radar (GPR) images for detecting damage in transportation infrastructure, including roads, bridges, tunnels, and airports. This dataset encompasses common types of damage found in transportation infrastructure, such as cracks, interlayer debonding, looseness, and voids. The images were collected from real-world surveys of highways, bridges, tunnels, and municipal roads in regions such as Guizhou, Jinhua, and Nanjing. The equipment used includes the 2D GPR equipment: IDS-FastWave and MALA GX750, as well as the 3D GPR equipment GeoScope 3D-Radar. The 2D GPR systems captured B-scan images with a resolution of 200 × 200 pixels, while the 3D GPR system provided both B-scan and C-scan images with a resolution of 320 × 320 pixels. The image dimensions correspond to actual infrastructure scales, with a length of 10 meters and a depth of 1 meter. This dataset integrates information from multiple damage types and views, supporting the development of deep learning models for the detection, classification, and segmentation of transportation infrastructure damage. It has the potential to enhance non-destructive testing and automated evaluation of transportation infrastructure.
TIGPR数据集是一套用于检测交通基础设施（含道路、桥梁、隧道及机场）损伤的探地雷达（ground-penetrating radar, GPR）图像综合集。该数据集涵盖交通基础设施常见损伤类型，包括裂缝、层间脱粘、松散及空洞。图像采集自贵州、金华、南京等地区的高速公路、桥梁、隧道及市政道路的实地勘测。所用设备包括2D探地雷达设备：IDS-FastWave与MALA GX750，以及3D探地雷达设备GeoScope 3D-Radar。2D探地雷达系统采集的B扫面（B-scan）图像分辨率为200×200像素；3D探地雷达系统则可提供B扫面与C扫面（C-scan）两种图像，分辨率均为320×320像素。图像尺寸对应实际基础设施尺度，单幅图像覆盖长度10米、深度1米。本数据集整合了多类损伤与多视角信息，可支撑用于交通基础设施损伤检测、分类与分割的深度学习模型开发，有望提升交通基础设施的无损检测与自动化评估水平。
