COCO-Bridge 2021+
收藏资源简介:
COCO-Bridge 2021+数据集由Eric Bianchi和Matt Hebdon创建,用于桥梁结构细节的检测。该数据集包含1470张标注的桥梁结构照片,其中10%用于测试集,90%用于训练集。数据集涵盖四种桥梁结构细节:平面外加劲肋、盖板终止、角钢连接和轴承。数据集的创建基于弗吉尼亚交通部提供的真实桥梁检查记录。该数据集主要用于评估和优化无人机辅助的桥梁检查中的深度学习模型,旨在提高检测精度和处理速度,确保桥梁的安全性和耐用性。
The COCO-Bridge 2021+ dataset was created by Eric Bianchi and Matt Hebdon for the detection of bridge structural details. It contains 1,470 annotated photographs of bridge structures, with 10% allocated to the test set and 90% for the training set. The dataset covers four types of bridge structural details: planar external stiffeners, cover plate terminations, angle steel connections, and bearings. This dataset was developed based on real bridge inspection records provided by the Virginia Department of Transportation. It is primarily used to evaluate and optimize deep learning models in drone-assisted bridge inspection, aiming to improve detection accuracy and processing speed, and ensure the safety and durability of bridges.

- 1Deep Learning Models for UAV-Assisted Bridge Inspection: A YOLO Benchmark Analysis越南胡志明市信息技术大学,越南国家大学胡志明市 · 2024年



