Augmented UT Data of Pipe Wall-Thinning

This dataset contains ultrasonic thickness(UT) measurement data used to train, validate and evaluate the deep learning models presented in the paper. UT data generated to address class imbalance during model training-reflecting the same error patterns and augmentation strategies-have been made publicly accesible, fully anonymized, and preprocessed for research use only. Contents:- Pipe Thickness Repeated UT Measurement Data.pkl- Repeated UT Measurement Data.xlsxFile format: .pkl, .xlsxTotal size: ~ 4 GB本数据集包含用于训练、验证与评估该论文所提出深度学习模型的超声测厚（Ultrasonic Thickness，UT）测量数据。为解决模型训练过程中的类别不平衡问题而生成的UT数据，保留了一致的误差模式与数据增强策略，现已完成完全匿名化与预处理，仅面向科研用途公开共享。数据集内容： - 管道厚度重复UT测量数据.pkl - 重复UT测量数据.xlsx 文件格式：.pkl、.xlsx 总大小：约4 GB