Centerline-Guided Weld Annotation for PCCP: Methods and Application Performance
- 1School of Mechanical Science and Engineering, Huazhong University of Science and Technology, Wuhan 430074, China
- 2Qinglong Pipe Industry Group Co., Ltd., Yinchuan 750002, China
Abstract
Semantic segmentation is a pivotal computer vision task, yet pixel-level annotation is labor-intensive and costly. For ribbon-like weld seams with ambiguous boundaries, polygon-based annotation suffers severe inter-annotator variability, an issue that has received limited attention in existing model-centric research. This study proposes a centerline-based annotation strategy for Prestressed Concrete Cylinder Pipe (PCCP) weld seams to improve efficiency, consistency, and model performance. Ground truth masks are generated via the Image-Guided Centerline-to-Mask Generation algorithm using slope-adaptive kernels to convert centerlines into pixel-level masks. During training and inference, multi-directional convolutional response maps are concatenated with RGB images as six-channel inputs and evaluated on UNet, R2UNet, AttUNet, UNet++, and R2AUNet. Centerline annotation reduces interactive clicks by 63.5% and the average per-image annotation time by approximately 68%, while improving inter-annotator consistency (Cohen’s kappa: 0.913 to 0.967). Across all five architectures, mean intersection over union improves by 8.68%, with residual-recurrent networks (R2UNet and R2AUNet) gaining up to 17.32% and 13.98%, respectively. In the evaluated setting, the strategy improves annotation efficiency, inter-annotator consistency, and segmentation accuracy; its engineering feasibility was preliminarily assessed through a three-month deployment on a PCCP steel-cylinder leak detection line, during which the missed-detection rate was 0.90% (95% CI: 0.19–2.62%).
Related
- CHACAM: a cell-cell interaction-guided hierarchical attention model for high-precision cell identity annotation of scRNA-seq data in early C. elegans embryogenesis
- napari-TurboBox: Interactive Multi-Viewer Bounding Box Annotation for AI-Guided Volumetric Bioimaging
- GoCo: structure-guided conformal calibration for error-controlled Gene Ontology annotation release
- Combined foliar application of urea and potassium humate on nursery performance of guava cv. Sardar (L-49)
- Enhancing physiological performance and antioxidant defense in Mexican lime (Citrus aurantifolia) under water-deficit: synergistic effects of PGPB and KNO₃ application
BibTeX
@article{chen2026centerlineguided,
title = {{Centerline-Guided Weld Annotation for PCCP: Methods and Application Performance}},
author = {Zhenyu Chen and Kuanmin Mao and Junjie Tang and Jiaping Li},
journal = {Applied Sciences},
year = {2026},
month = oct,
doi = {10.3390/app162010267},
url = {https://doi.org/10.3390/app162010267},
note = {CC BY},
}
RIS
TY - JOUR
TI - Centerline-Guided Weld Annotation for PCCP: Methods and Application Performance
AU - Zhenyu Chen
AU - Kuanmin Mao
AU - Junjie Tang
AU - Jiaping Li
PY - 2026
DA - 2026/10/26
JO - Applied Sciences
DO - 10.3390/app162010267
UR - https://doi.org/10.3390/app162010267
AB - Semantic segmentation is a pivotal computer vision task, yet pixel-level annotation is labor-intensive and costly. For ribbon-like weld seams with ambiguous boundaries, polygon-based annotation suffers severe inter-annotator variability, an issue that has received limited attention in existing model-centric research. This study proposes a centerline-based annotation strategy for Prestressed Concrete Cylinder Pipe (PCCP) weld seams to improve efficiency, consistency, and model performance. Ground truth masks are generated via the Image-Guided Centerline-to-Mask Generation algorithm using slope-adaptive kernels to convert centerlines into pixel-level masks. During training and inference, multi-directional convolutional response maps are concatenated with RGB images as six-channel inputs and evaluated on UNet, R2UNet, AttUNet, UNet++, and R2AUNet. Centerline annotation reduces interactive clicks by 63.5% and the average per-image annotation time by approximately 68%, while improving inter-annotator consistency (Cohen’s kappa: 0.913 to 0.967). Across all five architectures, mean intersection over union improves by 8.68%, with residual-recurrent networks (R2UNet and R2AUNet) gaining up to 17.32% and 13.98%, respectively. In the evaluated setting, the strategy improves annotation efficiency, inter-annotator consistency, and segmentation accuracy; its engineering feasibility was preliminarily assessed through a three-month deployment on a PCCP steel-cylinder leak detection line, during which the missed-detection rate was 0.90% (95% CI: 0.19–2.62%).
SN - 2076-3417
ER -