Related Experiment Video
Updated: Aug 15, 2026
10:25
Easy and Accurate Mechano-profiling on Micropost Arrays
Published on: November 17, 2015
HNC: a heatmap-guided normal centroid method for multi-line laser stripe centerline extraction with geometry-based
Optics Express
|August 14, 2026
Summary
This study introduces a new heatmap-guided normal centroid (HNC) method for precise multi-line laser stripe centerline extraction in 3D measurement. The HNC method significantly improves accuracy and stability, outperforming existing algorithms in industrial applications.
Area of Science:
- Computer Vision
- Metrology
- 3D Measurement
Background:
- Industrial 3D measurement faces challenges with multi-line laser stripe degradation.
- Specular reflection, noise, and surface variations cause unstable centerline extraction.
Purpose of the Study:
- To develop a robust subpixel extraction method for multi-line laser stripe centerlines.
- To enhance the accuracy and stability of 3D measurement in industrial settings.
Main Methods:
- A physics-based simulation in Blender was used for generating ground truth data.
- A lightweight regression network (EDR-Net) predicts heatmaps for centerline localization.
- Subpixel coordinates are decoded using structure-tensor normal estimation and centroid refinement.
Main Results:
- The HNC method achieved high performance in heatmap prediction (Dice: 96.67%, IoU: 93.56%).
- It demonstrated superior centerline extraction stability on real images, outperforming other methods.
- Significant error reductions were observed in train wheel measurements (e.g., flange thickness by 71.43%).
Conclusions:
- The proposed heatmap-guided normal centroid (HNC) method offers robust and efficient multi-line laser stripe centerline extraction.
- This technique has strong potential for high-precision industrial 3D measurement applications.
- The method addresses key challenges in laser stripe analysis for metrology.
