Related Experiment Video
Updated: May 25, 2025
10:53
Image-guided, Laser-based Fabrication of Vascular-derived Microfluidic Networks
Published on: January 3, 2017
9.8K
Laser Stripe Centerline Extraction Method for Deep-Hole Inner Surfaces Based on Line-Structured Light Vision Sensing
Huifu Du1, Daguo Yu1,2, Xiaowei Zhao1
1School of Mechanical Engineering, North University of China, Taiyuan 030051, China.
Sensors (Basel, Switzerland)
|February 26, 2025
Summary
This study introduces a novel method using minimum spanning tree (MST) and depth-first search (DFS) for accurate laser stripe centerline extraction from deep hole inner surfaces, overcoming noise and burr challenges.
Area of Science:
- Engineering
- Computer Science
- Optics
Background:
- Traditional image processing methods struggle with noise and burrs on complex surfaces.
- Accurate extraction of laser stripe centerlines is crucial for 3D reconstruction and parameter calculation.
Purpose of the Study:
- To develop a robust point cloud post-processing method for precise laser stripe centerline extraction from deep hole inner surfaces.
- To enhance accuracy and reliability compared to existing techniques.
Main Methods:
- Utilized 360° structured light for illumination and a sensor for image capture.
- Employed the Steger algorithm for sub-pixel point cloud extraction.
- Applied minimum spanning tree (MST) for point cloud connectivity and depth-first search (DFS) for pathfinding and noise reduction.
Main Results:
- Achieved high extraction accuracy with a Dice Similarity Coefficient (DSC) approaching 1.
- Reported a maximum Hausdorff Distance (HD) of 3.3821 pixels.
- Demonstrated superior performance over previous methods in complex environments.
Conclusions:
- The proposed MST and DFS-based method offers an efficient and reliable solution for complex laser stripe extraction.
- Provides a strong data foundation for subsequent 3D reconstruction and feature analysis.
- Significantly improves precision in challenging industrial inspection scenarios.
