Related Experiment Video
Updated: Oct 5, 2025
05:41
Author Spotlight: Integrating Ultrasound Imaging with Biochemical Markers for Thyroid Disease Diagnosis
Published on: February 9, 2024
766
3D reconstruction method based on second-order semiglobal stereo matching and fast point positioning Delaunay
Yongbing Xu1, Kaixiang Liu2, Jinyan Ni2
1Shandong Survey and Design Institute of Water Conservancy, Jinan, China.
Plos One
|January 25, 2022
Summary
This study introduces an improved 3D reconstruction method using enhanced stereo matching and Delaunay triangulation. The approach achieves higher accuracy and efficiency in generating 3D models from binocular vision data.
Area of Science:
- Computer Vision
- 3D Reconstruction
- Computational Geometry
Background:
- Binocular vision leverages parallax for 3D information acquisition, crucial for 3D reconstruction.
- Existing stereo matching methods face challenges in accuracy and efficiency.
- Point cloud noise and reconstruction speed impact 3D model quality.
Purpose of the Study:
- To enhance the accuracy and efficiency of 3D reconstruction from binocular vision.
- To develop a robust stereo matching algorithm for generating accurate disparity maps.
- To accelerate Delaunay triangulation for faster 3D surface reconstruction.
Main Methods:
- Improved second-order semiglobal matching with added smoothness constraints.
- Guided filtering applied to 3D point clouds for noise reduction and detail preservation.
- Accelerated Delaunay triangulation utilizing fast point location for surface reconstruction.
Main Results:
- The enhanced second-order semiglobal matching method demonstrated superior accuracy compared to existing stereo matching techniques.
- Guided filtering effectively removed noise while retaining critical details in the 3D point cloud.
- The proposed fast point location-based Delaunay triangulation significantly reduced reconstruction time.
Conclusions:
- The integrated method provides a more accurate and efficient solution for 3D reconstruction tasks.
- The improvements in stereo matching and triangulation contribute to higher quality 3D models.
- This approach offers a promising advancement for applications requiring precise and rapid 3D reconstruction.
