Related Experiment Video
Updated: Mar 12, 2026
14:14
Quantification of Strain in a Porcine Model of Skin Expansion Using Multi-View Stereo and Isogeometric Kinematics
Published on: April 16, 2017
12.0K
Seamless images stitching for 3D human models
Chao Lai1, Fangzhao Li1, Shiyao Jin1
1National Laboratory for Parallel and Distributed Processing, National University of Defense Technology, Changsha, China ; College of Computer, National University of Defense Technology, Changsha, China.
Springerplus
|November 1, 2016
Summary
This study presents a novel image stitching method for seamless 3D human model texturing. The approach optimizes texture mapping, improving realism in computer graphics and vision applications.
Area of Science:
- Computer Graphics
- Computer Vision
- 3D Reconstruction
Background:
- Realistic 3D human model reconstruction is crucial for computer graphics and vision.
- Texturing is a key stage in 3D model reconstruction, requiring seamless integration of color information.
Purpose of the Study:
- To develop an effective method for generating seamless, integrated, and smooth textures on 3D human models.
- To address challenges in texture mapping caused by depth data inaccuracies and multi-view misregistration.
Main Methods:
- Formulated texture mapping as an image stitching optimization problem using a Markov Random Field.
- Employed alpha-expansion optimization with 2D translation for adaptive mesh boundary color matching.
- Utilized Poisson blending in the gradient domain to resolve illumination variations.
- Applied K-Nearest Neighbor algorithm on a parameterized 2D surface projection to fill blank regions.
Main Results:
- The proposed method successfully generates seamless and smooth textures on 3D human models.
- The technique effectively compensates for misalignments and illumination variations between adjacent images.
- Quantitative and visual comparisons show superior performance against three advanced methods.
Conclusions:
- The developed image stitching optimization method provides superior texture quality for 3D human models.
- This approach enhances the realism and visual fidelity of reconstructed 3D human models.
- The method offers a robust solution for texture mapping challenges in 3D reconstruction.
