不连续性分类
相位展开
稳健性(进化)
结构光三维扫描仪
计算机科学
轮廓仪
投影(关系代数)
算法
影子(心理学)
计算机视觉
人工智能
结构光
迭代重建
相(物质)
相位恢复
光学
错误检测和纠正
图像处理
干涉测量
过程(计算)
图像(数学)
实体造型
一级
投影法
绝对相位
作者
Jiayi Qin,Yansong Jiang,Yiping Cao
标识
DOI:10.1109/tip.2025.3607619
摘要
Phase unwrapping is a critical step in fringe projection profilometry, essential for achieving accurate and efficient three-dimension (3D) imaging. Temporal phase unwrapping is the most widely utilized to improve robustness and the reconstruction quality. Unfortunately, due to abrupt phase discontinuities at boundaries, misalignment between the wrapped phases, and unreliable shadow regions, fringe order errors may occur. To address these challenges, this study presents a generalized multi-feature-guided progressive order correction algorithm (GMP-OCA) for high-quality 3D imaging. The algorithm integrates global coarse detection, incremental line-wise optimization, and regional precision scanning to progressively correct fringe orders. Static and dynamic experimental results demonstrate that GMP-OCA effectively eliminates the systematic errors inherent in various phase unwrapping methods, producing high-quality 3D imaging results.
科研通智能强力驱动
Strongly Powered by AbleSci AI