计算机科学
结构光三维扫描仪
帧速率
计算
计算机视觉
校准
光学
人工智能
图形处理单元
图像处理
投影(关系代数)
计算机图形学(图像)
算法
图像(数学)
物理
扫描仪
量子力学
操作系统
作者
Hieu Nguyen,Dung Nguyen,Zhaoyang Wang,Hien Kieu,Minh Le
出处
期刊:Applied Optics
[Optica Publishing Group]
日期:2014-11-11
卷期号:54 (1): A9-A9
被引量:129
摘要
In spite of the recent advances in 3D shape measurement and geometry reconstruction, simultaneously achieving fast-speed and high-accuracy performance remains a big challenge in practice. In this paper, a 3D imaging and shape measurement system is presented to tackle such a challenge. The fringe-projection-profilometry-based system employs a number of advanced approaches, such as: composition of phase-shifted fringe patterns, externally triggered synchronization of system components, generalized system setup, ultrafast phase-unwrapping algorithm, flexible system calibration method, robust gamma correction scheme, multithread computation and processing, and graphics-processing-unit-based image display. Experiments have shown that the proposed system can acquire and display high-quality 3D reconstructed images and/or video stream at a speed of 45 frames per second with relative accuracy of 0.04% or at a reduced speed of 22.5 frames per second with enhanced accuracy of 0.01%. The 3D imaging and shape measurement system shows great promise of satisfying the ever-increasing demands of scientific and engineering applications.
科研通智能强力驱动
Strongly Powered by AbleSci AI