TY - JOUR
T1 - “3-2-1” PMP
T2 - Adding an Extra Pattern to Dual-Band Phase-Shift Profilometry for Higher Precision 3D Imaging
AU - Wang, Jian
AU - Zhang, Guanyu
AU - Guo, Wei
AU - Zhang, Jun
AU - Zhang, Zonghua
AU - Chen, Liangzhou
AU - Li, Tukun
AU - Jiang, Jane
N1 - Funding Information:
This project is supported by National Natural Science Foundation of China (No. 52075206, U2341275), National Key Research & Development Program (No. 2023YFB4606000), and Knowledge Innovation Program of Wuhan-Basic Research.
Publisher Copyright:
© 1963-2012 IEEE.
PY - 2024/7/24
Y1 - 2024/7/24
N2 - Phase measurement profilometry, or PMP, is undergoing rapid development at present. Combined with defocused projection and advanced multiband phase shift (PS) algorithms, such as heterodyne and number theory, PMP with over thousand framerates has shown great potential in 3-D dynamic scene monitoring, e.g., manufacturing process monitoring and industrial digital twins. PMP's outstanding advantages include a combination of high lateral resolution, axial precision, and speed; however, the widely used dual-band PS algorithms have limited axial precision compared to low-efficiency tri-band algorithms. In this article, we propose '3-2-1' PMP, which optimizes the existing dual-band approach by adding an extra fringe pattern for projection acquisition to achieve similar precision performance as tri-band PMP. Associated stabilization algorithms in dealing with fringe blur, nonlinear camera responses, phase-jump errors, and related impact factors have also been investigated. Comparative simulation and experiments show that the '3-2-1' approach produces higher precision performance, with a 42%-85% root-mean-square error lower than general dual-band PS methods and famous Fourier transform profilometry (μ FTP), but using an equivalent projection pattern number of 6 or less.
AB - Phase measurement profilometry, or PMP, is undergoing rapid development at present. Combined with defocused projection and advanced multiband phase shift (PS) algorithms, such as heterodyne and number theory, PMP with over thousand framerates has shown great potential in 3-D dynamic scene monitoring, e.g., manufacturing process monitoring and industrial digital twins. PMP's outstanding advantages include a combination of high lateral resolution, axial precision, and speed; however, the widely used dual-band PS algorithms have limited axial precision compared to low-efficiency tri-band algorithms. In this article, we propose '3-2-1' PMP, which optimizes the existing dual-band approach by adding an extra fringe pattern for projection acquisition to achieve similar precision performance as tri-band PMP. Associated stabilization algorithms in dealing with fringe blur, nonlinear camera responses, phase-jump errors, and related impact factors have also been investigated. Comparative simulation and experiments show that the '3-2-1' approach produces higher precision performance, with a 42%-85% root-mean-square error lower than general dual-band PS methods and famous Fourier transform profilometry (μ FTP), but using an equivalent projection pattern number of 6 or less.
KW - 3D imaging
KW - Accuracy
KW - calibration
KW - Dual band
KW - Encoding
KW - fringe order error
KW - Fringe projection profilometry
KW - Heuristic algorithms
KW - pattern optimization
KW - Phase measurement
KW - Sensors
KW - Three-dimensional displays
KW - fringe order error (FOE)
KW - 3-D imaging
KW - fringe projection profilometry
UR - http://www.scopus.com/inward/record.url?scp=85198712124&partnerID=8YFLogxK
U2 - 10.1109/TIM.2024.3428592
DO - 10.1109/TIM.2024.3428592
M3 - Article
AN - SCOPUS:85198712124
VL - 73
JO - IEEE Transactions on Instrumentation and Measurement
JF - IEEE Transactions on Instrumentation and Measurement
SN - 0018-9456
M1 - 10599300
ER -