Nanomanufacturing and Metrology · 2024 · 32 citations · 64 references
EngineeringManufacturing EngineeringGeometric ModelingMachine VisionOptics3D Imaging3D ReconstructionForm MeasurementOptical MetrologyPhotogrammetryPhysical Optics3D MeasurementData AccuracyPhase InformationSpecular ReflectanceSpecular ReflectionSurface ReflectionsThree-dimensional MeasurementThree-dimensional Shape Measurement
Abstract Full-field three-dimensional (3D) measurement technology based on phase information has become an indispensable part of geometric dimension measurement in modern scientific research and engineering applications. This field has been developing and evolving for the study of highly reflective phenomena, diffuse reflections, and specular surfaces, and many novel methods have emerged to increase the speed of measurements, enhance data accuracy, and broaden the robustness of the system. Herein, we will discuss the latest research progress in full-field 3D shape measurement based on phase information systematically and comprehensively. First, the fundamentals of 3D shape measurement based on phase information are introduced, namely, phase-shifting and transform-based methods. Second, recent technological innovations are highlighted, including increases in measurement speed and automation and improvements in robustness in complex environments. In particular, the challenges faced by these technological advances in solving highly dynamic, composite surface measurement problems are presented, i.e., with multiexposure techniques proposed for high dynamics that extend the dynamic range of the camera to reduce the effects of overexposure but increase the cost of time and have high hardware requirements, fringe adaptive techniques that overcome light variations but are computationally complex, and multipolarized camera techniques that reduce the effects of light variations but are sensitive to the light source. Third, the phase-shifting method combined with coding is proposed to improve the measurement speed, but the accuracy is slightly reduced. Deep learning techniques are proposed to cope with measurements in complex environments, but the dataset computation process is cumbersome. Finally, future research directions are suggested, and the challenges are presented. Overall, this work provides a reference for researchers and engineers.
64
Fringe projection profilometry by conducting deep learning from its digital twin
Yi Zheng, Shaodong Wang, Qing Li et al. · Optics Express · 2020 · 138 citations · Full text
High-speed high dynamic range 3D shape measurement based on deep learning
Liang Zhang, Qian Chen, Chao Zuo et al. · Optics and Lasers in Engineering · 2020 · 125 citations
Real-time high dynamic range 3D measurement using fringe projection
Liang Zhang, Qian Chen, Chao Zuo et al. · Optics Express · 2020 · 61 citations · Full text