Before downloading any paper or material from our website, please carefully read and understand the Copyright disclaimer.
Journal Papers [2017] [2016] [2015] [2014] [2013] [2012] [2011] [2010] [Earlier]
Books / Book Chapters / Dissertations
X. Chen, M. Grossard, N. Kubota, D. Wollherr, S. X. Yang, and S. Zhang, "Introduction to the focused section on sensing and perception for autonomous and networked robotics," Int. J. Intell. Robot. Applic. 1 (4), 369–371 (2017); (Focused Section Guest Editorial) doi:10.1007/s41315-017-0040-8
Y. Wang, C. Jiang, and S. Zhang, "Double-pattern triangular pulse width modulation technique for high-accuracy high-speed 3D shape measurement," Opt. Express 25(24), 30177-30188 (2017); doi:10.1364/OE.25.03177
Using 1-bit binary patterns for 3D shape measurement has been demonstrated advantageous over using 8-bit sinusoidal patterns in terms of achievable speeds. However, the phase quality generated by binary pattern(s) typically is not high if only a small number of phase-shifted patterns is used. This paper proposes a method to improve the phase quality by representing each pattern with the difference of two binary patterns: the first binary pattern is generated by triangular pulse width modulation (TPWM) technique, and the second being $\pi$ shifted from the first pattern is also generated by TPWM technique. The phase is retrieved by applying a three-step phase-shifting algorithm to the difference patterns. Through optimizing the modulation frequency of the triangular carrier signal, we demonstrate that high-quality phase can be generated for a wide range of fringe periods (e.g., from 18 to 1140 pixels) with only six binary patterns. Since only 1-bit binary patterns are required for 3D shape measurement, this paper will present a real-time 3D shape measurement system that can achieve 30 Hz.
[106] T. Bell, B. Vlahov, J.P. Allebach, and S. Zhang, "Three-dimensional range geometry compression via phase encoding," Appl. Opt., 56(33), 9285-9292, (2017); doi: 10.1364/AO.56.009285
One of the state-of-the-art methods for three-dimensional (3D) range geometry compression is to encode 3D data within a regular 24-bit 2D color image. However, most existing methods use all three color channels to solely encode 3D data, leaving no room to store other information (e.g., texture) within the same image. This paper presents a novel method which utilizes geometric constraints, inherent to the structured light 3D scanning device, to reduce the amount of data which need be stored within the output image. The proposed method thus only requires two color channels to represent 3D data, leaving one channel free to store additional information (such as a texture image). Experimental results verify the overall robustness of the proposed method. For example, a compression ratio of 3038:1 can be achieved, versus the STL format, with a root-mean-square (RMS) error of 0.47% if the output image is compressed with JPEG 80%.
B. Li and S. Zhang, “Superfast, high-resolution absolute 3D recovery of a stabilized flapping flight process,” Opt. Express, 25(22), 27270-27282 (2017); doi:10.1364/OE.25.027270
Scientific research of a stabilized flapping flight process (e.g. hovering) has been of great interest to a variety of fields including biology, aerodynamics and bio-inspired robotics. Different from the current passive photogrammetry based methods, the digital fringe projection (DFP) technique has the capability of performing dense superfast (e.g. kHz) 3D topological reconstruction with the projection of defocused binary patterns, yet it is still a challenge to measure a flapping flight process with the presence of rapid flapping wings. This paper presents a novel absolute 3D reconstruction method for a stabilized flapping flight process. Essentially, the slow motion parts (e.g. body) and the fast-motion parts (e.g. wings) are segmented and separately reconstructed with phase shifting techniques and Fourier transform, respectively. The topological relations between the wings and the body are utilized to ensure absolute 3D reconstruction. Experiments demonstrate the success of our computational framework by testing a flapping wing robot at different flapping speeds.
C. Jiang and S. Zhang, “Absolute three-dimensional shape measurement with two-frequency square binary patterns,” Appl. Opt., 56(31), 8710-8718 (2017); doi:10.1364/AO.56.008710
This paper presents a novel method to achieve absolute three-dimensional (3D) shape measurement solely using square binary patterns. This method uses six patterns: three low-frequency phase-shifted patterns and three phase-shifted high-frequency patterns. The phase obtained from low-frequency phase temporally unwraps the phase obtained from high-frequency patterns. The projector is defocused such that the high-frequency patterns produce high-quality phase, but the phase retrieved from low-frequency patterns has large harmonic error that fails two-frequency temporal phase unwrapping process. In this paper, we develop a computational framework to address the challenge. The proposed computational framework includes four major approaches to alleviate the harmonic error problem: i) use more than one period of low-frequency patterns enabled by geometric constraint-based phase unwrapping method; ii) artificially apply a large Gaussian filter to low frequency patterns before phase computation; iii) create an error lookup table (LUT) to compensate for harmonic error; and iv) develop a boundary error correction method to alleviate problems associated with filtering. Both simulation and experimental results demonstrated the success of the proposed method.
[102] H. Sheng, J. Xu, and S. Zhang, "Dynamic projection theory for fringe projection profilometry," Appl. Opt., 56(30), 8452-8460 (2017); doi: 10.1364/AO.56.008452
Fringe projection profilometry (FPP) has been widely used for 3D reconstruction, surface measurement and reverse engineering. However, fringe projection profilometry is prone to overexposure if objects have a wide range of reflectance. In this paper, we propose a dynamic projection theory based on fringe projection profilometry to rapidly measure the overexposed region with an attempt to conquer this challenge. This theory modifies the projected fringe image to the next better measurement based on the feedback provided by the previously captured image intensity. Experiments demonstrated that the number of overexposed points can be drastically reduced after one or two iterations. Compared with the state-of-the-art methods, our proposed dynamic projection theory measures the overexposed region quickly and effectively, and thus broadens the applications of fringe projection profilometry.
C. Jiang and S. Zhang, “Absolute phase unwrapping for dual-camera system without embedding statistical features,” Opt. Eng. 56(9), 094114 (2017), doi: 10.1117/1.OE.56.9.094114.
This paper proposes an absolute phase unwrapping method for 3D measurement that uses two cameras and one projector. On the left camera image, each pixel has one wrapped phase value which corresponds to multiple projector candidates with different absolute phase values. We use geometric relationship of the system to map projector candidates into right camera candidates. By applying a series of candidate rejection criteria, a unique correspondence pair between two camera images can be determined. Then the absolute phase is obtained by tracing the correspondence point back to projector space. Experimental results demonstrate that the proposed absolute phase unwrapping algorithm can successfully work on both complex geometry and multiple isolated objects measurement.
[100] B. Li, T. Bell, and S. Zhang, "Computer-aided-design (CAD) model assisted absolute three-dimensional shape measurement," Appl. Opt. 56(24), 6770-6776 (2017); doi: 10.1364/AO.56.006770
Conventional three-dimensional (3D) shape measurement methods are typically generic to all types of objects. Yet, for many measurement conditions, such level of generality is inessential when having the pre-knowledge of object geometry. This paper introduces a novel adaptive algorithm for absolute 3D shape measurement with the assistance of the object CAD model. The proposed algorithm includes the following major steps: 1) export the 3D point cloud data from the CAD model; 2) transform the CAD model into the camera perspective; 3) obtain wrapped phase map from three phase-shifted fringe images; 4) retrieve absolute phase and 3D geometry assisted by CAD model. We demonstrate that if object CAD models are available, such algorithm is efficient in recovering absolute 3D geometries of both simple and complex objects with only three phase-shifted fringe images.
[99] J. -S. Hyun, B. Li, and S. Zhang, "High-speed high-accuracy three-dimensional shape measurement using digital binary defocusing method versus sinusoidal method," Opt. Eng. 56(7), 074102 (2017).
This paper presents our research findings on high-speed high-accuracy 3D shape measurement using digital light processing (DLP) technologies. In particular, we compare two different sinusoidal fringe generation techniques using the DLP projection devices: direct projection of 8-bit computer generated sinusoidal patterns (a.k.a., the sinusoidal method), and the creation of sinusoidal patterns by defocusing binary patterns (a.k.a., the binary defocusing method). This paper mainly examines their performance on high-accuracy measurement applications under precisely controlled settings. Two different projection systems were tested in this study: the commercially available inexpensive projector, and the DLP development kit. Experimental results demonstrated that the binary defocusing method always outperforms the sinusoidal method if a sufficient number of phase-shifted fringe patterns can be used.
Y. An and S. Zhang, "Three-dimensional absolute shape measurement by combining binary statistical pattern matching with phase-shifting methods," Appl. Opt., (2017); (accepted)
This paper presents a novel method that leverages the stereo geometric relationship between projector and camera for absolute phase unwrapping on a standard one-projector and one-camera structured light system. Specifically, we use only one additional binary random image and the epipolar geometric constraint to generate a coarse correspondence map between projector and camera images. The coarse correspondence map is further refined by using the wrapped phase as a constraint. We then use the refined correspondence map to determine a fringe order for absolute phase unwrapping. Experimental results demonstrated the success of our proposed method.
A. Wan, J. Xu, H. Chen, S. Zhang, and K. Chen, "Optimal path planning and control of assembly robots for hard measuring easy-deformation assemblies", IEEE Trans. Mechatronics, 22(4), 1600-1609, (2017); doi:10.1109/TMECH.2017.2671342
Assembly robots are widely used in the electronics and automotive industries. However, assembly robots still face formidable challenges for assembling large-scale heavy-weight components such as the tail of the plane. First, the largescale component is difficult to measure; thus, the optimal assembly path is difficult to obtain. To this end, a learning from demonstration-based optimal path planning method is developed and implemented. Second, the deformation caused by a heavy-weight component will lead to a large motion error and could cause damage to the component. To solve this problem, a Gaussian process regression (GPR)-based deformation prediction and compensation method is presented to improve the robot motion accuracy. The simulation results show that the proposed GPR-based deformation compensation method can achieve high accuracy. An experimental prototype was developed to evaluate the proposed methods, and the results demonstrate the effectiveness of the proposed methods. Therefore, the proposed methods provide a path toward hard-measuring easy deformation assembly task.
[95] H. Yun, B. Li, and S. Zhang, "Pixel-by-pixel absolute three-dimensional shape measurement with modified Fourier transform profilometry", Appl. Opt., 56(5), 1472-1480, (2017); doi: 10.1364/AO.56.001472
Single-pattern Fourier transform profilometry (FTP) method and double-pattern modified FTP method have great value on high-speed three-dimensional (3D) shape measurement, yet it is difficult to retrieve absolute phase pixel by pixel. This paper presents a method that can recover absolute phase pixel by pixel for the modified FTP method. The proposed method uses two images with different frequencies, and the recovered low frequency phase is used to temporally unwrap the high-frequency phase pixel by pixel. This paper also presents the computational framework to reduce noise impact for robust phase unwrapping. Experiments demonstrate the success of the proposed absolute phase recovery method using only two fringe patterns.
[93] J. -S. Hyun and S. Zhang, "Superfast 3D absolute shape measurement using five binary patterns," Opt. Laser Eng., 90, 217-224, 2017; 10.1016/j.optlaseng.2016.10.017
This paper presents a method that recovers high-quality 3D absolute coordinates point by point with only five binary patterns. Specifically, three dense binary dithered patterns are used to compute the wrapped phase; and the average intensity is combined with two additional binary patterns to determine fringe order pixel by pixel in phase domain. The wrapped phase is temporarily unwrapped point by point by referring to the fringe order. We further developed a computational framework to reduce random noise impact due to dithering, defocusing and random noise. Since only five binary fringe patterns are required to recover one 3D frame, extremely high speed 3D shape measurement can be achieved. For example, we developed a system that captures 2D images at 3,333Hz, and thus performs 3D shape measurement at 667 Hz.
[92] R. Chen, J. Xu, S. Zhang, H. Chen, Y. Guan, and K. Chen, "A self-recalibration method based on scale-invariaant registration for structured light measurement systems," Opt. Laser Eng., 88, 75-81 (2017); doi:10.1016/j.optlaseng.2016.07.003
The measurement accuracy of structured light measurement depends on delicate offline calibration. However, in some practical applications, the system is supposed to be reconfigured so frequently to track the target that an online calibration is required. To this end, this paper proposes a rapid and autonomous self-recalibration method. For the proposed method, first, the rotation matrix and normalized translation vector are attained from the fundamental matrix; second, the scale factor is acquired based on scale-invariant registration such that the actual translation vector is obtained. Experiments have been conducted to verify the effectiveness of our proposed method, and the results indicate a high degree of accuracy.
School of Mechanical Engineering, Purdue University. West Lafayette, Indiana. 47907.
