High Speed 3D Shape Measurement with Temporal Fourier Transform Profilometry

Fourier transform profilometry (FTP) based on fringe projection has high speed and high procession advantages, so it is widely used in industrial inspection, solid modeling, biomedicine, machine vision and so on. In this paper, we present a novel method, in which we project a bi-color fringe pattern that is consisted of two interlaced RGB base color fringe patterns with π phase difference onto the object using digital light projector and the deformed color pattern is captured by color digital camera. Then the deformed color fringe pattern is decoded into two individual fringe patterns with π phase difference using image-separating technique. After modifying their averages and contrast, we subtract one of the two fringe patterns from the other to eliminate zero spectra. Compared with π phase-shifting technique, the obvious advantage of this method is that only one fringe pattern is needed to remove the zero spectra and raise the measurable slope of height variation nearly three times while no phase shifting device needed. Computer simulation and experiments verify our method.

Application of grayscale expansion for accuracy improvement in phase-measuring profilometry

Fourier transform profilometry (FTP) is one of the frequently adopted three-dimensional (3D) shape measurement methods due to its ability to recover single-shot 3D shapes, yet it is challenging to retrieve the absolute phase map solely from one single grayscale fringe image. This paper presents a computational framework that overcomes this limitation of FTP with digital fringe projection (DFP). By using geometric constraints, an absolute phase map can be retrieved point-by-point from one single grayscale fringe image. Experiments demonstrate the success of our proposed framework with single-shot absolute 3D shape measurement capability.

The measurement system currently used in Phase Measuring Profilometry(PMP) consists of a Digital Light Projector(DLP), a CCD camera, and a computer system. However, the inherent gamma nonlinearity of the DLP and CCD camera can affect the output with a nonsinusoidal fringe image. In the same tine the systematic noise is an important error source. Some conventional filtering algorithms may either make the fringe more blurring or be inefficient for fringe images. Eventually, the obtained fringe image is non-sinusoidal and with systematic random noise inevitably. Aimed at these problems, a pre-processing method for the fringe image is presented in this paper. Firstly, an anti-deforming light model is designed and then projected by the DLP. Through the gamma nonlinear response of the whole system, the waveform is corrected. Secondly, an improved orientation filter-based method is designed to overcome the systematic random noise and can achieve better effect than other algorithm does. Experiment, according to aforementioned two steps, is carried out and preferable fringe images can be gained. The fringe waveform is more close to the ideal sinusoidal wave. Also the systematic noise is reduced effectively while the fringe image is still clear. In the paper, two steps of this method are detailed and some experimental results are also reported.

In this paper, a whole-field and high-speed optical three-dimensional (3D) shape measurement for the vibration pattern of a drumhead based on Fourier Transform Profilometry (FTP) is presented and verified by experiment. A projected sinusoidal fringe pattern on the surface of measured drumhead is dynamically deformed with the vibration of the membrane and grabbed by a high-speed camera. The 3D shape deformation of the drumhead at each sampling instant can be recovered from demodulating the phase of this obtained sequence of fringe patterns. The vibration of the membrane of a Chinese drum has been measured with a high-speed sampling rate (1,000 fps) and a standard deviation (0.075 mm). This method should be helpful in studying the acoustic characteristic and the manufacture technique of the percussion instrument.

Fringe projection profilometry (FPP) is a three-dimensional (3D) shape measurement method that involves projecting fringe patterns onto the object. A phase map retrieved from these fringe images is used for reconstructing the 3D surface of the object. Fourier transform and phase-shifting are two of the widely used fringe analysis techniques for performing 3D shape measurement using FPP. Fourier transform profilometry (FTP) has the advantage of performing high-speed measurement due to its single-shot nature. However, the reconstructed 3D surface has artifacts and high noise, specifically on the edges. On the other hand, phase-shifting profilometry (PSP) has the advantage of higher accuracy (relatively less noise level) but compromises on measurement speed. In this research, we propose a deep learning method to enhance the quality of the FTP phase maps using an efficient deep learning model called phase map enhancement net (PMENet). PMENet takes an FTP phase map as input and predicts a high-quality phase map in a supervised manner by using the phase maps obtained from 18-step PSP as ground truth. The training dataset was generated using a virtual FPP system. Validations were conducted on both the simulated data (generated by virtual FPP system) and the real-world data. The experimental results demonstrate that the trained neural network model can successfully improve the quality of the 3D geometric reconstruction with FTP and reduced the mean and root-mean-square error by 66% and 43%, respectively.

Palmprint and hand shape, as two kinds of important biometric characteristics, have been widely studied and applied to human identity recognition. The existing research is based mainly on 2D images, which lose the third-dimensional information. The biological features extracted from 2D images are distorted by pressure and rolling, so the subsequent feature matching and recognition are inaccurate. This paper presents a method to acquire accurate 3D shapes of palmprint and hand by projecting full-field composite color sinusoidal fringe patterns and the corresponding color texture information. A 3D imaging system is designed to capture and process the full-field composite color fringe patterns on hand surface. Composite color fringe patterns having the optimum three fringe numbers are generated by software and projected onto the surface of human hand by a digital light processing projector. From another viewpoint, a color CCD camera captures the deformed fringe patterns and saves them for postprocessing. After compensating for the cross talk and chromatic aberration between color channels, three fringe patterns are extracted from three color channels of a captured composite color image. Wrapped phase information can be calculated from the sinusoidal fringe patterns with high precision. At the same time, the absolute phase of each pixel is determined by the optimum three-fringe selection method. After building up the relationship between absolute phase map and 3D shape data, the 3D palmprint and hand are obtained. Color texture information can be directly captured or demodulated from the captured composite fringe pattern images. Experimental results show that the proposed method and system can yield accurate 3D shape and color texture information of the palmprint and hand shape.

In the three-dimensional (3D) measurement, phase measuring profilometry (PMP) requires no less than three deformed sinusoidal patterns, limiting its applications to static scenarios only. In this paper, a stroboscopic online PMP for fast rotating object with binary dithered patterns is proposed. N frames of novel circular binary grating patterns are designed to measure a rotating object. A defocused digital light projector projects the designed patterns onto the object with stroboscopic method, the corresponding deformed patterns can be “frozen” by charge-coupled device (CCD) camera. Owing to the repeatability and periodicity of rotating movement, a synchronization control unit is used to keep the object at the fixed position in each “frozen” deformed pattern. Every projected pattern is shifted from the previous pattern by a factor of 2π/N. This method makes the images of the object appear stationary in the deformed patterns. The experimental results show the accurate 3D shape of fast rotating object can be reconstructed with the proposed method.

A real-time tricolor phase measuring profilometry (RTPMP) based on charge coupled device (CCD) sensitivity calibration is proposed. Only one colour fringe pattern whose red (R), green (G) and blue (B) components are, respectively, coded as three sinusoidal phase-shifting gratings with an equivalent shifting phase of 2π/3 is needed and sent to an appointed flash memory on a specialized digital light projector (SDLP). A specialized time-division multiplexing timing sequence actively controls the SDLP to project the fringe patterns in R, G and B channels sequentially onto the measured object in one over seventy-two of a second and meanwhile actively controls a high frame rate monochrome CCD camera to capture the corresponding deformed patterns synchronously with the SDLP. So the sufficient information for reconstructing the three-dimensional (3D) shape in one over twenty-four of a second is obtained. Due to the different spectral sensitivity of the CCD camera to RGB lights, the captured deformed patterns from R, G and B channels cannot share the same peak and valley, which will lead to lower accuracy or even failing to reconstruct the 3D shape. So a deformed pattern amending method based on CCD sensitivity calibration is developed to guarantee the accurate 3D reconstruction. The experimental results verify the feasibility of the proposed RTPMP method. The proposed RTPMP method can obtain the 3D shape at over the video frame rate of 24 frames per second, avoid the colour crosstalk completely and be effective for measuring real-time changing object.

Digital projectors based on liquid crystal or micro-mirror technologies provide higher spatial resolution than the CRTs typically used in flight-simulator applications. However, their temporal response is limited by long hold-times. Hold-time is the interval that pixels are activated during each frame, and it is related to both the design of the driver circuitry and the display itself. Long hold-times result in the blurring of moving imagery, which can affect the detection and identification of simulated targets as well as the perceived realism of simulated scenes. Described here are the results of an evaluation of several digital projectors that implement various techniques for reducing pixel hold-time. The evaluation included estimates of both spatial and temporal resolution, as well as perceptual tests used to assess the validity of those estimates. The techniques described here can be used to assess the potential utility of candidate projectors for a given simulator application. The following projectors were tested: ferro-electric liquid crystal on silicon (FLCoS), liquid crystal on silicon (LCoS), liquid crystal (LCD), digital light projector (DLP, a digital micro-mirror device, or DMD), and cathode ray tube (CRT). The hold-time of the FLCoS and LCD projectors was varied using mechanical shutters, and the hold-time of the DLP projector was varied using an electronic shutter. The CRT and LCD projectors were used mainly for comparison and so their hold-times were not varied.

Free-space optical measurement systems can have a direct impact on evaluation systems operational in propagation paths. During propagation via optical fibers, light suffers scattering or interference, causing some output signal loss with an uncertainty outcome. Therefore, this study aims to explore the instant decisions related to the use of single- and multi-mode fiber optics and how they affect the gathering of data from high-speed optical measurement instrument links. The study also seeks to address a number of design methodology aspects and the empirical outcomes related to a surface topography measurement sensor based on fiber optics capable of surface roughness or step-height measurement. The study suggests that the Fourier transform profilometry method (FTP) can overcome the disadvantages of optical metrology sensors (e.g., bulkiness, challenging set-up, high costs, and low speed). However, despite eliminating vertical height problems, the Fourier transform profilometry (FTP) did have some shortcomings for every outcome related to core variables, including the dispersive optical fiber link sensor. The synthetic wavelength method enabled the dispersive optical fiber link sensor to calculate the vertical step height of the selected sample (1 µm). There was improved step-height repeatability, with satisfactory from 20 to 18 nm outcome improvement range. Additional investigations are necessary to establish the compatibility of single- or multi-mode optical fiber sensors with particular instruments, especially those currently preferred for embedded metrology applications.

In this paper, we propose a high-speed 3D shape measurement technique based on the optimized composite fringe patterns and stereo-assisted structured light system. Stereo phase unwrapping, as a new-fashioned method for absolute phase retrieval based on the multi-view geometric constraints, can eliminate the phase ambiguities and obtain a continuous phase map without projecting any additional patterns. However, in order to ensure the stability of phase unwrapping, the period of fringe is generally around 20, which limits the accuracy of 3D measurement. To solve this problem, we develop an optimized method for designing the composite pattern, in which the speckle pattern is embedded into the conventional 4-step phase-shifting fringe patterns without compromising the fringe modulation, and thus the phase measurement accuracy. We also present a simple and effective evaluation criterion for the correlation quality of the designed speckle pattern in order to improve the matching accuracy significantly. When the embedded speckle pattern is demodulated, the periodic ambiguities in the wrapped phase can be eliminated by combining the adaptive window image correlation with geometry constraint. Finally, some mismatched regions are further corrected based on the proposed regional diffusion compensation technique (RDC). These proposed techniques constitute a complete computational framework that allows to effectively recover an accurate, unambiguous, and distortion-free 3D point cloud with only 4 projected patterns. Experimental results verify that our method can achieve high-speed, high-accuracy, robust 3D shape measurement with dense (64-period) fringe patterns at 5000 frames per second.

With more than three decades of development, three-dimensional optical measurement technology has reached a mature stage in commercial applications, meanwhile new ones have continually arisen. Due to the development of Charge Coupled Device (CCD) array camera and digital projection technology, the applications of Phase Measurement Profilometry (PMP) become more and more broad. Among these, dual-frequency grating method has drawn many attentions because of its simplicity in principle and optical path, low requirement of equipment, high accuracy and level of automation comparing with other methods. The phase calculation is one of the key technologies in PMP. However, phase unwrapping algorithm in PMP is a difficult issue. A lot of new algorithm have been proposed, but neither one can solve all the problems, so how to set up new phase unwrapping algorithm becomes urgent. In this chapter, we systematically investigate the phase unwrapping method in dual-frequency grating method, and experimentally set up the system. To verify our method, we experimentally measure a three dimensional object which possesses complicated stair shapes on its surface. The results show that our dual-frequency grating method could achieve phase unwrapping without doing conventional phase unwrapping calculations, and it could also detect the detailed stair shapes on the surface of this three dimensional object automatically.

A fast phase measuring profilometry (PMP) using structured illumination with a digital micro-mirror device (DMD) has been extensively studied because of its advantage of the programmable projection. In the traditionary PMP, the phase-shift device is mechanical based and is controlled by step-motor which may cause phase errors more or less and several frames of deformed fringe patterns should be capture, which may take a longer time. In the system a DMD is used for projecting only one complex color distribution on the surface of the measured object. In the distribution three interlaced RGB based tricolor sinusoidal chroma two-dimensional fringe patterns are synchronously coded by which the phase difference of the patterns are 120 degree one another. A 24-bit color CCD digital camera is used as the detector to capture only one frame of the deformed complex color distributions. By using image separating techniques the three deformed interlaced tricolor sinusoidal chroma two-dimensional fringe patterns can be decoded individually. Using the 3-step phase shift algorithm the three-dimensional topography can be reconstructed from these decoded deformed fringe patterns. To improve the phase measurement accuracy the RGB tricolor based chroma transfer function (CTF) of the measuring system is introduced. The experiments show that by a reverse correction method with the CTF the phase measurement accuracy has been improved obviously.

In this presentation, we provide a thorough overview of recent advances in 3D surface imaging technologies. We focus particularly on these non‐ contact 3D surface measurement techniques that require structured illumination. The high speed and high resolution pattern projection capability offered by the digital light projection (DLP) technology, together with the recent advent of imaging sensor technologies, may enable new generation systems for of 3D surface measurement applications that provide much better functionality and performance than existing ones, in terms of speed, accuracy, resolution, size, cost, and ease of use. Performance indexes of 3D imaging system in general are discussed and various 3D surface imaging schemes are categorized, illustrated, and compared. Numerous medical applications of 3D technologies are discussed with several examples.