In this work, we demonstrate that by using a quad-camera multi-view fringe projection system and carefully arranging the relative spatial positions between the cameras and the projector, it becomes possible to completely eliminate the phase ambiguities in conventional three-step PSP patterns with high-fringe-density without projecting any additional patterns or embedding any auxiliary signals. Benefit from the position-optimized quadcamera system, stereo phase unwrapping can be efficiently and reliably performed by flexible phase consistency checks. Besides, redundant information of multiple phase consistency checks is fully used through a weighted phase difference scheme to further enhance the reliability of phase unwrapping. This paper explains the 3D measurement principle and the basic design of quad-camera system, and finally demonstrates that the resultant dynamic 3D sensing system can realize real-time 3D reconstruction with a depth precision of 50 μm.
Phase-shifting profilometry (PSP) is a widely used approach to high-accuracy three-dimensional shape measurements. However, when it comes to moving objects, phase errors induced by the movement often result in severe artifacts even though a high-speed camera is in use. From our observations, there are three kinds of motion artifacts: motion ripples, motion-induced phase unwrapping errors, and motion outliers. We present a novel motion-compensated PSP to remove the artifacts for dynamic measurements of rigid objects. The phase error of motion ripples is analyzed for the phaseshifting algorithm and is compensated using the statistical nature of the fringes. The phase unwrapping errors are corrected exploiting adjacent reliable pixels, and the outliers are removed by comparing the original phase map with a smoothed phase map. Compared with the three-step PSP, our method can improve the accuracy significantly for objects in motion
A distortion-free telecentric camera dose not have an optical center because of the orthogonal projection. However, the position of optical center should be known when the lens distortion is considered. Since the full-scale parameters are derived through an iterative algorithm, critical initial values of the optical center should be provided to avoid being trapped in local minima. In this paper, we proposed a two-step algorithm to estimate the optical center as the trustworthy initial value for the subsequent iteration process. The first step is directly calculating the pixel coordinates of the lateral distortion center using the extracted control points. The second step is optimizing both lateral and tangential coefficients considering the properties of the affine transformation in the imaging process. The effectiveness of our proposed method is proven by the measurement results using a new developed microscopic telecentric stereovision system.
An improved bi-frequency phase-shifting technique based on a multi-view fringe projection system is proposed, which significantly enhances the measurement precision without compromising the measurement speed. Based on the geometric constraints in a multi-view system, the unwrapped phase of the low-frequency (10-period) fringes can be obtained directly, which serves as a reference to unwrap the high-frequency phase map with a total number of periods of up to 160. Experiments on both static and dynamic scenes are performed, verifying that our method can achieve real-time and high-precision 3-D measurement with a precision of about 50 μm.
Microscopic 3-D shape measurement can supply accurate metrology of the delicacy and complexity of MEMS components of the final devices to ensure their proper performance. Fringe projection profilometry (FPP) has the advantages of noncontactness and high accuracy, making it widely used in 3-D measurement. Recently, tremendous advance of electronics development promotes 3-D measurements to be more accurate and faster. However, research about real-time microscopic 3-D measurement is still rarely reported. In this work, we effectively combine optimized binary structured pattern with number-theoretical phase unwrapping algorithm to realize real-time 3-D shape measurement. A slight defocusing of our proposed binary patterns can considerably alleviate the measurement error based on phase-shifting FPP, making the binary patterns have the comparable performance with ideal sinusoidal patterns. Real-time 3-D measurement about 120 frames per second (FPS) is achieved, and experimental result of a vibrating earphone is presented.
We propose an absolute 3D micro surface profile measurement technique based on a Greenough-type stereomicroscope. The camera and the projector are fixed on the stereomicroscope, facilitating a flexible 3D measurement of objects with different heights. Experiments of both calibration and measurements are conducted, and the results show that our proposed method works well for measuring different types of geometry like spheres, ramps and planes etc. The reconstruction accuracy can achieve 4.8 μm with a measurement depth about 3 mm.
We introduce a high-speed 3-D shape measurement technique based on composite phase-shifting fringes and a stereo camera system. Epipolar constraint is adopted to search the corresponding point independently without additional images. Meanwhile, by analysing the 3-D position and the main wrapped phase of the corresponding point, pairs with an incorrect 3-D position or considerable phase difference are effectively rejected. Then all the qualified corresponding points are corrected, and the unique one as well as the related period order is selected through the embedded triangular wave. Finally, considering that some points can only be captured by a single camera in some shading areas, the final period order of these points in one camera and the one of their corresponding points in another camera always have different values, so left-right consistency check is employed to eliminate those erroneous period orders in this case. Several experiments on both static and dynamic scenes are performed, verifying that our method can achieve a speed of 120 frames per second (fps) with 25-period fringe patterns for fast, dense, and accurate 3-D measurement.
We demonstrate lens-less quantitative phase microscopy and diffraction tomography based on a compact on-chip platform, using only a CMOS image sensor and a programmable color LED array. Based on multi-wavelength transport-of- intensity phase retrieval and multi-angle illumination diffraction tomography, this platform offers high quality, depth resolved images with a lateral resolution of ∼3.7μm and an axial resolution of ∼5μm, over wide large imaging FOV of 24mm2. The resolution and FOV can be further improved by using a larger image sensors with small pixels straightforwardly. This compact, low-cost, robust, portable platform with a decent imaging performance may offer a cost-effective tool for telemedicine needs, or for reducing health care costs for point-of-care diagnostics in resource-limited environments.
We design a holographic system which is lensless and compact. There is a beam expander in conventional holographic setup to produce parallel light and then with a beam splitter to separate the light into two parts. One is used to illuminate the objects and the other one as the reference light. In our system, instead of utilizing beam expander to generalize parallel beam, the laser is directly produced by a fiber, which provides a spherical wave with a center in the out port of fiber. For this reason, our system contains less optical components so that the setup would be more compact. The only needed processing is to eliminate the second-order aberration caused by different distance between two path and the off-axis to a small extent. An experiment of aberration compensation by using principle component analysis is given, and the result shows that the system works well.
Fourier ptychographic microscopy (FPM) is a newly developed super-resolution technique, which employs angularly varying illumination and a phase retrieval algorithm to surpass the diffraction limit of the objective lens. Specifically, FP captures a set of low-resolution (LR) images, under angularly varying illuminations, and stitches them together in the Fourier domain. However, because the requisite large number of incident illumination angles, the long capturing process becomes an obvious limiting factor. Furthermore, in order to acquire high-dynamic range images, the time can be increased several times over. In this work, utilizing the Hadamard code principle, we propose a highly efficient method, which applies coded multi-angular illumination for FPM, to shorten the exposure time of each raw image. High acquisition efficiency is achieved by employing an optimal multi-angular illumination scheme by using two set of Hadamard coded multiplexing patterns. Both simulation and experimental results indicate that the proposed multi-angular illumination process could shorten the acquisition time of conventional FPM.