Large depth-of-field (DOF) imaging with a high resolution is useful for applications ranging from robot vision to bio-imaging. However, it is challenging to construct an optical system with both a high resolution and large DOF. The common solution is to design relatively complex optical systems, but the setup of such systems is often bulky and expensive. In this paper, we propose a novel, compact, and low-cost method for large-DOF imaging. The core concept is to (1) design an aspherical lens with a depth-invariant point spread function to enable uniform image blurring over the whole depth range and (2) construct a deep learning network to reconstruct images with high fidelity computationally. The raw images captured by the aspherical lens are deblurred by the trained network, which enables large-DOF imaging at a smaller F number. Experimental results demonstrate that our end-to-end computational imager can achieve enhanced imaging performance. It can reduce loss by up to 46.5% compared to inherited raw images. With the capabilities of high-resolution and large-DOF imaging, the proposed method is promising for applications such as microscopic pathological diagnosis, virtual/augmented reality displays, and smartphone photography.

The gamma effect of phase-measuring profilometry systems yields nonlinear errors, which will substantially reduce the 3D shape measurement accuracy. Here, a robust and flexible gamma correction method based on the probability distribution function (PDF) of the wrapped phase is presented. First, a series of PDF curves are generated from the simulated wrapped phase distributions with different gamma values. Second, an experimental PDF curve will be produced after obtaining the wrapped phase from the captured three-step phase-shift fringe images. Then, a correlation procedure will be used to find the most similar PDF curve from the simulated PDF curves, and the gamma value of the matched PDF curve is that of the current system. Note that the gamma value detected by this method will be smaller than the true value due to the defocusing effect of the projection system with a large aperture. Therefore, an improved PDF-based algorithm, which projects two sets of three-step phase-shifting sinusoidal fringe patterns with different pre-coded gamma values and produces two PDF curves, is also added. Then after one more correlation procedure, a more accurate systematic gamma value could be calculated. It does not need large-step phase-shift images and 2×3 fringe images are quite enough. The experimental results show that the technique is very fast, easy to use and quite accurate.

In this paper, a fast rotary mechanical projector (RMP) is designed and manufactured for high-speed 3D shape measurement. Compared with the common high-speed projectors, RMP has a good performance in high-speed projection, which can obtain high quality projected fringes with shorter camera exposure time by using the error diffusion binary coding method and chrome plating technology. The magnitude, acceptability of systemic projection error is analyzed and quantified in detail. For the quantified error, the probability distribution function (PDF) algorithm is introduced to correct the error. Corrected projection error is reduced to more than one third of the original error. Subsequently, a monocular measurement system composed of the RMP and a single camera is constructed. The combination of the RMP device and PDF algorithm ensure the accuracy of a corresponding 3D shape measurement system. Experiments have demonstrated that the proposed solution has a good performance for the 3D measurement of high-speed scenes.

Three-dimensional (3-D) shape measurement using a novel encoded-phase grating is proposed. The projected sinusoidal fringe patterns are designed with wrapped and encoded phase instead of monotonic and unwrapped phase. Phase values of the projected fringes on the surface are evaluated by phase-shift technique. The absolute phase is then restored with reference to the encoded information, which is extracted from the differential of the wrapped phase. To solve the phase errors at some phase-jump areas, Hilbert transform is employed. By embedding the encoded information in the wrapped phase, there is no extra pattern that needs to be projected. The experimental results identify its feasibility and show the possibility to measure the spatially isolated objects. It will be promising to analyze dynamic objects.

In this Letter, the periodical errors, which are caused by the nonlinear effect of the commercial projector and camera, are analyzed as a more generic single-coefficient model. The probability density function of the wrapped phase distributions is used as a tool to find the compensation coefficient. When the compensation coefficient is detected, on the premise of ensuring accuracy, a correlation algorithm process is used to replace the traditional iterative process. Therefore, the proposed algorithm improves the efficiency of coefficient detection dramatically. Both computer simulation and experiment show the effectiveness of this method.

Single-shot fringe projection profilometry (FPP) is crucial for real-time or dynamic 3D measurement scenarios. In this regard, we propose a spatial composite fringe pattern projection profilometry (SCFPP), which creatively fuses the spatial feature in different frequency fringe patterns. Inspired by image inpainting techniques in computer vision, SCFPP employs a novel adaptive space-division encoding strategy (ASES) to divide fringe patterns into several basic blocks and re-aggregate partial basic blocks in the spatial domain to form a spatial composite fringe pattern. In the demodulation stage, we develop a deep learning network model, the phase inpainting network (PIN), to restore the spatial composite fringe pattern to complete phase information. The absolute phase map is then obtained with high accuracy by applying the conventional three-frequency heterodyne phase unwrapping (CTHPU) algorithm with three known wrapped phase maps. To our knowledge, SCFPP is the first successful use of spatial multiplexing strategy in FPP. Additionally, SCFPP can be seen as an application of image inpainting techniques in the fringe projection field, introducing new ideas for applying computer vision methods in fringe analysis.