This PDF file contains the front matter associated with SPIE Proceedings Volume 12435, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

Holographic near-eye displays are a promising technology to provide realistic and visually comfortable imagery with improved user experience, but their coherent light sources limit the image quality and restrict the types of patterns that can be generated. A partially-coherent mode, supported by emerging fast spatial light modulators (SLMs), has potential to overcome these limitations. However, these SLMs often have a limited phase control precision, which current computer-generated holography (CGH) techniques are not equipped to handle. In this work, we present a flexible CGH framework for fast, highly-quantized SLMs. This framework is capable of incorporating a wide range of content, including 2D and 2.5D RGBD images, 3D focal stacks, and 4D light fields, and we demonstrate its effectiveness through state-of-the-art simulation and experimental results.

Mid-infrared (MIR) hyperspectral imaging allows for spatially resolved chemical imaging, making it highly attractive for many branches. However, expensive MIR array detector technology limits its application. Here, we present a cost-effective MIR hyperspectral detection system based on a single-pixel imaging approach using an MIR-enhanced digital micromirror device. For illumination an MIR supercontinuum source was used completing the MIR hyperspectral microscope. It allows acquiring hyperspectral cubes with 64x64 pixels and 120 nm spectral resolution within ≈450 ms. The developed microscope provides fast MIR hyperspectral imaging with a tunable field of view and tunable spatial resolution based on cost-effective components. Thus, it will bring hyperspectral imaging into new fields and increase sample throughput due to its fast-imaging speed.

A new technique which combines the advantages of darkfield microscopy with those of confocal microscopy has been developed. The Massively Parallel Confocal darkfield method implemented on a DMD-based confocal platform allows for detection of non-fluorescing particles with dimensions below the diffraction limit. The lateral resolution and depth discrimination of three dimensional objects are improved relative to conventional darkfield microscopy.

Imaging through turbid media remains a relevant topic in biomedical imaging. In this contribution, we propose the combination of frequency domain imaging (SFDI) and single-pixel imaging (SPI) to image objects hidden by a turbid media. Firstly, the SFDI method allows to characterize the turbid media by projecting sinusoidal intensity patterns. Secondly, SPI technique provides images of the object through the areas of the turbid media with higher transmission of ballistic photons. The key elements of the system are a DMD to generate the sampling patterns and a LED array working as a programmable light source. Experimental results supporting this idea are shown.

Renal failure patients require lifesaving hemodialysis, two to three times weekly, for a few hours each time. The dialysis machine is connected to the patient via an arteriovenous (AV) fistula on an upper limb. The fistula must remain clear and stenosis (clogging) must be avoided. PatenSee provides a contactless, machine vision-based, monitoring system constructed by hybrid illumination DLP architecture with multiple coherent light sources and a fast CMOS sensor to capture multi-modal imaging of the fistula structure’s changes, sense sub-dermal information and read blood flow data for the early detection of fistula stenosis. The technology is designed to provide a simple way to streamline the workflow in dialysis centers, improve the quality of dialysis patient care and to support the busy nursing staff. The system is adaptable for home-use as well.

Research with soft materials, that is, polymeric gels, colloidal suspensions, liquid crystals, and most biomaterials often involves the need for microfabrication of confinement channels, cells, and lab-on-a-chip devices. Photolithography techniques are often chosen, as they offer the combination of versatility, precision, and quick delivery demanded by researchers. Beyond fabrication, stimulus-responsive systems, such as photosensitivity biomaterials, are the object of broad study within a very interdisciplinary community. Here, we show that a standard laboratory microscope can be quickly and economically transformed into a powerful maskless photofabrication/ photoexcitation station using off-the-shelf DMD development modules and simple optomechanical components allowing real time observation of the fabrication process.

In recent years, modulated illumination localization microscopy (MILM) methods have been proposed to provide around two-fold improvement in lateral localization precision over conventional single molecule localization microscopy methods with the same photon budget. However, MILM with laterally modulated illumination was so far reported in two-dimensional imaging modalities. To fully exploit its three-dimensional (3D) imaging potential, we propose a 3D Single-Molecule Modulated Illumination Localization Estimator (3D-SMILE) that uses the raw data measured from MILM, which has enabled a high localization precision that reaches the theoretical Cramér-Rao lower bound (CRLB) in all three dimensions. 3D-SMILE is based an optimal joint fitting algorithm implemented on a graphics processing unit (GPU) for acceleration. We have shown in simulations that the average lateral localization precision of 3D-SMILE has been improved by more than 3.5 folds over 3D-SMLM over an imaging depth range of around 1.2 μm.

Temperature calculations of Texas Instruments' DMD (Digital Micromirror Device) are well known under CW (continuous wave) optical exposure conditions and are described in existing DMD datasheets. However, application spaces are being developed which require the use of pulsed lasers as the illumination source, complicating the DMD temperature calculations. Certain combinations of laser power, pulse widths, and pulsing repetition rate result in acceptable mirror surface temperature while others can create an unreasonable thermal load well outside of acceptable DMD parameters. This paper details transient thermal calculations at the mirror surface, bulk mirror, and silicon die locations using the DMD thermal (RC) time constants along with laser peak power, pulse width, and repetition frequency to help define acceptable operating conditions and avoid conditions outside the capability of the DMD. Pulse widths spanning from continuous to femtosecond and the resulting maximum mirror surface temperature are shown along with the transient temperature profile to understand the response of the mirror.

Size and efficacy are two of the main challenges of an AR display. A compact form factor and low power consumption are highly desirable in AR glasses. Self-emitting display technologies such as OLED and MicroLED do not need illumination thus enabling simpler and smaller optics, but their efficacy is quite low, especially for pixel sizes of 5um or below. Spatial Light Modulator (SLM) based technologies, such as DMD, LCoS (Liquid Crystal on Silicon), and LBS (Laser Beam Scanning), take advantage of high efficacy light sources and can achieve a high brightness AR display combining low power consumption and high pixel density, which is highly preferable in many applications. The DMD, together with the efficient and ever-improving LED light source, has been highly favored in the battery-powered portable projection display market, due to its high efficacy in a compact form factor as well as its insensitivity to polarization. Here we present a compact optical engine architecture for AR glasses, taking into consideration both size and efficacy. The design simplifies the traditional DMD illumination to a compact size targeting a thin profile similar to a backlit LCD. This compact size is achieved with a small compromise to engine efficacy. Multiplying the pixel density with a glass plate actuator offers a performance improvement that easily justifies the slight increase in size. This architecture offers an excellent optical engine option for AR glasses which is both compact and high performance.

We present an electromagnetic simulation of a digital micromirror device (DMD) that models optical performance from the visible spectrum to the mid-wave infrared regime. We calculate DMD efficiency, including the effects of optical scatter and interference, over a wide range of focal ratios (f/2.8 to f/8) and wavelengths (0.4 μm to 5 μm). Furthermore, we investigate how contrast ratio varies with respect to wavelength, provided a set of operating parameters. The micromirror array structure of a DMD induces strong wavelength-dependent optical effects that impact the stray light and throughput of a system. To quantify this, we perform a three-dimensional electromagnetic finite-difference time-domain simulation where we illuminate the DMD with a focused, diffraction-limited beam; calculate the near-field electric field; and transform the distribution of light to the far-field. We characterize the performance of a DLP7000 device in three key wavelength regimes: the specular regime (λ < 1 μm), the transition regime (1 μm < λ < 3 μm), and the diffraction regime (3 μm < λ < 5 μm). Our results inform optical performance parameters and provide design constraints for the implementation of DMDs in sensitive optical instruments.

Dynamic three-dimensional (3D) surface imaging using phase-shifting fringe projection profilometry is currently driven by industrial manufacturing, archaeological inspection, and entertainment. However, existing techniques cannot simultaneously provide the robustness in solving spatially isolated 3D objects, the tolerance of large variation in surface reflectance, and the flexibility of tunable working distances with a cubic meter (m³)-level measurement volume at video rate. To overcome these limitations, we have developed multi-scale band-limited illumination profilometry (MS-BLIP). MS-BLIP implements multi-frequency fringe projection with the associated phase unwrapping, which enables robust 3D imaging of spatially isolated objects. Meanwhile, MS-BLIP adopts dual-level intensity projection to enhance its dynamic range, which allows recovering 3D information from surfaces with large reflectance variation. Moreover, MS-BLIP applies a newly developed iterative method for distortion compensation, which improves the 3D reconstruction quality over a cubic-meter-level measurement volume. Finally, a dove prism is used to adjust the orientation of the field of view, which helps MS-BLIP fit different experimental scenarios. With a measurement volume of up to 1.5 m³ and a working distance of up to 2.8 m, MS-BLIP showcases the video-rate 3D surface imaging of full human-body movements.

We demonstrated a real-time lidar system applying a Digital Micromirror Device (DMD) as a field of view (FOV) expander of a lidar receiver employing a 2D Multi-Pixel Photon Counter (MPPC). By temporally synchronizing the transitional state of micromirrors with returning photons from lidar, receiver FOV is diffractively steered to the targets’ direction enabled by nano-second pulse laser. With a nanosecond 905nm laser transmitter, time-of-flight (ToF) lidar images were captured across seven diffraction orders with the expanded 35 degrees full field of view lidar scanning range.

Micro Electro Mechanical System (MEMS) spatial light modulators enables adaptive and fast beam and image steering. For lidar applications, Texas Instruments Phase Light Modulator (TI-PLM) is paired with real-time calculation and display of Computer Generated Holograms (CGH) by CUDA-OpenGL interoperability assisted by YOLOv4-tiny network model for object detection and recognition. The real-time object recognition, CGH calculation, and display framework replaces conventional raster scanning with camera-input based and foveated beam steering while having a beam scan rate beyond the frame rate of TI-PLM. For Augmented Reality (AR) application, the same framework is used for image steering based on gaze information of eye. With Texas Instruments Digital Micromirror Device (TI-DMD), image is steered into a part of field of view by following movement of eye. The diffractive image steering enabled by TI-DMD increases FOV while not sacrificing resolution of the image displayed.

The purpose of computer-generated holography (CGH) is to generate arbitrary illumination patterns by modulating a coherent light beam. The possible applications of CGH include for example: optical testing, wavelength-selective switching, optical tweezers, and display technologies. For most applications, illumination patterns with high efficiency and a high contrast ratio are required. One of the major challenges related to CGH is the generation of a high signal to noise ratio to achieve the desired results of an effective and high-contrast illumination pattern. With emerging MEMS-based phase modulators, higher modulation speeds become accessible, resulting in higher frame rates than in existing LC based micro-displays. This high modulation speed enables the development of new algorithms for CGHs, which benefit from the high frame rate and utilize the high modulation speed to improve the contrast and image quality. This paper introduces a novel algorithm to generate time multiplex holograms (TMH). Using the TMH method, various illumination patterns for future automotive exterior projection scenarios are generated. To verify the applied method, numerical simulations and optical experiments are performed and indicate good consistency. Finally, the TMH results are compared with holograms generated using the Gerchberg-Saxton algorithm for computer-generated holograms (GS).