Welcome and Introduction to SPIE Photonics West OPTO conference 11698: Emerging Digital Micromirror Device Based Systems and Applications XIII

We present a fiber-based 3D printing system with the highest printing resolution. Fabricated lines and microstructures are presented too. A microlens (NA 0.9) is printed in the Nanoscribe 3D-printer and attached to the fiber. A femtosecond laser (780nm, 80MHz) is used for two-photon polymerization and the focused spot on the tip of the lens is generated by wavefront shaping with the Transmission Matrix method. The spot size (lens with NA 0.9) is compared to the spot size of the fiber (NA 0.29) and to the spot size obtained when a commercially available microlens is attached (NA 0.54). Exposure tests are conducted by printing lines or voxels on a photosensitive resin and finally a prepared high aspect ratio structure is used for demonstrating the utility of a fiber-3D printing system to reach confined areas for microfabrication.

First focus when developing imprint materials for optical applications is on their optical properties. However other properties are also key. Here we focus on the thermo-mechanical properties of polymer-based imprint materials and their importance for various applications. Special emphasis will be put on thermo-mechanical stress induced by thermal mismatch. We’ll compare thermo-mechanical bulk properties of such materials with results from FEM simulations and experimental tests.

When using DMD-based 3D printing methods to fabricate functional micro-devices, unwanted light scattering during the printing process is a significant challenge to achieve high resolution fabrication. We report the use of a deep neural network based machine learning technique to mitigate the scattering effects, where our neural network was employed to study the highly sophisticated relation between the input digital masks and their corresponding printed 3D structures. Furthermore, the neural network was used to model an inverse 3D printing process, where it took the desired printed structures as inputs and subsequently generate grayscale digital masks that optimized the light exposure dose according to the desired structures' local features. Verification results showed that using neural network generated grayscale digital masks yielded significant improvements in printing fidelity and resolution compared to using masks identical to the desired structures.

Volumetric additive manufacturing (VAM) is a novel technique where light images are projected into a rotating cylindrical vial filled with photopolymerizable resin to form 3D parts in a single exposure. This method presents new challenges and opportunities to identify the optimal optical configuration to form high-quality parts. We investigate three techniques to accommodate the strong lensing effect of the cylindrical resin-filled container and characterize the trade-offs in performance.

I will present our recent work on parallelization of the two-photon polymerization (TPP) process based on temporal focusing via a digital micromirror device, where programmable femtosecond light sheets are formed and used to substantially improve the fabrication rate without sacrificing resolution. In the experiments, we demonstrated the fabrication of arbitrarily complex structures at a record-breaking resolution and speed, i.e., lateral/axial resolution: 140 nm/175 nm at 10s mm3/min, which is approximately 3 orders of magnitude higher than any existing fabrication methods. Importantly, the throughput enhancement also effectively reduces the cost by 90% (~$1.5/mm3), which makes TPP an economic solution for nanofabrication.

Computed axial lithography (CAL) is a volumetric additive manufacturing method in which a three-dimensional light dose distribution is constructed in a photopolymer from the superposition of illumination patterns from many different angles. The technique’s advantages over layer-by-layer light printing methods stem from the fact that in CAL hydrodynamic stresses are effectively eliminated from the resin precursor material during printing. This key difference allows a wider range of materials to be processed, including high-viscosity or thermally gelled precursors, and allows polymeric objects to be printed around pre-existing solid objects (‘overprinting’). In this talk we describe some of the current limitations on spatial resolution, printing speed, and mechanical properties in CAL. We also describe a computationally efficient approach to modeling the occlusion of light by objects suspended in the printing volume, which supports the optimization of overprinting processes.

Hydrogels, due to their optical transparency and biocompatibility, have emerged as excellent alternative to conventional optical materials for biomedical applications. Advances in microfabrication techniques are now able to convert conventional hydrogels into functional diffraction optical elements (DOEs) or hDOEs. However, key challenges related to device customization and ease/speed of fabrication need to be addressed to enable widespread utility and acceptance of hDOEs for a range of target applications. Here, we report rapid printing of customized hDOEs on polyethylene glycol diacrylate (PEGDA) hydrogel using digital photopatterning; a novel method that combines simulated computer-generated hologram (SCGH) and projection photolithography. To showcase the versatility of this approach, a range of hDOEs are demonstrated, including 1D/2D diffraction gratings, Dammann grating, Fresnel zone plate (FZP) lens, fork-shaped grating and computer-generated hologram (CGH) of arbitrary pattern.

The ability to generate 3D angiogenesis models is central for tissue engineering and drug screening applications. However, existing bio-fabrication technologies have yet to attain precise guidance of capillary networks in 3D. Here, we present our latest results in fabricating capillary networks using a novel laser-assisted bioprinting technique named Laser Induced Side Transfer (LIST). We found that LIST-printed human umbilical vein endothelial cells (HUVECs) present negligible loss of viability and maintain their abilities to migrate, proliferate and form intercellular junctions. Furthermore, we showed that LIST enabled the formation of capillary-like networks in 3D with high spatial precision (50 μm) over a large volume (1 cm3). Those networks were validated as angiogenesis assays for pro- and anti-angiogenic compounds. LIST could be widely adapted for applications requiring multiscale bioprinting capabilities, like the development of 3D drug screening models and artificial tissues.

DLP-based 3D printing has recently allowed the fast and accurate printing of complex and physiologically relevant constructs using silk-based bioinks. The architecture, mechanics and organization of these constructs can be tailored to the specific needs of a given tissue. Here we investigated the potential of these constructs for bone, cartilage and osteochondral regeneration. We were able to use DLP printing to obtain silk-based nanocomposite structures containing bioactive morphogens or antibiotics, and sustaining the adhesion, proliferation and differentiation of human mesenchymal stem cells. These results suggest a highly promising future in the field of orthopedics and rheumatology.

Accurate characterization of the collagen microstructure of thick collagenous tissues has been challenging due to strong tissue scattering. We previously reported structured polarized light imaging (SPLI) to improve the imaging quality and quantification accuracy by suppressing the diffuse background from deeper tissue. However, current SPLI requires multiple sequential images (typically 12 images); thus, it is slow and precludes time-sensitive applications, such as dynamic tissue testing. In this study, we present real-time SPLI that is capable of imaging tissue dynamics at 55 frames per second, allowing continuous quantitative tracking of the deformation with improved contrast and accuracy.

Autofluorescence (AF) spectroscopy is used widely in biomedicine for disease diagnosis and screening. A major challenge in AF spectroscopy is low signal levels, dictating a need for high-throughput measurement systems. Highly efficient liquid crystal polarization gratings (LCPGs) are an attractive option for dispersing light, but unpolarized autofluorescent light will be split between two orders. To advance this technology, we present the design and demonstration of a high-throughput LCPG spectrometer, using a novel prism system used to merge the two orders into a single spectrum. We demonstrate minimal degradation of spectral resolution, and an increase in signal levels by up to 20% compared to using traditional diffraction gratings.

We propose a new deep compressed imaging modality depending on a learnable pattern-scanning. High-speed imaging with a high-quality reconstruction is effectuated by training an end-to-end auto-encoder framework composed by deep neural network and compressed sensing algorithm. By this framework, optimizing the illumination pattern and the corresponding reconstruction decoder faithfully rebuild the imaged object under a high subsampling rate. Comparing with the conventional single-pixel camera and point-scanning imaging system, we accomplish high-speed imaging with a reduced light dosage, albeit preserve an identical imaging quality evaluated by peak signal-to-noise ratio for error and Fourier ring correlation for spatial resolution.

We propose and demonstrate single-shot synthetic aperture quantitative phase microscopy to achieve high throughput imaging. A synthetic aperture technique is used to enhance the spatial resolution and a large pixel number camera is used to obtain a large field of view (FOV). 6 illumination angles are multiplexed in time and recorded within one single frame acquisition, thereby greatly speeding up the acquisition process. Our system can provide a FOV of 320 μm by 320 μm, an effective NA of 1.65 and an acquisition speed of 72 fps, which corresponds to a space-bandwidth-time product (SBP-T) value of more than 100 megapixels/s.

Full Field Optical Coherence Tomography (FF-OCT) provides higher lateral resolution than beam scanning OCT methods due to its use of a microscope objective. However, current systems suffer from reduced imaging quality when investigating transparent samples. The proposed design of a Michelson based FF-OCT microscope, when combined with high dynamic range illumination via a Digital Micro-Mirror Device (DMD), is capable of providing 3D visualization even of highly transparent structures within the cells as well as the surrounding membranes without any doping, staining, or labeling required. The unique combination of the DMD with FF-OCT provides a powerful microscopy platform for transparent samples.

We propose and demonstrate digital-micromirror-device (DMD) based laser-illumination Fourier ptychographic microscopy (FPM) for high-speed and high-resolution imaging. Using this method, we can simultaneously retrieve phase and amplitude images of low absorption samples, such as cancer tissue and cells. The system illumination is provided by a 532 nm laser, while two DMDs are utilized for achieving active illumination angle scanning and dynamic filtering. Our current system has achieved a field of view of around 0.60 mm by 0.34mm with a lateral resolution of 1.1 μm and an acquisition time of 0.188 second, which corresponds to a space-bandwidth-time-product (SBP-T) of 1 megapixels/second.

Cell instance segmentation is a critical task to perform for the quantitative analysis of 3D live-cell images. In this paper, we propose a novel cell instance segmentation approach based on point proposal for 3D cell imaging. Different from existing work, our model utilizes the nuclei of cells as point proposal and employ them as positive and negative point proposals. We constructed the 3D NIH3T3 dataset for training and evaluation, and examine the proposed model qualitatively on three independently gathered cells; HeLa, A549, and MDA-MB-231. Our model exhibits superior quantitative results; moreover, compared to previous methods, it properly predicts cell lines, which are not even well-annotated during training.

Spin Glasses (SG) are paradigmatic models for physical, computer science, biological and social systems. The problem of studying the dynamics for SG models is NP-hard, i.e., no algorithm solves it in polynomial time. Here we implement the optical simulation of an SG, exploiting the N segments of a Digital Micromirror Device to play the role of the spin variables, combining the interference at downstream of a scattering material to implement the random couplings and measuring the transmitted light intensity to retrieve the system energy. We demonstrate that This optical platform beats digital computation for large-scale simulation (N>12000).

A Digital Micromirror Mirror (DMD) based holographic lidar is reported that multiplexes coarse-steering by sawtooth phase modulation and fine-steering by binary amplitude modulation. Hybrid steering is achieved by overlaying displayed Computer-Generated Holograms (CGHs) with a sawtooth blazed grating phase mask, which the blaze angle programmed by synchronized short-pulse illumination of transitioning micromirrors creating the CGHs. The steering principle is demonstrated in a 905 nm lidar system with a 44° field-of-view, 0.9°×0.4° angular resolution, 1 meter range, 1.5 kHz sampling rate, and 7.8 FPS video frame rate. Scaling is proposed to improve maximum scanning distance and diffraction efficiency.

A frequency modulation technique is described to distinguish between particle signals and stray light in reticle inspection systems for semiconductor photolithography. In our frequency modulation approach, a digital micro-mirror spatial light modulator is used to illuminate the reticle with a sinusoidal pattern whose temporal frequency changes with position on the reticle surface. Light scattered by particles as well as stray light spatially overlap on the system’s detector. Fourier analysis is used to reconstruct the frequency content in each pixel individually. By comparing measured with projected frequencies, one can distinguish between light scattered by particles and stray light.

High-speed three-dimensional (3D) fringe projection profilometry (FPP) is widely used in many fields. Recently, researchers have successfully tested the feasibility of performing fringe analysis using deep convolutional networks (CNN). However, the existing methods require tremendous real-world scanning trials for model training, which is not trivial. In this work, we propose a framework to establish the digital twin of a real-world system in a virtual environment and a process to automatically generate 3D training data. Experiments are conducted to demonstrated that a physical system can adopt the CNN trained in the virtual environment to perform accurate real-world 3D shape measurements.

We report single-pixel three-dimensional (3D) dual photography. Inspired by single-pixel imaging, fringe projection profilometry (FPP), and dual photography, we propose a method for 3D imaging which allows for the synthesis of dual and relit images in a camera-free context, as well as the broadening of traditional dual photography image synthesis towards viewing and relighting of 3D images at user-selectable perspectives. By exploiting Helmholtz reciprocity and using a light-path symmetric single-pixel imaging system based on two digital micromirror devices (DMDs), our work extends the conceptual scope and imaging capability of dual photography.

We report on a novel kind of accelerating beams that follow parabolic paths in free space. In fact, this accelerating peculiar polygon beam (APPB) not only exhibits autofocusing property, but also possesses two types of accelerating intensity maxima, i.e., the cusp and spot structure, which does not exist in the previously reported accelerating beams. We also provide a detailed insight into the theoretical origin and characteristics of this spatially accelerating beam through catastrophe theory. Moreover, experimental results confirm the peculiar features presented in the theoretical findings, and the APPB is further verified to exhibit self-healing property during propagation with either obstructed cusp or spot-like main lobe reconstructing after a certain distance. Thus, we anticipate that the APPB will facilitate the applications in the areas of biomedical manipulation and optofludics.

Beside low cost, automobile design will be another major factor for the mass adoption of the LiDAR in the autonomous vehicles. Extra aperture in the automobile chassis for LiDAR operation will be undesirable for automobile designers. As a result, it will be advantageous to integrate the LiDAR together with the headlight such that the chassis design of the vehicle does not have to deviate from standard practices. This paper presents optical design of integrated LiDAR and smart headlight using a single DMD such that it replaces the current headlight without impact on the overall chassis design of the vehicle. For low cost considerations, the two functions share the same DMD. Preliminary designs of such system and results will be presented.

This paper discusses several advances made over the recent year in the areas of performance and reliability of the Texas Instruments MEMS-based phase-modulator device currently under development. Increase in the phase quantization from 3 to 4 bits, and the flattening of the mirrors have made significant improvements in efficiency and quality of the rendered image. The paper will discuss and give examples of the improved performance. Recent tests have which have shown phase level stability over more than 1000 hours of elevated temperature testing will be presented. Also, some techniques that are being developed to test this device in a production environment will also be presented.

One of the major issues related to CGH is the computational power required to produce high quality images. Compared to existing LC-based micro displays, emerging MEMS-based phase modulators offer a significantly higher modulation speed with a negligible cross-talk between pixels. This allows reducing the required computational load. A novel non-iterative algorithm is used to construct time-sequential phase only sub holograms (TPOSH). Based on the TPOSH we generate various high dynamic range light distributions for typical automotive headlamps scenarios. To verify the method, simulations and experiments are performed. Results from numerical simulations and optical experiments show a high degree of consistency.

Computer generated holography, a disruptive technology for projection displays, enables variable projection distance by combining variable lens hologram with image hologram and encoding it on a spatial light modulator. The basic structure of a holographic projector includes a spatial light modulator and a coherent light source (i.e., LASER). The use of coherent light source in a holographic projection system and the limitation of phase retrieval algorithms, which calculate holograms, give the projected content a speckly appearance. We demonstrate the use of fast-switching piston-mode spatial light modulator for temporal averaging method to reduce the speckle of a holographic projector.

This paper presents a smart, cognitive camera vision system for industrial and robotic application. The vision system comprises of a DMD-based structured light projection unit with high intensity laser illumination and a pair of high frame rate and high resolution imaging sensors. A proprietary algorithm was developed to generate high-dense point cloud while keeping the measurement time less than 100ms, in comparison to a few seconds for those available on the market. By combining geometric models and deep learning algorithms, we demonstrated that this cognitive 3D vision system can guide the robots to perform complex assembly tasks with autonomy, agility, and accuracy.

Based on the fringe projection profilometry, a very compact and flexible positionable measuring head can be combined with optical fiber bundles to perform in-situ inspection tasks in industrial applications. Surface geometries can be reconstructed and quantified in metric coordinates by means of a fast, non-contact and high-resolution geometry inspection. Defect segmentation, on the other hand, is rather complex with three-dimensional point clouds, since reference data is required or a deviation determination is ambiguous and susceptible to errors. Since each reconstructed object point corresponds to a camera pixel, it is possible to apply image processing algorithms or neural networks for defect segmentation. Because image based segmentation is more susceptible to poor illumination and deviating surface curvature or texture, a circular array of miniature LEDs has been coaxially arranged around the imaging optics of the camera's fiber to provide different illumination directions.

In the calibration of a telecentric fringe projection system, the magnification factor of a single lens is optimized for the whole measuring volume. If a measurement object is placed near the margin of the measuring volume, deviations in the 3D reconstruction can occur. In order to prevent imprecise results, the measuring volume is divided into multiple regions. Within these regions, a measurement object with calibrated dimensions is placed to determine a separate magnification factor for each region. These magnification factors are taken into account for further 3D reconstructions. First reconstructions have shown that this method allows achieving a reduced deviation within the margin of the measuring volume.

For the fast 3D inspection of hard-to-reach components such as turbine blades in partially disassembled aircraft engines, a borescopic fringe projection system was developed. With the rapid development of MIPI / smartphone cameras, new possibilities for miniaturization of these measuring systems are constantly evolving. This paper shows a methodology for the comparability of camera and projection units and quantifies the limits for the use of multimedia sensors in metrological applications. Therefore, the characteristic properties of the camera and projector as well as the whole measuring system are analyzed and a comparison is made regarding the reconstruction of calibrated features.

We report dual-view band-limited illumination profilometry (BLIP) with temporally interlaced acquisition (TIA), as a new technique on 3D phase-shifting fringe projection profilometry (PSFPP). TIA employs two cameras to capture complementary half a sequence separately, which eliminates the information redundancy in data acquisition. Meanwhile, a newly developed algorithm is used for robust pixel matching and 3D image reconstruction. The resultant TIA-BLIP system enables 3D imaging over 1 kHz on a field of view of up to 150×130 mm2. TIA-BLIP’s performance has been demonstrated by imaging glass vibration induced by sound and glass breakage by a hammer.

Digital micromirror device (DMD) allows very fast binary light modulation. For digital hologram displaying using DMD, binarization should be applied. For this task various non-iterative and iterative error diffusion binarization methods are considered. Quality of obtained reconstructed images were compared.