Speckle contrast evaluation: a new concept for optical inspectionAgour, Mostafa; Falldorf, Claas; Bergmann, Ralf B.
doi: 10.1117/12.3052137pmid: N/A
We present an overview of a new technique that uses speckle decorrelation in imaging with partially coherent light for 3D shape measurement of technical parts with optically rough surfaces, called SPICE profilometry. It exploits the limited spatial coherence of the illumination for depth discrimination, with speckle patterns decorrelating at different depths. SPICE enables rapid, non-mechanical depth scanning using optical light modulators such as electrically tunable lenses or micro-mirror devices to acquire all the necessary images in less than a second. Compared to traditional methods such as confocal and white light interferometry, SPICE is faster and more robust, ensuring eye safety and offering a depth of focus of several millimeters. The limitations of SPICE will be presented, a solution to these limitations will be discussed, and examples of applications will be shown.
Free form GPR scanning with augmented reality visualizationTanch, Scott A.; Huston, Dryver; Xia, Tian; Burns, Dylan
doi: 10.1117/12.3054929pmid: N/A
This paper presents a workflow in the field of ground penetrating radar. Using a combination of sensing technologies, including photogrammetry, encoder odometry, inertial measurement, and GPS tracking, the position of a radar unit can be tracked and recorded in real time by means of an extended Kalman filter. This allows for ungridded, free form scans or F-Scans which have increased precision over traditional linear gridded methods. The additional data stream is then provided along with conventional GPR data to an edge computing node where it is processed into a 3-D visualization of the survey. These 3-D visuals are made available to a custom-built augmented reality application for the Microsoft Hololens 2, allowing for new modes of interaction between users and data. By incorporating a visual target, here a QR code, into the initial data collection process, a shared virtual experience is created allowing multiple users to evaluate a holographic survey concurrently. The basic processing shown here can be extended into more sophisticated techniques leading towards more accurate and less abstract representations of subsurface conditions.
Speckle-free digital holographic microscopy with partially spatially coherent light: ultra-high spatial phase sensitivity, high temporal phase stability, and large space bandwidth productMehta, Dalip Singh
doi: 10.1117/12.3052205pmid: N/A
We demonstrate speckle-free digital holographic microscopy (DHM) for quantitative phase imaging of biological cells and 3D-surface profilometry of industrial objects with partially spatially coherent monochromatic light. By means of controlling the spatial coherence of laser light source we could obtain speckle-free interferograms with no spurious fringes in holographic microscopes. Due to partially spatially coherence monochromatic light we could obtain ultra-high spatial phase sensitivity, high temporal phase stability and large space bandwidth product which is important for holographic microscopy of realistic objects. The experimental results are compared with coherent light and partially spatially coherent light-based schemes, and we found an order of magnitude improvement in the accuracy of phase measurement of biological cells, such as, Red blood cells, He-La cells and sperm cells. We also obtained very high accuracy of step-height measurement of industrial objects, such as, silicon integrated circuits, waveguide structures, and industrial rough objects. Finally, we compared the various performance parameters of developed three types of holographic microscopes, such as, low coherence, high coherence and partially spatially coherent based systems and we found that partially spatially coherent light-based systems outperform the other two systems, in terms of space bandwidth product, spatial phase sensitivity, accuracy in phase and step-height measurement.
Real-time aquatic imaging using living animals as light sources for VR biologyIchikawa, Ryosuke; Takatsuka, Hiroki; Takiyama, Kazuaki; Ikeda, Takehide; Iwane, Toru; Suyama, Shiro; Yamamoto, Hirotsugu
doi: 10.1117/12.3054044pmid: N/A
We propose an aquatic display that forms real-time aquatic images by use of living animals as a light source. In behavioral studies of living animals, three-dimensional computer graphics (3DCG) and moving images have been used as visual stimuli. However, they require prior preparation of 3D rendering and video recording. The purpose of this research is to clarify a method for forming real-time aquatic imaging of stimuli without requiring special processing. Our proposed aquatic display uses living animals as the light source. The proposed optical system consists of the living animal, a beam splitter, a retro-reflector, and two water tanks. A living fish is illuminated to form a fish's aquatic image. Both in cylindrical and cubic water tanks, the living fish in one water tank can form a corresponding aquatic image in another tank. The proposed optical system enables the formation of aquatic image synchronized with the animal's movements as stimuli, without the need for preparing stimuli or performing special processing.
Photon-counting imaging with denoising diffusion modelsPark, Seonghwan; Moon, Inkyu
doi: 10.1117/12.3052241pmid: N/A
In this paper, we present a multispectral photon-counting imaging (PCI) method based on denoising diffusion models for multispectral visualization of virtually photon limited scenes. We measure the accuracy as well as the speed of denoising diffusion algorithms to estimate multispectral scenes at low light levels. Experimental results demonstrate that the proposed deep learning model achieves better performance in terms of peak-to-SNR (PSNR) and faster computation than variational autoencoders.
Confidence-driven planar element restoration for 3D volumetric dataFafard, Alex
doi: 10.1117/12.3053756pmid: N/A
We introduce a confidence-driven framework for addressing localized geometric distortions within 3D volumetric data through planar element restoration. Our method formulates a tripartite optimization objective: preserving high-confidence reference structures (HCRS) as spatial anchors, enforcing geometric plausibility via nearest-planar- point alignment of distorted coordinates, and minimizing global displacement energy in the corrected volume. A hybrid framework coupling robust RANSAC-based plane fitting with adaptive loss weighting enables nonlinear distortion correction under heterogeneous error models. Evaluation demonstrates efficacy against CT beam hardening artifacts and other small distortions affecting planarity through dual validation: (1) quantitative metrics on stochastic synthetic volumes, and (2) qualitative enhancement of PCB structure visibility.
A study of point cloud 3D modeling for multi-perspective image processing and visualizationHaggerty, Ryan; Shen, Xin
doi: 10.1117/12.3052628pmid: N/A
In this undergraduate research work, we focus on the study of point cloud 3D modeling methods for denoising using multiple perspective. The methodology implemented for multiple perspective denoising was to take a 3D model STL convert it to a point cloud and introduce noise to the model. The two noisy point cloud data (PCD)s are the run through the statistical outlier removal (SOR) process for denoising. The results of this experiment yielded that using 2 noisy 3D point cloud inputs in the SOR denoising method had a 4% noise reduction effectiveness increase from using one noisy 3D PCD input for denoising, there was also an observable increase in original point cloud data retention.
High-resolution quantitative phase imaging of live cells: an application of nanoscopic computational algorithm for nanoscale motion analysisBhatt, Sunil; Saxena, Kanchan; Saxena, Anuj; Mehta, Dalip Singh
doi: 10.1117/12.3057007pmid: N/A
We report label-free 3D imaging of live biological cells using digital holographic microscopy. We developed a speckle free DHM system using Linnik-type interferometric configuration by means of using significant speckle reduction technique. Live cell imaging such as U2OS cells and motion of sperms cells were analyzed using quantitative phase imaging. Fine movement of sperm head and tails, and cell-to cell interaction was monitored. But the speckle-free phase image was diffraction-limited hence fine details could not be visualized. In order to improve these fine details of live cells, we used nanoscopic computational algorithm MUltiple SIgnal Classification ALgorithm (MUSICAL) for nanoscale motion trace analysis. A significant improvement in the image reconstruction was achieved. Fine details of live cell images were reconstructed and 3× improvement in the resolution was achieved. Experimental results of motion trace of sperm cells and cell-to-cell interaction of U2OS cells will be presented.
Metasurfaces for 3D imaging and displays: potential and limitationsThibault, Simon
doi: 10.1117/12.3054052pmid: N/A
Over the years, technologies have emerged that have the potential to change the world of optics. Metasurfaces are a case in point. However, the use of metasurfaces has yet to deliver on its promise to revolutionize optical systems. We can only wonder how this type of surface could have changed 3D imaging and displays. Through a few examples, we'll discuss the limitations inherent in the use of metasurface in this field of applications.