Flat optics with dispersion-engineered metasurfaces

Nature Reviews Materials, Journal Year: 2020, Volume and Issue: 5(8), P. 604 - 620

Published: June 19, 2020

Language: Английский

---

Transport of intensity equation: a tutorial

Optics and Lasers in Engineering, Journal Year: 2020, Volume and Issue: 135, P. 106187 - 106187

Published: June 19, 2020

When it comes to "phase measurement" or "quantitative phase imaging", many people will automatically connect them with "laser" and "interferometry". Indeed, conventional quantitative imaging measurement techniques generally rely on the superposition of two beams a high degree coherence: complex interferometric configurations, stringent requirements environmental stabilities, associated laser speckle noise severely limit their applications in optical microscopy. On different note, as one most well-known retrieval approaches, transport intensity equation (TIE) provides new non-interferometric way access information through only measurement. Despite insufficiency for interferometry, TIE is applicable under partially coherent illuminations (like Köhler's illumination microscope), permitting optimum spatial resolution, higher signal-to-noise ratio, better image quality. In this tutorial, we give an overview basic principle, research fields, representative TIE, focus particularly imaging, metrology, The purpose tutorial twofold. It should serve self-contained introduction readers little no knowledge TIE. other hand, attempts recent developments field. These results highlight era which strict coherence interferometry are longer prerequisites diffraction tomography, paving toward generation label-free three-dimensional microscopy, all branches biomedicine.

Language: Английский

---

Multi-depth hologram generation using stochastic gradient descent algorithm with complex loss function

Optics Express, Journal Year: 2021, Volume and Issue: 29(10), P. 15089 - 15089

Published: April 20, 2021

The stochastic gradient descent (SGD) method is useful in the phase-only hologram optimization process and can achieve a high-quality holographic display. However, for current SGD solution multi-depth generation, time increases dramatically as number of depth layers object increases, leading to nearly impractical generation complicated three-dimensional object. In this paper, proposed uses complex loss function instead an amplitude-only process. This substitution ensures that total be obtained through only one calculation, reduced hugely. Moreover, since both amplitude phase parts are optimized, obtain relatively accurate distribution. defocus blur effect therefore matched with result from reconstruction. Numerical simulations optical experiments have validated effectiveness method.

Language: Английский

---

Computational neuromorphic imaging: principles and applications

Published: March 13, 2024

The widespread presence and use of visual data highlight the fact that conventional frame-based electronic sensors may not be well-suited for specific situations. For instance, in many biomedical applications, there is a need to image dynamic specimens at high speeds, even though these objects occupy only small fraction pixels within entire field view. Consequently, despite capturing them frame rate, resulting pixel values are uninformative therefore discarded during subsequent computations. Neuromorphic imaging, which makes an event sensor responds changes intensities, ideally suitable detecting such fast-moving objects. In this work, we outline principle detectors, demonstrate their computational imaging setting, discuss algorithms process variety applications.

Language: Английский

---

LightGAN: A Deep Generative Model for Light Field Reconstruction

IEEE Access, Journal Year: 2020, Volume and Issue: 8, P. 116052 - 116063

Published: Jan. 1, 2020

A light field image captured by a plenoptic camera can be considered sampling of distribution within given space. However, with the limited pixel count sensor, acquisition high-resolution sample often comes at expense losing parallax information. In this work, we present learning-based generative framework to overcome such tradeoff directly simulating distribution. An important module our model is high-dimensional residual block, which fully exploits spatio-angular By learning distribution, approach generate both high-quality sub-aperture images and densely-sampled fields. Experimental results on real-world synthetic datasets demonstrate that proposed method outperforms other state-of-the-art approaches achieves visually more realistic results.

Language: Английский

---

Holographic Classifier: Deep Learning in Digital Holography for Automatic Micro-objects Classification

2022 IEEE 20th International Conference on Industrial Informatics (INDIN), Journal Year: 2020, Volume and Issue: unknown, P. 515 - 520

Published: July 20, 2020

Micro-objects, such as microplastics and particulate pollution, need to be accurately observed detected by high-precision optical systems. Digital holography is a powerful tool detect microscopic objects. However, traditional digital requires additional image processing phase unwrapping, de-noising, refocusing, which costs lot of time does not have consistently better performance in micro-object detection. Here, we propose an intelligent holographic classifier, deep learning-based lensless inline system the directly on raw holograms show quantitative information micro-objects for individual hologram automatic object classification. In demonstration where capture particles, are easily confused with dust arrive at accuracy above 97%. Compared other leading classifiers, our method has shorter training time, faster classification analysis, higher accuracy, robustness. Furthermore, this system, only light-emitting diode (LED), sample slide, CMOS camera, can used portable low-cost counting tool, driving development detection ecological environment.

Language: Английский

---

Digital holographic phase imaging based on phase iteratively enhanced compressive sensing

Optics Letters, Journal Year: 2019, Volume and Issue: 44(6), P. 1395 - 1395

Published: March 8, 2019

Digital holography has been widely applied in quantitative phase imaging (QPI) for monolayer objects within a limited depth. For multilayer imaging, compressive sensing is employed to eliminate defocused images but with missing information. A iteratively enhanced (PIE-CS) algorithm proposed achieve and simultaneously. Linear filtering first the off-axis hologram Fourier domain, an intermediate reconstructed complex image obtained. periodic mask then superimposed on recover object The experimental recovery of amplitude two-layer sample as little 7% random measurement demonstrated. average error PIE-CS analyzed, results show feasibility QPI.

Language: Английский

---

Method of chromatic aberration elimination in holographic display based on zoomable liquid lens

Optics Express, Journal Year: 2019, Volume and Issue: 27(7), P. 10058 - 10058

Published: March 26, 2019

In this paper, we propose a method of chromatic aberration elimination in holographic display based on zoomable liquid lens. The lens is filled with two immiscible liquids and developed by using the principle electrowetting. shape at liquid-liquid interface changes voltage applied to lens, so focal length can be adjusted changing voltage. By system, position reconstructed image controlled. When three color lasers illuminate corresponding holograms accordingly, images coincide same location clearly. experimental results verify its feasibility.

Language: Английский

---

Digital imaging technology-assisted water-sensitivity assessment of asphalt mixtures: A comprehensive review

Measurement, Journal Year: 2024, Volume and Issue: unknown, P. 115871 - 115871

Published: Oct. 1, 2024

Language: Английский

---

Computational Light Field Generation Using Deep Nonparametric Bayesian Learning

IEEE Access, Journal Year: 2019, Volume and Issue: 7, P. 24990 - 25000

Published: Jan. 1, 2019

In this paper, we present a deep nonparametric Bayesian method to synthesize light field from single image. Conventionally, light-field capture requires special optical architecture, and the gain in angular resolution often comes at expense of reduction spatial resolution. Techniques for computationally generating image can be expanded further variety applications, ranging microscopy materials analysis vision-based robotic control autonomous vehicles. We treat as multiple sub-aperture views, compute novel viewpoints, our model contains three major components. First, convolutional neural network is used predicting depth probability map Second, multi-scale feature dictionary constructed within multi-layer learning network. Third, views are synthesized taking into account both probabilistic dictionary. The experiments show that outperforms several state-of-the-art view synthesis methods delivering good

Language: Английский

---