Deep convolutional neural network based video super-resolution (SR) models have achieved significant progress in recent years. Existing deep SR methods usually impose optical flow to wrap the neighboring frames for temporal alignment. However, accurate estimation of is quite difficult, which tends produce artifacts super-resolved results. To address this problem, we propose a novel end-to-end that dynamically generates spatially adaptive filters alignment, are constituted by local...

Optical systems have been crucial for versatile applications such as consumer electronics, remote sensing and biomedical imaging. Designing optical has a highly professional work due to complicated aberration theories intangible rules-of-thumb, hence neural networks are only coming into this realm until recent years. In work, we propose implement generic, differentiable freeform raytracing module, suitable off-axis, multiple-surface freeform/aspheric systems, paving the way toward deep...

The commonly used atmospheric model in image dehazing cannot hold real cases. Although deep end-to-end networks were presented to solve this problem by disregarding the physical model, transmission map contains significant haze density information, which simply be ignored. In article, we propose a novel hierarchical density-aware network, consists of densely connected pyramid encoder, generator, and Laplacian decoder. proposed network incorporates estimation but alleviates constraint model....

Lung cancer remains a major global health challenge, and accurate pathological examination is crucial for early detection. This study aims to enhance hyperspectral image analysis by refining annotations at the cell level creating high-quality dataset of lung tumors. We address challenge coarse manual in datasets, which limit effectiveness deep learning models requiring precise labels training. propose semi-automated annotation refinement method that leverages data diagnosis. Specifically, we...

In this work, we present a multifield direct design method for ultrashort throw ratio projection optics. The allows us to directly calculate two freeform mirror profiles, which are fitted by odd polynomials and imported into an optical program as excellent starting point. Afterward, these mirrors represented XY polynomial surfaces further optimization. final configuration consists of off-the-shelf lens greatly shorten the regular distance from 2 m 48 cm 78.3 inch diagonal screen. values...

Reflective imaging systems form an important part of photonic devices such as spectrometers, telescopes, augmented and virtual reality headsets or lithography platforms. optics provide unparalleled spectral performance can be used to reduce overall volume weight. So far, most reflective designs have focused on two three reflections, while four-reflection freeform deliver a higher light throughput (faster F-number) well larger field-of-view (FOV). However, advanced optical design strategies...

Optical aberration is a ubiquitous degeneration in realistic lens-based imaging systems. aberrations are caused by the differences optical path length when light travels through different regions of camera lens with incident angles. The blur and chromatic manifest significant discrepancies system changes. This work designs transferable effective image simulation simple lenses via multi-wavelength, depth-aware, spatially-variant four-dimensional point spread functions (4D-PSFs) estimation...

Proper initialization of the nonlinear optimization is important to avoid local minima in phase diversity wavefront sensing (PDWS). An effective neural network based on low-frequency coefficients Fourier domain has proved determine a better estimate unknown aberrations. However, relies significantly training settings, such as imaging object and optical system parameters, resulting weak generalization ability. Here we propose generalized Fourier-based PDWS method by combining an...

Phase diversity wavefront sensing (PDWS) has been a successful approach to quantifying aberrations with only few intensity measurements and nonlinear optimization. However, the inherent non-convexity of inverse problem may lead stagnation at local minimum far from true solution. Proper initialization optimization is important avoid minima improve retrieval accuracy. In this paper, we propose an effective neural network based on low-frequency coefficients in Fourier domain determine better...

Most existing techniques that are typically used by specialists to image the cornea based on point, slit, or annular scanning due a narrow field of view. The difficulty in achieving larger view comes from convex shape human eyeball. Field curvature for refractive imaging system with positive power is negative and thus concave surface. In order full sclera snapshot imaging, we calculate qualified two- three-mirror solutions Seidel aberration theory. A solution further optimized as...

The conventional optical system design employs combinations of different lenses to combat aberrations, which usually leads considerable volume and weight. In this Letter, a tailored scheme that exploits state-of-the-art digital aberration correction algorithms in addition traditional optics is investigated. particular, the proposed method applied refractive telescopes by shifting burden correcting chromatic aberrations software. By enforcing cross-channel information transfer post-processing...

Depth mapping from binocular endoscopy images plays an important role in stereoscopic surgical treatment. Owing to the development of deep convolutional neural networks (CNNs), depth estimation models have achieved many exciting results fields autonomous driving and machine vision. However, application these methods endoscopic imaging is greatly limited by fact that not only are rare, but also unsatisfying features such as no texture, ground truth, bad contrast, high gloss. Aiming at solving...

Including an entrance pupil in optical systems provides clear benefits for balancing the overall performance of freeform and/or rotationally symmetric imaging systems. Current existing direct design methods that are based on perfect few discrete ray bundles not well suited wide field view In this paper, a three-dimensional multi-fields approach is proposed to balance full two-surface lens. The path lengths and image points numerous fields calculated during procedures, wherefore very initial...

Freeform tailoring optics are core components in advanced illumination and laser technologies. In this work, we propose an efficient streamlined design method to face the challenges its implementation. We integrate ray-mapping least squares calculations with freeform surface optimization tools embedded commercial optical software. As a result, complexity of beam shaping is significantly reduced by only adapting built-in merit function. demonstrate feasibility our approach through three...

We propose digital correction methods for recovering aberrated spectra in a portable low-cost miniaturized grating-based spectrometer, which is modelled optical and numerical simulations. To realize the spectrum recovery, different Point Spread Function (PSF) modelling approaches wavelength-dependent PSFs are proposed implemented into four algorithms, namely inverse Fourier Transform (IFT) deconvolution, Wiener Lucy-Richardson (LR) Landweber iterative algorithm. Results show that both LR...

This paper proposes and demonstrates a linearized model for phase diversity wavefront sensing, facilitating real-time processing much less data required training. Specifically, we find that the low-frequency Fourier coefficients of point spread function images are linearly proportional to pupil aberration under certain conditions. Simulation experimental results show can greatly reduce time several milliseconds by merely requiring hundreds training samples while maintaining comparatively...

In this work, a multifields optical design method aiming to calculate two high-order aspheric lens profiles with an embedded entrance pupil is proposed. This direct algorithm capable of partially coupling more than three ray bundles that enter the same only surfaces. Both infinite and finite conjugate objectives can be designed approach. Additional constraints such as surface continuity smoothness are taken into account smooth accurate contours described by point clouds. The calculated...

We demonstrate a miniaturized broadband spectrometer employing reconstruction algorithm for resolution enhancement. use an opto-digital co-design approach, by firstly designing optical system with certain residual aberrations and then correcting these digital algorithm. The proposed design provides less than 1.7 nm in the VIS-channel (400-790 nm) 3.4 NIR-channel (760-1520 nm). Tolerance analysis results show that components are within commercial class, ensuring cost-efficient design. build...

A complete theoretical treatment of sine-wave photothermal deflection spectroscopy in a sample that includes surface absorption and bulk is given. The results can be used to separate from absorption, the by almost all optical coatings measured this method. have been verified experimentally.

Phase diversity phase retrieval (PDPR) has been a popular technique for quantitatively measuring wavefront errors of optical imaging systems by extracting the information from several designated intensity measurements. As problem is inverse and non-convex in general, accuracy robustness most such algorithms rely greatly on initial conditions. In this work, we propose new strategy that combines Limited-Memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) with points generated k-means clustering...

We developed the first freeform lens used to generate spherical wavefront required for digital lensless holographic microscopy illumination with a 0.3 numerical aperture. The proposal performance is validated samples containing epithelial cheek cells.

A complete and rigorous theoretical treatment of the continuous wave photothermal deflection spectroscopy in a flowing medium is given. The results have been verified experimentally.