RESUMO
Image dehazing is a typical low-level visual task. With the continuous improvement of network performance and the introduction of various prior knowledge, the ability of image dehazing is becoming stronger. However, the existing dehazing methods have problems such as the inability to obtain real shooting datasets, unreliable dehazing processes, and the difficulty to deal with complex lighting scenes. To solve these problems, we propose a new haze model combining the optical scattering model and the computer graphics rendering. Based on the new haze model, we propose a high-quality and widely applicable dehazing dataset generation pipeline that does not require paired-data training and prior knowledge. We reconstruct the three-dimensional fog space with array camera and remove haze by thresholding voxel deletion. We use the Unreal Engine 5 to generate simulation datasets and the real shooting in laboratory to verify the effectiveness and the reliability of our generation pipeline. Through our pipeline, we can obtain wonderful dehaze results and dehaze datasets under various complex outdoors lighting conditions. We also propose a dehaze dataset enhancement method based on voxel control. Our pipeline and data enhancement are suitable for the latest algorithm model, these solutions can obtain better visual effects and objective indicators.
RESUMO
Hyperspectral imaging is a critical tool for gathering spatial-spectral information in various scientific research fields. As a result of improvements in spectral reconstruction algorithms, significant progress has been made in reconstructing hyperspectral images from commonly acquired RGB images. However, due to the limited input, reconstructing spectral information from RGB images is ill-posed. Furthermore, conventional camera color filter arrays (CFA) are designed for human perception and are not optimal for spectral reconstruction. To increase the diversity of wavelength encoding, we propose to place broadband encoding filters in front of the RGB camera. In this condition, the spectral sensitivity of the imaging system is determined by the filters and the camera itself. To achieve an optimal encoding scheme, we use an end-to-end optimization framework to automatically design the filters' transmittance functions and optimize the weights of the spectral reconstruction network. Simulation experiments show that our proposed spectral reconstruction network has excellent spectral mapping capabilities. Additionally, our novel joint wavelength encoding imaging framework is superior to traditional RGB imaging systems. We develop the deeply learned filter and conduct actual shooting experiments. The spectral reconstruction results have an attractive spatial resolution and spectral accuracy.
RESUMO
In telescopic systems consisting of Alvarez lenses, chromatic aberrations vary with the magnifications and the fields of view. Computational imaging has developed rapidly in recent years, therefore we propose a method of optimizing the DOE and the post-processing neural network in 2 stages for achromatic aberrations. We apply the iterative algorithm and the gradient descent method to optimize the DOE, respectively, and then adopt U-Net to further optimize the results. The results show that the optimized DOEs improve the results, the gradient descent optimized DOE with U-Net performs the best and has a very robust and good performance in the case of simulated chromatic aberrations. The results also verify the validity of our algorithm.
RESUMO
Hyperspectral imaging attempts to determine distinctive information in spatial and spectral domain of a target. Over the past few years, hyperspectral imaging systems have developed towards lighter and faster. In phase-coded hyperspectral imaging systems, a better coding aperture design can improve the spectral accuracy relatively. Using wave optics, we post an equalization designed phase-coded aperture to achieve desired equalization point spread functions (PSFs) which provides richer features for subsequent image reconstruction. During the reconstruction of images, our raised hyperspectral reconstruction network, CAFormer, achieves better results than the state-of-the-art networks with less computation by substituting self-attention with channel-attention. Our work revolves around the equalization design of the phase-coded aperture and optimizes the imaging process from three aspects: hardware design, reconstruction algorithm, and PSF calibration. Our work is putting snapshot compact hyperspectral technology closer to a practical application.
RESUMO
Alvarez lenses are known for their ability to achieve a broad range of optical power adjustment by utilizing complementary freeform surfaces. However, these lenses suffer from optical aberrations, which restrict their potential applications. To address this issue, we propose a field of view (FOV) attention image restoration model for continuous zooming. In order to simulate the degradation of optical zooming systems based on Alvarez lenses (OZA), a baseline OZA is designed where the polynomial for the Alvarez lenses consists of only three coefficients. By computing spatially varying point spread functions (PSFs), we simulate the degraded images of multiple zoom configurations and conduct restoration experiments. The results demonstrate that our approach surpasses the compared methods in the restoration of degraded images across various zoom configurations while also exhibiting strong generalization capabilities under untrained configurations.
RESUMO
Atmospheric turbulence, a pervasive and complex physical phenomenon, challenges optical imaging across various applications. This paper presents the Alternating Spatial-Frequency (ASF)-Transformer, a learning-based method for neutralizing the impact of atmospheric turbulence on optical imaging. Drawing inspiration from split-step propagation and correlated imaging principles, we propose the Alternating Learning in Spatial and Frequency domains (LASF) mechanism. This mechanism utilizes two specially designed transformer blocks that alternate between the spatial and Fourier domains. Assisted by the proposed patch FFT loss, our model can enhance the recovery of intricate textures without the need for generative adversarial networks (GANs). Evaluated across diverse test mediums, our model demonstrated state-of-the-art performance in comparison to recent methods. The ASF-Transformer diverges from mainstream GAN-based solutions, offering a new strategy to combat image degradation introduced by atmospheric turbulence. Additionally, this work provides insights into neural network architecture by integrating principles from optical theory, paving the way for innovative neural network designs in the future.
RESUMO
Due to severe noise and extremely low illuminance, restoring from low-light images to normal-light images remains challenging. Unpredictable noise can tangle the weak signals, making it difficult for models to learn signals from low-light images, while simply restoring the illumination can lead to noise amplification. To address this dilemma, we propose a multi-stage model that can progressively restore normal-light images from low-light images, namely Dark2Light. Within each stage, We divide the low-light image enhancement (LLIE) into two main problems: (1) illumination enhancement and (2) noise removal. Firstly, we convert the image space from sRGB to linear RGB to ensure that illumination enhancement is approximately linear, and design a contextual transformer block to conduct illumination enhancement in a coarse-to-fine manner. Secondly, a U-Net shaped denoising block is adopted for noise removal. Lastly, we design a dual-supervised attention block to facilitate progressive restoration and feature transfer. Extensive experimental results demonstrate that the proposed Dark2Light outperforms the state-of-the-art LLIE methods both quantitatively and qualitatively.
RESUMO
Dynamic distortion is one of the most critical factors affecting the experience of automotive augmented reality head-up displays (AR-HUDs). A wide range of views and the extensive display area result in extraordinarily complex distortions. Existing methods based on the neural network first obtain distorted images and then get the predistorted data for training mostly. This paper proposes a distortion prediction framework based on the neural network. It directly trains the network with the distorted data, realizing dynamic adaptation for AR-HUD distortion correction and avoiding errors in coordinate interpolation. Additionally, we predict the distortion offsets instead of the distortion coordinates and present a field of view (FOV)-weighted loss function based on the spatial-variance characteristic to further improve the prediction accuracy of distortion. Experiments show that our methods improve the prediction accuracy of AR-HUD dynamic distortion without increasing the network complexity or data processing overhead.
RESUMO
This paper proposes an optimized design of the Alvarez lens by utilizing a combination of three fifth-order X-Y polynomials. It can effectively minimize the curvature of the lens surface to meet the manufacturing requirements. The phase modulation function and aberration of the proposed lens are evaluated by using first-order optical analysis. Simulations compare the proposed lens with the traditional Alvarez lens in terms of surface curvature, zoom capability, and imaging quality. The results demonstrate the exceptional performance of the proposed lens, achieving a remarkable 26.36% reduction in the maximum curvature of the Alvarez lens (with a coefficient A value of 4×10-4 and a diameter of 26 mm) while preserving its original zoom capability and imaging quality.
RESUMO
Thin devices with large areas have strong and omnidirectional absorption over a wide bandwidth and are in demand for applications such as energy harvesting, structural color, and vehicle LiDAR (laser detection and ranging). Despite persistent efforts in the design and fabrication of such devices, the simultaneous realization of all these desired properties remains a challenge. In this study, a 190-nm-thick metasurface with an area of 3 cm2, incorporating dielectric cylinder arrays, a chromium layer, a silicon nitride (SiNx) layer, and an aluminum layer is theoretically and experimentally demonstrated. The developed device achieves an average absorptivity of â¼99% (97% in the experiment) in the entire visible spectrum 400-700 nm. Moreover, it exhibits strong absorption over a wide range of incident angles (â¼91% and 90% at 60° in the calculation and experiment, respectively). Importantly, the feasibility of applying the developed metasurface absorber to solar thermophotovoltaics and vehicle LiDAR (laser detection and ranging) has been explored. Moreover, the photoresist can be replaced by other glues and easily scaled up to a large area using the roll-to-roll nanoimprinting process. With the excellent spectral properties and performance, this device is promising for large-area applications.
RESUMO
In mobile photography applications, limited volume constraints the diversity of optical design. In addition to the narrow space, the deviations introduced in mass production cause random bias to the real camera. In consequence, these factors introduce spatially varying aberration and stochastic degradation into the physical formation of an image. Many existing methods obtain excellent performance on one specific device but are not able to quickly adapt to mass production. To address this issue, we propose a frequency self-adaptive model to restore realistic features of the latent image. The restoration is mainly performed in the Fourier domain and two attention mechanisms are introduced to match the feature between Fourier and spatial domain. Our method applies a lightweight network, without requiring modification when the fields of view (FoV) changes. Considering the manufacturing deviations of a specific camera, we first pre-train a simulation-based model, then finetune it with additional manufacturing error, which greatly decreases the time and computational overhead consumption in implementation. Extensive results verify the promising applications of our technique for being integrated with the existing post-processing systems.
RESUMO
Diffractive optical elements play a crucial role in the miniaturization of the optical systems, especially in correcting achromatic aberration. Considering the rapidity and validity of the design method, we propose a fast method for designing broadband achromatic diffractive optical elements. Based on the direct binary search algorithm, some improvements have been made including the selection of the initial height map to mitigate the uncertainty, the reduction of the variations to accelerate the optimization and the increase of sampling rate to deal with the large operation bandwidth. The initial height map is calculated instead of random initial value. Due to different regions of the height map contributing to point spread functions differently, the variations are reduced to speed up the optimization. The large operation bandwidth is solved by increasing the sampling rate at unfitted wavelengths instead of setting weighting coefficients. We demonstrate via simulations that our method is effective through several examples. The design of broadband achromatic diffractive optical elements can be quickly achieved by our method.
RESUMO
Photonic integrated interferometric imaging (PIII) is an emerging technique that uses far-field spatial coherence measurements to extract intensity information from a source to form an image. At present, low sampling rate and noise disturbance are the main factors hindering the development of this technology. This paper implements a deep learning-based method to improve image quality. Firstly, we propose a frequency-domain dataset generation method based on imaging principles. Secondly, spatial-frequency dual-domain fusion networks (SFDF-Nets) are presented for image reconstruction. We utilize normalized amplitude and phase to train networks, which reduces the difficulty of network training using complex data. SFDF-Nets can fuse multi-frame data captured by rotation sampling to increase the sampling rate and generate high-quality spatial images through dual-domain supervised learning and frequency domain fusion. Furthermore, we propose an inverse fast Fourier transform loss (IFFT loss) for network training in the frequency domain. Extensive experiments show that our method improves PSNR and SSIM by 5.64 dB and 0.20, respectively. Our method effectively improves the reconstructed image quality and opens a new dimension in interferometric imaging.
RESUMO
Rotated rectangular aperture imaging has many advantages in large aperture telephoto systems due to its lower cost and lower complexity. This technology makes it possible to build super large aperture telescopes. In this paper, we combine the ideas of deblurring with rotated rectangular aperture imaging and propose an image synthesis algorithm based on multi-frame deconvolution. In the specific reconstruction process, Hyper-Laplacian priors and sparse priors are used, and an effective solution is developed. The simulation and real shooting experiments show that our algorithm has excellent performance in visual effect and objective evaluation. The synthetic images are significantly sharper than the results of the existing methods.
RESUMO
Mask based lensless imagers have huge application prospects due to their ultra-thin body. However, the visual perception of the restored images is poor due to the ill conditioned nature of the system. In this work, we proposed a deep analytic network by imitating the traditional optimization process as an end-to-end network. Our network combines analytic updates with a deep denoiser prior to progressively improve lensless image quality over a few iterations. The convergence is proven mathematically and verified in the results. In addition, our method is universal in non-blind restoration. We detailed the solution for the general inverse problem and conducted five groups of deblurring experiments as examples. Both experimental results demonstrate that our method achieves superior performance against the existing state-of-the-art methods.
RESUMO
Under-display imaging technique was recently proposed to enlarge the screen-to-body ratio for full-screen devices. However, existing image restoration algorithms have difficulty generalizing to real-world under-display (UD) images, especially to images containing strong light sources. To address this issue, we propose a novel method for building a synthetic dataset (CalibPSF dataset) and introduce a two-stage neural network to solve the under-display imaging degradation problem. The CalibPSF dataset is generated using the calibrated high dynamic range point spread function (PSF) of the under-display optical system and contains various simulated light sources. The two-stage network solves the color distortion and diffraction degradation in order. We evaluate the performance of our algorithm on our captured real-world test set. Comprehensive experiments demonstrate the superiority of our method in different dynamic range scenes.
RESUMO
Mask-based lensless imaging cameras have many applications due to their smaller volumes and lower costs. However, due to the ill-nature of the inverse problem, the reconstructed images have low resolution and poor quality. In this article, we use a mask based on almost perfect sequence which has an excellent autocorrelation property for lensless imaging and propose a Learned Analytic solution Net for image reconstruction under the framework of unrolled optimization. Our network combines a physical imaging model with deep learning to achieve high-quality image reconstruction. The experimental results indicate that our reconstructed images at a resolution of 512 × 512 have excellent performances in both visual effects and objective evaluations.
RESUMO
This paper presents an adaptive reconstruction method for coded aperture temporal compressive imaging. A pixel-wise equal-exposure coding strategy is first implemented to induce speckle-like features in the moving area of the measured image. A moving area detection method is then proposed to reconstruct the adaptively segmented moving areas into a series of video frames, which are filled back into the stationary clear background. Both simulation and experimental results demonstrate that the proposed method significantly reduces the time consumption of video reconstruction without degradation of image quality.
RESUMO
A new method for path transverse wind velocity survey was introduced by analyzing time lagged covariance function of different separation sub-apertures of Hartmann wavefront sensor. A theoretical formula was logically deduced for the light propagation path transverse wind velocity profile. According to the difference of path weighting function for different sub apertures spacing, how to select reasonable path weighting functions was analyzed. Using a Hartmann wavefront sensor, the experiment for measuring path transverse velocity profile along 1 000 m horizontal propagating path was carried out for the first time to our knowledge. The experiment results were as follows. Path transverse averaged velocity from sensor had a good consistency with transverse velocity from the wind anemometer sited near the path receiving end. As the path was divided into two sections, the path transverse velocity of the first section had also a good consistency with that of the second one. Because of different specific underlaying surface of light path, the former was greater than the later over all experiment period. The averaged values were 1.273 and 0.952 m x s(-1) respectively. The path transverse velocity of second section and path transverse averaged velocity had the same trend of decrease and increase with time. The correlation coefficients reached 0.86.
RESUMO
Nighttime image dehazing presents unique challenges due to the unevenly distributed haze caused by the color change of artificial light sources. This results in multiple interferences, including atmospheric light, glow, and direct light, which make the complex scattering haze interference difficult to accurately distinguish and remove. Additionally, obtaining pairs of high-definition data for fog removal at night is a difficult task. These challenges make nighttime image dehazing a particularly challenging problem to solve. To address these challenges, we introduced the haze scattering formula to more accurately express the haze in three-dimensional space. We also proposed a novel data synthesis method using the latest CG textures and lumen lighting technology to build scenes where various hazes can be seen clearly through ray tracing. We converted the complex 3D scattering relationship transformation into a 2D image dataset to better learn the mapping from 3D haze to 2D haze. Additionally, we improved the existing neural network and established a night haze intensity evaluation label based on the idea of optical PSF. This allowed us to adjust the haze intensity of the rendered dataset according to the intensity of the real haze image and improve the accuracy of dehazing. Our experiments showed that our data construction and network improvement achieved better visual effects, objective indicators, and calculation speed.