Astronomical Image Reconstruction Research Articles

Large-field Astronomical Image Restoration and Superresolution Reconstruction using Deep Learning

The existing astronomical image restoration and superresolution reconstruction methods have problems such as low efficiency and poor results when dealing with images possessing large fields of view. Furthermore, these methods typically only handle fixed-size images and require step-by-step processing, which is inconvenient. In this paper, a neural network called Res&RecNet is proposed for the restoration and superresolution reconstruction of astronomical images with large fields of view for direct imaging instruments. This network performs feature extraction, feature correction, and progressive generation to achieve image restoration and superresolution reconstruction. The network is constructed using fully convolutional layers, allowing it to handle images of any size. The network can be trained using small samples and can perform image restoration and superresolution reconstruction in an end-to-end manner, resulting in high efficiency. Experimental results show that the network is highly effective in terms of processing astronomical images with complex scenes, generating image restoration results that improve the peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) by 4.69 (dB)/0.073 and superresolution reconstruction results that improve the PSNR and SSIM by 1.97 (dB)/0.077 over those of the best existing algorithms, respectively.

A real-time lucky imaging algorithm based on Fourier transform and its implementation techniques

Abstract Lucky imaging is a high-angular-resolution astronomical image reconstruction technique which can effectively reduce or remove the effects of atmospheric turbulence. There have been successful real-time lucky imaging algorithms and corresponding implementations in the spatial domain. Although the results of lucky imaging algorithms in the frequency domain are better than in the spatial domain, until now there has been no corresponding real-time algorithm and implementation due to its complexity. This paper proposes a real-time lucky imaging algorithm based on Fourier transform that features real-time processing and dynamic updating and display. The algorithm is implemented on a small- or medium-scale field programmable gate array (FPGA) development board. In this system, a two-dimensional fast Fourier transform of the image is carried out by means of dimension reduction. In order to achieve real-time lucky imaging in the frequency domain, the sum of the squares of the real and imaginary parts are used to replace the amplitude for data selection, and the method of image grouping is used to speed up the data selection. Compared with other real-time lucky imaging algorithms in the spatial domain, this algorithm shows great advantages in image quality. Moreover, the system can carry out real-time lucky imaging processing in the frequency domain for more than 10000 frames of an original image of 512 × 512 pixels continuously, and dynamic display with an image of 128 × 128 pixels. The experimental results illustrate the validity of the proposed algorithm and the feasibility of the proposed implementation techniques for the FPGA-based algorithm.

射电天文图像重建算法研究综述

射电天文图像重建算法研究综述

A New Real-Time Lucky Imaging Algorithm and its Implementation Techniques

Lucky imaging is a high-angular resolution astronomical image reconstruction technique that can effectively reduce the impact of atmospheric turbulence on image quality and improve the imaging resolution of ground-based telescopes. Its key steps include image selection, registration and superposition. However, the lucky imaging algorithms based on a central processing unit (CPU) encounter difficulty accomplishing real-time processing; thus, they are post-processing methods and cannot meet the needs of on-site observers. Taking advantage of the parallelism and flexibility of the field programmable gate array (FPGA), this paper presents a new real-time lucky imaging algorithm that features real-time processing and dynamic updating and displaying. The algorithm consists of a dynamic sorting-free image selection algorithm, an improved registration and storage method, a parallel superposition algorithm, a parallel preprocessing method for noise suppression and cosmic ray removal, and a dynamic multithreshold display scheme. The simulation results show that the algorithm is feasible, effective and efficient. Compared with other lucky imaging algorithms based on FPGAs, this algorithm shows great advantages in clock consumption and on-chip resource consumption. Furthermore, it can be implemented on a small or medium-size development board of an FPGA. Moreover, the implemented FPGA system can perform real-time and dynamic lucky imaging for more than 10,000 frames of short-exposure images with an original format of 512 × 512 pixels continuously. The experimental results not only show the validity of the proposed algorithm but also demonstrate the feasibility of the proposed implementation techniques for the FPGA-based algorithm.

A parallel and automatically tuned algorithm for multispectral image deconvolution

ABSTRACT In the era of big data, radio astronomical image reconstruction algorithms are challenged to estimate clean images given limited computing resources and time. This article is driven by the need for large-scale image reconstruction for the future Square Kilometre Array (SKA), which will become in the next decades the largest low and intermediate frequency radio telescope in the world. This work proposes a scalable wide-band deconvolution algorithm called MUFFIN, which stands for ‘MUlti Frequency image reconstruction For radio INterferometry’. MUFFIN estimates the sky images in various frequency bands, given the corresponding dirty images and point spread functions. The reconstruction is achieved by minimizing a data fidelity term and joint spatial and spectral sparse analysis regularization terms. It is consequently non-parametric w.r.t. the spectral behaviour of radio sources. MUFFIN algorithm is endowed with a parallel implementation and an automatic tuning of the regularization parameters, making it scalable and well suited for big data applications such as SKA. Comparisons between MUFFIN and the state-of-the-art wide-band reconstruction algorithm are provided.

High noise astronomical image denoising via 2G-bandelet denoising compressed sensing

In deep space exploration, high resolution astronomical image captured is often contaminated by various cosmic noise signals during its shooting and long distance transmission, which has brought inconvenience to astronomical image analysis. The famous compressed sensing (CS) proposed by Candes et al. can successfully solve the problem of high resolution astronomical image compression and low noise reconstruction. In this paper, we further concern how to reconstruct a high quality image from a high resolution and high noise astronomical image. A 2G-bandelet denoising compressed sensing (BDCS) is first proposed based on the advantage of CS in image denoising and the superior ability of 2G-bandelet in sparse representation of astronomical images, then iterative bandelet thresholding (IBT-BTCS) algorithm based on BDCS is proposed for high resolution and high noise astronomical image reconstruction. Firstly, an iterative bandelet thresholding method is designed to obtain optimal approximation of original image; Secondly, to further improve the reconstructed image quality, group sparse total variation with stepsize constraints (GSTV-SC) method is proposed to adjust the reconstructed astronomical image in each iteration. The simulation results show that the proposed algorithm can quickly reconstruct a high quality astronomical image only using a few observations, preserve more astronomical image details and effectively solve high noise astronomical image denoising problem.

基于集中稀疏表示的天文图像超分辨率重建

基于集中稀疏表示的天文图像超分辨率重建

Fast compression and reconstruction of astronomical images based on compressed sensing

With the fast increase in the resolution of astronomical images, the question of how to process and transfer such large images has become a key issue in astronomy. We propose a new real-time compression and fast reconstruction algorithm for astronomical images based on compressive sensing techniques. We first reconstruct the original signal with fewer measurements, according to its compressibility. Then, based on the characteristics of astronomical images, we apply Daubechies orthogonal wavelets to obtain a sparse representation. A matrix representing a random Fourier ensemble is used to obtain a sparse representation in a lower dimensional space. For reconstructing the image, we propose a novel minimum total variation with block adaptive sensing to balance the accuracy and computation time. Our experimental results show that the proposed algorithm can efficiently reconstruct colorful astronomical images with high resolution and improve the applicability of compressed sensing.

Joint myopic deconvolution

Astronomical image reconstruction is an inverse problem based on the knowledge of the point-spread function (PSF). However, this knowledge is often only partial, and a myopic deconvolution process is required for reaching the estimation of the solution. In this paper we propose a new statistical model which incorporates the presence of noise both on the image of the object to retrieve and on the ‘measured’ PSF. This technique also takes into account the nonnegativity constraint on the solution and on the PSF. Deconvolution results are presented for simulated data. A comparison between the classical algorithms and that proposed in this paper is given. This method can also be extended when different measures of PSF with different sizes are available.

Iterative methods for the reconstruction of astronomical images with high dynamic range

In most cases astronomical images contain objects with very different intensities such as bright stars combined with faint nebulae. Since the noise is mainly due to photon counting (Poisson noise), the signal-to-noise ratio may be very different in different regions of the image. Moreover, the bright and faint objects have, in general, different angular scales. These features imply that the iterative methods which are most frequently used for the reconstruction of astronomical images, namely the Richardson–Lucy Method (RLM), also known in tomography as Expectation Maximization (EM) method, and the Iterative Space Reconstruction Algorithm (ISRA) do not work well in these cases. Also standard regularization approaches do not provide satisfactory results since a kind of adaptive regularization is required, in the sense that one needs a different regularization for bright and faint objects. In this paper we analyze a number of regularization functionals with this particular kind of adaptivity and we propose a simple modification of RLM and ISRA which takes into account these regularization terms. The preliminary results on a test object are promising.

Speckle imaging signal-to-noise ratio performance as a function of frame integration time

In the past 10 years astronomical image reconstruction based on a variety of speckle techniques has become a popular means of enhancing and improving the resolution of turbulence-degraded images. These techniques are based on the Fourier processing of a large number of turbulence-degraded snapshots or frames of the irradiance in the system’s image plane. The number of snapshots needed is largely a function of the signal-to-noise ratio (SNR) of the Fourier components of a single snapshot. Estimating these Fourier components is hampered by the inherent noise induced by the photon detection process and the random effect of the turbulent atmosphere. In most cases the SNR is much less than unity, and many frames are averaged to improve the overall SNR. It is well established that if the frames are uncorrelated then the SNR improves in comparison with the single-frame SNR by a factor of m, where m is the number of frames averaged. This fact implies that the smallest exposure time possible is desirable in order for m to be maximized for a given observation time. On the other hand, because of finite photon flux levels and read noise effects it can be shown that the exposure time should be increased at the cost of reducing m. Results from a detailed SNR analysis of estimating the modulus of an object’s Fourier spectrum from turbulence-distorted images are described. Unlike previous analyses, this work takes into account the proper temporal correlation properties of the atmosphere and the spatial frequency being estimated as well as the interframe correlations that degrade the SNR improvement factor from the ideal case of m.

Preliminary results of an astronomical image reconstruction experiment with speckle masking methodAAnS 33/3 (1992) 280–286

Preliminary results of an astronomical image reconstruction experiment with speckle masking methodAAnS 33/3 (1992) 280–286

A SIMPLE, PHOTOMETRICALLY ACCURATE ALGORITHM FOR DECONVOLUTION OF OPTICAL IMAGES

ABSTRACT a noise model, has been tested on simulated HST data. This straightforward algorithm displays many desirable features for astronomical image reconstruction, as shown by rigorous numerical tests: (1) The photometric scale of the data is preserved accurately. (2) Information on the local noise level due to Poisson statistics in the input data is preserved. (3) It can accommodate a spatially variable point-spread function. Some of these properties are not shared by any widely used deconvolution algorithm. Some considerations of computing speed, and application to ground-based images, are briefly discussed.

Atmospheric isoplanatism and astronomical image reconstruction on Mauna Kea

Atmospheric isoplanatism for visual wavelength image-reconstruction applications was measured on Mauna Kea in Hawaii. For most nights the correlation of the transform functions is substantially wider than the long-exposure transform function at separations up to 30 arcsec. Theoretical analysis shows that this is reasonable if the mean Fried parameter is approximately 30 cm at 5500 Å. Reconstructed image quality may be described by a Gaussian with a FWHM of λ/s0. Under average conditions, s0 (30 arcsec) exceeds 55 cm at 7000 Å. The results show that visual image quality in the 0.1–0.2-arcsec range is obtainable over much of the sky with large ground-based telescopes on this site.

Astronomical image reconstruction via slit aperture telescope

The interest of a ground based Slit Aperture Telescope (SAT) which operates with a long thin pupil has been underlined for applications in the field of high angular resolution observations by speckle-interferometric techniques. The SAT lies halfway between the Michelson stellar interferometer and the standard telescope with circular aperture. The authors present the principle of an image processing algorithm which could be used associated with a space-SAT for astronomical image reconstruction. A full coverage of the two dimensional Fourier plane of the astronomical object can be obtained by rotating the SAT around its optical axis. The method of image reconstruction is analogous to those used in computerized tomography and uses Fast Fourier Transform algorithm. The interest of the use of a rotating space-SAT among synthetic apertures is its capability of giving a reconstructed image of the astronomical object simultaneously in several light wavelengths, each colored image being comparable with the others for fruitful astrophysical applications. They suggest the ways to perform a ground based TPF for astronomical imagery.

Effects of photon noise on speckle image reconstruction with the Knox-Thompson algorithm

An analysis of the effects of photon noise on astronomical speckle image reconstruction using the Knox-Thompson algorithm is presented. It is shown that the quantities resulting from the speckle average are biased, but that the biases are easily estimated and compensated. Calculations are also made of the convergence rate for the speckle average as a function of the source brightness. An illustration of the effects of photon noise on the image recovery process is included.

Shift-and-dd-Type Algorithms and their Application to Capella

Bates and Cady (1980) proposed a method for astronomical image reconstruction that involves superimposing regions about the brightest pixels of each frame produced by speckle interferometric techniques. Unlike the method of Lynds, Worden, and Harvey (1976), no attempt is made to center apparent speckle images before superposition. This method is hereafter referred to as “shift-and-add” (SAA). The simplicity of this method makes it highly suitable for analysis and simulation.This paper summarizes the application of extensions of SAA to the problem of determining reliable double star intensity ratios. Theoretical and simulation results presented in a previous paper (hereafter called “paper I”) by Bagnuolo (1982) will be summarized. In addition, the results of the application of SAA-type algorithms to Capella will be presented.

Astronomical image reconstruction

Astronomical image reconstruction