Search

We introduce a novel technique to mitigate the adverse effects of atmospheric turbulence on astronomical imaging. Utilizing a video-to-image neural network trained on simulated data, our method processes a sliding sequence of short-exposure ($\sim$0.2s) stellar field images to reconstruct an image devoid of both turbulence and noise. We demonstrate the method with simulated and observed stellar fields, and show that the brief exposure sequence allows the network to accurately associate speckles to their originating stars and effectively disentangle light from adjacent sources across a range of seeing conditions, all while preserving flux to a lower signal-to-noise ratio than an average stack. This approach results in a marked improvement in angular resolution without compromising the astrometric stability of the final image., Comment: Accepted for publication in MNRAS (advance copy available at https://doi.org/10.1093/mnras/stae1018). Project website available at https://dancecam.info/ . 20 pages, 17 figures

The Hubble diagram of type-Ia supernovae (SNe-Ia) provides cosmological constraints on the nature of dark energy with an accuracy limited by the flux calibration of currently available spectrophotometric standards. The StarDICE experiment aims at establishing a 5-stage metrology chain from NIST photodiodes to stars, with a targeted accuracy of \SI{1}{mmag} in $griz$ colors. We present the first two stages, resulting in the calibration transfer from NIST photodiodes to a demonstration \SI{150}{Mpixel} CMOS sensor (Sony IMX411ALR as implemented in the QHY411M camera by QHYCCD). As a side-product, we provide full characterization of this camera. A fully automated spectrophotometric bench is built to perform the calibration transfer. The sensor readout electronics is studied using thousands of flat-field images from which we derive stability, high resolution photon transfer curves and estimates of the individual pixel gain. The sensor quantum efficiency is then measured relative to a NIST-calibrated photodiode. Flat-field scans at 16 different wavelengths are used to build maps of the sensor response. We demonstrate statistical uncertainty on quantum efficiency below \SI{0.001}{e^-/\gamma} between \SI{387}{nm} and \SI{950}{nm}. Systematic uncertainties in the bench optics are controlled at the level of \SI{1e-3}{e^-/\gamma}. Uncertainty in the overall normalization of the QE curve is 1\%. Regarding the camera we demonstrate stability in steady state conditions at the level of \SI{32.5}{ppm}. Homogeneity in the response is below \SI{1}{\percent} RMS across the entire sensor area. Quantum efficiency stays above \SI{50}{\percent} in most of the visible range, peaking well above \SI{80}{\percent} between \SI{440}{nm} and \SI{570}{nm}. Differential non-linearities at the level of \SI{1}{\percent} are detected. A simple 2-parameter model is proposed to mitigate the effect., Comment: accepted for publication in A&A, correct author list

In this work, we propose two convolutional neural network classifiers for detecting contaminants in astronomical images. Once trained, our classifiers are able to identify various contaminants, such as cosmic rays, hot and bad pixels, persistence effects, satellite or plane trails, residual fringe patterns, nebulous features, saturated pixels, diffraction spikes, and tracking errors in images. They encompass a broad range of ambient conditions, such as seeing, image sampling, detector type, optics, and stellar density. The first classifier, MaxiM ask , performs semantic segmentation and generates bad pixel maps for each contaminant, based on the probability that each pixel belongs to a given contaminant class. The second classifier, MaxiTrack , classifies entire images and mosaics, by computing the probability for the focal plane to be affected by tracking errors. We gathered training and testing data from real data originating from various modern charged-coupled devices and near-infrared cameras, that are augmented with image simulations. We quantified the performance of both classifiers and show that M axi M ask achieves state-of-the-art performance for the identification of cosmic ray hits. Thanks to a built-in Bayesian update mechanism, both classifiers can be tuned to meet specific science goals in various observational contexts., Comment: 25 pages, 14 figures, accepted in A&A journal, https://github.com/mpaillassa/MaxiMask

Observatories and satellites around the globe produce tremendous amounts of imaging data to study many different astrophysical phenomena. The serendipitous observations of Solar System objects are a fortunate by-product which have often been neglected due to the lack of a simple yet efficient identification algorithm. Meanwhile, the determination of the orbit, chemical composition, and physical properties such as rotation period and 3D-shape of Solar System objects requires a large number of astrometry and multi-band photometry observations. Such observations are hidden in current and future astrophysical archives, and a method to harvest these goldmines is needed. This article presents an easy-to-implement, light-weight software package which detects bodies of the Solar System in astronomical images and measures their astrometry and photometry. The ssos pipeline is versatile, allowing for application to all kinds of observatory imaging products. The sole principle requirement is that the images observe overlapping areas of the sky within a reasonable time range. Both known and unknown Solar System objects are recovered, from fast-moving near-Earth asteroids to slow objects in the Kuiper belt. The high-level pipeline design and two test applications are described here, highlighting the versatility of the algorithm with both narrow-field pointed and wide-field survey observations. In the first study, 2,828 detections of 204 SSOs are recovered from publicly available images of the GTC OSIRIS Broad Band DR1. The false-positive ratio of SSO detections ranges from 0%-23% depending on the pipeline setup. The second test study utilizes the images of the first data release of J-PLUS, a 12-band optical survey. 4,606 SSO candidates are recovered, with a false-positive ratio of (2.0 +- 0.2)%., Comment: Accepted for publication in Astronomy&Computing

We developed a Deep Convolutional Neural Network (CNN), used as a classifier, to estimate photometric redshifts and associated probability distribution functions (PDF) for galaxies in the Main Galaxy Sample of the Sloan Digital Sky Survey at z < 0.4. Our method exploits all the information present in the images without any feature extraction. The input data consist of 64x64 pixel ugriz images centered on the spectroscopic targets, plus the galactic reddening value on the line-of-sight. For training sets of 100k objects or more ($\geq$ 20% of the database), we reach a dispersion $\sigma_{MAD}$<0.01, significantly lower than the current best one obtained from another machine learning technique on the same sample. The bias is lower than 0.0001, independent of photometric redshift. The PDFs are shown to have very good predictive power. We also find that the CNN redshifts are unbiased with respect to galaxy inclination, and that $\sigma_{MAD}$ decreases with the signal-to-noise ratio (SNR), achieving values below 0.007 for SNR >100, as in the deep stacked region of Stripe 82. We argue that for most galaxies the precision is limited by the SNR of SDSS images rather than by the method. The success of this experiment at low redshift opens promising perspectives for upcoming surveys., Comment: Submitted to A&A, Comments welcome, Model and Example available at "https://github.com/jpasquet/Photoz

Large scale imaging surveys will increase the number of galaxy-scale strong lensing candidates by maybe three orders of magnitudes beyond the number known today. Finding these rare objects will require picking them out of at least tens of millions of images and deriving scientific results from them will require quantifying the efficiency and bias of any search method. To achieve these objectives automated methods must be developed. Because gravitational lenses are rare objects reducing false positives will be particularly important. We present a description and results of an open gravitational lens finding challenge. Participants were asked to classify 100,000 candidate objects as to whether they were gravitational lenses or not with the goal of developing better automated methods for finding lenses in large data sets. A variety of methods were used including visual inspection, arc and ring finders, support vector machines (SVM) and convolutional neural networks (CNN). We find that many of the methods will be easily fast enough to analyse the anticipated data flow. In test data, several methods are able to identify upwards of half the lenses after applying some thresholds on the lens characteristics such as lensed image brightness, size or contrast with the lens galaxy without making a single false-positive identification. This is significantly better than direct inspection by humans was able to do. (abridged), Comment: 22 pages, 16 figures, 4 tables, accepted version for A&A

Current constraints on models of galaxy evolution rely on morphometric catalogs extracted from multi-band photometric surveys. However, these catalogs are altered by selection effects that are difficult to model, that correlate in non trivial ways, and that can lead to contradictory predictions if not taken into account carefully. To address this issue, we have developed a new approach combining parametric Bayesian indirect likelihood (pBIL) techniques and empirical modeling with realistic image simulations that reproduce a large fraction of these selection effects. This allows us to perform a direct comparison between observed and simulated images and to infer robust constraints on model parameters. We use a semi-empirical forward model to generate a distribution of mock galaxies from a set of physical parameters. These galaxies are passed through an image simulator reproducing the instrumental characteristics of any survey and are then extracted in the same way as the observed data. The discrepancy between the simulated and observed data is quantified, and minimized with a custom sampling process based on adaptive Monte Carlo Markov Chain methods. Using synthetic data matching most of the properties of a CFHTLS Deep field, we demonstrate the robustness and internal consistency of our approach by inferring the parameters governing the size and luminosity functions and their evolutions for different realistic populations of galaxies. We also compare the results of our approach with those obtained from the classical spectral energy distribution fitting and photometric redshift approach.Our pipeline infers efficiently the luminosity and size distribution and evolution parameters with a very limited number of observables (3 photometric bands). When compared to SED fitting based on the same set of observables, our method yields results that are more accurate and free from systematic biases., Comment: 24 pages, 12 figures, accepted for publication in A&A

Astronomical images from optical photometric surveys are typically contaminated with transient artifacts such as cosmic rays, satellite trails and scattered light. We have developed and tested an algorithm that removes these artifacts using a deep, artifact free, static sky coadd image built up through the median combination of point spread function (PSF) homogenized, overlapping single epoch images. Transient artifacts are detected and masked in each single epoch image through comparison with an artifact free, PSF-matched simulated image that is constructed using the PSF-corrected, model fitting catalog from the artifact free coadd image together with the position variable PSF model of the single epoch image. This approach works well not only for cleaning single epoch images with worse seeing than the PSF homogenized coadd, but also the traditionally much more challenging problem of cleaning single epoch images with better seeing. In addition to masking transient artifacts, we have developed an interpolation approach that uses the local PSF and performs well in removing artifacts whose widths are smaller than the PSF full width at half maximum, including cosmic rays, the peaks of saturated stars and bleed trails. We have tested this algorithm on Dark Energy Survey Science Verification data and present performance metrics. More generally, our algorithm can be applied to any survey which images the same part of the sky multiple times., Comment: 17 pages, 6 figures. Accepted for publication in Astronomy and Computing

Deep optical images are often crowded with overlapping objects. This is especially true in the cores of galaxy clusters, where images of dozens of galaxies may lie atop one another. Accurate measurements of cluster properties require deblending algorithms designed to automatically extract a list of individual objects and decide what fraction of the light in each pixel comes from each object. We present new software called the Gradient And INterpolation based deblender (GAIN) as a secondary deblender to improve deblending the images of cluster cores. This software relies on using image intensity gradient and using an image interpolation technique usually used to correct flawed terrestrial digital images. We test this software on Dark Energy Survey coadd images. GAIN helps extracting unbiased photometry measurement for blended sources. It also helps improving detection completeness while introducing only a modest amount of spurious detections. For example, when applied to deep images simulated with high level of deblending difficulties, this software improves detection completeness from 91% to 97% for sources above the 10? limiting magnitude at 25.3 mag. We expect this software to be a useful tool for cluster population measurements., Comment: Published on PASP

Visualizing and navigating through large astronomy images from a remote location with current astronomy display tools can be a frustrating experience in terms of speed and ergonomics, especially on mobile devices. In this paper, we present a high performance, versatile and robust client-server system for remote visualization and analysis of extremely large scientific images. Applications of this work include survey image quality control, interactive data query and exploration, citizen science, as well as public outreach. The proposed software is entirely open source and is designed to be generic and applicable to a variety of datasets. It provides access to floating point data at terabyte scales, with the ability to precisely adjust image settings in real-time. The proposed clients are light-weight, platform-independent web applications built on standard HTML5 web technologies and compatible with both touch and mouse-based devices. We put the system to the test and assess the performance of the system and show that a single server can comfortably handle more than a hundred simultaneous users accessing full precision 32 bit astronomy data., Comment: Published in Astronomy & Computing. IIPImage server available from http://iipimage.sourceforge.net . Visiomatic code and demos available from http://www.visiomatic.org/

The Dark Energy Survey (DES) is a 5000 deg2 grizY survey reaching characteristic photometric depths of 24th magnitude (10 sigma) and enabling accurate photometry and morphology of objects ten times fainter than in SDSS. Preparations for DES have included building a dedicated 3 deg2 CCD camera (DECam), upgrading the existing CTIO Blanco 4m telescope and developing a new high performance computing (HPC) enabled data management system (DESDM). The DESDM system will be used for processing, calibrating and serving the DES data. The total data volumes are high (~2PB), and so considerable effort has gone into designing an automated processing and quality control system. Special purpose image detrending and photometric calibration codes have been developed to meet the data quality requirements, while survey astrometric calibration, coaddition and cataloging rely on new extensions of the AstrOmatic codes which now include tools for PSF modeling, PSF homogenization, PSF corrected model fitting cataloging and joint model fitting across multiple input images. The DESDM system has been deployed on dedicated development clusters and HPC systems in the US and Germany. An extensive program of testing with small rapid turn-around and larger campaign simulated datasets has been carried out. The system has also been tested on large real datasets, including Blanco Cosmology Survey data from the Mosaic2 camera. In Fall 2012 the DESDM system will be used for DECam commissioning, and, thereafter, the system will go into full science operations., Comment: 12 pages, submitted for publication in SPIE Proceeding 8451-12

The Dark Energy Survey Data Management (DESDM) system will process and archive the data from the Dark Energy Survey (DES) over the five year period of operation. This paper focuses on a new adaptable processing framework developed to perform highly automated, high performance data parallel processing. The new processing framework has been used to process 45 nights of simulated DECam supernova imaging data, and was extensively used in the DES Data Challenge 4, where it was used to process thousands of square degrees of simulated DES data., Comment: 4 pages, 1 figure; submit to ADASS XVII