Background–source separation in astronomical images with Bayesian probability theory – I. The method

A probabilistic technique for the joint estimation of background and sources with the aim of detecting faint and extended celestial objects is described. Bayesian probability theory is applied to gain insight into the co-existence of background and sources through a probabilistic two-component mixture model, which provides consistent uncertainties of background and sources. A multiresolution analysis is used for revealing faint and extended objects in the frame of the Bayesian mixture model. All the revealed sources are parametrized automatically providing source position, net counts, morphological parameters and their errors.
We demonstrate the capability of our method by applying it to three simulated data sets characterized by different background and source intensities. The results of employing two different prior knowledge on the source signal distribution are shown. The probabilistic method allows for the detection of bright and faint sources independently of their morphology and the kind of background. The results from our analysis of the three simulated data sets are compared with other source detection methods. Additionally, the technique is applied to ROSAT All-Sky Survey data.     

图像移位堆叠方法用于暗天然卫星的天体测量

暗弱天然卫星与主带小行星相比,具有亮度低、速度变化快的特点.在观测这类天体时,不能简单地延长曝光时间来提高其信噪比.尝试观测多幅短曝光的CCD (chargecoupled device)图像,采用移位堆叠(shift-and-add)方法,希望提高目标成像的信噪比,获得暗弱天然卫星的精确测量结果.使用2018年4月9—12日夜间,中国科学院云南天文台1 m望远镜(1 m望远镜)拍摄的木星5颗暗卫星的229幅CCD图像,实施了移位堆叠试验.为了验证结果的正确性,与相近日期中国科学院云南天文台2.4 m望远镜(2.4 m望远镜)观测的相同木卫图像的测量结果进行了比较和分析.位置归算采用了JPL (Jet Propulsion Laboratory)历表.结果表明,对CCD图像使用移位堆叠方法,通过叠加约10幅曝光时间100 s的图像, 1 m望远镜能观测暗至19等星的不规则天然卫星,而且测量的准确度与2.4 m望远镜的测量结果有良好的一致性.     

A new gravitational lens from the MUSCLES survey: ULAS J082016.1+081216

We present observations of a new double-image gravitational lens system, ULAS J082016.1+081216, of image separation 2.3 arcsec and high (∼6) flux ratio. The system is selected from the Sloan Digital Sky Survey (SDSS) spectroscopic quasar list using new high-quality images from the UKIRT (United Kingdom Infrared Telescope) Deep Sky Survey (UKIDSS). The lensed quasar has a source redshift of 2.024, and we identify the lens galaxy as a faint red object of redshift  0.803 ± 0.001  . Three other objects from the UKIDSS survey, selected in the same way, were found not to be lens systems. Together with the earlier lens found using this method, the SDSS–UKIDSS lenses have the potential to significantly increase the number of quasar lenses found in SDSS, to extend the survey to higher flux ratios and lower separations, and to give greater completeness which is important for statistical purposes.     

基于亚像素图像平移叠加提高大倾角同步卫星信噪比的方法研究

大倾角地球同步轨道(GEO)卫星相对地面并非完全静止,其星下点做南北方向的"8"字周期运动,倾角越大运动范围越大.这会降低地面望远镜对其进行凝视观测时的有效曝光时间,无法获得较高的信噪比.对地球同步轨道的监测中,在硬件条件不变的前提下,提出亚像素图像平移叠加方法.根据目标的运动速度和相邻帧图像的时间间隔,在亚像素尺度平移并对齐多幅图像,使GEO目标星象在图像序列中的位置重合,通过叠加多幅序列图像来提高目标信噪比,从而达到提升整个系统探测能力的目的.实测图像叠加结果表明,该方法可以显著提高该类目标的信噪比.叠加5幅图像时,整数像素叠加图像的信噪比约为原图像的1.7倍,而亚像素叠加图像的信噪比是原图像的2倍左右.     

Digital stacking of photographic plates with SuperCOSMOS

Photographic Schmidt plates are among the most effective tools in wide-field astronomy. One of the principal difficulties of photographic plates when compared to modern detectors is lack of image depth. We present a technique for stacking plates digitized using the SuperCOSMOS microdensitometer, aimed at maximizing the signal-to-noise ratio in faint images. The efficacy of several image combination algorithms is tested by stacking plate frames in the presence of spurious images. We find that an 'average sigma clipping' type pixel rejection in conjunction with our weighting scheme is most effective in delivering a clean, high signal-to-noise ratio stack. The gain in limiting magnitude obtained from stacking is found to be consistent with that expected: Δ M ∼1.5 for a stack of 16 good-quality plates.     

Use of Mathematic Morphological Operators for Processing the Smeared Images of A Camera without Shutter

In the observation of space objects with a full-frame transfer CCD camera, the camera shutter is often removed, and therefore the smeared images are produced, which affect seriously the object detection and location. Based on the difference of geometrical morphology between star images and smears, a mathematical morphology method is proposed to process the smeared images. By comparing the position accuracies of stars and space objects before and after processing, it is veri?ed that the morphological method can effectively remove the smears from astronomical images, and can signi?cantly improve the detection rates and position accuracies of stars and moving objects. The result indicates that this is an effective method to eliminate the smearing effect.     

Solar System Science with LSST

The Large Synoptic Survey Telescope (LSST) will provide a unique tool to study moving objects throughout the solar system, creating massive catalogs of Near Earth Objects (NEOs), asteroids, Trojans, TransNeptunian Objects (TNOs), comets and planetary satellites with well-measured orbits and high quality, multi-color photometry accurate to 0.005 magnitudes for the brightest objects. In the baseline LSST observing plan, back-to-back 15-second images will reach a limiting magnitude as faint as r = 24.7 in each 9.6 square degree image, twice per night; a total of approximately 20,000 square degrees of the sky will be imaged in multiple filters, with revisits about every 3 nights over several months of each year.     

Exploring three faint source detections methods for aperture synthesis radio images

Wide-field radio interferometric images often contain a large population of faint compact sources. Due to their low intensity/noise ratio, these objects can be easily missed by automated detection methods, which have been classically based on thresholding techniques after local noise estimation. The aim of this paper is to present and analyse the performance of several alternative or complementary techniques to thresholding. We compare three different algorithms to increase the detection rate of faint objects. The first technique consists of combining wavelet decomposition with local thresholding. The second technique is based on the structural behaviour of the neighbourhood of each pixel. Finally, the third algorithm uses local features extracted from a bank of filters and a boosting classifier to perform the detections. The methods’ performances are evaluated using simulations and radio mosaics from the Giant Metrewave Radio Telescope and the Australia Telescope Compact Array. We show that the new methods perform better than well-known state of the art methods such as SExtractor, SAD and DUCHAMP at detecting faint sources of radio interferometric images.     

Automatic morphological classification of galaxy images

We describe an image analysis supervised learning algorithm that can automatically classify galaxy images. The algorithm is first trained using manually classified images of elliptical, spiral and edge-on galaxies. A large set of image features is extracted from each image, and the most informative features are selected using Fisher scores. Test images can then be classified using a simple Weighted Nearest Neighbour rule such that the Fisher scores are used as the feature weights. Experimental results show that galaxy images from Galaxy Zoo can be classified automatically to spiral, elliptical and edge-on galaxies with an accuracy of ∼90 per cent compared to classifications carried out by the author. Full compilable source code of the algorithm is available for free download, and its general-purpose nature makes it suitable for other uses that involve automatic image analysis of celestial objects.     

A search for nearby counterparts to the moving objects in the Hubble Deep Field

Ibata et al. have recently discovered very faint, moving objects in the Hubble Deep Field (HDF). The number, apparent magnitudes and proper motions of these objects are consistent with old white dwarfs making up part of the Galactic dark halo. We review a number of ground-based proper motion surveys in which nearby dark-halo white dwarfs might be present, if they have the colours and absolute magnitudes proposed. No such objects have been found, whereas we argue here that several times more would be expected than in the HDF. We conclude that it is unlikely that hydrogen-atmosphere white dwarfs make up a significant fraction of the Galactic dark matter. No limits can be placed as yet on helium-atmosphere dwarfs from optical searches.     

同幅双速跟踪成像CCD相机数字控制器的设计

在介绍了同幅双速跟踪成像CCD相机的工作原理及其对相机数字控制器的要求之后,阐述了该数字控制器的设计思想,基本结构,并给出了仿真及测试结果。针对仿真及测试过程中出现的问题作了较为详细的分析,并提出了解决的方法。通过软件仿真及硬件测试,表明其设计是基本正确的。     

基于大视场反卷积的天文图像叠加方法

地基望远镜在成像过程中,由于受大气湍流、望远镜静态像差、跟踪误差、指向误差及视场变化的影响,不同视场区域的PSF (Point Spread Function)具有差异;同时,不同望远镜获取的图像PSF也存在差异.将多个望远镜获取的星象直接叠加至相同的区域后,图像质量受像质最差的望远镜限制,最终观测分辨率和灵敏度均会受到影响.通过图像复原,可以提高图像质量,进而提高叠加效果.根据该思路提出了1种基于PSF分区的迭代图像复原方法:该方法首先通过SOM (Self-organizing Maps)对PSF进行聚类分析,利用同类别PSF的平均PSF进行反卷积,再将反卷积结果按PSF聚类结果分割为不同大小的子图,最后将子图进行拼接.图像复原在提高图像质量的同时,降低了PSF不一致性对图像叠加带来的影响.将几个望远镜在同一时刻获取的图像经反卷积处理之后利用图像配准算法进行矫正并叠加,可获得高信噪比图像.对实际望远镜获取的数据处理后的结果表明:图像在进行复原和叠加过程中,星象目标信噪比不断提升,提高了成像系统对暗星的探测能力.     

Unbiased image reconstruction as an inverse problem

An unbiased method for improving the resolution of astronomical images is presented. The strategy at the core of this method is to establish a linear transformation between the recorded image and an improved image at some desirable resolution. In order to establish this transformation only the actual point spread function and a desired point spread function need be known. No image actually recorded is used in establishing the linear transformation between the recorded and improved image.
This method has a number of advantages over other methods currently in use. It is not iterative, which means it is not necessary to impose any criteria, objective or otherwise, to stop the iterations. The method does not require an artificial separation of the image into 'smooth' and 'point-like' components, and thus is unbiased with respect to the character of structures present in the image. The method produces a linear transformation between the recorded image and the deconvolved image, and therefore the propagation of pixel-by-pixel flux error estimates into the deconvolved image is trivial. It is explicitly constrained to preserve photometry and should be robust against random errors.     

Combining Gaia Windows: The Method and Simulations

During the 5-year-observational phase of Gaia several one-dimensional observations will be obtained from the same objects at different epochs and transit angles. Combining the information from all the transits to produce a single two-dimensional image is scientifically important as this will allow to detect much fainter stars in the vicinity of the main source than otherwise possible from single observations on-board the satellite or from individual transmitted patches. Most of the emerging objects will be physical companions, but some background/foreground objects will also be found. The combination procedure further increases the image resolution in the across scan-direction. The procedure consists of combining patches that are stretched back to 2-D windows, only possible since the centroid positions of the windows are known. The initial image window from which the patches originate is not square, but elongated along the scan-direction if the Astro AF11 field of view is used in the current calculations. Transit angles can take any value between 0^∘ and 360^∘. The window combining method is a simple mapping procedure that is fast and should handle all realistic complexities occurring during the Gaia data reduction. The method is not intended to be a perfect image restoration technique, but its purpose is to fulfill the scientific requirement of detecting faint neighboring sources existing in the close vicinity of the main targets. In this study we show examples of the combined windows and illustrate how faint companions around primary targets could be detected. We investigate the performances and limitations of the method and discuss possible complexities that might occur during the final data analysis.     

CLASS B2108+213: a new wide-separation gravitational lens system

We present observations of CLASS B2108+213, the widest separation gravitational lens system discovered by the Cosmic Lens All-Sky Survey. Radio imaging using the VLA at 8.46 GHz and MERLIN at 5 GHz shows two compact components separated by 4.56 arcsec with a faint third component in between which we believe is emission from a lensing galaxy. 5-GHz VLBA observations reveal milliarcsecond-scale structure in the two lensed images that is consistent with gravitational lensing. Optical emission from the two lensed images and two lensing galaxies within the Einstein radius is detected in Hubble Space Telescope imaging. Furthermore, an optical gravitational arc, associated with the strongest lensed component, has been detected. Surrounding the system is a number of faint galaxies which may help explain the wide image separation. A plausible mass distribution model for CLASS B2108+213 is also presented.     

同幅双速跟踪成像CCD相机模拟电路系统设计

讨论了同幅双速跟踪成像技术对CCD相机系统的特殊要求,详细地介绍了该相机前端模拟系统的设计思想、系统结构和主要的电路模块。对系统中一些关键电路模块进行了仿真和测试,并对仿真和测试结果进行比较分析,以检验电路系统设计的合理性。     

Relative astrometric and photometric measurements of visual binaries made with the Nice 76‐cm refractor in 2008

We present relative astrometric and photometric measurements of visual binaries made in 2008, with the 76‐cm refractor of Côte d'Azur Observatory. Our observing list contains orbital couples as well as binaries whose motion is still uncertain. Three different techniques were used for obtaining measurements: Lucky imaging, speckle interferometry and the Direct Vector Autocorrelation method. We obtained 2420 new measurements of the relative position of 2225 objects, with angular separations in the range 0″.07‐11″.5, and an average accuracy of 0″.02. The mean error on the position angles is 0°.6. We managed to observe faint systems (m_V ≈ 12) with large magnitude difference (up to Δm_V ≈ 5). We have thus been able to measure many systems containing red dwarf stars that had been poorly monitored since their discovery. We also measured the difference of magnitude of the two components of 376 objects with an estimated error of 0.1 mag. Finally, we provide a list of 33 newly discovered components of multiple or binary systems, which is a significant complement to our GII catalog (© 2012 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Multi-imaging and Bayesian estimation for photon counting with EMCCDs

A multi-imaging strategy is proposed and experimentally tested to improve the accuracy of photon counting with an electron multiplying CCD (EMCCD), by taking into account the random nature of its on-chip gain and the possibility of multiple photodetection events on one pixel. This strategy is based on Bayesian estimation on each image, with a priori information given by the sum of the images. The method works even for images with large dynamic range, with more improvement in the low light level areas. In these areas, two thirds of the variance added by the EMCCD in a conventional imaging mode are removed, making the physical photon noise predominant in the detected image.     

Research on Improving SNR of Large Angle Synchronous Satellite Based on Sub-pixel Translation and Superposition

Geosynchronous satellite with big inclination is not completely stationary relative to the ground. Its sub-satellite point moves periodically in the north-south direction with ground track of shape like “8”. The larger the inclination angle is, the larger the range of motion is. Such effect makes the effective exposure time of Geosynchronous satellite limited when observe it by optical telescope with stare mode, and the SNR (Signal Noise Ratio) of Geosynchronous satellite image could not be increased with long exposure time. In the monitoring of Geosynchronous orbit without changing hardware, a method called sub-pixel translating superposition of successive frames of images is proposed. By translating and aligning images in the sub-pixel scale according to the moving speed of the object and the time interval of the adjacent frame images, the positions of the Geosynchronous satellite in the image sequences are coincident, and then by superposition of these translational images, the SNR of moving Geosynchronous objects and the detection capability of the whole system can be improved. Results of superposition of real observed images show that the method can remarkably improve the SNR of moving. When five images are superimposed, compared with the SNR of the original image, the SNR of the superimposed image by integer pixels is increased by around 1.7 times, and the SNR of the superimposed image by sub-pixel is increased by about 2 times.     

Normalized and asynchronous mirror alignment for Cherenkov telescopes

Imaging Atmospheric Cherenkov Telescopes (IACTs) need imaging optics with large apertures and high image intensities to map the faint Cherenkov light emitted from cosmic ray air showers onto their image sensors. Segmented reflectors fulfill these needs, and as they are composed from mass production mirror facets they are inexpensive and lightweight. However, as the overall image is a superposition of the individual facet images, alignment is a challenge. Here we present a computer vision based star tracking alignment method, which also works for limited or changing star light visibility. Our method normalizes the mirror facet reflection intensities to become independent of the reference star’s intensity or the cloud coverage. Using two CCD cameras, our method records the mirror facet orientations asynchronously of the telescope drive system, and thus makes the method easy to integrate into existing telescopes. It can be combined with remote facet actuation, but does not require one to work. Furthermore, it can reconstruct all individual mirror facet point spread functions without moving any mirror. We present alignment results on the 4 m First Geiger-mode Avalanche Cherenkov Telescope (FACT).