光学系统参数确定带宽约束的天文图像盲反卷积

大气湍流极大限制了地基大口径望远镜观测天文目标图像的空间分辨率.根据最大似然估计原理,提出了用实际光学带宽约束的可有效减小天文观测图像中大气湍流影响的盲反卷积方法,通过共轭梯度优化算法使卷积误差函数趋向最小.建立了望远镜光学系统参数和图像频域带宽的关系,采用变量正性约束、点扩散函数带宽有限约束,提高算法的收敛性.为避免图像处理中有效傅立叶变换频率超出截止频率,要求采集望远镜焦面图像时单个成像单元(如CCD像素单元)应小于四分之一衍射斑直径.算法中未用目标支持域约束,所提出的方法适用于全视场天文图像恢复.用计算机模拟和对实际天文目标双鱼座图像数据的恢复结果验证了所提出方法的有效性.     

高分辨率太阳图像重建方法

高空间分辨率的太阳观测数据有助于深入研究太阳大气现象、太阳物理基本问题。地基大口径太阳望远镜常通过自适应光学技术和图像重建技术获取高空间分辨率图像。目前太阳图像重建技术主要有斑点成像术和斑点相位差法两类。介绍了斑点成像术中几类方法的原理,阐述了斑点成像术重建太阳像的流程以及几个关键步骤,介绍了多帧盲反卷积和相位差法的原理,比较了斑点成像术和斑点相位差法的特点,最后阐述了它们在太阳高分辨率观测中的应用和发展趋势。     

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.     

太阳色球图像的选帧处理

讲述抚仙湖太阳观测站的色球图像的选帧处理,主要针对观测数据的网上发布,处理过程主要由选帧和简单位移叠加(SAA)组成。使用在太阳光球图像处理中提出的斑点干涉术选帧法对图像进行选取。对太阳等扩展目标,简单位移叠加的位移量由图像和参考图像的相关极大值位置确定,参考图像选像质最好的一帧。选帧处理过程算法简单,处理速度快,相比斑点掩模法可以节省大量的重建时间,通过观察得到的处理结果还可以确定数据是否有用斑点掩模法进行进一步处理的必要。     

基于自然导引星的自适应光学系统非等晕性分析

采用波前各阶zernike成分孔径滤波函数方法,推导了当观测目标与信标不在同一方向时所产生的非等晕误差的表达式,同时利用该表达式进行了对自适应光学系统倾斜校正后的波前残余误差数值模拟.此外还首次对天文干涉仪中大气湍流活塞效应引起的活塞角非等晕误差进行了研究分析,为以后实际利用自适应光学校正工作提供理论参考.     

面向ONSET实时数据处理的图像选帧GPU技术实现

光学和近红外太阳爆发监测望远镜每天可以获得大量的太阳图像数据,对这些观测数据进行实时选帧处理,一方面可以减轻存储压力,另一方面也可以提高后续图像重建的质量。针对观测过程中的选帧要求,设计并实现了一套基于图形处理器的图像选帧实时处理模块,当前的模块已经实现了平均梯度法和谱比法选帧两种算法的高速并行处理。对模块的实现进行了细致的讨论,并比较了两种选帧方法的加速比。实验表明,该模块运行稳定可靠;从执行效率来看,针对近全日面图像的选帧总体执行时间最快为1.2 s,比原有串行实现提升了7倍;局部面图最快为0.7 s,平均提升了5倍。整体模块的实现与当前性能已经可以满足实时观测与处理的要求。     

改进的帧差法在空间运动目标检测中的应用

针对空间运动目标检测易受光照、云层等因素的干扰,导致在没有出现运动目标的情况下,错误地判断为检测到运动目标,设计了一种改进的帧差算法,把帧差法和背景减除法相结合,周期性地把当前帧更新为背景帧。先对待检测图像进行二值化处理,有效消除了光照、云层等噪声因素,并且强化了空间目标的图像,然后通过帧差法检测出空间目标。实验表明,该方法有效降低了空间运动目标的误判率,改进的算法不需要把每一帧图像作为背景帧,提高了运行速度,也不需要对背景进行统计建模,简化了背景的建立过程,算法易于实现,操作简单,资金投入少,灵敏度高,具有较高的实用价值。     

基于盲退卷积的AIA图像增强方法

太阳图像中存在各种不同尺度、亮度和结构的物理活动现象,由于太阳日冕高动态活动和传感器设备等因素的影响,太阳图像成像质量不佳。根据太阳动力学天文台(Solar Dynamic Observatory,SDO)的大气成像仪(Atmospheric Imaging Assenbly,AIA)拍摄不同波段数据结构的动态范围大、噪声大、结构相对模糊等特点,提出一种基于盲退卷积的图像增强方法。首先对图像进行去噪和降低动态范围的处理,基于图像功率谱的分布假设,从原图中估计点扩散函数(Point Spread Function,PSF)的功率谱;然后使用相位提取算法恢复点扩散函数的相位,再退卷积得出较高质量的目标图像;最后通过轮廓切片分析、功率谱分析以及点扩散函数分析对增强结果进行定量和定性评价。实验结果表明,相比现有的图像增强方法,该方法在有效增强太阳日冕图像细节结构的同时,能够复原原图中因模糊无法识别的结构。     

多星中天时角法精密测定天文方位角

在高精度天文方位角测定方法中,北极星任意时角法只适用于北半球中纬度地区,子午星对法则受限于选星配对条件.提出可适用于全球地区的多星中天时角法,观测多颗任意中天位置的南星和北星,通过线性回归模型修正测站经度误差和计时误差对方位角解算的影响.该方法只需已知测站的概略位置,对观测天体的中天位置没有限制.实测数据表明,与北极星任意时角法相比,在相同精度指标要求下,观测量可以减少一半.     

丽江观测站1.8米望远镜自适应光学系统性能初步理论估计

为了探讨丽江1.8m自适应光学望远镜在天文或天体物理研究领域的可能应用,使用误差基本标度理论对其性能进行了初步的理论分析.结果表明,一般观测条件下,系统分辨力可以达到近似衍射极限.同时对在其上配置激光导星系统后可获得的性能进行了预测,结果表明,激光导星自适应光学系统将大幅度提高系统的天空覆盖率.     

Adaptive Optics Image Restoration Based on Frame Selection and Multi-frame Blind Deconvolution

Restricted by the observational condition and the hardware, adaptive optics can only make a partial correction of the optical images blurred by atmospheric turbulence. A postprocessing method based on frame selection and multi-frame blind deconvolution is proposed for the restoration of high-resolution adaptive optics images. By frame selection we mean we first make a selection of the degraded (blurred) images for participation in the iterative blind deconvolution calculation, with no need of any a priori knowledge, and with only a positivity constraint. This method has been applied to the restoration of some stellar images observed by the 61-element adaptive optics system installed on the Yunnan Observatory 1.2m telescope. The experimental results indicate that this method can effectively compensate for the residual errors of the adaptive optics system on the image, and the restored image can reach the diffraction-limited quality.     

Adaptive optics for large telescopes

The performance of large telescopes is determined both by their angular resolution and by their collection area. It is, therefore, important to achieve as high an angular resolution as possible by site selection, by avoiding image deterioration by the telescope and its environment, and by real time image restoration by adaptive optics. We summarize the principles of adaptive optics, their predicted performance and the current programs underway to implement adaptive optics for astronomical purposes.Paper presented at the Symposium on the JNLT and Related Engineering Developments, Tokyo, November 29–December 2, 1988.     

High-Spatial-Resolution Imaging Combining High-Order Adaptive Optics,Frame Selection,and Speckle Masking Reconstruction

We present, for the first time, high-spatial-resolution observations combining high-order adaptive optics (AO), frame selection, and post-facto image correction via speckle masking. The data analysis is based on observations of solar active region NOAA 10486 taken with the Dunn Solar Telescope (DST) at the Sacramento Peak Observatory (SPO) of the National Solar Observatory (NSO) on 29 October 2003. The high Strehl ratio encountered in AO corrected short-exposure images provides highly improved signal-to-noise ratios leading to a superior recovery of the object’s Fourier phases. This allows reliable detection of small-scale solar features near the diffraction limit of the telescope. Speckle masking imaging provides access to high-order wavefront aberrations, which predominantly originate at high atmospheric layers and are only partially corrected by the AO system. In addition, the observations provided qualitative measures of the image correction away from the lock point of the AO system. We further present a brief inspection of the underlying imaging theory discussing the limitations and prospects of this multi-faceted image reconstruction approach in terms of the recovery of spatial information, photometric accuracy, and spectroscopic applications.The editors apologize to the authors: due to a misunderstanding during the editorial process, the publication of this paper has been delayed.     

Near Real-Time Image Reconstruction

In recent years, post-facto image-processing algorithms have been developed to achieve diffraction-limited observations of the solar surface. We present a combination of frame selection, speckle-masking imaging, and parallel computing which provides real-time, diffraction-limited, 256×256 pixel images at a 1-minute cadence. Our approach to achieve diffraction limited observations is complementary to adaptive optics (AO). At the moment, AO is limited by the fact that it corrects wavefront abberations only for a field of view comparable to the isoplanatic patch. This limitation does not apply to speckle-masking imaging. However, speckle-masking imaging relies on short-exposure images which limits its spectroscopic applications. The parallel processing of the data is performed on a Beowulf-class computer which utilizes off-the-shelf, mass-market technologies to provide high computational performance for scientific calculations and applications at low cost. Beowulf computers have a great potential, not only for image reconstruction, but for any kind of complex data reduction. Immediate access to high-level data products and direct visualization of dynamic processes on the Sun are two of the advantages to be gained.     

Spada: An Array of Spad Detectors For Astrophysical Applications

Astrophysical studies require accurate, sensitive and fast detectors to detect faint sources with high variability. Recently an array of Single Photon Avalanche Diodes (SPAD), SPADA, has been developed. This array is suitable for competitive adaptive optics operations and fast transient image acquisition at a fraction of the current cost of imaging arrays. The fabricated solid-state photon counters are rugged, easily integrated with the optics, free from readout noise, and have very fast frame rates (> 10 kHz, for visible corrections) with nanosecond electronic gating. In this paper, the following are described: the development of silicon monolithic arrays of 60 photon-counters, the detection electronics (based on integrated active quenching circuits for each pixel of the array), the real-time data-processing board implemented into FPGA and some aspects of the mechanical housing.     

High Resolution Solar Speckle Imaging With the Extended Knox–Thompson Algorithm

We present an implementation of the extended Knox-Thompson (EKT) speckle reconstruction algorithm dedicated to solar observations. EKT speckle imaging yields nearly diffraction-limited images from bursts of short exposure solar observations under a wide range of seeing conditions. Our implementation supports field dependent amplitude calibration to permit analyzing data obtained with a partially compensating adaptive optics systems. The principles of the method and some technical details of our implementation are discussed. We have performed various tests using simulated data of representative solar scenes. The simulations include the effects of seeing and noise with the exception of anisoplanatism. The expected photometric error of a reconstructed image amounts to a few percent of the mean intensity under seeing conditions ranging from poor to excellent. We also present sample reconstructions of real data and discuss issues arising from anisoplanatism.     

High resolution near-IR imaging with arrays

The availability of large format, low noise detector arrays has opened the 1 to 20 m region for sub-arcsecond imaging. Using recent results in the investigation of Galactic star forming regions, the Galactic center and external galaxies as examples, we discuss the various techniques that have been employed. These range from image selection and image sharpening, over speckle techniques and adaptive optics to lunar occultation. These examples demonstrate that sub-arcsecond imaging and spatial interferometry in the near- and mid-infrared has a great future potential.     

Improving the seeing with wide-field adaptive optics in the near-infrared

In this paper, we present simulation results of a ground-layer correction adaptive optics system (GLAO), based on four laser guide stars and a single deformable mirror. The goal is to achieve a seeing improvement over an 8-arcmin field of view, in the near-infrared (from 1.06 to 2.2 μm). We show results on the scaling of this system (number of subapertures, frame rates), and the required number of tip-tilt stars. We investigate the use for GLAO of both sodium and Rayleigh guide stars. We also show that if the lasers can be repositioned, the performance of the adaptive optics can be tailored to the astronomical observations.     

Image Quality in High-resolution and High-cadence Solar Imaging

Broad-band imaging and even imaging with a moderate bandpass (about 1 nm) provides a photon-rich environment, where frame selection (lucky imaging) becomes a helpful tool in image restoration, allowing us to perform a cost-benefit analysis on how to design observing sequences for imaging with high spatial resolution in combination with real-time correction provided by an adaptive optics (AO) system. This study presents high-cadence (160 Hz) G-band and blue continuum image sequences obtained with the High-resolution Fast Imager (HiFI) at the 1.5-meter GREGOR solar telescope, where the speckle-masking technique is used to restore images with nearly diffraction-limited resolution. The HiFI employs two synchronized large-format and high-cadence sCMOS detectors. The median filter gradient similarity (MFGS) image-quality metric is applied, among others, to AO-corrected image sequences of a pore and a small sunspot observed on 2017 June 4 and 5. A small region of interest, which was selected for fast-imaging performance, covered these contrast-rich features and their neighborhood, which were part of Active Region NOAA 12661. Modifications of the MFGS algorithm uncover the field- and structure-dependency of this image-quality metric. However, MFGS still remains a good choice for determining image quality without a priori knowledge, which is an important characteristic when classifying the huge number of high-resolution images contained in data archives. In addition, this investigation demonstrates that a fast cadence and millisecond exposure times are still insufficient to reach the coherence time of daytime seeing. Nonetheless, the analysis shows that data acquisition rates exceeding 50 Hz are required to capture a substantial fraction of the best seeing moments, significantly boosting the performance of post-facto image restoration.     

Solar GLAO and TAO:performance comparisons

Multi-conjugate adaptive optics(MCAO),consisting of several deformable mirrors(DMs),can significantly increase the adaptive optics(AO)correction field of view.Current MCAO can be realized by either star-oriented or layer-oriented approaches.For solar AO,ground-layer adaptive optics(GLAO)can be viewed as an extreme case of layer-oriented MCAO in which the DM is conjugated to the ground,while solar tomography adaptive optics(TAO)that we proposed recently can be viewed as star-oriented MCAO with only one DM.Solar GLAO and TAO use the same hardware as conventional solar AO,and therefore it will be important to see which method can deliver better performance.In this article,we compare the performance of solar GLAO and TAO by using end-to-end numerical simulation software.Numerical simulations of TAO and GLAO with different numbers of guide stars are conducted.Our results show that TAO and GLAO produce the same performance if the DM is conjugated to the ground,but TAO can only generate better performance when the DM is conjugated to the best height.This result has important application in existing one-DM solar AO systems.