Automated determination of the orientation of dust devil tracks in mars orbiter images

The paper presents and evaluates three methods for automatically estimating the main orientation of Martian dust devil tracks in MOC and HiRISE images. Inferring such information about dust devils from their tracks is important to better understand the near surface wind. The methods considered were based on gradient direction, directional openings and morphological granulometry. The accuracy of the methods was asserted by comparing the results to a set of directions estimated visually and assumed to be the ground truth. The higher accuracy was reached using directional openings. Besides, the directions inferred by this method were compared to those predicted by the GCM and the results agreed.     

基于视频数据的机场跑道外来物检测

提出了一种基于视频数据的机场跑道外来物(FOD,Foreign Object Debris)检测算法,该算法包括几何校正、背景差分、杂波抑制和伪装消除等4个步骤.其中,几何校正排除了摄像头轻微抖动造成的图像差异;背景差分利用每个像素对应的颜色向量信息,建立背景模型并进行定时更新;杂波抑制和伪装消除利用差分图像和原始图像信息,建立马尔科夫随机场和概率统计模型,在降低虚警的同时提高了检测率.将本算法应用于两组不同光照条件下获取的视频图像,检测不同形状和颜色的FOD目标;采用本算法对机场实测数据进行测试,验证了算法的有效性.     

一种基于FPGA的图像自动增强算法与实现

图像灰度分布不均匀是影响画质的重要因素，集中表现在对空间场景成像时，图像中存在大量过亮或过暗的区域.同时，算法的实时性和嵌入式平台的可实现性是设计实现空间应用载荷需要着重考虑的问题.针对上述问题，提出一种适合于FPGA实现的图像自动增强算法.算法以分段线性变换方法为基础，采用K均值聚类对视频图像的灰度直方图进行动态的区间划分；构建分段函数系数与直方图之间的定量关系，自动计算线性变换系数；采用高效的并行流水线结构设计，实现了基于FPGA的硬件处理系统.仿真和成像试验结果表明，该FPGA硬件系统实时性好，适应性强，针对不同的场景均取得较好的处理效果，具有广阔的应用前景.      

The MAGO experiment for dust environment monitoring on the Martian surface

Among the main directions identified for future Martian exploration, the study of the properties of dust dispersed in the atmosphere, its cycle and the impact on climate are considered of primary relevance. Dust storms, dust devils and the dust “cycle” have been identified and studied by past remote and in situ experiments, but little quantitative information is available on these processes, so far. The airborne dust contributes to the determination of the dynamic and thermodynamic evolution of the atmosphere, including the large-scale circulation processes and its impact on the climate of Mars. Moreover, aeolian erosion, redistribution of dust on the surface and weathering processes are mostly known only qualitatively. In order to improve our knowledge of the airborne dust evolution and other atmospheric processes, it is mandatory to measure the amount, mass-size distribution and dynamical properties of solid particles in the Martian atmosphere as a function of time. In this context, there is clearly a need for the implementation of experiments dedicated to study directly atmospheric dust. The Martian atmospheric grain observer (MAGO) experiment is aimed at providing direct quantitative measurements of mass and size distributions of dust particles, a goal that has never been fully achieved so far. The instrument design combines three types of sensors to monitor in situ the dust mass flux (micro balance system, MBS) and single grain properties (grain detection system, GDS + impact sensor, IS). Technical solutions and science capabilities are discussed in this paper.     

基于长椭球面波函数去卷积方法的高动态范围太阳射电成像

利用综合孔径射电望远镜对太阳进行观测时,通过对图像中存在的明亮扩展源进行准确建模并移除,可以更好地观测视场内的微弱源并提高图像的动态范围。在射电天文领域,主要利用CLEAN算法对图像中的明亮源进行移除,以显示微弱的背景。然而,使用图像像素作为基函数的CLEAN算法的固有限制导致其对扩展源的建模效果较差。为了克服这种限制,将基于长椭球面波函数（Prolate Spheroidal Wave Functions,PSWF）的去卷积方法应用于太阳射电成像。PSWF最优正交基由脏图中的感兴趣区域（Region of Interest,ROI）和UV覆盖共同决定。为了验证该方法的有效性,基于PSWF正交基对均匀圆环阵观测的太阳射电图像进行去卷积,并从动态范围和保真度两个方面定量化对比了CLEAN算法和基于PSWF正交基方法的性能。基于PSWF正交基去卷积方法剩余脏图中的微弱源更接近真实情况且动态范围更高。     

What is the role of magnetic null points in large flares?

We have performed the analysis of the magnetic topology of active region NOAA 10486 before two large flares occurring on October 26 and 28, 2003. The 3D extrapolation of the photospheric magnetic field shows the existence of magnetic null points when using two different methods. We use TRACE 1600 Å and 195 Å brightenings as tracers of the energy release due to magnetic reconnections. We conclude on the three following points:

1. The small events observed before the flares are related to low lying null points. They are long lasting and associated with low energy release. They are not triggering the large flares.

2. On October 26, a high altitude null point is found. We look for bright patches that could correspond to the signatures of coronal reconnection at the null point in TRACE 1600 Å images. However, such bright patches are not observed before the main flare, they are only observed after it.

3. On October 28, four ribbons are observed in TRACE images before the X17 flare. We interpret them as due to a magnetic breakout reconnection in a quadrupolar configuration. There is no magnetic null point related to these four ribbons, and this reconnection rather occurs at quasi-separatrix layers (QSLs).

We conclude that the existence of a null point in the corona is neither a sufficient nor a necessary condition to give rise to large flares.     

A Novel Ship Wake Detection Algorithm Based on WTHT and Radon Transform

This paper proposes a novel ship wake detection algorithm based on the White Top-hat Transform (WTHT) and the Radon transform, which aims to improve the contrast between the ship wake and the background so as to improve the detection performance on Synthetic Aperture Radar (SAR) images. The proposed algorithm includes two major processes, and one is to improve the contrast and another one is to locate the ship wake. In high sea state conditions, the contrast of ship wake and background can be very low, which makes it difficult to detect. In the first step, the proposed contrast improvement algorithm is applied to improving the contrast which helps for improving the detection performance. An attribute filter based on edge detection result is adopted here. In the second step the contrast improved image is transformed into the Radon domain followed by peak extraction process to find the wake, the WTHT is used once more in this step. Finally, in the last step, the wake is overlapped on the original image. Experimental results on Tiangong-2 Interferometric Imaging Radar Altimeter (InIRA) images are presented and compared with that obtained by using the classical algorithm, and in this way, the better performance of our algorithm is demonstrated.      

基于非下采样轮廓波变换的红外和SAR图像融合算法

针对现有红外和合成孔径雷达(Synthetic Aperture Radar,SAR)图像的融合算法融合质量差、边缘轮廓不清晰、效率低下、可视性差,目标检测效率低等问题,提出一种基于非下采样轮廓波变换的融合算法。首先采用非下采样轮廓波变换对预处理的红外和SAR图像进行分解,获得各自低频和带通方向图像,接着根据红外和SAR图像的特征选取其含重要目标信息的频带进行低频图像和带通方向图像融合。为了检验本文所提出算法性能的优越性,采用两组红外和SAR图像进行融合实验,与其他图像融合算法进行对比,并对融合图像进行目标检测,证明了该融合算法能有效提高多源图像目标检测率。     

Ground-based observations of the dust emission from comet Halley

A preliminary analysis of the dust emission from comet Halley is presented based on large scale observations of its dust tail. Selected images obtained between February 22 and May 10, 1986 are compared to synchrone-syndyne graphs to infer the history of the dust production and the properties of the dust, at least qualitatively. Quantitative modeling of the dust tall has also been initiated and preliminary results are shown for the cases of isotropic and anisotropic (jet) dust production.     

基于GCF-SB视觉注意模型的红外目标检测算法

针对复杂背景、低对比度条件下的红外目标检测,提出了一种基于灰度对比度特征 相似性贝叶斯(GCF SB)模型的红外显著性目标检测算法.建立了一种灰度对比度特征(GCF)模型,该模型利用两个通道分别提取红外图像的灰度特征和对比度特征,然后通过特征融合获得初级显著图;建立了一种基于相似性的贝叶斯(SB)模型,该模型根据初级特征图分别计算目标和背景的先验概率和似然函数,然后利用贝叶斯公式获得最终显著图,进而实现红外图像的显著性目标检测.实验结果表明,所提出算法能够有效抑制复杂背景、低对比度红外图像的噪声,增强对比度,具有较高的检测精度和鲁棒性.     

Systematic examination of the geomagnetic storm sudden commencement using multi resolution analysis

Geomagnetic storm sudden commencements (SSC) contain a wealth of information which is useful in many applications. It is important to point out that the SSCs used in this study are sampled at the rate of one sample per second in order get use of such high resolution data. In this paper, two studies are made on the geomagnetic SSC and an SSC onset automatic detection algorithm is introduced. The first study is about finding the relationship between the SSC rise time and its amplitude. Where it is found that there is a positive correlation between the amplitude and the amplitude gradient which is the amplitude divided by rise time. The second study is checking the spectrum of the SSC, starting from its onset until the end of the SSC rise time. This check had proved that the SSC contains both low and high frequency regions. This led us to introduce a new term, namely the SSC variation rate (VR). This VR is defined as the maximum rate of change of the field in the higher-frequency region of the SSC. These two studies were the guide to build an SSC automatic detector of one sample per second data using multi resolution analysis (MRA) of the discrete wavelet transform (DWT). The data set contains 134 SSCs with different VRs that were collected from the Circum-pan Pacific Magnetometer Network (CPMN). It is found that the standard deviation of the detection error is 41 s and that the average error is 9 s. From the calculated error distribution function, it is found that the detection error is within the range of −1.5 to 3 min. The detection process, as will be shown in the article, takes 70 s for one station and 3 min if the decision is related to the detection(s) of other stations. These results demonstrate the superiority of the proposed algorithm over other algorithms in which the detection error ranges between −8 and 5 min and the detection process takes 2–10 min. In addition to being faster and more accurate than the other algorithms, the proposed algorithm is the first algorithm that automatically detects the SSC onset times from high-resolution data unlike previous studies that focused on determining the SSC times automatically using one-minute resolution data.     

一种图像缩放算法的SoC协同加速设计方法

针对无人机自主着陆的跑道检测、识别、跟踪等视觉算法中需要对大量图像进行缩放处理以便后续计算,但又对实时性要求比较高的情况,根据输入输出像素点的映射关系提出了一种适用于硬件加速的图像缩放算法,简化算法结构的同时利用现场可编程门阵列进行模块硬件功能的设计对算法加速,并采用软硬件协同的体系结构搭建实时图像处理系统。实验结果表明,该缩放算法处理精度高、耗时少,且用硬件逻辑实现后,可以进一步提速171倍,硬化后的系统可以通过摄像头获取图像数据,实时处理后在显示器中显示,达到30帧/s的处理速度,可以应用于实时性要求较高的图像处理算法中。      

The automated prediction of solar flares from SDO images using deep learning

In the last few years, there has been growing interest in near-real-time solar data processing, especially for space weather applications. This is due to space weather impacts on both space-borne and ground-based systems, and industries, which subsequently impacts our lives. In the current study, the deep learning approach is used to establish an automated hybrid computer system for a short-term forecast; it is achieved by using the complexity level of the sunspot group on SDO/HMI Intensitygram images. Furthermore, this suggested system can generate the forecast for solar flare occurrences within the following 24 h. The input data for the proposed system are SDO/HMI full-disk Intensitygram images and SDO/HMI full-disk magnetogram images. System outputs are the “Flare or Non-Flare” of daily flare occurrences (C, M, and X classes). This system integrates an image processing system to automatically detect sunspot groups on SDO/HMI Intensitygram images using active-region data extracted from SDO/HMI magnetogram images (presented by Colak and Qahwaji, 2008) and deep learning to generate these forecasts. Our deep learning-based system is designed to analyze sunspot groups on the solar disk to predict whether this sunspot group is capable of releasing a significant flare or not. Our system introduced in this work is called ASAP_Deep. The deep learning model used in our system is based on the integration of the Convolutional Neural Network (CNN) and Softmax classifier to extract special features from the sunspot group images detected from SDO/HMI (Intensitygram and magnetogram) images. Furthermore, a CNN training scheme based on the integration of a back-propagation algorithm and a mini-batch AdaGrad optimization method is suggested for weight updates and to modify learning rates, respectively. The images of the sunspot regions are cropped automatically by the imaging system and processed using deep learning rules to provide near real-time predictions. The major results of this study are as follows. Firstly, the ASAP_Deep system builds on the ASAP system introduced in Colak and Qahwaji (2009) but improves the system with an updated deep learning-based prediction capability. Secondly, we successfully apply CNN to the sunspot group image without any pre-processing or feature extraction. Thirdly, our system results are considerably better, especially for the false alarm ratio (FAR); this reduces the losses resulting from the protection measures applied by companies. Also, the proposed system achieves a relatively high scores for True Skill Statistics (TSS) and Heidke Skill Score (HSS).     

Developing a dust storm detection method combining Support Vector Machine and satellite data in typical dust regions of Asia

Enhancing the dust storm detection is a key part for the environmental protection, human healthy and economic development. The goal of this paper is to propose a new Support Vector Machine (SVM)-based method to automatically detect dust storms using remote sensing data. Existing methods dealing with this problem are usually threshold-based that are of great complexity and uncertainty. In this paper we propose a simple and reliable method combining SVM with MODIS L1 data and explore the optimal band combinations used as the feature vectors of SVM. The developed method was evaluated by MODIS and OMI data qualitatively and quantitatively on three study sites located in the Arabian Desert, Gobi Desert and Taklimakan Desert, and it was also compared to three other traditional methods based on their accuracy, complexity, reliability and sensitivity to thresholds. The detection results demonstrated that the combination of (Band7 − Band3)/(Band7 + Band3) ((B7 − B3)/(B7 + B3)), Band20 − Band31 (B20 − B31), and Band31/Band32 (B31/B32) can detect the dust storms more precisely than other individual bands or their combination. The comparison among those cases indicated that the proposed automatic method exhibited an advantage of minimizing the uncertainty and complexity, which were the limits of defining thresholds based on the threshold-based methods. The conclusions can provide references for studies that focus on statistical-based dust storm detection.     

A fine mist of very small comet dust particles

A comet nucleus considered as an aggregate of interstellar dust would produce a mist of very finely divided (radius ～ 0.01 μm) particles of carbon and metal oxides accompanying the larger dust grains. These small particles which are very abundant in the interstellar dust size spectrum would provide substantial physical effects because of their large surface area. They may show up strongly in particle detectors on the Halley probes. A strong basis for serious consideration of these particles comes from the other evidence that interstellar dust grains are the building blocks of comets; e.g. (1) the explanation of the “missing” carbon in comets; (2) The S₂ molecule detection which suggests that the comet solid ice materials have been previously subjected to ultraviolet radiation (as are interstellar grains) before aggregation into the comet; (3) the predicted dust to gas ratio.     

基于加权全变差最小化的中子外部CT重建算法

针对中子外部CT检测需求，提出了基于加权方向全变差（WDTV）最小化的中子外部CT重建算法。首先，利用平行束数据对称性原理将正弦图中旋转中心另一侧缺失的投影数据进行补充；其次，采用传统FBP算法进行了外部CT扫描模式下的CT重建。为了抑制传统FBP、SART算法重建图像中径向边缘伪影，采用WDTV算法计算沿径向和周向方向的局部方向差分并作加权和。另外，在重建模型的WDTV项中引入2个权重参数以控制径向边缘和切向边缘的不同响应强度。最后，采用所提算法研究了重建图像质量与投影数据量的关系。计算机仿真和真实冷中子实验表明:基于WDTV最小化的中子外部CT重建算法能够有效抑制径向边缘伪影，获得高质量重建图像。      

基于边缘信息的灰度目标跟踪算法

提出了一种基于边缘信息的跟踪算法,其可以实现对剧烈变化的灰度目标的精确跟踪.首先,利用基于双同心圆窗口算子的非线性边缘检测算法得到高质量的边缘信息;其次,为了解决单一边缘特征空间不能充分表征目标的难题,提出了一种通过组合边缘图像构建特征空间的方法,以便为目标建模提供充分信息;再次,在构建的特征空间中使用核估计方法对目标进行建模;在目标定位阶段,利用Kalman滤波器对目标进行预估后,再由Mean Shift算法在预估位置邻近区域实现目标定位;最后,提出了一种基于形态学的动态模型更新策略,使得算法不仅可以获得精确的目标区域,还可以实现对目标尺寸和形状变化的自适应.实验结果表明,本算法不仅可以有效跟踪剧烈变化的灰度目标,而且跟踪窗口可以实现对目标尺寸和形状的自适应.      

一种基于自动特征学习的陨石坑区域检测算法

基于陨石坑的视觉导航技术成为一种新颖的高精度空间探测自主导航方式,如何从导航图像中精确地提取陨石坑区域是实现基于陨石坑视觉导航的首要条件。针对这一问题,根据陨石坑导航图像特点,提出了一种基于自动特征学习的陨石坑区域检测算法。首先,基于最大稳定极值区域检测算法提取陨石坑候选区域;其次,利用卷积神经网络（CNN）自动学习提取候选区域的特征;最后,通过支持向量机（SVM）实现候选区域的精确分类,得到真实的陨石坑区域。大量的仿真实验表明:与传统的基于人工特征的陨石坑区域检测算法相比,提出的基于自动特征学习的陨石坑区域检测算法具有更高的检测精度和更好的鲁棒性,在通用火星表面陨石坑数据集上,所提算法的F₁度量指标较于传统算法高出8%,可以广泛地应用于基于陨石坑的视觉导航算法中的陨石坑区域提取,为基于陨石坑视觉导航算法提供精确的导航路标输入。      

Spectroscopic observation of Mars

Visible and near-infrared reflectance spectroscopy has proven a powerful tool for exploring the geology of Mars. Most of this data has been obtained from Earth, but the technique is ideally suited to orbital application, as proposed for the U.S. Mars Geoscience/Climatology Orbiter mission. Spectral reflectance in the near-UV and visible is highly diagnostic of ferric-iron mineralogy, and has shown that Fe³⁺ in the ubiquitous bright dust and soil is amorphous or poorly-crystalline. Other iron-oxide minerals, indicative of other modes or episodes of crustal alteration, may be found in spatially localized deposits. Clay minerals (hydroxylated silicates) have diagnostic vibrational absorptions throughout the near-infrared. While some form of bound water and/or OH has been known on Mars for many years, a new result presented here is the identification of structural OH in a dilute or poorly crystalline magnesian clay. Salts such as carbonates, sulfates, and nitrates have not yet been detected in martian soils but have diagnostic spectral features in the 3- to 4-μm region, best suited to Mars-orbital observation. Analysis of reflectance spectra of low-albedo regions is a primary source of evidence for a basaltic or ultramafic crust, with identification of abundant clinopyroxene and possible detection of other mafic minerals. The distinctive near-infrared spectral shape of dark regions indicates that the dark materials commonly consist of relatively unaltered rocks or rock fragments very thinly coated by (or mixed with) bright oxidized material similar to the global dust. Visible and near-infrared reflectance spectroscopy is also a sensitive technique for detecting and analyzing water ice, as has been demonstrated on Mars by observations of the north polar cap.     

Adjustment of systematic errors in ALS data through surface matching

Surface matching is a well researched topic in both Computer Vision (CV) and terrestrial laser scanning (TLS) or ground based light detection and ranging (LiDAR), but the extent of the range images derived from these technologies is typically orders of magnitude smaller than those derived from airborne laser scanning (ALS), also known as airborne LiDAR. Iterative closest point (ICP) and its variants have been successfully used to align and register multiple overlapping views of the range images for CV and TLS applications. However, many challenges are encountered in applying the ICP approach to ALS data sets. In this paper, we address these issues, explore the possibility of automating the algorithm, and present a technique to adjust systematic discrepancies in overlapping strips, using geometrical attributes in a given terrain. In this method, the ALS point samples used in the algorithm are selected depending on their ability to constrain the relative movement between the overlapping laser strips. The points from overlapping strips are matched through modified point to plane based on the ICP method.