The attitude inversion method of geostationary satellites based on unscented particle filter

The attitude information of geostationary satellites is difficult to be obtained since they are presented in non-resolved images on the ground observation equipment in space object surveillance. In this paper, an attitude inversion method for geostationary satellite based on Unscented Particle Filter (UPF) and ground photometric data is presented. The inversion algorithm based on UPF is proposed aiming at the strong non-linear feature in the photometric data inversion for satellite attitude, which combines the advantage of Unscented Kalman Filter (UKF) and Particle Filter (PF). This update method improves the particle selection based on the idea of UKF to redesign the importance density function. Moreover, it uses the RMS-UKF to partially correct the prediction covariance matrix, which improves the applicability of the attitude inversion method in view of UKF and the particle degradation and dilution of the attitude inversion method based on PF. This paper describes the main principles and steps of algorithm in detail, correctness, accuracy, stability and applicability of the method are verified by simulation experiment and scaling experiment in the end. The results show that the proposed method can effectively solve the problem of particle degradation and depletion in the attitude inversion method on account of PF, and the problem that UKF is not suitable for the strong non-linear attitude inversion. However, the inversion accuracy is obviously superior to UKF and PF, in addition, in the case of the inversion with large attitude error that can inverse the attitude with small particles and high precision.     

Latitudinal distribution of solar coronal active regions

We studied the cyclic evolution of the latitudinal distribution of solar coronal active regions based on daily images from SOHO EIT for the period 1995–2017. Fully automated software was used, which included the following steps: initial preparation of images in the data series, normalization of histograms and correction of limb brightening, segmentation of images using threshold intensity values obtained from their histograms, scanning of segmented images in heliographic coordinates and obtaining profiles of latitudinal distribution of coronal active regions for each image of the data series. From the output data, we obtained a temporary change in the latitudinal distribution profiles and the migration of activity centers on the solar disk. From the period of minimum activity to the next minimum in both hemispheres, activity centers begin to migrate from high latitudes towards the equator. At the same time, the general center of activity repeatedly changes the direction of migration. The latitudinal distribution of the so-called presence factor of coronal active regions closely resembles the magnetic butterfly diagram, which proves their direct causal relationships. Variations in the presence factor of coronal active regions are correlated with cyclic variations in the sunspot daily numbers.     

应用最优化理论中的投影定理计算平行正投影

拟应用最优化理论中的投影定理,从理论上严格证明物体平行正投影的存在和唯一性,并在此基础上建立一个全新的平行正投影计算方法.内容包括:①应用数学语言阐明三维物体平行正投影的含义;②应用最优化理论中的相关理论依次论证与投影定理有关的几个命题,并在此基础上严格证明物体平行正投影的存在和唯一性,以及利用傅里叶级数形式建立平行正投影计算式;③简要分析这种计算方法的特点.      

空间材料科学实验中的X光三维成像仿真

用于空间材料科学研究的X射线观察设备已列入空间站材料科学实验规划,仿真实现X光三维成像的流程可为X射线观察设备的研制提供设计依据.仿真流程主要包括三个步骤:使用锥束X光照射被扫描对象并旋转对象,获取多角度下的投影影像;根据投影影像和几何信息重建扫描对象,获取对象空间点数据;将扫描对象三维显示出来.使用线衰减系数不同且相互重叠的多个椭球组成的三维Shepp-Logan模型对算法进行了仿真测试,仿真结果能很好地将模型重建并显示出来.      

基于相关滤波器的嵌入式动态目标跟踪系统设计

针对空间非合作运动目标监控,提出了一种基于嵌入式技术的动态目标实时检测与跟踪方法,并完成了跟踪系统的研制设计和测试。该视觉目标跟踪系统以嵌入式处理器为控制器,通过相机模组获取实时视频图像,图像经过视觉检测与跟踪算法程序分析,得到目标的位置信息后,控制伺服运动系统调整相机姿态实现对动态目标跟踪。提出了基于相关滤波原理的跟踪算法,搭建了树莓派嵌入式测试系统,完成实时检测与跟踪的实验评估。实验结果表明,嵌入式处理器中检测与跟踪程序的平均运行帧率达到25FPS,伺服机构在水平360°和俯仰±60°范围内实现了对移动速度90°/s的目标的实时、稳定监控与跟踪。     

An object-level strategy for pan-sharpening quality assessment of high-resolution satellite imagery

Current satellite imaging systems offer a trade-off between high spatial and high spectral resolution providing panchromatic images at a higher spatial resolution and multispectral images at a lower spatial resolution but rich in spectral information while a wide range of applications need the highest level of this information, simultaneously. Image fusion techniques as means of enhancing the information content of initial panchromatic and multispectral images produce new images, titled pan-sharpened, which inherent the advantages of the initial images. Considering the impact of fusion accuracy on the quality of corresponding applications, it is necessary to evaluate the quality of these processed images. During the last decade, a lot of quality evaluation metrics have been proposed which are mostly inspired by traditional image quality metrics. These methods are mostly based on applying quality metrics at the pixel level and evaluating final quality value based on averaging of obtained metric values through the whole image. However, obtained results clearly show that the behaviour of image fusion quality is inconsistent amongst different image objects. In this article, by applying image fusion quality metrics (IFQMs) to image objects, an object-level strategy for quality assessment of the image fusion process is proposed. The proposed strategy is applied to different satellite imagery covering residential and agricultural areas. Experimental results show higher capabilities of object-level quality assessment strategy in the quality assessment of the fusion process. Evaluating fusion quality at the object level provides the potential of fusion quality assessment for each individual image object in compliance with different parameters such as the type of objects and the effective size of objects in data set.     

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。     

使用深度神经网络检测Cassini ISS图像中圆盘状星体轮廓

Cassini空间探测器光学成像系统（ISS）拍摄的图像中,很多卫星呈现为面元,其轮廓检测是天体测量的重要工作.使用神经网络方法进行ISS图像中面元轮廓检测.每个ISS图像的像素分为轮廓边缘和非轮廓两类.使用神经网络框架TensorFlow,输入每个像素的9个特征,输出每个像素的分类.利用约3.6万个像素训练该网络,通过380幅ISS图像进行测试.与人工标记结果相比,轮廓像素检测的平均精确率为78.26%,平均召回率为73.32%.以检出轮廓像素作为输入,通过椭圆拟合得到面元的轮廓,所得轮廓与面元真实轮廓吻合良好.研究结果表明该方案能够有效检测出面元轮廓,进而给出假图像星的排除范围.      

基于卫星遥感的岛礁影像多类目标智能化提取

卫星遥感影像具有背景复杂、目标尺度不一、观测方向各异、纹理不清晰等特点,主流的深度学习目标检测算法不能直接适用于卫星遥感影像的目标检测。改进了RetinaNet,使其适用于卫星遥感影像。首先设计了一种新的特征融合方式,融合ResNet50输出的特征图,使得融合后的特征图同时具有高层语义信息和低层纹理细节信息。为了减弱遥感影像复杂背景对目标特征的影响,设计了特征感知模块,在减弱噪声对特征图影响的同时增强有用特征。挑选DOTA数据集中船只、飞机和存储罐图像进行训练和测试。改进的算法与RetinaNet相比,飞机、船只和存储罐的平均精度分别提高了41%、25%、24%。基于高分二号卫星（GF 2）真实影像数据的试验结果表明,提出的算法能够用于卫星遥感岛礁影像的多类目标智能化提取。     

空间目标RCS序列的分形分析

空间运动目标雷达散射截面(RCS)序列为非平稳时间序列, 常用时间序列分析方法很难对其进行分析和特征提取。针对部分空间运动目标 RCS序列的变化规律, 首次引入了分数布朗运动模型对其 RCS序列进行分析和特征提取。实测数据处理结果表明,其RCS序列具有分形特性, 其分形维可作为空间目标识别的有效特征。     

真实场景水下语义分割方法及数据集

随着水下生物抓取技术的不断发展,高精度的水下物体识别与分割成为了挑战。已有的水下目标检测技术仅能给出物体的大体位置,无法提供物体轮廓等更加细致的信息,严重影响了抓取效率。为了解决这一问题,标注并建立了真实场景水下语义分割数据集DUT-USEG,该数据集包含6 617张图像,其中1 487张具有语义分割和实例分割标注,剩余5 130张图像具有目标检测框标注。基于该数据集,提出了一个关注边界的半监督水下语义分割网络(US-Net),该网络通过设计伪标签生成器和边界检测子网络,实现了对水下物体与背景之间边界的精细学习,提升了边界区域的分割效果。实验表明:所提方法在DUT-USEG数据集的海参、海胆和海星3个类别上相较于对比方法提升了6.7%,达到了目前最好的分割精度。      

Segmentation of extreme ultraviolet solar images via multichannel fuzzy clustering

The study of the variability of the solar corona and the monitoring of its traditional regions (Coronal Holes, Quiet Sun and Active Regions) are of great importance in astrophysics as well as in view of the Space Weather and Space Climate applications. Here we propose a multichannel unsupervised spatially constrained fuzzy clustering algorithm that automatically segments EUV solar images into Coronal Holes, Quiet Sun and Active Regions. Fuzzy logic allows to manage the various noises present in the images and the imprecision in the definition of the above regions. The process is fast and automatic. It is applied to SoHO–EIT images taken from February 1997 till May 2005, i.e. along almost a full solar cycle. Results in terms of areas and intensity estimations are consistent with previous knowledge. The method reveal the rotational and other mid-term periodicities in the extracted time series across solar cycle 23. Further, such an approach paves the way to bridging observations between spatially resolved data from imaging telescopes and time series from radiometers. Time series resulting form the segmentation of EUV coronal images can indeed provide an essential component in the process of reconstructing the solar spectrum.     

基于免疫抗体编码机理的卫星图像云探测

针对卫星遥感图像中云与下垫面的复杂性和多样性,提出了一种新的在复杂背景下常用目标图像描述方法.通过类比生物免疫抗体特异性与其构成单元氨基酸性质的相关性,得到目标图像的免疫基元集合形式及其分类方法.借鉴生物免疫抗体编码顺序中氨基酸结合能量最小原则,统计分析出训练样本的免疫基元亲和度计算公式,实现了目标图像描述的有限多特征优化组合.对于云检测问题,最终提出云的免疫抗体编码方法,定义了云的免疫亲和度计算公式,并成功构建了云免疫抗体.经200幅IKONOS卫星图像测试,验证了该方法在识别率和运行时间效率方面的有效性.      

基于区域小波变换的序列显微图像融合

利用图像处理技术实现序列显微图像的融合显示是解决显微成像焦深有限的有效方法.针对空域融合算法在处理离焦步长较大和透明显微图像时存在的块效应和重影现象,基于图像像素间高度的相关性和小波变换多分辨率分析的特点,提出了一种基于区域小波变换的序列显微图像融合算法:在小波变换域内进行序列显微图像融合时,不同频段选择不同的融合算子;在确定融合的小波系数时,不仅考虑到相应位置的小波系数,还考虑了与它相邻的小波系数,使得融合后图像克服了空域算法存在的块效应和重影现象.同时,研究了不同的小波滤波器对图像融合质量的影响.实验证明,该方法具有很好的融合效果.      

面向低视角场面监视的移动目标速度测量

为构建有效的机场场面视觉监视系统,提出了一种基于特征点持续跟踪与分析的移动目标速度测量方法。首先,利用场面几何特征对摄像机进行标定;然后,基于光流场对图像运动区域的特征点进行持续跟踪,在此基础上通过特征点轨迹聚类区分不同移动目标;最后,根据特征点高度与运动距离完成速度测量。所提方法能够利用机场场面摄像机获取的低视角单目视频图像,对移动目标的运动速度进行准确测量。基于广州白云国际机场的场面运行视频进行了仿真分析,验证了所提方法在低视角速度测量方面的可行性与优势。      

一种用于分析MCS-51目标码堆栈深度的方法

在嵌入式软件中,针对目标码的堆栈分析是堆栈检查的常用手段．提出了一种用于MCS-51系列处理器目标码的堆栈深度分析方法,该方法可分析最坏情况下的堆栈深度,并考虑了不同优先级下中断服务程序对堆栈的影响．利用该方法可开发出分析MCS-51目标码的堆栈分析工具,其分析结果对堆栈安全检查和优化具有参考意义．     

基于序列图像的空间目标三维重建

空间目标三维重建对空间态势感知和理论研究具有重要意义。针对空间目标图像存在的由纹理重复导致的错误重建问题,提出了一种新的基于运动信息恢复三维场景结构策略。该策略将序列目标图像的成像时间顺序作为先验信息,顺序地加入新图像进行迭代,以避免因目标结构对称、纹理重复所导致的重建错误。同时针对空间目标成像数据匮乏的问题,进行了空间目标图像仿真,并开展了空间目标地面模拟成像实验研究。结果表明:运动分析结果精确,对噪声有较强的鲁棒性,恢复出的目标三维点云能在一定程度上表达目标的结构信息。同时给出了进行空间目标三维重建时图像序列应满足的边界条件。      

Spin motion determination of the Envisat satellite through laser ranging measurements from a single pass measured by a single station

The Satellite Laser Ranging (SLR) technology is used to accurately determine the position of space objects equipped with so-called retro-reflectors or retro-reflector arrays (RRA). This type of measurement allows to measure the range to the spacecraft with high precision, which leads to determination of very accurate orbits for these targets. Non-active spacecraft, which are not attitude controlled any longer, tend to start to spin or tumble under influence of the external and internal torques and forces.If the return signal is measured for a non-spherical non-active rotating object, the signal in the range residuals with respect to the reference orbit is more complex. For rotating objects the return signal shows an oscillating pattern or patterns caused by the RRA moving around the satellite’s centre of mass. This behaviour is projected onto the radial component measured by the SLR.In our work, we demonstrate how the SLR ranging technique from one sensor to a satellite equipped with a RRA can be used to precisely determine its spin motion during one passage. Multiple SLR measurements of one target over time allow to accurately monitor spin motion changes which can be further used for attitude predictions. We show our solutions of the spin motion determined for the non-active ESA satellite Envisat obtained from measurements acquired during years 2013–2015 by the Zimmerwald SLR station, Switzerland. All the necessary parameters are defined for our own so-called point-like model which describes the motion of a point in space around the satellite centre of mass.     

基于窄带雷达RCS序列的空间目标尺寸估计

从窄带雷达横截面积(RCS)序列中提取目标尺寸特征对空间目标的分类识别具有重要价值. 本文在研究椭球体RCS序列特性的基础上, 提出了一种空间目标形状估计方法, 并进一步研究了空间目标尺寸估计问题, 提出了一种新的尺寸估计方法. 实测数据的验证表明, 该方法能有效估计卫星、碎片等空间目标的形状和二维尺寸.      

Improved detection of faint streaks based on a streak-like spatial filter

This paper presents an algorithm to detect very faint object streaks on CCD images acquired with an optical system. The proposed detection method uses image filters simulating the geometrical form and orientation of possible streaks on the CCD image. The method searches for a matching between streak and filter evaluating the convolutions of the image with all possible filters. Based on the statistics of the image background a threshold is applied in order to accept, respectively reject, the possible streak candidates. The detection probabilities and the effect of different parameter settings are estimated with tests on simulated images, while subframes of real images are used to evaluate the applicability of the algorithm to real cases. The detection probability of this method depends on the length and on the signal-to-noise ratio of the streak. For long streaks, a detection for signal-to-noise values around 0.5 is achieved. The further characterization of the detected streak (e.g. centroid and length), which is not performed in the current algorithm, and the reduction of the computation time, which is relatively high for full acquired frames, as well as possible improvements are briefly addressed.