【作者】 江振治；

【导师】 韩玲； 汤晓涛；

【作者基本信息】 长安大学 ， 摄影测量与遥感， 2009， 硕士

【摘要】 测绘卫星是目前获取境外目标空间信息的主要技术手段,为实现摄影定位卫星无地面控制条件下的目标定位,在卫星对地摄影的同时,必须对星空的恒星进行拍摄,通过处理恒星影像来获取地像片摄影瞬间的姿态角元素。这些姿态数据在卫星摄影测量定位中将起到重要的控制作用:能有效地抑制系统误差的累积,或在无地面控制时,与轨道数据一同使用,通过构建摄影测量控制网和区域网的平差计算,达到精确测定目标点和控制网点坐标的任务。像片姿态的计算原理是双矢量定姿原理以及摄影测量中的共线方程。通过以下步骤求得地像片的姿态角:即根据精确量测所得的恒星影像坐标,经解析定向,将星像点坐标归化到框标坐标系中;利用快速找星算法从星表数据库中查找与其相对应的恒星天球坐标,并按标准历元进行恒星视位置化算;利用归化过的星像点坐标和相应的恒星天球坐标(经过视位置化算)按共线方程组成误差方程,采用最小二乘法平差答解求出摄影瞬间星像片的姿态角;通过建立星地相机空间数据转换模型,确定出地像片姿态角,为摄影定位卫星提供高精度的姿态数据。主要成果和结论如下:(1)利用图像处理技术自动判认恒星像片框标点和星像点,并精确量其像点坐标,解决了恒星影像的框标点和星像点的自动判认和量测问题。(2)根据矢量定姿原理,建立了高精度星像片姿态计算模型,并综合利用J2000星表数据库和恒星视位置的计算方法,使判星能力得到提高,并显著提高了星像片姿态的计算精度和可靠性。(3)研究了快速实用的自动找星算法。能够根据星像点的量测坐标,采用快速查找算法,从星表数据库的数十万条记录进行筛选,找出与星像点相对应的恒星。(4)建立了星地相机空间数据转换模型,研究开发了基于数字影像的卫星像片姿态计算系统。(5)实际处理了63张恒星影像,所判的星像点数量多且分布均匀,满足了整体可靠性和整体精度指标要求,姿态角精度平均为m_α,m_δ≤0.8″,m_κ≤5″。(6)卫星像片姿态数据应用于卫星像片空中三角测量的加密结果,得知有像片姿态角控制比无像片姿态角控制的加密精度要好得多,表明像片姿态角控制能有效地消除系统误差累积,提高了整体精度。更多还原

【Abstract】 The surveying and mapping satellite is the main technology means to acquire the outer-boundary targets’ space information.In order to complement the photographic position satellites’ targets’ position without ground control conditions,Simultaneity when the satellite photographs the ground,it must shoot the star and then after the process of the star’s image can acquire the attitude and angle elements of the ground photo’s instant photography.And these attitude data will be important in controlling in the satellite photogrammetry position:effectively suppressing the system errors’ accumulation,or without ground controls, being used with the orbit data,by constructing photogrammetry controlling grid and the block adjustment calculation,it can achieve the assignment that accurately survey the target points and control the grid points’ coordinates.It is possible to do the big block adjustment calculation.The photo attitude’s computed principle is the double vectors attitude principle and the photogrammetry’s collinear equation.Through the followed steps it can get the photo’s attitudes and angles:through the star’s image coordinates which is acquired from accurately surveyed,after analytical orientation,normalizing the stellar points’ coordinates to the collimating mark coordinate system;By the rapid looking for star algorithms getting the correspondent star’s celestial coordinates from the star catalogs database,then normalizing the star’s apparent place by the standard epoch. Applying the normalized stellar points’ coordinates and the correspondent star celestial coordinates (normalized by the apparent place),then constituting the error equations by the co-linearity equations, it can solve the photo’s attitude angle of the instant photographic.By constructing the satellite-ground camera’s space data conversion model,determining the ground photo’s attitudes and angles,it can offer high accurate attitude data for the photogrammetry and position satellites.The major progenies and conclusions are generalized as follows:(1)Utilizing the image process technology to automatically recognize the star photo collimating mark points and the stellar points and accurately survey the photo coordinates,at last solving out the problem that the automatic recognizing and surveying of the star photo collimating mark points and the stellar points.(2)According to the vector surveying principle,constituting a high accuracy stellar photo’s attitude computation model,and synthetically utilizing J2000 star catalogs database and the star apparent place’s computation methods,making the capacity of the star recognition improved,simultaneity obviously increasing the computation accuracy and reliability of the stellar photos’ attitude.(3)Researching the rapid and practical automatic looking for stars algorithms,by the stellar’s surveying coordinates,applying rapid looking for algorithms,it can find the correspondent star of the stellar sieving from star catalogs database’s 1000 thousands of records.(4)Constituting the stellar cameras’ space data conversion model,in researches and develops the satellite photo’s attitude computation system based on the digital image.(5)Practically processing 63 pieces of star images,the quantity of the recognized stellar is many and uniform distribution,it is content for the index requirement of the total reliability and total accuracy,the attitude angle’s average accuracy is m_α,m_δ≤0.8″,m_κ≤5″.(6)When the satellite photo attitude data applied in the satellite photo aerial triangulation’s densification results,it can draw a conclusion that the densification accuracy is better with photo attitude angle control than without photo attitude angle control,especially it indicates that the photo attitude angle can effectively eliminate the system errors’ accumulation and increase the whole accuracy.更多还原