航天三线阵传感器在轨几何检校及其区域网平差技术研究
发布时间:2018-04-10 07:37
本文选题:在轨几何检校 切入点:区域网平差 出处:《武汉大学》2014年博士论文
【摘要】:遥感对地观测技术是当前人类获取地球表层信息的重要手段。随着航天遥感技术的不断发展与进步,越来越多的遥感卫星被发射升空,其地面分辨率和几何定位精度也不断的提高。2014年8月13日发射的WorldView-3是美国Digital Global公司开发的高分辨率遥感卫星WorldView系列第三颗卫星,其地面分辨率已经可以达到0.31m,直接几何定位精度预计优于3.5m,是目前世界范围内最先进的商业遥感卫星。我国也于2010年5月全面启动实施高分专项(高分辨率对地观测系统重大专项)计划,高分辨率遥感卫星对地观测技术已经成为当前遥感领域的研究热点之一,而遥感卫星几何定位技术又是对地观测技术的重要课题,提高遥感卫星几何定位精度有利于遥感卫星数据的广泛应用。 影响遥感卫星几何定位精度的因素主要包括轨道定位精度,姿态定位精度,星上时问基准以及后期的传感器几何定标精度和平差处理精度等,受硬件水平及定轨定姿技术的限制,大部分国产遥感卫星的直接对地定位精度无法满足高精度应用的需求。因此,传感器几何定标技术以及区域网平差技术则成为了提高国产遥感卫星对地几何定位精度的重要手段。 本文主要介绍了当前国内外主要的航天线阵传感器系统及其相关的在轨几何检校和区域网平差技术,并在此基础上确立了以下几个方面的研究内容。 (1)基于海量中等精度控制信息的多条带影像联合在轨几何检校 研究了线阵传感器的成像几何以及定位模型,并针对其成像原理列出了相关的在轨几何检校参数,同时分析上述参数对传感器定位精度的影响以及各类参数之间的相关性,并提出了消除或者削弱其相关性的一些策略。结合当前地面控制点信息缺乏以及在轨几何检校过度依赖检校场的现状,本文提出利用海量中等精度控制信息自动提取技术,采用多条带影像联合在轨几何检校技术,使用大量的资源三号三线阵影像数据,进行相关的在轨几何检校试验,试验结果表明,仅利用自动匹配的中等精度地面控制点的数据,可以达到利用高精度检校场的检校结果。不仅提高了定位精度,减少了检校成本,而且由于地面控制点是自动匹配,因此整个检校的过程已经完全自动化,提高了检校效率。 (2)基于海量中等精度控制信息的线阵影像自检校区域网平差 研究了自检校区域网平差技术的基本原理以及方法理论,并介绍了相关的自检校参数选择办法以及自检校模型的建立(美国布朗教授提出的包含四类改正项的共21个参数模型,德国Ebner教授提出的纯数学意义的共12个附加参数)。自检校参数的选择应当避开各类参数之间的相关性。选择具有几何或者物理意义的参数,能够更好的补偿系统误差。本文针对国产CBERS-02B高分辨率相机的内部成像几何特点,分别采用不同组合的自检校参数,并利用一个条带的CBERS-02B高分辨率影像数据,进行自检校区域网平差试验,试验结果表明,引入了内方位元素改正参数后,平差结果精度比未引入这些参数时的精度要提高了2-3个像素。 (3)带地面控制点的航天线阵影像区域网平差 介绍了经典的粗差检测与定位方法以及相关的选权策略。并选取合适的平差模型和粗差检测与选权策略,分别对天绘一号三线阵影像以及资源三号三线阵影像进行区域网平差试验,试验结果表明,对于长轨道数据(轨道长度超过1000公里),采用合适的同一地面控制点布设方案(本文采用14个地面控制点),定向片模型的平差结果最好。针对资源三号卫星三线阵影像数据,采用系统误差补偿模型,用自动匹配的中等精度地面点作为控制,仅补偿一次项参数即可得到平面不超过15m,高程不超过12m的精度。 (4)基于无地面控制点的大范围多条带航天线阵影像区域网平差 研究了无地面控制点平差的算法原理以及相关的定权方法,采用大范围多条带影像的联合区域网平差的策略,通过各轨道间的重叠度来约束其位置和姿态数据在平差系统中的自由度,提高无控平差的稳定性。分析了对于超大范围数据产生的大数据量存储于运算的问题以及超大法方程的求逆解算问题。采用总共13个条带的资源三号三线阵影像数据进行了基于无地面控制点的区域网平差试验,试验结果表明,资源三号卫星无地面控制点区域网平差的精度相对于直接对地定位的精度,有了较大的提升,达到平面8.3米,高程5.0米,该精度可以满足国家1:5万地形图的绘制要求。根据超大区域网数据处理能力分析结果,超大范围(无缝覆盖整个中国)的区域网平差处理,具备实际可行性。
[Abstract]:The earth observation technology is an important means to obtain the information of the current human earth surface. With the continuous development and progress of remote sensing technology, remote sensing satellites are increasingly being launched, the ground resolution and the geometric positioning accuracy is also improved in August 13th.2014 launched WorldView-3 is a high resolution satellite remote sensing satellite developed by American WorldView series third Digital Global, the ground resolution can reach 0.31m, is expected to direct geometric positioning accuracy is better than 3.5m, is a commercial remote sensing satellite in the world at present. The most advanced in China in May 2010 started the implementation of special high (high resolution earth observation system of major projects) plan, high resolution satellite remote sensing has become one of the hot spots the remote sensing of earth observation, remote sensing and satellite positioning technology and geometry of Earth Observation Technology The important task of the operation is to improve the precision of the remote sensing satellite's geometric positioning, which is beneficial to the wide application of the remote sensing satellite data.
Influence factors of remote sensing satellite geometric positioning accuracy including track positioning accuracy, attitude accuracy, and later asked the reference sensor geometric calibration precision and poor processing precision of star, by the hardware level and the orbit and attitude determination technology is limited, most of the domestic satellite to ground positioning accuracy can not meet the high precision application the demand. Therefore, the sensor geometry calibration and regional adjustment technology has become the important means to improve the domestic satellite remote sensing geometric positioning accuracy.
This paper mainly introduces the current domestic and foreign main line space array sensor system and its related on orbit geometric calibration and regional adjustment technology, and established on the basis of the following research contents.
(1) multi band image and on orbit geometric calibration of mass medium precision control based on information
The study of linear sensor imaging geometry and positioning model, and lists the on orbit geometric calibration parameters related to the imaging principle, analyzing the influence of parameters on sensor positioning accuracy and the correlation between the various parameters at the same time, and puts forward some strategies to eliminate or weaken the correlation. According to the current situation of ground control points and the lack of information over reliance on orbit geometric calibration calibration field, this paper proposes the use of mass medium precision control automatic information extraction technology, using multi band image and on orbit geometric calibration technology, using a lot of resources three, three line array image data, for on orbit geometric calibration test, test results show that with only moderate accuracy of ground control points automatic data matching, can be achieved by using high precision calibration field calibration results. Not only improves the positioning accuracy and reduce the The calibration cost, and because the ground control points are automatically matched, so the whole process has been fully automated calibration, improve the calibration efficiency.
(2) regional network adjustment of linear image self-examination Campus Based on mass medium precision control information
Study on the basic principle of self calibration block adjustment technology and method of theory, and introduces the establishment of self calibration parameters related to the choice of methods and self calibration model (a total of 12 additional parameters of 21 parameters including four kinds of correction model proposed by Professor Brown, the pure mathematical significance put forward by Professor Ebner in Germany.). Self calibration parameters should avoid all kinds of correlation between parameters. Parameter selection with geometric or physical meaning, can compensate the system error better. Based on the domestic CBERS-02B high resolution camera imaging geometric characteristics, self calibration parameters of different combinations were used, and the use of a strip of high resolution CBERS-02B image the data were self calibration block adjustment test, test results show that the introduction of elements of interior orientation correction parameters, the adjustment results than without introducing these parameters The accuracy of the time is increased by 2-3 pixels.
(3) regional network adjustment of space linear array images with ground control points
The classic gross error detection and location method and choose the right strategy. And select the appropriate adjustment model and outlier detection and choose the right strategy, respectively on the day painted a No. three line array images and resources three, three line array images of regional network adjustment test, test results show that for long track data (track length more than 1000 km), using the same ground control points (using appropriate 14 ground control points), directional model adjustment results. The best resources for the three satellite three line array image data, a systematic error compensation model, with medium precision ground point automatic matching as a control, only one can get the compensation parameters plane does not exceed 15m, height of not more than 12m precision.
(4) regional network adjustment of large range multi strip space linear array images based on no ground control points
Study on the algorithm of point adjustment principle and method related to the right of ground control, using a large range of multiple joint area network with image adjustment strategy, the degree of overlap between each track to constrain the data of position and attitude adjustment in degrees of freedom in the system, improve the stability of control adjustment the analysis of large data. For the large range of data stored in the production operation and large equation inverse solution. The area network without ground control point adjustment test based on a total of 13 bands of the No. three resources three line array image data, test results show that the resources of No. three the satellite ground control points of regional network adjustment with respect to the accuracy of direct precision of positioning, have been greatly improved, the plane reached 8.3 meters, height of 5 meters, the precision can meet the national 1:5 million topographic map drawing requirements. According to the large area The analysis results of the data processing capability of the domain network, the area network adjustment processing with a wide range (covering the whole of China), has practical feasibility.
【学位授予单位】:武汉大学
【学位级别】:博士
【学位授予年份】:2014
【分类号】:P237;P207.2
【参考文献】
相关期刊论文 前10条
1 杨博;王密;;资源一号02C卫星全色相机在轨几何定标方法[J];遥感学报;2013年05期
2 张永军;郑茂腾;王新义;黄心蕙;;“天绘一号”卫星三线阵影像条带式区域网平差[J];遥感学报;2012年S1期
3 袁修孝;余翔;;高分辨率卫星遥感影像姿态角系统误差检校[J];测绘学报;2012年03期
4 李德仁;;我国第一颗民用三线阵立体测图卫星——资源三号测绘卫星[J];测绘学报;2012年03期
5 李德仁;王密;;“资源三号”卫星在轨几何定标及精度评估[J];航天返回与遥感;2012年03期
6 张永生;;高分辨率遥感测绘嵩山实验场的设计与实现——兼论航空航天遥感定位精度与可靠性的基地化验证方法[J];测绘科学技术学报;2012年02期
7 祝小勇;张过;唐新明;祝欣欣;翟亮;;资源一号02B卫星影像几何外检校研究及应用[J];地理与地理信息科学;2009年03期
8 赵德文;;ADS40数字传感器的摄影测量处理与应用[J];铁道勘察;2009年02期
9 袁修孝;;一种补偿POS定位测姿系统误差的新方法[J];自然科学进展;2008年08期
10 袁修孝;余俊鹏;;高分辨率卫星遥感影像的姿态角常差检校[J];测绘学报;2008年01期
,本文编号:1730285
本文链接:https://www.wllwen.com/kejilunwen/dizhicehuilunwen/1730285.html