缺失飞行数据的雷达原始数据成像和运动补偿

发布时间：2018-03-19 23:04

本文选题：航空和卫星雷达　切入点：成像　出处：《武汉大学》2013年博士论文　论文类型：学位论文

【摘要】：美国国家航空航天局(NASA)戈达德航天中心(GSFC)最近研发了一款先进的L波段航空雷达——波束成形合成孔径雷达(DBSAR)。为了集成多模式的雷达测量技术,该雷达传感器采用了一些高级雷达技术,该雷达可以自定义数据获取模式,也可以实现实时在线处理。DBSAR将被作为一个测试平台用于开发、实施和测试波束成形合成孔径雷达技术用于地球科学和行星测量。本研究的目的是将DBSAR原始数据处理成单视复数图像(SLC),并且尽可能的考虑运动补偿及干涉测量应用,如生产数字高程模型(DEMs)和二路的差分解缠相位图。然而,本研究的DBSAR数据缺乏高精度飞行数据,这对高清晰和高相位保真地成像处理造成了一些特殊困难。因为：1)飞行的前向速度和近距长度等运动参数对成像来说是必不可少的。缺乏这些参数将无法对DBSAR成像;2)航空合成孔径雷达(SAR)通常都会受到气流抖动的影响,这会使得传感器的航线偏离正常的直线飞行轨迹。缺乏这些运动偏离参数将无法对这些运动做相应的补偿,并成出高清晰和高相位保真的图像。为了处理DBSAR的原始数据,作者开发了一款针对航空雷达原始数据成像和运动补偿软件。对雷达成像处理来说,选择一个高精度和高效的成像算法对最终聚焦的图像产品的清晰度和相位保真度很重要,特别是当客户对清晰度和相位保真度要求较高时。基于若干实验对比,作者从多个成像算法(如range-Doppler算法,chirp scaling算法和wavenumber domain算法)中选择了chirp scaling这一高效和高精度的算法,并将其用于聚焦DBSAR原始数据。为了提高成像质量和相位保真度,成像过程中必须将传感器的运动轨迹纠正到理想的直线运动轨迹。传感器的运动偏离误差可以从惯性导航系统(INS)和全球定位系统(GPS)数据,或原始数据中获得。由于本研究中DBSAR的飞行数据不可用,这些运动误差数据只能从原始数据中估计。为了估计这些运动误差数据,本文提出了一种基于chirp scaling算法的运动误差参数提取方法。该方法可以很好地分离出方位向的前向速度和视线方向(LOS)上的偏移量。该方法分两步：第一步,从原始数据中提取出即时的多普勒调频率；第二步,基于这些提取的调频率,分离出方位向的前向速度和视线方向上的偏移量。为了实施运动补偿,本文提出一种基于chirp scaling算法的运动误差补偿方法。该方法在距离方向上和方位向分两步补偿从原始数据中提取的运动误差。首先,在距离方向上补偿由视线方向上的偏移量引起的相位误差和信号包络偏移；然后,在方位向和range-Doppler域补偿由于前向速度变化引起的多普勒调频率变化,完成方位向压缩后,再重采样聚焦后的图像来完成地物在方位向的位置偏移纠正。与使用统一速度和近距离,而不考虑前向速度和方位向偏移量的传统的chirp scaling算法,结合了运动补偿的chirp scaling算法聚焦出了明显质量更高的图像。本文的另一个强调的重点是在通过实施雷达干涉测量,来验证这些用从原始数据中估计的参数聚焦出的SLC图像的相位保真度。由于DBSAR缺乏高精度飞行数据,本研究不能完全补偿由运动误差造成的散焦情况,相位也不能完全保真。而且,我们仅有一幅DBSAR数据,不能使用仅有的一幅图像开展重复轨道的干涉测量研究。下一代的DBSAR数据中将包含高精度的飞行数据,并将开展二路或多路的重复轨道干涉。为了准备下一代的DBSAR研发,本文提出并运行了一个全面的评估方案,来评估用从一对Advanced Land Observing Satellite/Phased Array type L-band Synthetic Aperture Radar (ALOS/PALSAR)原始数据中提取的各项参数来聚焦图像,并产生数字高程模型(DEMs)和2路差分干涉解缠相位图的可能性。初步的实验结果表明对卫星雷达来说,从原始数据中提取的各项参数来聚焦图像,并产生数字高程模型(DEMs)和2路差分干涉解缠相位图是可行的,特别是当精度要求不太高时。也就是说,用从原始数据中提取的各项参数聚焦出的SLC图像相位是足够保真的,这进一步证明了本文提出的运动补偿方案是可行的。该评估为下一代DBSAR研发提供了一个积极的参考。这些聚焦后的SLC图像将帮助NASA/GSFC的其他科研人员测量北美森林的三维结构(树高和稠密度)。
[Abstract]:The United States National Aeronautics and Space Administration (NASA) Goddard Space Center (GSFC) recently developed an advanced synthetic aperture radar L band airborne radar - beamforming (DBSAR) radar technology. In order to integrate multi mode, the radar sensor adopts some advanced radar technology, the radar can customize the data acquisition mode, can also be to achieve real-time online processing of.DBSAR will be used as a test platform for the development, implementation and testing of beamforming synthetic aperture radar technology for Earth Science and planetary geodesy. The purpose of this study is to DBSAR the original data processing into a single look complex (SLC), and as much as possible to consider the motion compensation and interference measurement applications, such as the production of digital elevation model (DEMs) and two differential phase unwrapping.
However, the lack of DBSAR data in the study of high precision flight data, which caused some special difficulties of high definition and high phase fidelity imaging. Because: 1) the flight parameters to the speed and length of short-range motion on the imaging is essential. You will not be able to DBSAR imaging is the lack of these parameters; 2) aviation synthetic aperture radar (SAR) are usually affected by airflow jitter, which makes the sensor deviation linear normal flight trajectories. The lack of these parameters will not deviate from the movement of these sports do the corresponding compensation, and a high resolution image and high phase fidelity.
In order to deal with the original data of DBSAR, the author developed a software for radar original data imaging and motion compensation.
For radar imaging, clarity and fidelity of phase selection of a high precision and efficient imaging algorithm focusing on the final image of the product is very important, especially when the customer for clarity and fidelity phase higher requirements. Some experiments based on the author, from a plurality of imaging algorithms (such as range-Doppler algorithm, chirp scaling algorithm and wavenumber domain chose chirp scaling) algorithm in the high efficiency and high accuracy of the algorithm, and the DBSAR used to focus on the original data.
In order to improve the imaging quality and phase fidelity, the imaging process must be corrected trajectory sensor to linear motion trajectory. The ideal movement of the sensor deviation from the inertial navigation system (INS) and global positioning system (GPS) data, or the original data. Because the flight data of DBSAR in this study is not available. The motion error of data can only be estimated from the original data. In order to estimate the motion error of the data, this paper proposes a extraction method of kinematic error parameters of chirp based on scaling algorithm. This method can be effectively separated prior to the range speed and direction of the line of sight (LOS) offset. This method consists of two steps the first step: extract the Doppler frequency immediately from the original data; the second step, the extraction rate of FM Based on isolated range prior to the speed and direction of the line of sight to the partial Shift. In order to implement the motion compensation, this paper proposes a motion compensation method of chirp based on scaling algorithm. The motion error of this method is in the range direction and azimuth on the two steps of compensation from the original data extraction. Firstly, the phase error and the signal envelope offset caused by the offset along the line of sight of the compensation in the distance direction; then, in azimuth and range-Doppler because of the Doppler frequency domain compensation to changes caused by the change in speed, complete azimuth compression, re sampling the image after focusing to complete the object in the azimuth position offset correction. With the use of uniform speed and short distance, without considering the forward velocity and azimuth to offset the traditional chirp scaling algorithm, chirp scaling algorithm combines focused motion compensation out of the image significantly higher quality.
Another emphasis of this paper is through the implementation of radar interferometry, to verify the fidelity of phase out of the SLC image focusing parameters estimated from the original data of these. Due to lack of DBSAR high precision flight data, this research could not completely compensate the defocusing caused by motion error, phase can not be completely fidelity. Moreover, we only one DBSAR data, can not carry out repeat track interferometry using an image only. DBSAR data will be the next generation contains flight data with high precision, and will carry out repeat orbit two path or multi-path interference. In order to prepare for the next generation of DBSAR research, proposed and run a comprehensive evaluation scheme to assess from a pair of Advanced Land Observing Satellite/Phased Array type L-band Synthetic Aperture Radar (ALOS/PALSAR) in the extraction of raw data The parameters to focus image and digital elevation model (DEMs) interference possibility of unwrapping phase map and 2. Preliminary experimental results show that the satellite radar, the parameters extracted from the raw data to produce a focused image and digital elevation model (DEMs) and 2 differential interferometric phase unwrapping the phase diagram is feasible, especially when the accuracy requirement is not too high. That is to say, using SLC image phase parameters of focusing the extraction from the original data is sufficient fidelity, which proves that the motion compensation scheme is proposed in this paper. It is feasible to provide a positive reference for the evaluation generation of DBSAR is developed. The 3D structure of SLC images of the focus will help NASA/GSFC other researchers measured North American forest (tree height and density).

【学位授予单位】：武汉大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：P23;P225.1

【相似文献】