星載SAR影像定位與InSAR中應(yīng)用有理函數(shù)模型的研究
發(fā)布時(shí)間:2018-03-05 06:34
本文選題:星載SAR影像 切入點(diǎn):距離-多普勒模型 出處:《解放軍信息工程大學(xué)》2013年碩士論文 論文類型:學(xué)位論文
【摘要】:星載SAR系統(tǒng)的技術(shù)水平日益先進(jìn),星載SAR影像的空間分辨率和時(shí)間分辨率也不斷提高,在對(duì)地觀測(cè)領(lǐng)域扮演著越來(lái)越重要的角色。星載SAR影像的無(wú)控定位技術(shù)可以獲取測(cè)繪困難地區(qū)重點(diǎn)目標(biāo)的三維坐標(biāo),是星載SAR影像的重要應(yīng)用之一;有理函數(shù)模型(Rational Funtion Model, RFM)在光學(xué)遙感影像和SAR影像的攝影測(cè)量處理中得到了廣泛應(yīng)用,將RFM引入星載InSAR數(shù)據(jù)處理流程,也具有重要的理論和實(shí)踐意義。 本文以高分辨率SAR影像的應(yīng)用為主線,分析了SAR影像的嚴(yán)密幾何模型和通用幾何模型,分別對(duì)稀少(無(wú))控制條件下星載SAR影像的目標(biāo)定位技術(shù)和RFM在InSAR數(shù)據(jù)處理流程中的應(yīng)用進(jìn)行了研究,主要工作包括以下內(nèi)容: 1)從載荷技術(shù)、成像特點(diǎn)和應(yīng)用技術(shù)三個(gè)方面,總結(jié)了SAR和InSAR技術(shù)以及有理函數(shù)模型的研究和發(fā)展現(xiàn)狀,確定了論文的研究目的和意義。 2)基于距離-多普勒模型,系統(tǒng)闡述了星載SAR影像的無(wú)地面控制信息下的定位原理。利用高分辨率TerraSAR影像進(jìn)行了相關(guān)實(shí)驗(yàn),無(wú)控定位精度在方位向達(dá)到了70m,距離向達(dá)到了50m。利用少量控制點(diǎn)精化模型定向參數(shù),將定位精度提高至方位向25m,距離向20m。提出了一種新的目標(biāo)高程歸化至赤道半徑的計(jì)算方法,通過(guò)推導(dǎo)和仿真,分析得出在高緯度地區(qū),地形起伏將產(chǎn)生最大60米左右的定位誤差。 3)從InSAR基本幾何原理出發(fā),建立了InSAR嚴(yán)密幾何模型和通用幾何模型。對(duì)比了光學(xué)攝影測(cè)量和InSAR的共性和差異,闡明了將RFM引入InSAR數(shù)據(jù)處理的可行性。 4)對(duì)InSAR干涉影像對(duì),分別建立了用于主輔影像配準(zhǔn)、去平地效應(yīng)和相位-高程轉(zhuǎn)換的RFM方程;基于全球低分辨率DEM,設(shè)計(jì)了一種無(wú)需初值和迭代的RFM參數(shù)求解方法。以SRTM模擬干涉數(shù)據(jù)和天山地區(qū)模擬干涉數(shù)據(jù),進(jìn)行了相位-高程轉(zhuǎn)換實(shí)驗(yàn),,二階RFM的替代精度分別達(dá)到了0.1941米和0.000014米。為解決RFM參數(shù)求解時(shí)的病態(tài)問(wèn)題,分別采取了直接計(jì)算方法改正和間接誤差補(bǔ)償改正兩種方法,SRTM模擬數(shù)據(jù)直接計(jì)算改正的精度提高了17.9%,間接誤差補(bǔ)償?shù)木忍岣吡?4.8%。
[Abstract]:The technology level of spaceborne SAR system is becoming more and more advanced, and the spatial and temporal resolution of spaceborne SAR image is also improving. It is one of the important applications of spaceborne SAR images to obtain the 3D coordinates of the key targets in difficult areas by using the uncontrolled localization technology of spaceborne SAR images, which plays an increasingly important role in the field of Earth observation. Rational Funtion Model (RFM) has been widely used in photogrammetric processing of optical remote sensing and SAR images. It is also of great theoretical and practical significance to introduce RFM into spaceborne InSAR data processing process. Based on the application of high-resolution SAR images, the strict geometric model and the general geometric model of SAR images are analyzed in this paper. The target location technology of spaceborne SAR images and the application of RFM in InSAR data processing are studied respectively under rare (no) control conditions. The main work includes the following contents:. 1) the research and development status of SAR and InSAR technology and rational function model are summarized from three aspects of load technology, imaging characteristics and application technology, and the purpose and significance of this paper are determined. 2) based on the range-Doppler model, the positioning principle of spaceborne SAR images without ground control information is systematically expounded. Experiments are carried out using high-resolution TerraSAR images. The accuracy of uncontrolled positioning reaches 70 m in azimuth direction and 50 m in distance direction. By using a small number of control point refinement model orientation parameters, the positioning accuracy is raised to 25 m in azimuth direction and 20 m in distance direction. A new method for calculating the elevation of target to equatorial radius is proposed. Through derivation and simulation, it is concluded that the maximum positioning error is about 60 meters in the high latitude area. 3) based on the basic geometry principle of InSAR, the strict geometric model of InSAR and the general geometric model of InSAR are established, the commonness and difference between optical photogrammetry and InSAR are compared, and the feasibility of introducing RFM into InSAR data processing is expounded. 4) for the InSAR interference image pair, the RFM equations are established for the main and auxiliary image registration, levelling effect and phase-height conversion. Based on global low resolution demm, a method of solving RFM parameters without initial value and iteration is designed. Using SRTM simulated interference data and Tianshan region simulated interference data, phase to height conversion experiments are carried out. The substitution accuracy of second-order RFM is 0.1941m and 0.000014m respectively. In order to solve the ill-posed problem of RFM parameters, Two methods, direct calculation correction and indirect error compensation correction, are adopted to improve the accuracy of direct calculation correction of SRTM simulation data and the accuracy of indirect error compensation by 17.9% and 44.8% respectively.
【學(xué)位授予單位】:解放軍信息工程大學(xué)
【學(xué)位級(jí)別】:碩士
【學(xué)位授予年份】:2013
【分類號(hào)】:P228
【參考文獻(xiàn)】
相關(guān)期刊論文 前4條
1 周金萍,唐伶俐,李傳榮;星載SAR圖像的兩種實(shí)用化R-D定位模型及其精度比較[J];遙感學(xué)報(bào);2001年03期
2 朱彩英,徐青,吳從暉,池天河,何建邦;機(jī)載SAR圖像幾何糾正的數(shù)學(xué)模型研究[J];遙感學(xué)報(bào);2003年02期
3 陳爾學(xué);李增元;盧穎;田昕;;三顆高分辨率星載SAR的定位模型構(gòu)建及其定位精度評(píng)價(jià)[J];遙感信息;2010年02期
4 劉佳音;韓冰;洪文;;一種新的SAR圖像斜距多普勒定位模型的直接解法[J];遙感技術(shù)與應(yīng)用;2012年05期
本文編號(hào):1569134
本文鏈接:http://sikaile.net/kejilunwen/dizhicehuilunwen/1569134.html
最近更新
教材專著