【摘要】：合成孔徑雷達(dá)(Synthetic Aperture Radar,SAR)具有全天時(shí)、全天候、觀測(cè)范圍廣的特點(diǎn),在遙感領(lǐng)域具有越來(lái)越廣泛的應(yīng)用。內(nèi)波(Internal Waves,IW)是指在流體內(nèi)部發(fā)生的一種波動(dòng)形式,而海洋內(nèi)波就是一種典型的內(nèi)波形式。由于海洋內(nèi)波具有較大的幅度,傳遞巨大的能量,在海洋開(kāi)發(fā)、船只航線規(guī)劃乃至國(guó)防軍事領(lǐng)域都具有重要的研究意義。并且,隨著分辨率的不斷提高,SAR對(duì)海面的觀察能力逐漸提升,包含的信息也更加的豐富。由于電磁波無(wú)法深入到海洋中直接對(duì)海洋內(nèi)波進(jìn)行觀測(cè),所以僅能通過(guò)海洋表面的變化,推導(dǎo)海洋海洋內(nèi)波的相關(guān)參數(shù)。但是,由于從海洋內(nèi)波到海洋表面,再到接收機(jī)的過(guò)程會(huì)引入較多的干擾因素,使得在運(yùn)算量和估計(jì)的準(zhǔn)確性上都存在較大的可改進(jìn)性。基于上面的原因,本課題在分析內(nèi)波的形成機(jī)理、SAR對(duì)海洋表面的成像模型基礎(chǔ)上,著眼于利用SAR數(shù)據(jù),探究在SAR參數(shù)對(duì)內(nèi)波圖像的影響、利用圖像處理的方法在SAR圖像中進(jìn)行在圖像內(nèi)定位內(nèi)波的位置、使用期望最大化算法對(duì)統(tǒng)計(jì)直方圖進(jìn)行模型擬合以及在SAR圖像中進(jìn)行內(nèi)波參數(shù)提取的方法。本文的主要研究?jī)?nèi)容如下:論文首先建立海洋內(nèi)波產(chǎn)生的模型,并推導(dǎo)其所滿足的動(dòng)力學(xué)方程,求取滿足方程的穩(wěn)態(tài)解。對(duì)海洋表面在SAR觀測(cè)情況下的三種不同成像模型進(jìn)行了介紹;同時(shí),針對(duì)接收幅度過(guò)小的情況,引入調(diào)制深度的概念,利用仿真工具和特定的內(nèi)波模型以及成像模型,通過(guò)針對(duì)沿內(nèi)波傳播方向上極值點(diǎn)隨著參數(shù)變化的趨勢(shì),分析參數(shù)對(duì)成像效果帶來(lái)的影響,尋找最佳的雷達(dá)觀測(cè)參數(shù)組合。其次,利用簡(jiǎn)單圖像處理的方法對(duì)SAR圖像進(jìn)行預(yù)處理,分離出海洋區(qū)域;利用基于馬爾科夫隨機(jī)場(chǎng)的分割方法對(duì)含有內(nèi)波條紋的圖像進(jìn)行紋理增強(qiáng),并借助基于Radon變換的紋理檢測(cè)方法,從大場(chǎng)景的SAR圖像中,快速定位內(nèi)波在圖像中的位置;針對(duì)SAR圖像分辨率提升帶來(lái)統(tǒng)計(jì)直方圖出現(xiàn)多模態(tài)、拖尾嚴(yán)重的現(xiàn)象,結(jié)合之前學(xué)者提出的廣義混合模型,利用期望最大化的方法進(jìn)行分布模型的參數(shù)估計(jì),為SAR圖像的分類、降噪和目標(biāo)檢測(cè)提供輔助作用;從估計(jì)結(jié)果與實(shí)際結(jié)果的均方差衡量估計(jì)的結(jié)果,通過(guò)迭代次數(shù)衡量方法的計(jì)算量。最后,利用已知的先驗(yàn)?zāi)Ｐ秃瞳@得的圖像剖面數(shù)據(jù),采用曲線擬合的方式,對(duì)模型中的參數(shù)進(jìn)行確定;同時(shí)針對(duì)內(nèi)波圖像剖面的非線性特點(diǎn),利用基于經(jīng)驗(yàn)?zāi)B(tài)分解的方法,分解得到內(nèi)波分量,并利用該分量對(duì)內(nèi)波參數(shù)進(jìn)行估計(jì);根據(jù)干涉的概念以及地面運(yùn)動(dòng)速度和干涉相位的關(guān)系,利用干涉相位提取海洋表面的速度信息;提出基于仿真迭代的參數(shù)估計(jì)方法,對(duì)內(nèi)波的參數(shù)進(jìn)行迭代估計(jì)。
[Abstract]:Synthetic Aperture Radar (Synthetic Aperture Radar,SAR) has been widely used in the field of remote sensing because of its wide range of observation and all-weather. Internal wave (Internal Waves,IW) is a form of internal wave occurring in fluid, and ocean wave is a typical form of internal wave. Because of the large amplitude of ocean internal wave and the transmission of huge energy, it is of great significance in the field of ocean exploitation, ship route planning and even national defense and military affairs. Moreover, with the improvement of resolution, the ability of SAR to observe sea surface is improved gradually, and the information is more abundant. Because the electromagnetic wave can not penetrate into the ocean directly to observe the ocean internal wave, it can only deduce the relevant parameters of the ocean internal wave by the variation of the ocean surface. However, the process from the ocean internal wave to the ocean surface, and then to the receiver, will introduce more interference factors, which makes it possible to improve the computational complexity and the accuracy of the estimation. Based on the above reasons, based on the analysis of the formation mechanism of internal waves and the imaging model of SAR on the ocean surface, we focus on the use of SAR data to explore the influence of SAR parameters on the internal wave images. The method of image processing is used to locate the position of the internal wave in the SAR image, the model fitting of the statistical histogram and the extraction of the internal wave parameters in the SAR image are carried out by using the expectation maximization algorithm. The main contents of this paper are as follows: firstly, the model of ocean internal wave generation is established, and the satisfied dynamic equation is derived, and the steady-state solution of the satisfied equation is obtained. Three different imaging models of ocean surface under SAR observation are introduced. At the same time, the concept of modulation depth is introduced in the case of too small receiving amplitude. By using the simulation tool, the specific internal wave model and the imaging model, the trend of the extreme point changing with the parameters along the direction of the internal wave propagation is analyzed. The effect of the parameters on the imaging effect is analyzed and the best combination of radar observation parameters is found. Secondly, the SAR image is preprocessed by the simple image processing method, and the ocean area is separated. The segmentation method based on Markov random field is used to enhance the texture of the image with internal wave fringes. With the help of texture detection method based on Radon transform, the position of internal wave in large scene SAR image is quickly located. Aiming at the phenomenon of multi-modal and serious trailing in the statistical histogram caused by the resolution enhancement of SAR image, combined with the generalized mixed model proposed by the previous scholars, the parameter estimation of the distribution model is carried out by using the method of expectation maximization, which is the classification of the SAR image. Noise reduction and target detection provide auxiliary effect; The estimated results are measured from the RMS of the estimated results and the actual results, and the computational complexity of the method is measured by the number of iterations. Finally, using the known prior model and the obtained image profile data, the parameters in the model are determined by the way of curve fitting. At the same time, according to the nonlinear characteristics of the internal wave image profile, the internal wave component is obtained by using the empirical mode decomposition method, and the internal wave parameters are estimated by this component. According to the concept of interference and the relationship between the velocity of ground motion and the phase of interference, the velocity information of ocean surface is extracted by using the phase of interference, and a parameter estimation method based on simulation iteration is proposed to estimate the parameters of internal wave iteratively.
【學(xué)位授予單位】：哈爾濱工業(yè)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TN958

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 范開(kāi)國(guó);顧艷鎮(zhèn);陳春濤;許可;陳鵬;;希爾伯特—黃變換在遙感圖像中的實(shí)例研究與應(yīng)用展望[J];海洋技術(shù)學(xué)報(bào);2015年04期

2 陳捷;陳標(biāo);陶榮華;許素芹;;SAR圖像海洋內(nèi)波參數(shù)自動(dòng)提取方法[J];海洋技術(shù)學(xué)報(bào);2014年06期

3 楊剛;陳標(biāo);張本濤;;EMD在海洋內(nèi)波研究中的應(yīng)用[J];艦船電子工程;2014年03期

4 丁燦;張本濤;高國(guó)興;陳標(biāo);吳寶勤;;合成孔徑雷達(dá)圖像中海洋內(nèi)波的特征檢測(cè)[J];海洋測(cè)繪;2012年05期

5 汪雄良;王春玲;朱炬波;梁甸農(nóng);;基于EEMD的SAR海洋內(nèi)波參數(shù)反演[J];測(cè)試技術(shù)學(xué)報(bào);2012年01期

6 宋海斌;拜陽(yáng);董崇志;宋洋;;南海東北部?jī)?nèi)波特征——經(jīng)驗(yàn)?zāi)B(tài)分解方法應(yīng)用初探[J];地球物理學(xué)報(bào);2010年02期

7 宋海斌;Luis Pinheiro;王東曉;董崇志;宋洋;拜陽(yáng);;海洋中尺度渦與內(nèi)波的地震圖像[J];地球物理學(xué)報(bào);2009年11期

8 陳捷;陳標(biāo);許素芹;;一種新的SAR圖像內(nèi)波檢測(cè)與參數(shù)估計(jì)方法[J];中國(guó)圖象圖形學(xué)報(bào);2009年01期

9 陳標(biāo);侯海平;陶榮華;丁燦;;艦船尾跡SAR圖像特征的譜分析研究[J];艦船科學(xué)技術(shù);2008年06期

10 甘錫林;黃韋艮;楊勁松;周長(zhǎng)寶;史愛(ài)琴;金為民;;基于希爾伯特-黃變換的合成孔徑雷達(dá)內(nèi)波參數(shù)提取新方法[J];遙感學(xué)報(bào);2007年01期

相關(guān)會(huì)議論文 前2條

1 高建虎;陳杰;張履謙;;基于壓縮感知和EMD的SAR海洋內(nèi)波探測(cè)方法[A];第九屆全國(guó)信息獲取與處理學(xué)術(shù)會(huì)議論文集Ⅱ[C];2011年

2 何宜軍;趙進(jìn)平;;TOPEX/POSEIDON衛(wèi)星高度計(jì)海洋潮波分析[A];第十三屆全國(guó)遙感技術(shù)學(xué)術(shù)交流會(huì)論文摘要集[C];2001年

相關(guān)碩士學(xué)位論文 前1條

1 李海艷;利用合成孔徑雷達(dá)研究海洋內(nèi)波[D];中國(guó)海洋大學(xué);2004年

，

本文編號(hào)：2300663