國(guó)家坐標(biāo)系下機(jī)載傳感器的直接對(duì)地目標(biāo)定位
發(fā)布時(shí)間：2019-06-02 06:27
【摘要】：自上世紀(jì)末GPS/INS組合導(dǎo)航技術(shù)被成功地引入航空攝影測(cè)量領(lǐng)域以來,對(duì)地目標(biāo)定位的數(shù)據(jù)處理流程出現(xiàn)了相當(dāng)大的變化。在嚴(yán)格檢校出機(jī)載成像傳感器、GPS和INS之間的幾何關(guān)系之后,機(jī)載POS系統(tǒng)就能夠高精度地輸出成像傳感器的外方位元素并直接用于地形圖測(cè)圖,這種方法通常被稱為直接對(duì)地目標(biāo)定位。與傳統(tǒng)的間接對(duì)地目標(biāo)定位方法相比,直接對(duì)地目標(biāo)定位不再需要布測(cè)地面控制點(diǎn),也不需要進(jìn)行空中三角測(cè)量,因此可以大幅度降低航空攝影測(cè)量的內(nèi)外業(yè)工作量。然而,由于缺少地面控制點(diǎn)和像方連接點(diǎn)的約束,利用直接對(duì)地目標(biāo)定位方法計(jì)算出的地面點(diǎn)坐標(biāo)非常容易受到系統(tǒng)誤差的影響。 航空攝影測(cè)量的各種數(shù)據(jù)成果最終都需要納入到某種形式的國(guó)家坐標(biāo)系,因此內(nèi)業(yè)數(shù)據(jù)處理時(shí)一般直接選取該坐標(biāo)系作為對(duì)地目標(biāo)定位的計(jì)算坐標(biāo)系。然而,由于國(guó)家坐標(biāo)系并不是一種笛卡爾空間直角坐標(biāo)系,而通常是采用等角地圖投影,對(duì)地目標(biāo)定位過程會(huì)不可避免地受到多種幾何變形因素的影響。特別是對(duì)于直接對(duì)地目標(biāo)定位方法,由于具有數(shù)據(jù)外推的性質(zhì),各種幾何變形會(huì)直接傳播到地面點(diǎn)坐標(biāo)中。因此,為了提高國(guó)家坐標(biāo)系下直接對(duì)地目標(biāo)定位結(jié)果的精度,必須對(duì)這些幾何變形進(jìn)行高精度的改正。 本文圍繞國(guó)家坐標(biāo)系下機(jī)載傳感器的直接對(duì)地目標(biāo)定位這一核心問題展開研究,主要內(nèi)容包括： 1)機(jī)載POS數(shù)據(jù)向國(guó)家坐標(biāo)系轉(zhuǎn)換方法的理論分析和精度評(píng)定 為了實(shí)現(xiàn)國(guó)家坐標(biāo)系下的直接對(duì)地目標(biāo)定位,首先需要將機(jī)載POS數(shù)據(jù)處理軟件輸出的軌跡數(shù)據(jù)高精度地轉(zhuǎn)換到國(guó)家坐標(biāo)系下。本文詳細(xì)推導(dǎo)了有代表性的三種姿態(tài)轉(zhuǎn)換方法(旋轉(zhuǎn)矩陣法、補(bǔ)償矩陣法和坐標(biāo)轉(zhuǎn)換法)的計(jì)算公式,并設(shè)計(jì)了高精度的試驗(yàn)方法對(duì)這些姿態(tài)轉(zhuǎn)換算法和兩種商業(yè)軟件(Applanix公司POSPac軟件和Leica公司的IPAS CO軟件)進(jìn)行精度評(píng)定。試驗(yàn)結(jié)果顯示旋轉(zhuǎn)矩陣法的姿態(tài)轉(zhuǎn)換誤差完全可以忽略不計(jì),而且計(jì)算代價(jià)通常小于其他方法,因此是最為推薦的機(jī)載POS姿態(tài)轉(zhuǎn)換方法。而兩種商業(yè)軟件的姿態(tài)轉(zhuǎn)換結(jié)果并不十分理想,最大誤差已經(jīng)超過了機(jī)載POS硬件系統(tǒng)的理論測(cè)量誤差。 2)國(guó)家坐標(biāo)系下直接對(duì)地目標(biāo)定位的幾何變形模型 本文對(duì)國(guó)家坐標(biāo)系下直接對(duì)地目標(biāo)定位過程中出現(xiàn)的幾何變形進(jìn)行了系統(tǒng)的分析,歸納出四類共七種幾何變形因素,具體包括一種尺度變形(基準(zhǔn)尺度變形)、一種高程變形(地球曲率變形)、兩種長(zhǎng)度變形(空間直線與大地線之間的長(zhǎng)度變形和大地線與投影線之間的長(zhǎng)度變形)和三種角度變形(標(biāo)高差變形、截面差變形和方向改化),其中基準(zhǔn)尺度變形、標(biāo)高差變形和截面差變形是現(xiàn)有參考文獻(xiàn)中沒有研究過的。 3)國(guó)家坐標(biāo)系下機(jī)載激光雷達(dá)數(shù)據(jù)直接對(duì)地目標(biāo)定位的幾何變形改正算法 本文推導(dǎo)了國(guó)家坐標(biāo)系下機(jī)載激光雷達(dá)數(shù)據(jù)直接對(duì)地目標(biāo)定位地圖投影變形改正的高精度和實(shí)用計(jì)算公式。試驗(yàn)結(jié)果證明實(shí)用公式的計(jì)算精度遠(yuǎn)高于傳統(tǒng)公式,而只需要多付出25%的計(jì)算代價(jià)；高精度改正公式的計(jì)算殘差則幾乎可以忽略不計(jì)。 4)國(guó)家坐標(biāo)系下航空影像直接對(duì)地目標(biāo)定位的幾何變形改正算法 本文推導(dǎo)了國(guó)家坐標(biāo)系下航空影像直接對(duì)地目標(biāo)定位幾何變形改正幾種傳統(tǒng)算法(傳統(tǒng)地球曲率改正方法、改變航高法和改變主距法兩種長(zhǎng)度變形改正方法)的計(jì)算公式并設(shè)計(jì)了數(shù)種改進(jìn)算法：等效豎直影像法、改變像方坐標(biāo)法、改變物方坐標(biāo)法和前方交會(huì)預(yù)測(cè)法。試驗(yàn)結(jié)果證明等效豎直影像法完全不受影像傾角變化的影響而且不會(huì)增加任何的計(jì)算量,性能優(yōu)于傳統(tǒng)的地球曲率改正方法；改變物方坐標(biāo)法基本不會(huì)受到地形起伏的影響,計(jì)算精度通常優(yōu)于兩種傳統(tǒng)的長(zhǎng)度變形改正方法(改變航高法和改變主距法)；前方交會(huì)預(yù)測(cè)法可以將幾何變形改正殘差限制到可以忽略的程度,能夠滿足高精度的應(yīng)用需求。
[Abstract]:Since the GPS/ INS combined navigation technology has been successfully introduced into the field of aerophotogrammetry since the end of the last century, there has been a considerable change in the data processing flow of the location of the ground target. After rigorous inspection of the geometrical relationship between the airborne imaging sensors, GPS and INS, the on-board POS system can output the outer bearing elements of the imaging sensor with high precision and directly used for topographic mapping, which is often referred to as a direct-to-ground target location. Compared with the traditional indirect target positioning method, the direct-to-ground target positioning is no longer required to test the ground control point, and the aerial triangulation is not required, so that the workload of the internal and external industries of the aerial photography measurement can be greatly reduced. However, due to the lack of the constraint of the ground control point and the image square connection point, the ground point coordinate calculated by the direct-to-ground target positioning method is very easy to be affected by the system error. The various data results of aerial photogrammetry will eventually need to be incorporated into some form of national coordinate system, so that the coordinate system is directly selected as the calculation coordinate of the positioning of the target in the inner industry data processing. However, due to the fact that the national coordinate system is not a Cartesian space rectangular coordinate system, it is usually an equal-angle map projection, and the positioning process of the ground object is inevitably affected by various geometric deformation factors. in particular for a direct-to-ground target location method, various geometric distortion can be spread directly to that ground point coordinate due to the nature of the extrapolation of the data, Therefore, in order to improve the accuracy of the direct-to-ground target positioning results in the national coordinate system, it is necessary to change the geometric deformation with high precision In this paper, the core problem of the direct-to-ground target location of the on-board sensors in the national coordinate system is studied, mainly Capacity includes:1) The theory of the conversion of the on-board POS data to the national coordinate system In order to realize the direct-to-ground target location under the national coordinate system, it is necessary to convert the trace data output by the on-board POS data processing software with high precision. In this paper, the representative three-position conversion methods (rotation matrix method, compensation matrix method and coordinate conversion) are derived in detail in this paper. The calculation formula of the method is given, and the high-precision test method is designed for these attitude conversion algorithms and the IPAS CO software of the two kinds of commercial software. The results show that the attitude transition error of the rotation matrix method is completely negligible, and the calculation cost is usually less than that of other methods, so it is the most recommended on-board PO The attitude transition of the two commercial software is not ideal, and the maximum error has exceeded the on-board POS hardware system The theoretical measurement error of the theory. In this paper, the geometric deformation model of the target location is analyzed and summarized in the paper on the geometric deformation of the direct-to-ground target positioning in the national coordinate system. The four types of seven geometric deformation factors, including one kind of scale deformation (reference scale deformation), one kind of elevation deformation (earth curvature deformation), two length deformation (the length deformation between the space straight line and the earth line and the length deformation between the earth line and the projection line) and the three The deformation of the reference scale, the deformation of the elevation difference and the deformation of the cross-section are the prior art. in that absence of a study in the reference (3) the airborne lidar data on the national coordinate system is directly In this paper, the direct-to-ground target location map projection of the airborne lidar data under the national coordinate system is derived in this paper. The experimental results show that the calculation accuracy of the practical formula is much higher than that of the traditional formula, but only 25% of the calculation cost is required, and the high-precision correction of the utility model the calculation residual of the formula is almost negligible.4) The aerial image in the national coordinate system In this paper, the geometric deformation correction algorithm based on the direct-to-ground target location is derived, and the geometric deformation of the target positioning of the aerial image under the national coordinate system is derived. Several improvement algorithms have been designed to correct several traditional algorithms (the traditional method for correcting the curvature of the earth, the method of changing the high-altitude method and the method of changing the two-length deformation of the main-distance method), and several improved algorithms are designed: the equivalent vertical image method, the change of the image The results of the test show that the equivalent vertical image method is not affected by the change of the image inclination, and no calculation amount is added, and the performance is better than that of the traditional method for correcting the curvature of the earth. The variable-object coordinate method is not affected by the relief of the terrain, and the calculation accuracy is generally superior to the two traditional methods of length deformation correction (change of the high-altitude method and the change of the main-distance method), and the forward intersection prediction method can limit the geometric deformation correction residual to the available method.
【學(xué)位授予單位】：武漢大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2014
【分類號(hào)】：P231
本文編號(hào)：2490895
[Abstract]:Since the GPS/ INS combined navigation technology has been successfully introduced into the field of aerophotogrammetry since the end of the last century, there has been a considerable change in the data processing flow of the location of the ground target. After rigorous inspection of the geometrical relationship between the airborne imaging sensors, GPS and INS, the on-board POS system can output the outer bearing elements of the imaging sensor with high precision and directly used for topographic mapping, which is often referred to as a direct-to-ground target location. Compared with the traditional indirect target positioning method, the direct-to-ground target positioning is no longer required to test the ground control point, and the aerial triangulation is not required, so that the workload of the internal and external industries of the aerial photography measurement can be greatly reduced. However, due to the lack of the constraint of the ground control point and the image square connection point, the ground point coordinate calculated by the direct-to-ground target positioning method is very easy to be affected by the system error. The various data results of aerial photogrammetry will eventually need to be incorporated into some form of national coordinate system, so that the coordinate system is directly selected as the calculation coordinate of the positioning of the target in the inner industry data processing. However, due to the fact that the national coordinate system is not a Cartesian space rectangular coordinate system, it is usually an equal-angle map projection, and the positioning process of the ground object is inevitably affected by various geometric deformation factors. in particular for a direct-to-ground target location method, various geometric distortion can be spread directly to that ground point coordinate due to the nature of the extrapolation of the data, Therefore, in order to improve the accuracy of the direct-to-ground target positioning results in the national coordinate system, it is necessary to change the geometric deformation with high precision In this paper, the core problem of the direct-to-ground target location of the on-board sensors in the national coordinate system is studied, mainly Capacity includes:1) The theory of the conversion of the on-board POS data to the national coordinate system In order to realize the direct-to-ground target location under the national coordinate system, it is necessary to convert the trace data output by the on-board POS data processing software with high precision. In this paper, the representative three-position conversion methods (rotation matrix method, compensation matrix method and coordinate conversion) are derived in detail in this paper. The calculation formula of the method is given, and the high-precision test method is designed for these attitude conversion algorithms and the IPAS CO software of the two kinds of commercial software. The results show that the attitude transition error of the rotation matrix method is completely negligible, and the calculation cost is usually less than that of other methods, so it is the most recommended on-board PO The attitude transition of the two commercial software is not ideal, and the maximum error has exceeded the on-board POS hardware system The theoretical measurement error of the theory. In this paper, the geometric deformation model of the target location is analyzed and summarized in the paper on the geometric deformation of the direct-to-ground target positioning in the national coordinate system. The four types of seven geometric deformation factors, including one kind of scale deformation (reference scale deformation), one kind of elevation deformation (earth curvature deformation), two length deformation (the length deformation between the space straight line and the earth line and the length deformation between the earth line and the projection line) and the three The deformation of the reference scale, the deformation of the elevation difference and the deformation of the cross-section are the prior art. in that absence of a study in the reference (3) the airborne lidar data on the national coordinate system is directly In this paper, the direct-to-ground target location map projection of the airborne lidar data under the national coordinate system is derived in this paper. The experimental results show that the calculation accuracy of the practical formula is much higher than that of the traditional formula, but only 25% of the calculation cost is required, and the high-precision correction of the utility model the calculation residual of the formula is almost negligible.4) The aerial image in the national coordinate system In this paper, the geometric deformation correction algorithm based on the direct-to-ground target location is derived, and the geometric deformation of the target positioning of the aerial image under the national coordinate system is derived. Several improvement algorithms have been designed to correct several traditional algorithms (the traditional method for correcting the curvature of the earth, the method of changing the high-altitude method and the method of changing the two-length deformation of the main-distance method), and several improved algorithms are designed: the equivalent vertical image method, the change of the image The results of the test show that the equivalent vertical image method is not affected by the change of the image inclination, and no calculation amount is added, and the performance is better than that of the traditional method for correcting the curvature of the earth. The variable-object coordinate method is not affected by the relief of the terrain, and the calculation accuracy is generally superior to the two traditional methods of length deformation correction (change of the high-altitude method and the change of the main-distance method), and the forward intersection prediction method can limit the geometric deformation correction residual to the available method.
【學(xué)位授予單位】：武漢大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2014
【分類號(hào)】：P231
【相似文獻(xiàn)】
相關(guān)期刊論文 前10條
1 ;國(guó)家測(cè)繪局制定啟用2000國(guó)家坐標(biāo)系實(shí)施方案[J];城市勘測(cè);2008年05期
2 雷偉偉;姜斌;;國(guó)家坐標(biāo)系與城市坐標(biāo)系轉(zhuǎn)換方法的探討[J];測(cè)繪科學(xué);2010年01期
3 蔣碩;;對(duì)假設(shè)坐標(biāo)系與國(guó)家坐標(biāo)系轉(zhuǎn)換問題的淺析[J];科技與企業(yè);2014年09期
4 雷勇;;國(guó)家坐標(biāo)系與施工坐標(biāo)系的轉(zhuǎn)換方法探討[J];畢節(jié)學(xué)院學(xué)報(bào);2009年04期
5 賈志強(qiáng);;1980國(guó)家坐標(biāo)系向某市地方獨(dú)立坐標(biāo)系的轉(zhuǎn)換[J];測(cè)繪與空間地理信息;2010年01期
6 趙祖軍;張瑋;;建筑坐標(biāo)系和國(guó)家坐標(biāo)系特殊關(guān)系的解析[J];電力勘測(cè)設(shè)計(jì);2010年02期
7 姜德才;張文;羅京輝;陳繼偉;高峰;段彩梅;;不同國(guó)家坐標(biāo)系下子午線收斂角計(jì)算公式及其精度分析[J];測(cè)繪與空間地理信息;2013年10期
8 何佳;韓燕;張獻(xiàn)州;羅文彬;;國(guó)家坐標(biāo)系與基于工程投影面的地方坐標(biāo)系坐標(biāo)變換通用計(jì)算模型研究[J];城市勘測(cè);2007年03期
9 饒德慶;李金平;劉長(zhǎng)星;;礦區(qū)獨(dú)立坐標(biāo)系的建立及同國(guó)家坐標(biāo)系的轉(zhuǎn)換[J];科技信息;2010年23期
10 趙雙明;左正立;陸宇紅;徐軍;;集成POS的機(jī)載三線陣傳感器幾何定位[J];測(cè)繪與空間地理信息;2008年04期
相關(guān)重要報(bào)紙文章 前1條
1 立實(shí);國(guó)家測(cè)繪局制定啟用2000國(guó)家坐標(biāo)系實(shí)施方案[N];中國(guó)測(cè)繪報(bào);2008年
相關(guān)博士學(xué)位論文 前1條
1 沈翔;國(guó)家坐標(biāo)系下機(jī)載傳感器的直接對(duì)地目標(biāo)定位[D];武漢大學(xué);2014年
，本文編號(hào)：2490895
本文鏈接：http://sikaile.net/kejilunwen/dizhicehuilunwen/2490895.html
最近更新
教材專著
