CN102735216B - CCD stereoscopic camera three-line imagery data adjustment processing method - Google Patents

Abstract

The invention discloses a three-line image data adjustment processing method of a CCD stereo camera. The method comprises the following steps: step 1, obtaining lunar surface image data and corresponding ephemeris data and attitude data; step 2, using image automatic matching technology to extract The image data of the same track with different viewing angles and the image points of the same name of the adjacent track image data; step 3, use the above data and the image points of the same name, and use the independent model method block adjustment to perform the survey area adjustment process; step 4, in the survey area The full moon adjustment is performed on the basis of the adjustment. The method of the invention does not need to calculate the coordinates of the outer orientation elements and encrypted points in the adjustment processing, and obviously simplifies the adjustment processing process on the premise of ensuring the accuracy of data processing.

Description

Adjustment processing method for three-line array image data of CCD stereo camera

technical field

The invention relates to image processing methods and application fields, in particular to an adjustment processing method suitable for three-line array image data of a lunar CCD stereo camera.

Background technique

The CCD stereo camera is an area array camera with only one set of optical lenses and a detector size of 1024×1024. On the area array, the 11th, 512th and 1013th lines perpendicular to the flight direction are read, respectively. As an image array with three different viewing angles of front view, front view and rear view, the number of pixels in each line array is 512 columns, and the viewing angle difference between adjacent line arrays of front view, front view and rear view is 16.7°. The segment is the visible light band of 0.5 μm to 0.75 μm. According to the on-orbit flight parameters predetermined by the satellite, the scanning speed of the camera is set to 11.89 frames per second to ensure that the three linear arrays of the satellite are acquired by push-broom in the three directions of forward, downward and backward at the same time on the orbit at a height of 200km. The three two-dimensional image strips can all image 100% of the lunar surface that the satellite flies over. The three image strips are acquired almost at the same time, with a width of 60km and 100% overlap in the direction of satellite flight. The overlap of adjacent orbit image strips near the equator is about 41%, and the overlap is greater in high latitudes. It provides enough image information for the three-dimensional terrain geometry reconstruction of the lunar surface. Figure 2 shows the process of CCD stereo camera image data acquisition.

The main purpose of the adjustment processing of the three-line array image data of the CCD stereo camera is to realize the seamless mosaic and absolute orientation of the image data acquired by the satellite at different times and different positions within the whole month. According to the principle of satellite photogrammetry, the adjustment processing The first task is to calculate the outer orientation elements of each line array of the three line array image data (that is, to obtain the spatial position and attitude information of the satellite at the moment of the line array image). Since the CCD stereo camera has only three images of front view, front view and rear view in the same scanning cycle, affected by changes in the outer orientation elements, terrain fluctuations, etc., it is used to calculate the same-named image points of the outer orientation elements of the line array (at least 6), it is impossible to fall on these three images, so it is impossible to geometrically solve the outer azimuth elements of this scanning period, and some simplified algorithms need to be used for approximate processing. For satellite photography, the platform is relatively stable, and the outer orientation elements do not change much. Hofmann, Wang Renxiang and others proposed to use the time with a large distance between them, that is, the orientation time or EFP time, to discretize the route model, and approximate the expression of the route model and the outer position. Orientation element.

In the 1980s, German scholar Hofmann and others first created the beam adjustment method for processing three-line array CCD images in the MOMS project, that is, the "directional image method". This method solves the problem that the outer orientation elements of each sampling cycle (moment) of the CCD image cannot be solved only by the three-line array image itself, and only solves the outer orientation elements of the "directional image" moment, while the outer orientation element of any CCD image sampling moment Elements are interpolated from their adjacent oriented image values. Therefore, the adjustment calculation only needs to solve the exterior orientation elements of these oriented images, thereby reducing the number of exterior orientation elements and the amount of calculation.

The equivalent photo method is a processing method proposed by Wang Renxiang et al. This algorithm transforms a set of linear array images at a certain time interval to generate an equivalent image of the plane center projection, and then uses the plane center projection algorithm to process . This approach also reduces the number of exterior orientation elements and computation.

Although the processing methods such as the oriented slice method and the equivalent photographic method obviously reduce the number of external orientation elements in the adjustment process, the coordinates of the encrypted points (picture points with the same name) used to solve the external orientation elements are also used in the adjustment process. The number of unknowns to be solved is huge, which increases the workload and difficulty of adjustment processing.

Contents of the invention

The purpose of the present invention is to, by referring to the methods of approximate processing of the entire route image data such as the directional film method and the EFP (equivalent photo) method, and adopting the theory and method of the independent model method area network aerial triangulation, a kind of method is proposed. The method of adjusting the lunar three-line array image data of the CCD stereo camera overcomes the problem that the azimuth elements outside each scanning line of the three-line array image of the CCD stereo camera cannot be solved geometrically.

The CCD stereo camera three-line array image data adjustment processing method of the present invention comprises the steps: Step 1, obtaining lunar surface image data and corresponding ephemeris data and attitude data; Step 2, utilizing image automatic matching technology to extract images of different viewing angles on the same track Data and image points with the same name of adjacent track image data; step 3, use the above data and image points with the same name, and use the independent model method block adjustment to perform survey area adjustment; step 4, perform survey area adjustment on the basis of survey area adjustment Full lunar adjustment processing.

Preferably, the entire lunar surface is divided into multiple mapping areas, and each mapping area is further divided into multiple survey areas, and each survey area has several aerial images.

Preferably, the different viewing angles include front view, front view and rear view.

Preferably, the matching algorithm adopts a combination of scale invariant feature transformation SIFT feature matching and least square matching.

Preferably, the SIFT feature matching algorithm is used to provide the initial position of the feature, and then the least squares matching is used to achieve precise image matching.

Preferably, the independent model is constructed in the following manner: within the scope of the measurement area, the sequence of directional slices on the front view image is obtained, and the directional slices with the same name on the front view and the front view, the front view and the back view, and the front view and the back view form an directional slice pair , each pair of oriented slices undergoes relative orientation processing to construct an independent model, which is the smallest unit of the block adjustment of the independent model method.

The method of the present invention does not need to solve the coordinates of the outer azimuth elements and encrypted points in the adjustment process, but solves another set of parameters (relative orientation parameters and absolute orientation parameters), which are used for the three-line array image data in the whole month range The seamless mosaic and absolute orientation in the interior, the number of unknowns to be solved is obviously less than the above two methods in the prior art, which obviously simplifies the adjustment process, but the data processing accuracy does not decrease.

Description of drawings

Fig. 1 is the processing flowchart of the area adjustment algorithm and the global adjustment algorithm;

Fig. 2 is a schematic diagram of the acquisition process of the three-line array image data of the lunar surface of the CCD stereo camera;

Fig. 3 is a schematic diagram of the division of directional slices on the front view image of the CCD stereo camera;

Figures 4a and 4b are schematic diagrams of the comparison of the results before and after the linear error correction of the independent model;

Figures 5a and 5b are schematic diagrams of the spatial distribution of the CCD stereo camera adjustment processing measuring area division;

Figures 6a and 6b are schematic diagrams of the spatial distribution of lunar surface control points for survey area adjustment and global adjustment;

Fig. 6c is a partially enlarged view of Fig. 6b.

detailed description

In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be described in further detail below in conjunction with specific embodiments and with reference to the accompanying drawings.

Referring to the flow chart of Fig. 1, the adjustment processing method of the three-line array image data of the lunar CCD stereo camera of the present invention is described in detail. The precondition that the adjustment processing method of the present invention relies on is that the satellite photography platform is relatively stable, and the external azimuth elements do not change much.

Referring to Fig. 1, when implementing the adjustment processing method, data preparation and organization are performed first.

Taking CE-1 (Chang'e No. 1) as an example, the satellite obtained 1,000 multi-track image data during its lifetime, and selected 628 of them (this data is only exemplary, and any suitable number can be selected according to the actual situation) Image data with better image quality, such as uniform track tone and small difference in lighting conditions, covering the entire lunar surface. In addition, satellite ephemeris data and attitude data corresponding to the image data must be prepared for adjustment processing. The satellite ephemeris data is the satellite position data obtained by the measurement and control department from tracking and measuring the CE-1 satellite based on ground observations. The attitude is determined by the satellite The satellite platform roll, pitch, yaw and other data measured by the equipment carried on board.

In order to ensure the consistency of the adjustment results on a global scale, it is necessary to realize the seamless mosaic and absolute orientation of the above-mentioned selected image data on a global scale, and carry out the adjustment processing on a full-month scale. In order to carry out the global adjustment processing, the whole world is divided into several survey areas, the survey area adjustment process is performed first, and then the global adjustment process is performed.

Divide the entire lunar surface into several mapping areas, for example, three mapping areas of S70°-N70°, S70°-S90° and N70°-N90°, and each mapping area is further divided into several surveying areas. 202 survey areas, about 20 aerial images for each survey area. In order to ensure the connection conditions between survey areas, adjacent survey areas overlap about 4° along the longitude direction and about 2° along the latitude direction when clipping the data.

See Figures 4a and 4b for the division results of the survey area. Figure 4a is a schematic diagram of the survey area division of the S70°-N70° middle and low latitude mapping area. There are 170 survey areas in total, with a latitude span of 14° and a total of 10 latitude zones. The longitude span varies with different latitudes; Figure 4b is S70°-S90° and N70°-N90° polar mapping area survey area division diagram, a total of 32 survey areas. The number information of each survey area is given in the figure.

In this step, the acquired lunar surface image data, as well as the corresponding ephemeris data and attitude data, etc. are cut according to the survey area and managed in folders named according to the survey area number, as the input for subsequent image matching and adjustment processing data.

Afterwards, using the image data of the lunar surface acquired in the above steps, the automatic matching and matching processing steps of the three-line array images are performed.

Image matching is mainly to realize the extraction of the image points with the same name in the overlapping area between the front-sight, front-sight and back-sight images of the same track in each survey area, as well as between adjacent track image data.

The matching algorithm adopts the combination of SIFT (ScaleInvariantFeatureTransform) feature matching and least squares matching. Feature matching based on SIFT descriptors has been successfully applied to many fields because of its invariance to image scale, brightness, contrast, rotation, translation, and small affine, and it is currently the best feature descriptor. SIFT feature matching is not only invariant to the scale change and rotation of the image, but also has strong adaptability to the change of illumination and image deformation, and the DOG difference operator is mainly used in the calculation process, and most of the found features are "blobs". "(circular point), which is just suitable for the feature of lunar image with poor texture and many tiny impact craters. The spatial resolution of the CCD image is 120m, and the impact craters with a diameter of less than 500m appear as bright round spots on the image due to the strong reflection at the edge. Image fine matching is mainly realized by least squares matching. The most difficult problem of the least squares matching algorithm is to determine the initial position. Here we use the SIFT feature matching algorithm to provide the initial position of the feature, and then use the least squares matching. Realize image fine matching. The image matching accuracy is better than 0.3 pixel, and the spatial resolution of the CCD stereo camera image data is 120m.

The automatic matching of the three-line array image includes the automatic matching between the image data of the same track (including the front-view, front-view and rear-view images) and the adjacent track image data. The main purpose of the automatic matching is to extract the adjustment parameters ( That is, the image coordinates of the encrypted point (or the image point with the same name) calculated by the relative orientation parameter and the absolute orientation parameter).

According to the matching method of image points with the same name above, select the number and distribution of the images that meet the requirements of relative orientation, survey area adjustment, and global adjustment on the front-sight, front-sight and back-sight images of the same flight zone, adjacent flight zone, and adjacent survey area. The image points with the same name are the relative orientation tie points, absolute orientation tie points and global adjustment tie points shown in Figure 1. Then, manually inspect and repair these image points with the same name in a stereoscopic environment. These image points with the same name are the input data for subsequent area adjustment and global adjustment.

After obtaining the above-mentioned image points with the same name, the area adjustment processing is carried out.

Referring to Figure 1, the steps in the dotted box are the process of area adjustment. In order to realize the seamless mosaic and absolute orientation of the three-line array image data within the whole month, the whole world is divided into several survey areas, and the survey area adjustment processing is performed first, and then the global adjustment processing is performed. Both the area adjustment and the global adjustment adopt the block network adjustment technology using the independent model method.

According to the principle of directional slice proposed by Hofmann, Wang Renxiang, etc., the directional time is selected every 50 lines on the orthoscopic image strip within the survey area, and 160 lines are taken as the width of the directional slice with the oriented time as the center, and the directional slice sequence on the orthoscopic image is obtained. , the orientation slice boundaries on the front-view and back-view images are determined by image matching. Orientation slices with the same name on front view and front view, front view and back view, front view and back view form an orientation slice pair, and each orientation slice pair undergoes relative orientation processing to build an independent model.

Fig. 3 is a schematic diagram of the division of directional slices on the front view image of the CCD stereo camera. The horizontal direction in the figure represents the height of the front view image strip of the CCD stereo camera, the height is 512 pixels, and the length is consistent with the actual number of image scanning lines. The black solid line is the center line of the independent model (marked with a dotted line on the left in the figure), and the line array data acquisition time corresponding to the center line is the orientation time of the model, and the distance between adjacent center lines is 50 lines. The height of the independent model is consistent with the height of the image strip, that is, 512 pixels, centered on the center line of the independent model, and the width of the model is 160 lines. symbol Represents the position of the image point with the same name on the orientation sheet, which is generated by image matching. The independent model is the smallest unit of the block adjustment of the independent model method. The purpose of the survey area adjustment is to seamlessly connect the independent models constructed by the directional patch pairs within the survey area to form a ground three-dimensional model of the entire survey area.

The survey area adjustment process is completed in the following steps:

(1) Orientation on-chip orientation:

Convert the image coordinates of the image point with the same name into the image plane coordinates with the principal point of the oriented image as the coordinate origin. The transformation formula is as follows:

$\left.\begin{array}{c}{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; ds</span>}\\ {\mathrm{<span\; class="notranslate">\; the\; y</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; ds</span>}\\ {\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\end{array}\right\}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Wherein, x ', y ' are image plane coordinate values (unit is mm), and d _x is the image coordinate value (unit is pixel) of x direction in the orientation sheet; x, y are original image row, column number; ds=14 μ m Be the detector pixel size; x ₀ =0.8050, y ₀ =-0.7990 is the image plane coordinate value (unit is pixel) of the principal point; L ₀ is the position of the scanning line in the CCD array (front-sight L ₀ = 11; front view L ₀ =512, back view L ₀ =1013); L _i is the line number of the center scanning line of the directional slice in the image strip.

Through the orientation processing in the photo, the image coordinates of the original image point with the same name are converted into the image plane coordinates with the principal point of the image as the coordinate origin, and the coordinates will be used as the input data for the relative orientation processing of the model.

(2) Relative orientation processing: using the coordinates of the same-named image point of the orientation slice pair, the error equation is established through the coplanar condition, and the relative orientation elements of the orientation slice are solved. The equations that satisfy the coplanar condition are as follows:

in, kappa, ω', κ' are the relative orientation elements to be found; B is the baseline length, the front-sight and front-sight or front-sight and rear-sight model B=60km, the front-sight and rear-sight model B=120km; u, v, w are the left image The left image space coordinates of the point, u′, v′, w′ are the right image space coordinates of the right image point, which are obtained from the image plane coordinates of the image point with the same name on the left and right directional slices through their rotation matrix coordinate conversion, and the rotation matrix is obtained by the directional slice’s Outer azimuth element construction.

The above formula (Formula (2)) is expanded by Taylor series and the first-order term is taken to obtain the error equation:

Among them, v _r is the residual error of the coplanar condition equation at the image point with the same name;

The initial value computed for introducing relative orientation element approximations; Δω′, Δκ, Δκ′ are the correction numbers of the approximate values of the relative orientation elements, that is, the correction numbers of the unknowns to be solved; (x, y), (x′, y′) are the image coordinates of the left and right image points, according to the formula ( 1) Calculated; f is the focal length of the camera.

Considering that the directional slice is an approximate process, the error equation only uses the first-degree term of Taylor expansion. In order to improve the calculation accuracy, about 150 uniformly distributed orientation points are selected on each independent model. The orientation points are image points with the same name obtained through image matching, and the image coordinates of these image points are taken as observation numbers. According to equation (3) Construct the error equation, construct the normal equation according to the least square theory, and solve the incremental value of the relative orientation parameter Δκ, Δω′, Δκ′, the initial value of the relative orientation element plus the corresponding incremental value as the value of the relative orientation element. This process requires iterative operations until the precision requirements are met.

Through the relative orientation processing, the relative orientation elements of the independent model are obtained, which are used to calculate the model coordinates of the orientation point.

(3) Model coordinate calculation.

After the relative orientation, the rotation matrices R, R′ and baseline length B of the left and right image pairs used to construct the independent model are known. At this time, the same-named beam pairs corresponding to the same-named image points on the left and right images intersect, and the position of the intersection point is The image point with the same name corresponds to the model point on the surface of the independent model. The coordinate value of the model point in the model coordinate system can be calculated according to the relative orientation element and the set baseline length. The calculation formula is as follows:

U=U _s +Mu _t , V=V _s +Mv _t , W=W _s -Mf(4)

$\left.\begin{array}{c}{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}\frac{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}\mathrm{<span\; class="notranslate">\; f</span>}}{{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}\mathrm{<span\; class="notranslate">\; f</span>}}\\ {\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}\frac{{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}\mathrm{<span\; class="notranslate">\; f</span>}}{{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}\mathrm{<span\; class="notranslate">\; f</span>}}\\ {{\mathrm{<span\; class="notranslate">\; u</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}\frac{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\mathrm{<span\; class="notranslate">\; f</span>}}{{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\mathrm{<span\; class="notranslate">\; f</span>}}\end{array}\right\}$

$\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; B</span>}}{{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}$

Among them, U _s , V _s , W _g are the coordinate values of the projection center of the left photo in the model coordinates; U, V, W are the coordinate values of the model points in the model coordinates; a _i , b _i , c _i and a ′ _i , b′ _i , c′ _i (i=1, 2, 3) are the 9 matrix elements of the left and right image rotation matrices R and R′ respectively; (x, y), (x′, y′) are The image coordinates of the left and right image points; f is the focal length of the camera; u _t , v _t , u′ _t and M are intermediate variables constructed to simplify the expression of the formula.

Through model coordinate calculation, we get the model coordinate value of each model point, which will be used as input data for absolute orientation processing.

(4) Model coordinate correction.

The experimental results show that the independent model has obvious systematic errors in the X direction, and this error changes linearly in the X direction. To solve this problem, linear error correction in the X direction is added to the adjustment algorithm. The results before and after error correction are shown in Figures 4a and 4b. Surface A and surface B in the figure are constructed using the three-dimensional coordinates of uniformly distributed model points on an independent model in the image space coordinate system. Figure 4a is the adjustment result of the model without linear error correction, and there is an obvious linear change in the X direction; Figure 4b is the adjustment result after linear error correction, and the systematic error is obviously eliminated. The coordinate correction equation is as follows:

U'=U+aU(5)

Among them, U is the X coordinate of the model. a is the coordinate correction coefficient.

The absolute orientation process will use the corrected model coordinates as input.

(5) Introduce laser altimeter data as elevation control information.

During the process of block adjustment, the elevation information of the control points is obtained from the DEM data with a spatial resolution of 3 kilometers, which is made by using the detection data of the laser altimeter of the Chang'e-1 probe. The height control point error equation is as follows:

${\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; Z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; Z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; dem</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; Z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; dem</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; Y</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 4</span>}}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; Y</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 5</span>}}{\mathrm{<span\; class="notranslate">\; Y</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}$

In the formula, νi is the Z _i coordinate to be solved for the image point with the same name and the Z _i coordinate obtained from the DEM data difference in coordinates. (X _i , Y _i , Z _i ) are the lunar surface coordinates of the image point with the same name obtained by forward intersection, and obtained from the DEM data according to the lunar surface latitude and longitude coordinates corresponding to (X _i , Y _i , _Zi ) value. a ₀ , a ₁ , a ₂ , a ₃ , a ₄ , a ₅ are quadric surface coefficients, the surface is a 3×3 grid near the position of (X _i , Y _i , _Zi ) on the DEM built within the DEM data. i is the number of the elevation control point.

In the process of absolute orientation processing, the error equation will be used as a constraint equation to participate in the adjustment, eliminating the elevation system deviation value between the absolute orientation result and the reference DEM data, and improving the accuracy of the absolute orientation.

(6) Absolute Orientation

Through the above relative orientation process, we have calculated the relative position and attitude between the left and right photos (in the model coordinate system), but it has not been converted to the absolute position and attitude of the photo, and has nothing to do with whether there are ground control points. In order to find the lunar surface coordinates of these model points, we need to scale, translate and rotate the independent model as a whole, and consider the connection between adjacent models and the lunar surface control points. The absolute orientation of the model.

After the model point undergoes absolute orientation transformation, the coordinates of the lunar surface can be obtained. The transformation adopts the spatial similarity transformation (seven parameter transformation) formula as follows:

$\left[\begin{array}{c}\mathrm{<span\; class="notranslate">\; x</span>}\\ \mathrm{<span\; class="notranslate">\; Y</span>}\\ \mathrm{<span\; class="notranslate">\; Z</span>}\end{array}\right]<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&lambda;R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&Phi;</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; \&Omega;</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; )</span>\left[\begin{array}{c}\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; u</span>}\\ \mathrm{<span\; class="notranslate">\; V</span>}\\ \mathrm{<span\; class="notranslate">\; W</span>}\end{array}\right]<span\; class="notranslate">\; +</span>\left[\begin{array}{c}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}\\ {\mathrm{<span\; class="notranslate">\; Y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}\\ {\mathrm{<span\; class="notranslate">\; Z</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}\end{array}\right]<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 7</span>}<span\; class="notranslate">\; )</span>$

Among them, a is the linear correction coefficient; (X, Y, Z) is the lunar surface photogrammetric coordinates of the model point; (U, V, W) is the model coordinate of the model point; λ is the model scale parameter to be obtained, and R(Φ , Ω, K) is a rotation matrix constructed by rotation angle parameters, Φ, Ω, K are three rotation angle parameters; X _t , Y _t , Z _t are three translation parameters. The lunar surface photogrammetric coordinate system is a tangent plane coordinate system with the center of the survey area as the origin, and the conversion matrix between the tangent plane coordinates and the lunar surface photogrammetric coordinates is constructed according to the lunar fixed coordinates of the origin.

Expand the above formula with Taylor series and take the primary term to get the error equation:

${\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; \&Phi;</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;\&Phi;</span>}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; \&Omega;</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;\&Omega;</span>}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; K</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;K</span>}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}}{\mathrm{<span\; class="notranslate">\; \&Delta;X</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;\&lambda;</span>}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; a</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;a</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; Y</span>}}<span\; class="notranslate">\; =</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; \&Phi;</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;\&Phi;</span>}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; \&Omega;</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;\&Omega;</span>}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; K</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;K</span>}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; Y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}}{\mathrm{<span\; class="notranslate">\; \&Delta;Y</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;\&lambda;</span>}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; a</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;a</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 8</span>}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; Z</span>}}<span\; class="notranslate">\; =</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; \&Phi;</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;\&Phi;</span>}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; \&Omega;</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;\&Omega;</span>}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; K</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;K</span>}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; Z</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}}{\mathrm{<span\; class="notranslate">\; \&Delta;Z</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;\&lambda;</span>}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&PartialD;</span>\mathrm{<span\; class="notranslate">\; a</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;a</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Z</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; Z</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; )</span>$

In the formula, v _X , v _Y , v _Z are the residuals of lunar surface coordinates, [X], [Y], [Z] are the calculation results obtained by substituting approximate values of unknowns into formula (7); Δλ, ΔΦ, ΔΩ , ΔK, ΔX _t , ΔY _t , ΔZ _t , Δa are the correction numbers of the undetermined values, that is, the correction numbers of the unknowns to be solved; is the coefficient of the error equation.

After image matching is performed on the three-line array data, and a certain number of evenly distributed connection points of adjacent directional slices are determined (that is, the same-named image points in the overlapping area of directional slices), the model coordinates of the connection points can be calculated according to formula (4), as The connection conditions of the adjacent independent models are then substituted into the error equation (8), joint equations (6) and (8), and the absolute orientation parameters X _t , Y _t , Z _t , Φ, Ω can be solved by using the method of least squares , K, λ and model coordinate correction coefficient a. The selection of connection points should ensure the connection between the models in the flight zone (the image strip along the satellite flight direction) and the connection between the models in the adjacent flight belts. In this work, 5 rows (vertical to the flight direction) and 9 columns (along the flight direction) connection points that are evenly distributed are selected as model connection points for each model in the air strip, and the overlapping area of adjacent models has 6 Column connection points; in the overlapping area of the model of adjacent flight strips, evenly distributed 3-column and 4-row connection points are selected as the connection points of the flight strips.

Absolute orientation parameters X _t , Y _t , Z _t , Φ, Ω, K, λ and model coordinate correction coefficient a are the input parameters for calculating the lunar surface coordinates of each model point.

(7) Moon point coordinate calculation

Using absolute orientation elements (X _t , Y _t , Z _t , Φ, Ω, K, λ) and model coordinate correction coefficient a, substituting equation (7) to solve the tangent coordinate value (lunar surface photogrammetry coordinates). After the above absolute orientation and global adjustment, what we get is the coordinate value in the photogrammetric coordinate system, that is, the coordinate value in the tangential plane coordinate system with the center of the survey area as the origin. According to the external orientation elements of the independent model composed of the front-sight and rear-sight oriented slices at this position, we can establish the conversion matrix between the tangent plane coordinates and the moon-fixed coordinates, realize the transformation from the tangent plane coordinates to the moon-fixed coordinates, and obtain all The lunar surface coordinate value of the model point.

So far, the relative orientation parameters and absolute orientation parameters of each independent model in each independent survey area are obtained. These parameters can ensure the seamless connection of all independent models in the survey area, thereby realizing the construction of the ground three-dimensional model of the entire survey area. The surface model of the survey area will be used as an independent model for global adjustment processing, and it is the smallest unit of global adjustment processing.

According to the above steps, the area adjustment processing is completed, and then the global adjustment processing is performed.

The survey area adjustment realizes the seamless connection of each independent model in the survey area. In order to realize the seamless mosaic of global data, global adjustment must also be carried out. The entire lunar surface is divided into several blocks (or survey areas, see Figure 5b) according to certain principles, and on the basis of the survey area adjustment, the global adjustment is performed with the survey area as a unit. The global adjustment also adopts the above-mentioned independent model block adjustment method. The only difference is that the independent model of the survey area adjustment is a ground model constructed by dividing directional slice pairs, while the independent model of the global adjustment is the entire survey area corresponding ground model. The input data of the survey area adjustment is the connection points between adjacent independent models (constructed by oriented slices), and these connection points ensure the seamless mosaic of data in the entire survey area; the input data of the global adjustment is the adjacent survey area These connection points ensure a seamless mosaic of all survey areas on a global scale.

Both survey area adjustment and global adjustment must be carried out with the support of control points, and the elevation control points are selected from the DEM with a spatial resolution of 3 kilometers, see the above formula (6). The coordinates of the plane control points are calculated from a certain number and distribution of image points with the same name according to the forward intersection solution. See Figure 6a and 6b for the distribution of the control points of the survey area adjustment and the global adjustment. Control points, 9 are evenly selected in the entire survey area, distributed on the first model, middle model and last model of the first track, middle track and last track; dots represent elevation control points, distributed in the survey area The center position of each ground model; the black solid line represents the boundary line of the survey area, and the number information of the adjacent survey area is given in the figure. Figure 6b shows the distribution map of the lunar surface control points of the global adjustment. The triangles represent the height control points, and 9 are evenly arranged around the world; The black dots represent the connection points between adjacent survey intervals, about 23,766 in total, evenly distributed in the overlapping areas of the survey areas. Figure 6c is a partial enlarged view of the oblique area in Figure 6b, and the meanings of the symbols in the figure are the same as those in Figure 6b. The weight of the height control point is set to 20.0, and the height control point is set to 3.0. The elevation value is obtained from the global DEM data with a spatial resolution of 3km produced by the "Chang'e-1 CE-1" altimeter.

Using the internal orientation parameters of the CCD stereo camera, the survey area adjustment and the global adjustment process, the relative orientation parameters, absolute orientation parameters, and survey area absolute orientation parameters of the independent model are obtained, according to formula (1), formula (4), formula (5 ), Formula (7), calculate the lunar surface coordinates of any image point with the same name on the three-line array image. In addition, in the survey area adjustment process, a sufficient number and distribution of connection points are selected in the overlapping areas of adjacent independent models to ensure the seamless mosaic between independent models in the survey area. A sufficient number and distribution of connection points are also selected for the overlapping area, ensuring seamless mosaic between adjacent survey areas on a global scale. Therefore, the problem of global connection of data is well solved through survey area adjustment and global adjustment, and processing parameters are provided for the subsequent production of digital elevation model and orthophoto.

The specific embodiments described above have further described the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above descriptions are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims (3)

1. A three-line array image data adjustment processing method of a CCD stereo camera, the method comprising steps: Step 1. Obtain the image data of the lunar surface and the corresponding ephemeris data and attitude data. In this step, the entire lunar surface is divided into multiple mapping areas, and each mapping area is further divided into multiple survey areas. Each survey area has Several air strip images; Step 2, use the automatic image matching technology to extract the image data of different viewing angles of the same track and the image points of the same name of the adjacent track image data, the different viewing angles include front view, front view and rear view. In this step, the matching algorithm adopts scale-invariant Feature transformation SIFT feature matching and least squares matching are combined, using SIFT feature matching algorithm to provide the initial position of the feature, and then using least squares matching to achieve image fine matching; Step 3, using the above-mentioned data and the image points with the same name, use the independent model method block adjustment to carry out survey area adjustment processing; Step 4, on the basis of the area adjustment, carry out the whole moon adjustment processing; The independent model is constructed in the following way: Obtain the sequence of directional slices on the front view image within the scope of the measurement area, and the directional slices with the same name on the front view and the front view, the front view and the back view, the front view and the back view form the directional slice pairs, and each Oriented slice pairs are relatively oriented to construct an independent model, which is the smallest unit of the block adjustment of the independent model method; Step 3 further includes the steps of: Convert the image coordinates into the image plane coordinates with the principal point of the oriented film image as the origin of the coordinates; Using the same-name image point coordinates of the directional slice pair, the error equation is established through the coplanar condition, and the relative directional element of the directional slice is solved; Use relative orientation elements to solve the model coordinates of each image point with the same name on the independent model constructed by the orientation slice pair; Calculate the lunar surface coordinates of the image point with the same name, Among them, the image coordinates are converted into image plane coordinates with the principal point of the oriented image as the coordinate origin, and the transformation formula is as follows: $\left.\begin{array}{c}{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}\\ {\mathrm{<span\; class="notranslate">\; the\; y</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}\\ {\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\end{array}\right\}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$ Among them, x', y' are the coordinate values of the oriented image plane, the unit is μm; x, y are the coordinate values of the original image, the unit is pixel; ds is the pixel size of the detector; x ₀ , y ₀ are the principal points of the image Image plane coordinate value, the unit is pixel, L ₀ is the position of the scan line in the CCD array; L _i is the line number of the scan line in the image strip in the center of the directional film; d _x is the line number of the directional film, Using the coordinates of the same-named image point of the oriented slice pair, the error equation is established through the coplanar condition, and the relative oriented elements of the oriented slice are solved. The equations satisfied by the coplanar condition are as follows: where, ω, kappa, ω', κ' are the relative orientation elements to be found; B is the baseline length; u, v, w are the left image space coordinates of the left image point, u', v', w' are the right image space coordinates of the right image point , obtained by transforming the image plane coordinates of the same-named image points on the left and right directional slices through their rotation matrix coordinates, and the rotation matrix is constructed by the outer azimuth elements of the directional slices, Equation (2) is expanded by Taylor series and the first-order term is used to obtain the error equation: In the formula, f is the focal length of the camera, The model coordinates of each image point with the same name on the independent model constructed by the oriented slice pair are calculated by using the relative oriented elements, which are calculated by the following formula: U=U _s +Mu _t , V=V _s +Mv _t , W=W _s -Mf(4) $\left.\begin{array}{c}{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}\frac{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}\mathrm{<span\; class="notranslate">\; f</span>}}{{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}\mathrm{<span\; class="notranslate">\; f</span>}}\\ {\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}\frac{{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}\mathrm{<span\; class="notranslate">\; f</span>}}{{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}\mathrm{<span\; class="notranslate">\; f</span>}}\\ {\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}\frac{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\mathrm{<span\; class="notranslate">\; f</span>}}{{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\mathrm{<span\; class="notranslate">\; f</span>}}\end{array}\right\}$ $\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; B</span>}}{{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}$ Among them, U _s , V _s , W _s are the coordinate values of the projection center of the left directional sheet under the model coordinates; U, V, W are the coordinate values of the model points under the model coordinates; a _i , b _i , c _i and a ′ _i , b′ _i , c′ _i are the 9 matrix elements of the left and right oriented sheet rotation matrices R, R′ respectively, where i=1, 2, 3; (x, y), (x′, y′) are The image plane coordinates of the left and right oriented slices, and (x, y) is the result obtained by substituting the original image coordinates of the image point with the same name on the left oriented slice into the formula (1), and (x′, y′) is calculated according to the image with the same name Substituting the original image coordinates of the point on the right oriented slice into the result calculated by formula (1); f is the focal length of the camera; u _t , v _t , u′ _t and M are intermediate variables constructed to simplify the expression of the formula. 2. method according to claim 1, it is characterized in that, the whole moon surface is divided into some mapping areas according to certain principles, each mapping area is divided into a plurality of surveying areas again, on the basis of surveying area adjustment, The global adjustment is carried out with the survey area as the unit. 3. The method according to claim 2, wherein said step 4 further comprises: Using image automatic matching technology, a certain number of image points with the same name are selected in the overlapping area of image data in adjacent survey areas as the connection points of the survey area; Introduce laser lunar survey data for joint adjustment to improve the absolute control accuracy of the control network; Select the control point data for global adjustment processing, and solve the absolute orientation parameters of the survey area.