High-precision positioning method and system

Abstract

The invention discloses a high-precision positioning method and system, and relates to the technical field of satellite positioning, and the method comprises the steps: obtaining the image data of a target region through an optical remote sensing satellite platform, and obtaining the ephemeris attitude auxiliary data of a satellite; performing radiation correction, geometric correction and noise suppression processing on the acquired image data in sequence; constructing an imaging geometric model of the target area; performing segmented calibration on internal parameters of the satellite sensor; factors influencing satellite image positioning precision are attributed to an orthogonal rotation matrix, and a sensor external parameter calibration model is established; constructing an object space virtual grid, and solving parameters of the rational function model; and adjusting and optimizing the ground point coordinates of the image by using a bundle method. By arranging the data preprocessing module, the parameter calibration module and the coordinate optimization module, the direct positioning precision of the satellite image is improved, and accidental errors and residual system errors in adjustment are effectively weakened.

Description

Technical Field

[0001] The present invention relates to the field of satellite positioning technology, and in particular to a high-precision positioning method and system. Background Art

[0002] With the continuous progress of aerospace technology, computer technology, sensor technology and data processing technology, modern satellite remote sensing technology has achieved unprecedented development, and high-resolution earth observation systems have become an important means of obtaining geospatial information. High-precision geometric positioning of satellite images is the basis for geometric processing of remote sensing images and obtaining various geospatial information. It is an important prerequisite for the wide application of high-resolution remote sensing satellite images and the basic guarantee for the use of satellite images to measure and produce topographic maps of various scales.

[0003] The national economy and national defense construction, especially national defense construction, urgently need to accurately grasp the target's geographic spatial information and other characteristics. How to use satellite images to accurately obtain the target's three-dimensional information will inevitably involve the high-precision geometric positioning of satellite images. Summary of the invention

[0004] In order to solve the above technical problems, a high-precision positioning method and system are provided. This technical solution solves the problems raised in the above background technology.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0006] A high-precision positioning method, comprising:

[0007] Use the optical remote sensing satellite platform to obtain image data of the target area and obtain satellite ephemeris attitude auxiliary data;

[0008] The acquired image data is sequentially subjected to radiation correction, geometric correction and noise suppression processing;

[0009] Based on the imaging method of the optical remote sensing satellite platform and the parameters of the sensor, an imaging geometric model of the target area is constructed;

[0010] Using a constant model and a polynomial model to calibrate the internal parameters of the satellite sensor in sections, the internal parameters including focal length, principal point coordinates and distortion coefficient;

[0011] The factors affecting the positioning accuracy of satellite images are attributed to an orthogonal rotation matrix, and a sensor external parameter calibration model is established, wherein the external parameters include the attitude angle and position of the satellite;

[0012] An image grid is established according to the image size, the elevation range of the image coverage area is evenly layered, and an object-space virtual grid is constructed. The rational function model parameters are solved by coordinate normalization to obtain a terrain-independent rational function model.

[0013] The bundle adjustment is used to optimize the ground point coordinates of the image and weaken the systematic and accidental errors in the rational function model.

[0014] Preferably, the sequentially performing radiation correction, geometric correction and noise suppression processing on the acquired image data specifically includes:

[0015] Acquiring atmospheric parameters of the target area, wherein the atmospheric parameters include atmospheric path radiation, atmospheric scattered radiation and atmospheric transmittance;

[0016] Substitute the atmospheric parameters of the target area into the radiation correction formula and calculate the surface reflectivity of the target area;

[0017] Substitute the surface reflectance of the target area into the radiation brightness formula, and calculate the radiation brightness of the surface of the target area;

[0018] Select at least one ground control point on the image;

[0019] Based on the ground control points, a transformation model between the pixel coordinates of the image and the geographic coordinates is established;

[0020] Use the transformation model to geometrically correct the image data;

[0021] Calculate the histogram of grayscale values ​​of image data;

[0022] According to the histogram, determine the distribution range of grayscale values;

[0023] Based on the grayscale value, linear transformation is performed on the points in the image data;

[0024] Determine whether the difference in grayscale values ​​of adjacent points after transformation is lower than a preset threshold value. If so, output the transformed image data as the image data to be processed. If not, re-perform linear transformation on the points in the image data.

[0025] The radiation correction formula is:

[0026]

[0027] Where ρ is the surface reflectivity of the target area, L s is the radiation value received by the sensor, L p is the atmospheric path radiation, L a is the atmospheric scattered radiation, t is the atmospheric transmittance, E is the solar irradiance, and θ1 is the solar zenith angle;

[0028] The radiance formula is:

[0029]

[0030] Where, L r is the radiation brightness of the target area, and θ2 is the observation zenith angle.

[0031] Preferably, the imaging geometric model of the target area based on the imaging mode of the optical remote sensing satellite platform and the parameters of the sensor specifically includes:

[0032] The sensor coordinate system is constructed with the projection center of the sensor as the origin, the flight direction of the sensor as the x-axis, the image scanning direction as the y-axis, and the z-axis defined by the right-hand rule;

[0033] The satellite body coordinate system is constructed with the satellite center of mass as the origin, the satellite flight direction as the x-axis, the satellite horizontal axis direction as the y-axis, and the z-axis defined by the right-hand rule;

[0034] The satellite orbit coordinate system is constructed with the satellite mass center as the origin, the z-axis points from the center of the earth to the satellite mass center outward, the x-axis is in the satellite orbit plane and consistent with the satellite flight direction, and the y-axis is determined by the right-hand rule and is perpendicular to the satellite orbit plane;

[0035] With the center of mass of the earth as the origin, the z-axis points to the north pole of the earth, the x-axis points to the vernal equinox, and the y-axis is perpendicular to the xOz plane to form a right-handed rectangular coordinate system, thus constructing a fixed reference coordinate system in space;

[0036] With the center of mass of the earth as the origin, the z-axis points to the origin agreed upon by the international convention, the equatorial plane of the earth becomes the flat equatorial plane, the x-axis points from the origin to the intersection of the Greenwich meridian plane and the flat equatorial plane, and the y-axis is determined by the right-hand rule and is located in the flat equatorial plane, thus constructing the Earth-centered Earth-fixed coordinate system;

[0037] Get the transformation matrix between each coordinate system;

[0038] Use satellite imagery imaging formulas to construct an imaging geometry model of the target area;

[0039] The satellite image imaging formula is:

[0040]

[0041] Where (X, Y, Z) is the coordinate of the ground point in the Earth-centered Earth-fixed coordinate system, (X s ,Y s ,Z s ) is the coordinate of the projection center in the Earth-centered Earth-fixed coordinate system, (0, y, -f) is the coordinate of the image point in the sensor coordinate system, f is the camera principal distance, λ is the scale factor, R1 is the transformation matrix from the satellite body coordinate system to the satellite orbit coordinate system, R2 is the transformation matrix from the satellite body coordinate system to the space-fixed reference coordinate system, R3 is the transformation matrix from the space-fixed reference coordinate system to the Earth-centered Earth-fixed coordinate system, R4 is the transformation matrix from the sensor coordinate system to the satellite body coordinate system, [D x Dy D z ] T is the offset vector of the phase center in the satellite body coordinate system, [d x d y d z ] T is the offset vector of the camera node in the satellite body coordinate system.

[0042] Preferably, the step of calibrating the internal parameters of the satellite sensor by segment using the constant model and the polynomial model specifically includes:

[0043] Determine whether the frequency of change of the internal parameters of the satellite sensor is greater than a preset frequency threshold. If so, use a polynomial model to perform segmented calibration on the internal parameters of the satellite sensor. If not, use a constant model to perform segmented calibration on the internal parameters of the satellite sensor.

[0044] The internal parameters of the satellite sensor are described as a polynomial function that varies with time, and the imaging characteristics of the sensor are described by fitting the polynomial function.

[0045] The image data to be processed are segmented based on different time periods and different imaging areas, and each segment uses independent constant parameters to describe the internal parameters.

[0046] Preferably, the factors affecting the satellite image positioning accuracy are attributed to an orthogonal rotation matrix, and the establishment of the sensor external parameter calibration model specifically includes:

[0047] Obtaining factors affecting positioning accuracy, wherein the factors affecting positioning accuracy include measurement errors of sensor position and attitude, installation errors between each sensor and the satellite body, and model processing errors;

[0048] The factors that affect the positioning accuracy are summarized into an orthogonal rotation matrix, and the parameters in the orthogonal rotation matrix are constrained by using a set of constraint equations;

[0049] The orthogonal rotation matrix is ​​introduced to calibrate the external parameters of the sensor;

[0050] The sensor external parameter calibration model is used to describe the relationship between the sensor external parameters and the satellite image positioning accuracy;

[0051] Determine the specific form of the external parameters and obtain the change function of the external parameters during the imaging process;

[0052] Using the known ground control point information and the image data to be processed, the optimal value of the external parameters is solved by the least square method;

[0053] The constraint equations are:

[0054]

[0055] In the formula, a ij is the parameter of the i-th row and j-th column in the orthogonal rotation matrix, a tj is the parameter of the tth row and jth column in the orthogonal rotation matrix, N j is the total number of columns of the orthogonal rotation matrix.

[0056] Preferably, the step of establishing an image grid according to the image size, evenly stratifying the elevation range of the image coverage area, constructing an object-space virtual grid, solving the rational function model parameters by coordinate normalization, and obtaining a terrain-independent rational function model specifically includes:

[0057] Based on the actual size of the remote sensing satellite image, it is divided into at least one grid of equal size, and each grid point corresponds to a pixel on the image;

[0058] Get the coordinates of the image point;

[0059] The elevation range of the geographic area covered by the image is divided into layers with the same elevation coordinates;

[0060] Construct a virtual grid in actual geographic space based on imaging geometry model, image grid and elevation layering information;

[0061] Normalize the coordinates of the grid points and convert the coordinate values ​​to a uniform scale;

[0062] By using the correspondence between the image grid and the object virtual grid, the parameters of the rational function model are solved by mathematical methods to obtain a terrain-independent rational function model.

[0063] Preferably, the method of optimizing the ground point coordinates of the image by using bundle adjustment to weaken the systematic error and accidental error in the rational function model specifically includes:

[0064] Collect at least one ground control point data from the image data to be processed;

[0065] The bundle adjustment model is established using the collinearity condition equation;

[0066] Substitute the control point data into the bundle adjustment model for adjustment calculation to solve the ground point coordinates of the optimized image;

[0067] The collinearity condition equation is:

[0068]

[0069] Where Δx, Δy, and Δz are the horizontal, vertical, and vertical coordinate errors of the control point, respectively.

[0070] Furthermore, a high-precision positioning system is proposed, which is used to implement the high-precision positioning method as described above, comprising:

[0071] A data preprocessing module, wherein the data preprocessing module is used to obtain image data of the target area using the optical remote sensing satellite platform, and obtain satellite ephemeris attitude auxiliary data, perform radiation correction, geometric correction and noise suppression processing on the obtained image data in sequence, and construct an imaging geometric model of the target area based on the imaging mode of the optical remote sensing satellite platform and the parameters of the sensor;

[0072] A parameter calibration module, wherein the parameter calibration module is used to perform segmented calibration on the internal parameters of the satellite sensor using a constant model and a polynomial model, attribute the factors affecting the satellite image positioning accuracy to an orthogonal rotation matrix, and establish a sensor external parameter calibration model;

[0073] The coordinate optimization module is used to establish an image grid according to the image size, evenly layer the elevation range of the image coverage area, construct an object-space virtual grid, solve the rational function model parameters through coordinate normalization, obtain a terrain-independent rational function model, use the bundle method to optimize the ground point coordinates of the image, and weaken the systematic errors and accidental errors in the rational function model.

[0074] Optionally, the data preprocessing module specifically includes:

[0075] A data acquisition unit, the data acquisition unit is used to acquire image data of the target area using an optical remote sensing satellite platform, and acquire satellite ephemeris attitude auxiliary data;

[0076] A data correction unit, the data correction unit is used to sequentially perform radiation correction, geometric correction and noise suppression processing on the acquired image data;

[0077] A geometric model building unit, wherein the geometric model building unit is used to build an imaging geometric model of the target area based on the imaging mode of the optical remote sensing satellite platform and the parameters of the sensor;

[0078] The parameter calibration module specifically includes:

[0079] An internal parameter calibration unit, the internal parameter calibration unit is used to perform segmented calibration on the internal parameters of the satellite sensor using a constant model and a polynomial model;

[0080] An external parameter calibration unit, wherein the external parameter calibration unit is used to attribute factors affecting satellite image positioning accuracy to an orthogonal rotation matrix and establish a sensor external parameter calibration model;

[0081] The coordinate optimization module specifically includes:

[0082] A function model building unit, wherein the function model building unit is used to establish an image grid according to the image size, evenly stratify the elevation range of the image coverage area, build an object space virtual grid, solve the rational function model parameters through coordinate normalization, and obtain a terrain-independent rational function model;

[0083] The model error optimization unit is used to optimize the ground point coordinates of the image by using the bundle adjustment method to weaken the systematic error and accidental error in the rational function model.

[0084] Compared with the prior art, the present invention has the following beneficial effects:

[0085] We have conducted systematic and in-depth research on key issues such as the correction of systematic errors of imaging models involved in high-precision earth positioning of optical remote sensing satellite images, the construction of rational function models of object-space virtual grids, on-orbit geometric calibration of photogrammetric parameters, and bundle adjustment with the assistance of ephemeris attitude data. We have formed a relatively complete theoretical system and technical methods for processing optical remote sensing satellite images, which have improved the direct positioning accuracy of satellite images, effectively allocated accidental errors in adjustment, and eliminated residual systematic errors in adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Figure 1 It is a flow chart of the high-precision positioning method of the present invention;

[0087] Figure 2 It is a flow chart of a method for sequentially performing radiation correction, geometric correction and noise suppression processing on acquired image data according to the present invention;

[0088] Figure 3 A flow chart of a method for constructing an imaging geometric model of a target area according to the present invention;

[0089] Figure 4 It is a flow chart of the method for segmentally calibrating the internal parameters of a satellite sensor of the present invention;

[0090] Figure 5 A flow chart of the method for establishing a sensor external parameter calibration model of the present invention;

[0091] Figure 6 A flow chart of a method for solving rational function model parameters to obtain a terrain-independent rational function model according to the present invention;

[0092] Figure 7 The present invention is a flow chart of a method for optimizing ground point coordinates of an image by using bundle adjustment. DETAILED DESCRIPTION

[0093] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0094] Reference Figure 1 As shown, a high-precision positioning method includes:

[0095] Use the optical remote sensing satellite platform to obtain image data of the target area and obtain satellite ephemeris attitude auxiliary data;

[0096] The acquired image data is sequentially subjected to radiation correction, geometric correction and noise suppression processing;

[0097] Based on the imaging method of the optical remote sensing satellite platform and the parameters of the sensor, an imaging geometric model of the target area is constructed;

[0098] Using a constant model and a polynomial model to calibrate the internal parameters of the satellite sensor in sections, the internal parameters including focal length, principal point coordinates and distortion coefficient;

[0099] The factors affecting the positioning accuracy of satellite images are attributed to an orthogonal rotation matrix, and a sensor external parameter calibration model is established, wherein the external parameters include the attitude angle and position of the satellite;

[0100] An image grid is established according to the image size, the elevation range of the image coverage area is evenly layered, and an object-space virtual grid is constructed. The rational function model parameters are solved by coordinate normalization to obtain a terrain-independent rational function model.

[0101] The bundle adjustment is used to optimize the ground point coordinates of the image and weaken the systematic and accidental errors in the rational function model.

[0102] Reference Figure 2 As shown, the acquired image data is sequentially subjected to radiation correction, geometric correction and noise suppression processing, specifically including:

[0103] Acquiring atmospheric parameters of the target area, wherein the atmospheric parameters include atmospheric path radiation, atmospheric scattered radiation and atmospheric transmittance;

[0104] Substitute the atmospheric parameters of the target area into the radiation correction formula and calculate the surface reflectivity of the target area;

[0105] Substitute the surface reflectance of the target area into the radiation brightness formula, and calculate the radiation brightness of the surface of the target area;

[0106] Select at least one ground control point on the image;

[0107] Based on the ground control points, a transformation model between the pixel coordinates of the image and the geographic coordinates is established;

[0108] Use the transformation model to geometrically correct the image data;

[0109] Calculate the histogram of grayscale values ​​of image data;

[0110] According to the histogram, determine the distribution range of grayscale values;

[0111] Based on the grayscale value, linear transformation is performed on the points in the image data;

[0112] Determine whether the difference in grayscale values ​​of adjacent points after transformation is lower than a preset threshold value. If so, output the transformed image data as the image data to be processed. If not, re-perform linear transformation on the points in the image data.

[0113] The radiation correction formula is:

[0114]

[0115] Where ρ is the surface reflectivity of the target area, L s is the radiation value received by the sensor, L p is the atmospheric path radiation, L a is the atmospheric scattered radiation, t is the atmospheric transmittance, E is the solar irradiance, and θ1 is the solar zenith angle;

[0116] The radiance formula is:

[0117]

[0118] Where, L r is the radiation brightness of the target area, and θ2 is the observation zenith angle.

[0119] The surface reflectivity refers to the ratio of the amount of radiation reflected by the ground to the amount of incident radiation, which represents the ground's ability to absorb and reflect solar radiation. The greater the reflectivity, the less solar radiation the ground absorbs, and the smaller the reflectivity, the more solar radiation the ground absorbs. The radiation brightness is a physical quantity that represents the intensity of radiation at a certain point on a surface radiation source in a certain direction. It refers to the energy radiated by the surface radiation source per unit area and per unit solid angle in a direction perpendicular to the surface element normal in unit time, that is, the radiation flux of the radiation source per unit projection area and per unit solid angle.

[0120] Reference Figure 3 As shown in the figure, based on the imaging method of the optical remote sensing satellite platform and the parameters of the sensor, the imaging geometric model of the target area is constructed, which specifically includes:

[0121] The sensor coordinate system is constructed with the projection center of the sensor as the origin, the flight direction of the sensor as the x-axis, the image scanning direction as the y-axis, and the z-axis defined by the right-hand rule;

[0122] The satellite body coordinate system is constructed with the satellite center of mass as the origin, the satellite flight direction as the x-axis, the satellite horizontal axis direction as the y-axis, and the z-axis defined by the right-hand rule;

[0123] The satellite orbit coordinate system is constructed with the satellite mass center as the origin, the z-axis points from the center of the earth to the satellite mass center outward, the x-axis is in the satellite orbit plane and consistent with the satellite flight direction, and the y-axis is determined by the right-hand rule and is perpendicular to the satellite orbit plane;

[0124] With the center of mass of the earth as the origin, the z-axis points to the north pole of the earth, the x-axis points to the vernal equinox, and the y-axis is perpendicular to the xOz plane to form a right-handed rectangular coordinate system, thus constructing a fixed reference coordinate system in space;

[0125] With the center of mass of the earth as the origin, the z-axis points to the origin agreed upon by the international convention, the equatorial plane of the earth becomes the flat equatorial plane, the x-axis points from the origin to the intersection of the Greenwich meridian plane and the flat equatorial plane, and the y-axis is determined by the right-hand rule and is located in the flat equatorial plane, thus constructing the Earth-centered Earth-fixed coordinate system;

[0126] Get the transformation matrix between each coordinate system;

[0127] Use satellite imagery imaging formulas to construct an imaging geometry model of the target area;

[0128] The satellite image imaging formula is:

[0129]

[0130] Where (X, Y, Z) is the coordinate of the ground point in the Earth-centered Earth-fixed coordinate system, (X s ,Y s ,Z s ) is the coordinate of the projection center in the Earth-centered Earth-fixed coordinate system, (0, y, -f) is the coordinate of the image point in the sensor coordinate system, f is the camera principal distance, λ is the scale factor, R1 is the transformation matrix from the satellite body coordinate system to the satellite orbit coordinate system, R2 is the transformation matrix from the satellite body coordinate system to the space-fixed reference coordinate system, R3 is the transformation matrix from the space-fixed reference coordinate system to the Earth-centered Earth-fixed coordinate system, R4 is the transformation matrix from the sensor coordinate system to the satellite body coordinate system, [D x D y D z ] T is the offset vector of the phase center in the satellite body coordinate system, [d x d y d z ] T is the offset vector of the camera node in the satellite body coordinate system.

[0131] At the 1960 Helsinki Congress of the International Union of Geodesy and Geophysics, a resolution was passed to take the average position of the Earth's rotation axis measured by the five stations of the International Latitude Bureau from 1900 to 1905 as the terrestrial pole, that is, the origin of polar shift. This is commonly known as the "international agreed origin". It is a type of fixed parallel pole. The constant changes in the direction of the Earth's coordinate system cause many difficulties in practical work. Therefore, as early as 1967, the International Astronomical Union and the International Geodetic Institute recommended the use of the average polar position determined by the average latitude from 1900 to 1905 by five international latitude service stations as the reference point, usually called the international agreed origin.

[0132] Reference Figure 4 As shown in FIG. 1 , the internal parameters of the satellite sensor are calibrated in sections using the constant model and the polynomial model, including:

[0133] Determine whether the frequency of change of the internal parameters of the satellite sensor is greater than a preset frequency threshold. If so, use a polynomial model to perform segmented calibration on the internal parameters of the satellite sensor. If not, use a constant model to perform segmented calibration on the internal parameters of the satellite sensor.

[0134] The internal parameters of the satellite sensor are described as a polynomial function that varies with time, and the imaging characteristics of the sensor are described by fitting the polynomial function.

[0135] The image data to be processed are segmented based on different time periods and different imaging areas, and each segment uses independent constant parameters to describe the internal parameters.

[0136] By simulating the operating environment of satellite sensors in a laboratory environment and observing the changes in internal parameters, the appropriate threshold range can be found by adjusting different parameters and conditions.

[0137] Reference Figure 5 As shown in the figure, the factors affecting the positioning accuracy of satellite images are attributed to an orthogonal rotation matrix, and the sensor external parameter calibration model is established, which specifically includes:

[0138] Obtaining factors affecting positioning accuracy, wherein the factors affecting positioning accuracy include measurement errors of sensor position and attitude, installation errors between each sensor and the satellite body, and model processing errors;

[0139] The factors that affect the positioning accuracy are summarized into an orthogonal rotation matrix, and the parameters in the orthogonal rotation matrix are constrained by using a set of constraint equations;

[0140] The orthogonal rotation matrix is ​​introduced to calibrate the external parameters of the sensor;

[0141] The sensor external parameter calibration model is used to describe the relationship between the sensor external parameters and the satellite image positioning accuracy;

[0142] Determine the specific form of the external parameters and obtain the change function of the external parameters during the imaging process;

[0143] Using the known ground control point information and the image data to be processed, the optimal value of the external parameters is solved by the least square method;

[0144] The constraint equations are:

[0145]

[0146] In the formula, a ij is the parameter of the i-th row and j-th column in the orthogonal rotation matrix, a tj is the parameter of the tth row and jth column in the orthogonal rotation matrix, N j is the total number of columns of the orthogonal rotation matrix.

[0147] The least squares method, also known as the least square method, is a mathematical optimization technique that finds the best function match for data by minimizing the sum of squares of errors. The least squares method can be used to easily obtain unknown data and minimize the sum of squares of errors between the obtained data and the actual data. The least squares method can also be used for curve fitting. Some other optimization problems can also be expressed using the least squares method by minimizing energy or maximizing entropy.

[0148] Reference Figure 6 As shown in the figure, an image grid is established according to the image size, the elevation range of the image coverage area is evenly layered, and an object-space virtual grid is constructed. By normalizing the coordinates, the rational function model parameters are solved to obtain a terrain-independent rational function model, which specifically includes:

[0149] Based on the actual size of the remote sensing satellite image, it is divided into at least one grid of equal size, and each grid point corresponds to a pixel on the image;

[0150] Get the coordinates of the image point;

[0151] The elevation range of the geographic area covered by the image is divided into layers with the same elevation coordinates;

[0152] Construct a virtual grid in actual geographic space based on imaging geometry model, image grid and elevation layering information;

[0153] Normalize the coordinates of the grid points and convert the coordinate values ​​to a uniform scale;

[0154] By using the correspondence between the image grid and the object virtual grid, the parameters of the rational function model are solved by mathematical methods to obtain a terrain-independent rational function model.

[0155] Elevation refers to the distance from a point along the plumb line to the absolute base surface, which is called absolute elevation, or elevation for short. The distance from a point along the plumb line to a hypothetical leveling base surface is called assumed elevation. Normalization is a dimensionless processing method that converts the absolute value of a physical system's numerical value into a relative value relationship, which is an effective way to simplify calculations and reduce the value.

[0156] Reference Figure 7 As shown in the figure, the bundle adjustment is used to optimize the ground point coordinates of the image and weaken the systematic errors and accidental errors in the rational function model. Specifically, the following are included:

[0157] Collect at least one ground control point data from the image data to be processed;

[0158] The bundle adjustment model is established using the collinearity condition equation;

[0159] Substitute the control point data into the bundle adjustment model for adjustment calculation to solve the ground point coordinates of the optimized image;

[0160] The collinearity condition equation is:

[0161]

[0162] Where Δx, Δy, and Δz are the horizontal, vertical, and vertical coordinate errors of the control point, respectively.

[0163] a in the collinearity condition equation 11 Parameters such as are taken from the parameters of corresponding rows and columns in the orthogonal rotation matrix; bundle adjustment has the characteristics of rigorous theory and high precision. It can process multiple control point data at the same time, improving the stability and reliability of the positioning results. In addition, bundle adjustment can also adapt to different types of remote sensing satellite images and imaging geometric models, and has wide applicability.

[0164] Furthermore, based on the same inventive concept as the above-mentioned high-precision positioning method, the present solution also proposes a high-precision positioning system, comprising:

[0165] A data preprocessing module, wherein the data preprocessing module is used to obtain image data of the target area using the optical remote sensing satellite platform, and obtain satellite ephemeris attitude auxiliary data, perform radiation correction, geometric correction and noise suppression processing on the obtained image data in sequence, and construct an imaging geometric model of the target area based on the imaging mode of the optical remote sensing satellite platform and the parameters of the sensor;

[0166] A parameter calibration module, wherein the parameter calibration module is used to perform segmented calibration on the internal parameters of the satellite sensor using a constant model and a polynomial model, attribute the factors affecting the satellite image positioning accuracy to an orthogonal rotation matrix, and establish a sensor external parameter calibration model;

[0167] The coordinate optimization module is used to establish an image grid according to the image size, evenly layer the elevation range of the image coverage area, construct an object-space virtual grid, solve the rational function model parameters through coordinate normalization, obtain a terrain-independent rational function model, use the bundle method to optimize the ground point coordinates of the image, and weaken the systematic errors and accidental errors in the rational function model.

[0168] The data preprocessing module specifically includes:

[0169] A data acquisition unit, the data acquisition unit is used to acquire image data of the target area using an optical remote sensing satellite platform, and acquire satellite ephemeris attitude auxiliary data;

[0170] A data correction unit, the data correction unit is used to sequentially perform radiation correction, geometric correction and noise suppression processing on the acquired image data;

[0171] A geometric model building unit, wherein the geometric model building unit is used to build an imaging geometric model of the target area based on the imaging mode of the optical remote sensing satellite platform and the parameters of the sensor;

[0172] The parameter calibration module specifically includes:

[0173] An internal parameter calibration unit, the internal parameter calibration unit is used to perform segmented calibration on the internal parameters of the satellite sensor using a constant model and a polynomial model;

[0174] An external parameter calibration unit, wherein the external parameter calibration unit is used to attribute factors affecting satellite image positioning accuracy to an orthogonal rotation matrix and establish a sensor external parameter calibration model;

[0175] The coordinate optimization module specifically includes:

[0176] A function model building unit, wherein the function model building unit is used to establish an image grid according to the image size, evenly stratify the elevation range of the image coverage area, build an object space virtual grid, solve the rational function model parameters through coordinate normalization, and obtain a terrain-independent rational function model;

[0177] The model error optimization unit is used to optimize the ground point coordinates of the image by using the bundle adjustment method to weaken the systematic error and accidental error in the rational function model.

[0178] Furthermore, the present solution also proposes a computer-readable storage medium on which a computer-readable program is stored, and when the computer-readable program is called, the above-mentioned high-precision positioning method is executed.

[0179] It is understandable that the storage medium may be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid state drive (SSD).

[0180] In summary, the advantages of the present invention are: systematic and in-depth research is conducted on key issues such as the correction of system errors of imaging models involved in high-precision earth positioning of optical remote sensing satellite images, the construction of rational function models of object-space virtual grids, on-orbit geometric calibration of photogrammetric parameters, and bundle adjustment under the auxiliary conditions of ephemeris attitude data, forming a relatively complete theoretical system and technical method for processing optical remote sensing satellite images, improving the direct positioning accuracy of satellite images, effectively allocating accidental errors in adjustment, and eliminating residual systematic errors in adjustment.

[0181] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A high-precision positioning method, characterized in that: include: Use the optical remote sensing satellite platform to obtain image data of the target area and obtain satellite ephemeris attitude auxiliary data; The acquired image data is sequentially subjected to radiation correction, geometric correction and noise suppression processing; Based on the imaging method of the optical remote sensing satellite platform and the parameters of the sensor, an imaging geometric model of the target area is constructed; Using a constant model and a polynomial model to calibrate the internal parameters of the satellite sensor in sections, the internal parameters including focal length, principal point coordinates and distortion coefficient; The factors affecting the positioning accuracy of satellite images are attributed to an orthogonal rotation matrix, and a sensor external parameter calibration model is established, wherein the external parameters include the attitude angle and position of the satellite; An image grid is established according to the image size, the elevation range of the image coverage area is evenly layered, and an object-space virtual grid is constructed. The rational function model parameters are solved by coordinate normalization to obtain a terrain-independent rational function model. The bundle adjustment is used to optimize the ground point coordinates of the image and weaken the systematic and accidental errors in the rational function model.

2. A high-precision positioning method according to claim 1, characterized in that: The sequentially performing radiation correction, geometric correction and noise suppression processing on the acquired image data specifically includes: Acquiring atmospheric parameters of the target area, wherein the atmospheric parameters include atmospheric path radiation, atmospheric scattered radiation and atmospheric transmittance; Substitute the atmospheric parameters of the target area into the radiation correction formula and calculate the surface reflectivity of the target area; Substitute the surface reflectance of the target area into the radiation brightness formula, and calculate the radiation brightness of the surface of the target area; Select at least one ground control point on the image; Based on the ground control points, a transformation model between the pixel coordinates of the image and the geographic coordinates is established; Use the transformation model to geometrically correct the image data; Calculate the histogram of grayscale values ​​of image data; According to the histogram, determine the distribution range of gray value; Based on the grayscale value, linear transformation is performed on the points in the image data; Determine whether the difference in grayscale values ​​of adjacent points after transformation is lower than a preset threshold value. If so, output the transformed image data as the image data to be processed. If not, re-perform linear transformation on the points in the image data. The radiation correction formula is: Where ρ is the surface reflectivity of the target area, L s is the radiation value received by the sensor, L p is the atmospheric path radiation, L a is the atmospheric scattered radiation, t is the atmospheric transmittance, E is the solar irradiance, and θ1 is the solar zenith angle; The radiance formula is: Where, L r is the radiation brightness of the target area, and θ2 is the observation zenith angle.

3. A high-precision positioning method according to claim 2, characterized in that: The imaging geometric model of the target area based on the imaging mode of the optical remote sensing satellite platform and the parameters of the sensor specifically includes: The sensor coordinate system is constructed with the projection center of the sensor as the origin, the flight direction of the sensor as the x-axis, the image scanning direction as the y-axis, and the z-axis defined by the right-hand rule; The satellite body coordinate system is constructed with the satellite center of mass as the origin, the satellite flight direction as the x-axis, the satellite horizontal axis direction as the y-axis, and the z-axis defined by the right-hand rule; The satellite orbit coordinate system is constructed with the satellite mass center as the origin, the z-axis points from the center of the earth to the satellite mass center outward, the x-axis is in the satellite orbit plane and consistent with the satellite flight direction, and the y-axis is determined by the right-hand rule and is perpendicular to the satellite orbit plane; With the center of mass of the earth as the origin, the z-axis points to the north pole of the earth, the x-axis points to the vernal equinox, and the y-axis is perpendicular to the xOz plane to form a right-handed rectangular coordinate system, thus constructing a fixed reference coordinate system in space; With the center of mass of the earth as the origin, the z-axis points to the origin agreed upon by the international convention, the equatorial plane of the earth becomes the flat equatorial plane, the x-axis points from the origin to the intersection of the Greenwich meridian plane and the flat equatorial plane, and the y-axis is determined by the right-hand rule and is located in the flat equatorial plane, thus constructing the Earth-centered Earth-fixed coordinate system; Get the transformation matrix between each coordinate system; Use satellite imagery imaging formulas to construct an imaging geometry model of the target area; The satellite image imaging formula is: Where (X, Y, Z) is the coordinate of the ground point in the Earth-centered Earth-fixed coordinate system, (X s ,Y s ,Z s ) is the coordinate of the projection center in the Earth-centered Earth-fixed coordinate system, (0, y, -f) is the coordinate of the image point in the sensor coordinate system, f is the camera principal distance, λ is the scale factor, R1 is the transformation matrix from the satellite body coordinate system to the satellite orbit coordinate system, R2 is the transformation matrix from the satellite body coordinate system to the space-fixed reference coordinate system, R3 is the transformation matrix from the space-fixed reference coordinate system to the Earth-centered Earth-fixed coordinate system, R4 is the transformation matrix from the sensor coordinate system to the satellite body coordinate system, [D x D y D z ] T is the offset vector of the phase center in the satellite body coordinate system, [d x d y d z ] T is the offset vector of the camera node in the satellite body coordinate system.

4. A high-precision positioning method according to claim 3, characterized in that: The segmented calibration of the internal parameters of the satellite sensor using the constant model and the polynomial model specifically includes: Determine whether the frequency of change of the internal parameters of the satellite sensor is greater than a preset frequency threshold. If so, use a polynomial model to perform segmented calibration on the internal parameters of the satellite sensor. If not, use a constant model to perform segmented calibration on the internal parameters of the satellite sensor. The internal parameters of the satellite sensor are described as a polynomial function that varies with time, and the imaging characteristics of the sensor are described by fitting the polynomial function. The image data to be processed are segmented based on different time periods and different imaging areas, and each segment uses independent constant parameters to describe the internal parameters.

5. A high-precision positioning method according to claim 4, characterized in that: The factors affecting the positioning accuracy of satellite images are attributed to an orthogonal rotation matrix, and the establishment of the sensor external parameter calibration model specifically includes: Obtaining factors affecting positioning accuracy, wherein the factors affecting positioning accuracy include measurement errors of sensor position and attitude, installation errors between each sensor and the satellite body, and model processing errors; The factors that affect the positioning accuracy are summarized into an orthogonal rotation matrix, and the parameters in the orthogonal rotation matrix are constrained by using a set of constraint equations; The orthogonal rotation matrix is ​​introduced to calibrate the external parameters of the sensor; The sensor external parameter calibration model is used to describe the relationship between the sensor external parameters and the satellite image positioning accuracy; Determine the specific form of the external parameters and obtain the change function of the external parameters during the imaging process; Using the known ground control point information and the image data to be processed, the optimal value of the external parameters is solved by the least square method; The constraint equations are: In the formula, a ij is the parameter of the i-th row and j-th column in the orthogonal rotation matrix, a tj is the parameter of the tth row and jth column in the orthogonal rotation matrix, N j is the total number of columns of the orthogonal rotation matrix.

6. A high-precision positioning method according to claim 5, characterized in that: The image grid is established according to the image size, the elevation range of the image coverage area is evenly layered, the object space virtual grid is constructed, and the rational function model parameters are solved by coordinate normalization to obtain the terrain-independent rational function model. Specifically, the following steps are performed: Based on the actual size of the remote sensing satellite image, it is divided into at least one grid of equal size, and each grid point corresponds to a pixel on the image; Get the coordinates of the image point; The elevation range of the geographic area covered by the image is divided into layers with the same elevation coordinates; Construct a virtual grid in actual geographic space based on imaging geometry model, image grid and elevation layering information; Normalize the coordinates of the grid points and convert the coordinate values ​​to a uniform scale; By using the correspondence between the image grid and the object virtual grid, the parameters of the rational function model are solved by mathematical methods to obtain a terrain-independent rational function model.

7. A high-precision positioning method according to claim 6, characterized in that: The method of optimizing the ground point coordinates of the image by using the bundle adjustment method and weakening the systematic error and accidental error in the rational function model specifically includes: Collect at least one ground control point data from the image data to be processed; The bundle adjustment model is established using the collinearity condition equation; Substitute the control point data into the bundle adjustment model for adjustment calculation to solve the ground point coordinates of the optimized image; The collinearity condition equation is: Where Δx, Δy, and Δz are the horizontal, vertical, and vertical coordinate errors of the control point, respectively.

8. A high-precision positioning system, used to implement a high-precision positioning method as claimed in any one of claims 1 to 7, characterized in that: include: A data preprocessing module, wherein the data preprocessing module is used to obtain image data of the target area using the optical remote sensing satellite platform, and obtain satellite ephemeris attitude auxiliary data, perform radiation correction, geometric correction and noise suppression processing on the obtained image data in sequence, and construct an imaging geometric model of the target area based on the imaging mode of the optical remote sensing satellite platform and the parameters of the sensor; A parameter calibration module, wherein the parameter calibration module is used to perform segmented calibration on the internal parameters of the satellite sensor using a constant model and a polynomial model, attribute the factors affecting the satellite image positioning accuracy to an orthogonal rotation matrix, and establish a sensor external parameter calibration model; The coordinate optimization module is used to establish an image grid according to the image size, evenly layer the elevation range of the image coverage area, construct an object-space virtual grid, solve the rational function model parameters through coordinate normalization, obtain a terrain-independent rational function model, use the bundle method to optimize the ground point coordinates of the image, and weaken the systematic errors and accidental errors in the rational function model.

9. A high-precision positioning system according to claim 8, characterized in that: The data preprocessing module specifically includes: A data acquisition unit, the data acquisition unit is used to acquire image data of the target area using an optical remote sensing satellite platform, and acquire satellite ephemeris attitude auxiliary data; A data correction unit, the data correction unit is used to sequentially perform radiation correction, geometric correction and noise suppression processing on the acquired image data; A geometric model building unit, wherein the geometric model building unit is used to build an imaging geometric model of the target area based on the imaging mode of the optical remote sensing satellite platform and the parameters of the sensor; The parameter calibration module specifically includes: An internal parameter calibration unit, the internal parameter calibration unit is used to perform segmented calibration on the internal parameters of the satellite sensor using a constant model and a polynomial model; An external parameter calibration unit, wherein the external parameter calibration unit is used to attribute factors affecting satellite image positioning accuracy to an orthogonal rotation matrix and establish a sensor external parameter calibration model; The coordinate optimization module specifically includes: A function model building unit, wherein the function model building unit is used to establish an image grid according to the image size, evenly stratify the elevation range of the image coverage area, build an object space virtual grid, solve the rational function model parameters through coordinate normalization, and obtain a terrain-independent rational function model; The model error optimization unit is used to optimize the ground point coordinates of the image by using the bundle adjustment method to weaken the systematic error and accidental error in the rational function model.

Images