CN112270714A - In-orbit progressive calibration method for parameters of satellite linear array stereo camera - Google Patents

Abstract

The invention relates to an in-orbit progressive calibration method for parameters of a satellite linear array stereo camera, belonging to the technical field of photogrammetry and remote sensing. The method comprises the steps of firstly obtaining initial values of elements of an external orientation of a satellite in-orbit camera and laboratory calibration parameters of the camera. Then, measuring the image points of the satellite image; obtaining the ground coordinates (X) of the contact point in the test field range_i,Y_i,Z_i) I-1, 2, …, n-3 as control data; calibrating the focal length of the camera independently to obtain a preliminary calibration result of the focal length; and replacing a laboratory calibration value with the preliminarily calibrated focal length, performing overall parameter solution on the focal length of the camera, the longitudinal coordinate of the principal point, the intersection angle of the stereo camera and the included angle of the satellite-ground camera, and calculating a final camera parameter on-orbit calibration result IO'. Based on the adjustment theory of the stereoscopic image beam method, the invention systematically analyzes the correlation between the focal length of the camera and other photogrammetric parameters, and provides a progressive calibration method for firstly resolving the focal length of the camera and then integrally resolving the focal length of the cameraThe problem that system errors exist in a calibration result due to the fact that the aspect ratio of the satellite linear array image is too small is solved.

Description

In-orbit progressive calibration method for parameters of satellite linear array stereo camera

Technical Field

The invention relates to an in-orbit progressive calibration method for parameters of a satellite linear array stereo camera, belonging to the technical field of photogrammetry and remote sensing.

Background

The on-orbit calibration of camera parameters is an effective way and an important means for improving the positioning precision of satellite images. Currently, most remote sensing satellites adopt calibration methods such as internal calibration, external calibration, self-checking based on additional parameters and the like, but the methods mainly aim at a single camera. For stereo mapping cameras (such as twin or three line arrays), the intersection angle between the cameras is an important factor affecting the elevation accuracy. If the camera parameters are calibrated in an on-orbit mode only by starting from a single camera, the integrity of a stereo camera is ignored, and the later positioning precision is influenced.

The adjustment of the light beam method is also an effective way for realizing the on-orbit calibration of the camera parameters, and the calibration parameters are integrally solved through the adjustment of the light beam method of the three-dimensional image, so that the on-orbit calibration of the camera parameters can be realized. For example, the method is adopted by the German MOMS-2P and the China Tianzhu satellite I to carry out integral solution on the calibration parameters. However, in practical engineering application to the sky plot one satellite, it is found that the method cannot completely eliminate the systematic error existing in the camera parameter variation because the aspect ratio (ratio of image width to orbit height) of the satellite image is too small (about 1/10).

Disclosure of Invention

The purpose of the invention is: the invention provides an in-orbit progressive calibration method for parameters of a satellite linear array stereo camera, which aims to solve the problem of system errors in a calibration result caused by an excessively small aspect ratio of a satellite linear array image.

The technical scheme of the invention is as follows: the in-orbit progressive calibration method for the parameters of the satellite linear array stereo camera comprises the following steps:

the method comprises the following steps: obtaining initial values of external orientation elements of satellite in-orbit camera

And laboratory calibration parameters IO of the camera, including a camera focal length, a principal point ordinate, a stereo camera intersection angle and a satellite-ground camera included angle.

Step two: measuring image points of the satellite image to obtain coordinates (x) of the same-name image points of the orientation points_li,y_li)、(x_ri,y_ri) 1,2, …, n, co-named image point coordinates (x ') of the connection points'_li,y'_li)、(x'_ri,y'_ri),i＝1,2,…,n-3；

Step three: obtaining the ground coordinates (X) of the contact point in the test field range_i,Y_i,Z_i) I-1, 2, …, n-3 as control data;

step four: according to the ground coordinates of the connecting points, the exterior orientation elements and the image point measurement data, independently calibrating the focal length of the camera to obtain a preliminary calibration result of the focal length;

step five: and (3) replacing a laboratory calibration value with the preliminarily calibrated focal length, repeating the second step and the third step, then carrying out parameter integral solution on the focal length of the camera, the longitudinal coordinate of the principal point, the intersection angle of the stereo camera and the included angle of the satellite-ground camera, and calculating a final camera parameter on-orbit calibration result IO'.

The invention has the beneficial effects that: when an error equation is constructed by a beam method, because the aspect ratio is too small, the coefficient matrix value (y/f) corresponding to the focal length is too small, and further, when the calibration parameters are directly solved integrally, the focal length value cannot be effectively corrected, so that a certain system error exists in the positioning error. Based on the adjustment theory of the stereo image beam method, the invention systematically analyzes the correlation between the focal length of the camera and other photogrammetric parameters, provides a progressive calibration method for resolving the focal length of the camera firstly and then solving the overall solution, and effectively solves the problem of system errors in the calibration result caused by the over-small aspect ratio of the satellite linear array image.

Drawings

FIG. 1 is a schematic diagram of performing a pixel measurement on two satellite images in the steps of the embodiment.

DETAILED DESCRIPTION OF EMBODIMENT (S) OF INVENTION

In this embodiment, taking a satellite three-line-array camera as an example, the in-orbit progressive calibration of camera parameters includes the following steps:

the method comprises the following steps: initial value of exterior orientation element

See table 1:

TABLE 1 initial values of exterior orientation elements

The laboratory calibration parameters of the three-line-array camera are as shown in table 2:

TABLE 2 laboratory calibration parameter values

(f_l,f_v,f_r) Laboratory calibration values of the focal lengths of the front-view, front-view and rear-view cameras respectively, (alpha)_l,α_r) Laboratory calibration values for the angles of front and front, front and back cameras, respectively, (y)_ol,y_ov,y_or) Respectively are laboratory calibration values of longitudinal coordinates of main points of a front-view camera, a front-view camera and a rear-view camera,

respectively, are laboratory calibration values of the satellite-ground camera included angle.

Step two: on 2 base lines (about 470 km in length), 63 orientation points and 60 connection points were selected and measured, and the specific coordinate values are shown in tables 3 and 4.

TABLE 3 Directional Point coordinates (Unit: Pixel)

TABLE 4 connection points coordinates of image points (Unit: Pixel)

i	x′li	y′li	x′vi	y′vi	x′ri	y′ri
							1	3968.398	10482.895	3421.492	10632.166	3012.930	10532.668
2	4038.555	5888.384	3504.891	5999.500	3108.867	5940.380
							3	4193.508	949.182	3683.828	1018.833	3311.094	1002.406
…
							58	93773.518	10435.482	93089.104	10576.435	92541.934	10481.878
59	93741.422	5624.286	93070.899	5725.101	92536.760	5672.608
							60	93681.334	1002.314	93035.700	1064.700	92525.702	1052.527

Step three: the coordinates of the ground coordinate points of the connecting points are obtained, see table 5:

table 5 connection point ground point coordinates (unit: meter)

i	Xi	Yi	Zi
				1	-2580697.48443	3449526.14637	4687878.85496
2	-2559931.83233	3460281.00769	4691337.00794
				3	-2537672.57680	3472129.50374	4694668.25714
…
				58	-2683220.19149	3770655.80625	4374595.88045
59	-2661363.80939	3781386.84531	4378657.19069
				60	-2640298.08738	3791573.60100	4382551.81543

Step four: focal length of the camera (f)_l,f_v,f_r) Separate calibration calculations were performed and the results are shown in table 6.

Table 6 satellite three-line-array camera focal length in-orbit calibration value

Step five: and (5) replacing the laboratory calibration value with the focus preliminarily calibrated in the fourth step, integrally calibrating the parameters of the satellite three-line-array camera, and obtaining the calculation result shown in the table 7.

TABLE 7 satellite three-line-array camera parameter on-orbit calibration value

In this embodiment, in order to verify the in-orbit progressive calibration method for the parameters of the satellite three-linear-array camera, 24 check points and 41 check points are respectively selected by using two foreign precision detection fields, and the positioning precision of the original calibration method and the progressive calibration method is counted, wherein the counting result is shown in table 8.

TABLE 8 positioning accuracy statistics

In Table 8. mu._XRoot mean square error of X coordinate, mu_YRoot mean square error of Y coordinate, mu_hFor the root mean square error in elevation, mu_pThe horizontal position root mean square error. It can be seen from the results that the positioning accuracy after the progressive scaling is significantly improved.

Claims (1)

1. The in-orbit progressive calibration method for the parameters of the satellite linear array stereo camera comprises the following steps:

the method comprises the following steps: obtaining initial values of external orientation elements of satellite in-orbit camera

And laboratory calibration parameters IO of the camera, including a camera focal length, a principal point ordinate, a stereo camera intersection angle and a satellite-ground camera included angle.

Step two: measuring image points of the satellite image to obtain coordinates (x) of the same-name image points of the orientation points_li,y_li)、(x_ri,y_ri) 1,2, …, n, co-named image point coordinates (x ') of the connection points'_li,y'_li)、(x'_ri,y'_ri),i＝1,2,…,n-3；

Step three: obtaining the ground coordinates (X) of the contact point in the test field range_i,Y_i,Z_i) I-1, 2, …, n-3 as control data;

step four: according to the ground coordinates of the connecting points, the exterior orientation elements and the image point measurement data, independently calibrating the focal length of the camera to obtain a preliminary calibration result of the focal length;

step five: and (3) replacing a laboratory calibration value with the preliminarily calibrated focal length, repeating the second step and the third step, then carrying out parameter integral solution on the focal length of the camera, the longitudinal coordinate of the principal point, the intersection angle of the stereo camera and the included angle of the satellite-ground camera, and calculating a final camera parameter on-orbit calibration result IO'.

Images