CN114608540B - Measurement net type determining method for digital photogrammetry system - Google Patents

Abstract

The invention provides a measurement network type determining method of a digital photogrammetry system. The method comprises the following steps: and simulating by using a beam method adjustment model, solving error equations corresponding to all measurement points after imaging at each shooting station by using a least square method to obtain current optimal object point coordinates and current optimal net shape, fitting the current optimal object point coordinates with a preset theoretical digital model to obtain current measurement errors, sequentially adjusting each parameter to be optimized according to a preset optimization sequence, and iteratively solving the measurement errors until the measurement errors in the adjustment process of each parameter to be optimized meet corresponding convergence conditions to obtain the optimal measurement net shape. The whole method can determine the most suitable measurement net type according to the field actual measurement conditions, so that the measurement error of the digital photogrammetry system is stable, and the measurement accuracy is high.

Description

A Measuring Network Type Determination Method for Digital Photogrammetry System

technical field

The invention belongs to the fields of computer vision and computer graphics, and relates to a method for determining a measurement network type of a digital photogrammetry system.

Background technique

Digital photogrammetry is a measurement technology that has developed very rapidly in recent years. It mainly uses an optical camera to take a series of photos, and obtains the three-dimensional coordinates of the points to be measured through computer image matching and related mathematical calculations. The measurement principle of the digital photogrammetry system is the same as that of the theodolite system, which uses the triangular measurement principle.

Different from conventional measurement methods, the digital photogrammetry system cannot intuitively obtain measurement data from the measurement equipment, but first needs to take a series of photos of the sample to be tested, and then solve the target points on the sample to be tested by image processing and mathematical methods. Based on this, the spatial network formed by all shooting points, points to be measured and photographing light is called the measurement network type. The measurement network type mainly includes the shooting position and attitude distribution of the camera in the shooting station (ie, the shooting station distribution) , the target distribution of the sample to be tested (that is, the distribution of measuring points) and the position of the ruler (reference layout). The setting of the measurement network type will not only affect the pixel accuracy of the photo, but also an important input parameter in the solution model. Therefore, from the perspective of the measurement method, the measurement network type is the most influential factor on the measurement accuracy.

At present, the determination of the measurement network type of the digital photogrammetry system mainly depends on the engineering experience of the operator, and it is impossible to accurately determine the measurement network type under various conditions such as different accuracy requirements and different size measurement objects. When the operator determines the measurement network type, it is more based on the measurement principle of the triangular method. Since each point to be measured needs to be intersected by at least two photographic light beams, it can be solved if the photographic light rays intersected at the point to be measured are increased. Improving the measurement accuracy of digital photogrammetry systems. However, the actual engineering application environment is more complex, especially for large-aperture antennas or panel samples in environmental chambers, which are limited by the auxiliary measurement platform and the measurement space range, making it difficult to achieve an ideal measurement network type. Under extreme measurement conditions, when the photographic light cannot form an ideal intersection angle, the measurement data in the field of view of a single photo is small, or more photos need to be spliced to form a complete reflective surface coordinate information, etc., the digital photogrammetry system The measurement accuracy of the digital photogrammetry system is very sensitive to the position and attitude of the camera. At this time, only relying on the engineering experience of the operator cannot accurately determine the appropriate measurement network type, which leads to unstable measurement errors and low measurement accuracy of the digital photogrammetry system. .

Contents of the invention

Aiming at the deficiencies in the prior art, the present invention provides a method for determining the measurement network type of a digital photogrammetry system, including:

According to the on-site measurement conditions, determine the initial distribution of the station where the camera is located, the initial station distribution includes the initial position of the station where the camera is located, and the initial attitude angle of the station;

The beam adjustment model is used for simulation to obtain the error equation corresponding to each measurement point on the sample to be measured after imaging at each camera station, wherein each measurement point on the sample to be measured is according to the preset density The parameters are evenly distributed on the surface of the sample to be tested;

Use the least square method to solve the error equations corresponding to all measurement points after imaging at each camera station, and obtain the current optimal three-dimensional coordinates of the object space point, as well as the current optimal position of the camera station and the current optimal attitude angle of the camera station;

Fitting the current optimal three-dimensional coordinates of the object space point with the preset theoretical digital simulation to obtain the current measurement error;

According to the current optimal position of the camera station, the current optimal attitude angle of the camera station and the density parameter, the initial measurement network type is set, and each parameter to be optimized is adjusted in turn according to the preset optimization sequence, and the measurement error is solved iteratively until each The measurement errors in the adjustment process of the parameters to be optimized all satisfy the corresponding convergence conditions, and the optimal measurement network type is obtained, and the optimal measurement network type includes the optimal position of the camera station, the optimal attitude angle of the camera station and the optimal density parameter.

Further, the initial distribution is any one of ring shape, column shape, zigzag shape and cross shape.

Further, the simulation is carried out by using the beam adjustment model to obtain the error equation corresponding to each measurement point on the sample to be measured after imaging at each camera station, including:

The error equation corresponding to each measurement point on the sample to be tested after imaging at each camera station is obtained by the following formula:

为摄站姿态角构成的矩阵，由摄站的姿态角(R_x,R_y,R_z)转换得到，X、Y、Z为测量点的三维坐标，X₀、Y₀、Z₀为摄站位置，f为相机焦距，x、y为测量点在摄站成像后所对应的像点的图像坐标，x₀、y₀为像主点坐标，Δx_r、Δx_d、Δx_b、Δy_r、Δy_d、Δy_b为相机的畸变参数。Among them, v _x and v _y are errors,

is the matrix formed by the attitude angle of the camera station, which is obtained by converting the attitude angle of the camera station (R _x , R _y , R _z ), X, Y, and Z are the three-dimensional coordinates of the measurement point, and X ₀ , Y ₀ , and Z ₀ are the camera station position, f is the focal length of the camera, x, y are the image coordinates of the image point corresponding to the measurement point after imaging at the camera station, x ₀ , y ₀ are the principal point coordinates of the image, Δx _r , Δx _d , Δx _b , Δy _r , Δy _d , Δy _b are the distortion parameters of the camera.

Further, the parameters to be optimized include shooting distance, camera pointing, shooting station distribution density and measurement point distribution density.

Further, the initial measurement network type is set according to the current optimal position of the camera station, the current optimal attitude angle of the camera station, and the density parameter, and each parameter to be optimized is sequentially adjusted according to the preset optimization sequence, and iteratively solves the problem. Measurement errors until the measurement errors in the adjustment process of each parameter to be optimized meet the corresponding convergence conditions, and the optimal measurement network type is obtained, including:

Setting an initial measurement network type according to the current optimal position of the camera station, the current optimal attitude angle of the camera station, and the density parameter;

Adjusting the photographing distance, iteratively solving the measurement error until the measurement error meets the first preset convergence condition, and obtaining the first intermediate mesh type;

Adjust the camera pointing, iteratively solve the measurement error, until the measurement error meets the second preset convergence condition, and obtain the second intermediate network type;

Adjusting the distribution density of the photographing stations, iteratively solving the measurement error, until the measurement error meets the third preset convergence condition, and obtaining the third intermediate network type;

The distribution density of the measurement points is adjusted, and the measurement error is iteratively solved until the measurement error satisfies the fourth preset convergence condition, and an optimal measurement network type is obtained.

The beneficial effects of the present invention are: use the beam method adjustment model to simulate, and use the least squares method to solve the error equations corresponding to all measurement points after imaging at each camera station, and obtain the current optimal object space point coordinates and current The optimal network type, the current measurement error is obtained after fitting the current optimal object space point coordinates with the preset theoretical digital simulation, and the parameters to be optimized are adjusted in turn according to the preset optimization order, and the measurement error is iteratively solved until each The measurement errors in the process of parameter adjustment to be optimized all meet the corresponding convergence conditions, and the optimal measurement network type is obtained. In this way, the present invention can determine the most suitable measurement network type according to the actual measurement conditions on site, thereby making the measurement error of the digital photogrammetry system more stable and the measurement accuracy higher.

Description of drawings

In order to illustrate the technical solution of the present invention more clearly, the accompanying drawings that need to be used in the embodiments will be briefly introduced below. Obviously, for those of ordinary skill in the art, on the premise of not paying creative work, they can also Additional figures can be derived from these figures.

FIG. 1 is a schematic flowchart of a method for determining a measurement network type of a digital photogrammetry system provided by an embodiment of the present invention;

Fig. 2 is the schematic diagram of bundle adjustment model;

Fig. 3 is a schematic diagram of adjusting the shooting distance, camera pointing and shooting station distribution density provided by the embodiment of the present invention;

FIG. 4 is a schematic diagram of a measurement network type optimization model iteration method provided by an embodiment of the present invention;

FIG. 5 is a schematic flow chart corresponding to the method for determining the measurement network type of the digital photogrammetry system provided by the embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

Aiming at the deficiencies in the prior art, an embodiment of the present invention provides a method for determining a measurement network type of a digital photogrammetry system. Fig. 1 exemplarily shows a schematic flowchart of a method for determining a measurement network type of a digital photogrammetry system provided by an embodiment of the present invention, as shown in Fig. 1 , specifically including the following steps:

101: According to the on-site measurement conditions, determine the initial distribution of the shooting stations where the cameras are located.

Wherein, the initial station distribution includes the initial position of the station where the camera is located, and the initial attitude angle of the station.

Specifically, the initial distribution can be any one of circular, column, zigzag and cross.

That is to say, the initial camera station distribution should be combined with the on-site measurement conditions, and can be selected from conventional camera station distribution schemes such as circular, row-shaped (airline), meter-shaped, and cross-shaped.

102: Use the beam adjustment model to simulate, and obtain the error equations corresponding to each measurement point on the sample to be measured after imaging at each camera station.

Wherein, each measurement point on the sample to be tested is evenly distributed on the surface of the sample to be tested according to a preset density parameter.

Specifically, the error equation corresponding to each measurement point on the sample to be tested after imaging at each camera station can be obtained by formula (1):

为摄站姿态角构成的矩阵，由摄站的姿态角(R_x，R_y，R_z)转换得到，X、Y、Z为测量点的三维坐标，X₀、Y₀、Z₀为摄站位置，f为焦距，x、y为测量点在摄站成像后所对应的像点的图像坐标，x₀、y₀为像主点坐标，Δx_r、Δx_d、Δx_b、Δy_r、Δy_d、Δy_b为相机的畸变参数。In formula (1), v _x and v _y are errors,

is the matrix formed by the attitude angle of the camera station, which is obtained by converting the attitude angle of the camera station (R _x , R _y , R _z ), X, Y, and Z are the three-dimensional coordinates of the measurement point, and X ₀ , Y ₀ , and Z ₀ are the camera station position, f is the focal length, x, y are the image coordinates of the image point corresponding to the measurement point after imaging at the camera station, x ₀ , y ₀ are the principal point coordinates of the image, Δx _r , Δx _d , Δx _b , Δy _r , Δy _d and Δy _b are the distortion parameters of the camera.

The process of simulating the bundle adjustment model is introduced below.

The principle of the beam adjustment model is shown in Figure 2. Assume that a certain measurement point on the sample to be measured is P, and its three-dimensional coordinates in the global coordinate system are (X, Y, Z), and p is P after imaging at a certain camera station. For the corresponding image point, the image coordinates of p are (x, y), and the station coordinates of the camera include the station position (X ₀ , Y ₀ , Z ₀ ) and the attitude angle (R _x , R _y , R _z ), The coordinates of point P in the camera station coordinate system are (X′, Y′, Z′), and f is the focal length of the camera, then there is the relationship shown in the following formula (2):

In the formula (2), x and y are the image coordinates of the image point p, X', Y' and Z' are the coordinates of point P in the camera station coordinate system, and f is the focal length of the camera.

The coordinates X', Y', and Z' of point P in the camera station coordinate system can be expressed by the following formula (3):

According to formula (2) and formula (3), the following formula (4) can be obtained:

为摄站姿态角构成的矩阵，由摄站的姿态角(R_x，R_y，R_z)转换得到，X、Y、Z为测量点的三维坐标，X₀、Y₀、Z₀为摄站位置，f为相机焦距，x、y为像点p的图像坐标。In formula (3) and formula (4),

is the matrix formed by the attitude angle of the camera station, which is obtained by converting the attitude angle of the camera station (R _x , R _y , R _z ), X, Y, and Z are the three-dimensional coordinates of the measurement point, and X ₀ , Y ₀ , and Z ₀ are the camera station position, f is the focal length of the camera, and x, y are the image coordinates of the image point p.

Considering the position of the principal point of the image and the distortion of the image, the formula (4) is transformed as follows:

为摄站姿态角构成的矩阵，由摄站的姿态角(R_x,R_y,R_z)转换得到，X、Y、Z为测量点的三维坐标，X₀、Y₀、Z₀为摄站位置，f为相机焦距，x、y为像点p的图像坐标，x₀、y₀为像主点坐标，Δx_r、Δx_d、Δx_b、Δy_r、Δy_d、Δy_b为相机的畸变参数。In formula (5),

is the matrix formed by the attitude angle of the camera station, which is obtained by converting the attitude angle of the camera station (R _x , R _y , R _z ), X, Y, and Z are the three-dimensional coordinates of the measurement point, and X ₀ , Y ₀ , and Z ₀ are the camera station position, f is the focal length of the camera, x, y are the image coordinates of the image point p, x ₀ , y ₀ are the coordinates of the principal point of the image, Δx _r , Δx _d , Δx _b , Δy _r , Δy _d , Δy _b are the camera’s Distortion parameters.

Among them, the radial distortion correction amount formula can be expressed by formula (6):

The formula of tangential distortion correction amount can be expressed by formula (7):

The image plane correction formula can be expressed by formula (8):

In addition, some parameters in formula (6), formula (7) and formula (8) are calculated by formula (9):

In formula (6), formula (7), formula (8) and formula (9), Δx _r , Δy _r are radial distortion correction amounts, K ₁ , K ₂ , K ₃ are radial distortion parameters, x, y is the image coordinate of the image point p, x ₀ and y ₀ are the principal point coordinates of the image, Δx _d , Δy _d are the tangential distortion correction values, P ₁ and P ₂ are the tangential distortion parameters, Δx _b , Δy _b are the image planes Correction amount, b ₁ and b ₂ are image plane distortion parameters.

In this way, the error equation shown in formula (1) can be obtained according to formula (5).

103: Use the least square method to solve the error equations corresponding to all measurement points after imaging at each camera station, and obtain the current optimal three-dimensional coordinates of the object space point, as well as the current optimal position of the camera station and the current optimal attitude angle of the camera station.

Wherein, the current optimal three-dimensional coordinates of the object space point include the calculated three-dimensional coordinates of all measurement points on the sample to be measured.

Specifically, for any camera station, the beam adjustment method is used to simulate and establish the error equation of any measurement point on the sample to be measured at the camera station.

Then, according to the error equations of all measurement points at each camera station, the current optimal position and current optimal attitude angle of each camera station can be obtained by using the least square method, as well as the three-dimensional coordinates of each measurement point.

104: Fitting the current optimal three-dimensional coordinates of the object space point with the preset theoretical numerical model to obtain the current measurement error.

Among them, the sample to be tested is processed according to the theoretical digital model.

105: Set the initial measurement network type according to the current optimal position of the camera station, the current optimal attitude angle of the camera station, and the density parameters, adjust each parameter to be optimized in sequence according to the preset optimization sequence, and iteratively solve the measurement error until each parameter to be optimized is adjusted The measurement errors in the process all meet the corresponding convergence conditions, and the optimal measurement network type is obtained.

Among them, the optimal measurement network type includes the optimal position of the camera station, the optimal attitude angle of the camera station and the optimal density parameters.

Specifically, the parameters to be optimized include photographing distance, camera pointing, camera station distribution density and measurement point distribution density.

Fig. 3 exemplarily shows a schematic diagram of adjustment of photographing distance, camera pointing and photographing station distribution density provided by the embodiment of the present invention. Adjustment, different camera station distribution density means that the distance between camera stations has been adjusted, and different camera orientations means that the camera attitude angle of each camera station has been adjusted.

The optimization order can be set as photographing distance, camera pointing, camera station distribution density and measurement point distribution density from first to last.

Further, the optimal measurement network type can be obtained through the following steps:

Step 1. Set the initial measurement network type according to the current optimal position of the camera station, the current optimal attitude angle of the camera station, and the density parameters.

Step 2: Adjust the photographing distance, iteratively solve the measurement error until the measurement error meets the first preset convergence condition, and obtain the first intermediate mesh type.

Wherein, the photography distance can be reflected by the location coordinates of the photography station.

Specifically, during adjustment, the adjustment may be performed according to a preset parameter increment.

The first preset convergence condition may be that the measurement accuracy reaches a peak value within a preset parameter threshold range, that is, it is considered to be converged.

Step 3, adjust the camera pointing, iteratively solve the measurement error until the measurement error meets the second preset convergence condition, and obtain the second intermediate mesh type.

Specifically, during adjustment, the adjustment may be performed according to a preset parameter increment.

The second preset convergence condition may be that the measurement accuracy reaches a peak value within a preset parameter threshold range, that is, the convergence is considered.

Step 4, adjust the distribution density of the camera stations, and iteratively solve the measurement error until the measurement error satisfies the third preset convergence condition, and obtain the third intermediate network type.

Wherein, the distribution density of the camera stations can also be reflected by the location coordinates of the camera stations.

Specifically, during adjustment, the adjustment may be performed according to a preset parameter increment.

The third preset convergence condition may be that the ratio between the measurement accuracy increment and the parameter increment is smaller than the design value within the preset parameter threshold range. The design value can be determined according to the requirements of measurement accuracy and the conditions of on-site measurement.

Step 5, adjusting the distribution density of measurement points, iteratively solving the measurement error until the measurement error meets the fourth preset convergence condition, and obtaining the optimal measurement network type.

Specifically, during adjustment, the adjustment may be performed according to a preset parameter increment.

The fourth preset convergence condition may be that within a preset parameter threshold range, the ratio between the measurement accuracy increment and the parameter increment is smaller than the design value. The design value can be determined according to the requirements of measurement accuracy and the conditions of on-site measurement.

In addition, other optimization sequences may also be adopted to adjust each parameter to be optimized, which is not specifically limited.

Fig. 4 exemplarily shows a schematic diagram of the measurement network type optimization model iteration method provided by the embodiment of the present invention. As shown in Fig. 4, the optimization iteration order provided by the embodiment of the present invention is photographing distance, camera pointing, photographing station distribution density, The distribution density of measurement points, the iterative convergence conditions corresponding to the camera distance adjustment and camera pointing adjustment are all set parameter thresholds, and the peak accuracy is measured within the threshold range, which is considered to be converged. The iterative convergence conditions corresponding to the adjustment of the distribution density of camera stations and the adjustment of the distribution density of measurement points are to set the parameter threshold, whether the ratio between the measurement accuracy increment and the parameter increment is less than the design value within the threshold range, and the design value can be Determined according to measurement accuracy requirements and site conditions.

In order to more clearly illustrate the measurement network type determination method provided by the embodiment of the present invention, Fig. 5 exemplarily shows the specific flow chart corresponding to the measurement network type determination method of the digital photogrammetry system provided by the embodiment of the present invention, as shown in Fig. 5 As shown, the specific process is: complete the benchmark layout of the measurement network, determine the station distribution according to the boundary of the field measurement conditions, and determine the image point error according to the station distribution and measurement point distribution, and adjust the group error of the imaging model according to the beam method Equation, and use the least squares iterative solution, and then use ICP iterative calculation surface error calculation and data statistics to obtain the measurement error of the imaging model and the sample model, and judge whether it meets the convergence conditions of the iterative model. If not, continue to adjust the layout Parameters such as method, measurement point distribution, and camera station distribution, until the measurement error meets the convergence condition of the iterative model, and the optimal shooting network type is output.

The beneficial effects of the present invention are: use the beam method adjustment model to simulate, and use the least squares method to solve the error equations corresponding to all measurement points after imaging at each camera station, and obtain the current optimal object space point coordinates and current The optimal network type, the current measurement error is obtained after fitting the current optimal object space point coordinates with the preset theoretical digital simulation, and the parameters to be optimized are adjusted in turn according to the preset optimization order, and the measurement error is iteratively solved until each The measurement errors in the process of parameter adjustment to be optimized all meet the corresponding convergence conditions, and the optimal measurement network type is obtained. In this way, the present invention can determine the most suitable measurement network type according to the actual measurement conditions on site, thereby making the measurement error of the digital photogrammetry system more stable and the measurement accuracy higher.

The present invention has been described in detail above in conjunction with specific implementations and exemplary examples, but these descriptions should not be construed as limiting the present invention. Those skilled in the art understand that without departing from the spirit and scope of the present invention, various equivalent replacements, modifications or improvements can be made to the technical solutions and implementations of the present invention, all of which fall within the scope of the present invention. The protection scope of the present invention shall be determined by the appended claims.

Claims (2)

1. A method for determining the measurement network type of a digital photogrammetry system, characterized in that it comprises: According to the on-site measurement conditions, determine the initial distribution of the station where the camera is located, the initial station distribution includes the initial position of the station where the camera is located, and the initial attitude angle of the station; The beam adjustment model is used for simulation to obtain the error equation corresponding to each measurement point on the sample to be measured after imaging at each camera station, wherein each measurement point on the sample to be measured is according to the preset density The parameters are evenly distributed on the surface of the sample to be tested; Use the least square method to solve the error equations corresponding to all measurement points after imaging at each camera station, and obtain the current optimal three-dimensional coordinates of the object space point, as well as the current optimal position of the camera station and the current optimal attitude angle of the camera station; Fitting the current optimal three-dimensional coordinates of the object space point with the preset theoretical digital simulation to obtain the current measurement error; According to the current optimal position of the camera station, the current optimal attitude angle of the camera station and the density parameter, the initial measurement network type is set, and each parameter to be optimized is adjusted in turn according to the preset optimization sequence, and the measurement error is solved iteratively until each The measurement errors in the adjustment process of the parameters to be optimized all meet the corresponding convergence conditions, and the optimal measurement network type is obtained, and the optimal measurement network type includes the optimal position of the camera station, the optimal attitude angle of the camera station and the optimal density parameter; Wherein, the simulation is carried out by using the beam adjustment model to obtain the error equation corresponding to each measurement point on the sample to be measured after imaging at each camera station, including: The error equation corresponding to each measurement point on the sample to be tested after imaging at each camera station is obtained by the following formula:

Among them, v _x and v _y are errors,

is the matrix formed by the attitude angle of the camera station, which is obtained by transforming the attitude angles R _x , R _y , R _z of the camera station, X, Y, and Z are the three-dimensional coordinates of the measurement point, and X ₀ , Y ₀ , and Z ₀ are the positions of the camera station , f is the focal length of the camera, x, y are the image coordinates of the image point corresponding to the measurement point after imaging at the camera station, x ₀ , y ₀ are the coordinates of the principal point of the image, Δx _r , Δx _d , Δx _b , Δy _r , Δy _d , Δy _b is the distortion parameter of the camera; The parameters to be optimized include photographing distance, camera pointing, photographing station distribution density and measurement point distribution density; The initial measurement network type is set according to the current optimal position of the camera station, the current optimal attitude angle of the camera station, and the density parameter, and each parameter to be optimized is sequentially adjusted according to a preset optimization sequence, and the measurement error is iteratively solved, Until the measurement errors in the adjustment process of each parameter to be optimized meet the corresponding convergence conditions, the optimal measurement network type is obtained, including: Setting an initial measurement network type according to the current optimal position of the camera station, the current optimal attitude angle of the camera station, and the density parameter; Adjusting the photographing distance, iteratively solving the measurement error until the measurement error meets the first preset convergence condition, and obtaining the first intermediate mesh type; Adjust the camera pointing, iteratively solve the measurement error, until the measurement error meets the second preset convergence condition, and obtain the second intermediate network type; Adjusting the distribution density of the photographing stations, iteratively solving the measurement error, until the measurement error meets the third preset convergence condition, and obtaining the third intermediate network type; The distribution density of the measurement points is adjusted, and the measurement error is iteratively solved until the measurement error satisfies the fourth preset convergence condition, and an optimal measurement network type is obtained. 2. The method according to claim 1, characterized in that, the initial distribution is any one of a ring shape, a column shape, a zigzag shape and a cross shape.

Images