A target positioning method based on dual-camera space front intersection

Abstract

The application relates to a target positioning method based on a two-camera space forward intersection, which comprises the following steps: constructing a correlation of a regional object space coordinate system and a camera observation coordinate system; solving a target unit vector in the camera observation coordinate system; and calculating a target position coordinate in the regional object space coordinate system. The target positioning method does not need to use high-cost professional camera equipment, can realize target positioning based on a space forward intersection principle and a coordinate linear transformation by using an ordinary camera with known imaging parameters and an observation pose, can flexibly set an observation baseline, and is more suitable for high-precision positioning in a large space range scene. The target positioning method based on the two-camera space forward intersection does not need to establish a complicated coordinate system, a target position coordinate calculation model is constructed by a base coordinate linear transformation method, the imaging state can be described, the method can be widely applied to different scenes, and the accurate target position in the regional object space coordinate system can be obtained.

Description

Technical Field

[0001] This invention belongs to the field of visual positioning technology, specifically relating to a target positioning method based on the spatial forward intersection of two cameras. Background Technology

[0002] Currently, location services are widely used in people's daily lives, such as determining ground location based on GNSS (Global Navigation Satellite System); cellular positioning methods based on mobile communication networks; and WiFi positioning methods based on WiFi signals. However, these methods often rely on specialized equipment and signal characteristics, resulting in high costs and signal transmission being easily limited by the environment. With the development of computer technology, visual positioning methods have emerged that use cameras and image processing technology to determine the spatial location of targets. This method does not require receiving external signals, acquires rich information, and is more flexible and practical. Depending on the number of sensors required, visual positioning technology is divided into monocular, binocular, or multi-view based methods. Binocular positioning methods determine the object space coordinates of a target by using the interior and exterior orientation elements of two cameras and the image coordinate measurements of corresponding image points, and are one of the research hotspots in the field of visual positioning.

[0003] The existing binocular positioning methods mainly have the following problems: (1) Most of them use professional binocular positioning cameras. Due to the short baseline of these binocular cameras, the observation range and positioning accuracy are limited; (2) Most methods avoid the rigorous theoretical derivation of the geometric imaging principle. They generally use regression analysis based on multiple known points to solve the relationship between the spatial coordinate system, the camera observation coordinate system and the target, resulting in a large amount of preliminary work and limited positioning accuracy; (3) Collinearity is an important theoretical basis for various photogrammetric positioning methods. It generally requires the construction of a large number of two-dimensional or three-dimensional spatial coordinate systems, which is a bit complicated; (4) Rational function models can be used to establish the relationship between the image point and the target spatial coordinates, but many parameters in the model do not have clear physical meaning and are difficult to explain reasonably; (5) There is a lack of a positioning method that can be applied to ordinary cameras, has an unrestricted observation baseline, can accurately and rigorously derive the geometric imaging mechanism, and is more generalizable. Summary of the Invention

[0004] The purpose of this invention is to provide a target spatial coordinate calculation method based on the forward intersection principle, which utilizes two ordinary cameras to simultaneously observe the target and the known pose of the two cameras. This method is based on the geometric imaging principle and linear algebra theory to rigorously derive the relationship between the spatial coordinate system, the camera observation coordinate system, and the target, in order to solve a series of problems such as the limited positioning accuracy of dual-camera observation and the large workload of regression calculation using known points.

[0005] The objective of this invention is achieved through the following technical solution:

[0006] A target localization method based on dual-camera spatial forward intersection includes the following steps:

[0007] S1. Construction of the association between the regional object space coordinate system and the camera observation coordinate system:

[0008] S11. Establishment of the regional object space coordinate system;

[0009] S12. Acquisition of optical center coordinates of dual cameras;

[0010] S13. Establishment of the camera observation coordinate system;

[0011] S14. Dual-camera observation pose determination;

[0012] S15. Obtain the coordinates of the target image point;

[0013] S16. Determination of target vector pose deviation coefficient;

[0014] S2. Solving for the target unit vector in the camera observation coordinate system;

[0015] S3. Calculation of target position coordinates in the regional object space coordinate system;

[0016] S31. Solving for the target unit vector in the equivalent coordinate system of the regional object space;

[0017] S32. Construction of the target position coordinate solution model in the regional object space coordinate system.

[0018] Further, step S11 specifically involves: establishing a regional object space coordinate system with the positive directions of due east, due north, and zenith as the X, Y, and Z axes, respectively, and using unit vectors originating from the origin and pointing to the positive directions of the X, Y, and Z axes as the basis α1, α2, and α3 of the regional object space coordinate system.

[0019] Further, step S12 specifically involves: setting the camera position according to the positioning observation requirements, and recording the coordinates x of the optical centers of cameras A and B in the regional object space coordinate system. A x B .

[0020] Further, step S13 specifically involves: establishing a camera observation coordinate system with the optical centers of cameras A and B as the origin, the rightward direction of the camera when it is normally stationary as the positive X-axis, the camera's line of sight as the positive Y-axis, and the upward direction as the positive Z-axis, and using unit vectors originating from the camera's optical center and pointing to the positive X, Y, and Z axes as the basis vectors β1, β2, and β3 of the camera observation coordinate system.

[0021] Further, step S14 specifically involves: determining the camera's observation azimuth angle according to the positioning observation requirements. Set the pitch angle ω, and record the azimuth angles of cameras A and B in the regional object space coordinate system. and pitch angle ω A ω B The dual-camera observation pose is determined by the association between the regional object space coordinate system and the camera observation coordinate system; the linear transformation matrix D under basis α1, α2, α3 is the transition matrix between basis α1, α2, α3 and basis β1, β2, β3.

[0022] Further, step S15 specifically involves: based on the dual-camera pose determination, using the dual cameras to take pictures of the target to be located and acquire images, establishing a plane coordinate system for the image with the center of the image as the origin, the horizontal direction to the right as the positive direction of the horizontal axis, and the vertical direction downward as the positive direction of the vertical axis; and reading the horizontal coordinate u and vertical coordinate v of the target image point in the image.

[0023] Further, step S16 specifically involves: acquiring camera imaging observation parameters, including the horizontal field of view (FOV). r Vertical field of view (FOV) c Horizontal pixel count N of the image r and vertical pixel count N c ; and calculate the pose deviation coefficients of the target vector in the camera observation coordinate system, base β1, β2, β3, by using the target image point coordinates, including the horizontal deflection angle δ and the vertical deflection angle θ.

[0024] Further, step S2 specifically involves: the state of the target unit vector in the camera observation coordinate system basis β1, β2, β3 is determined by the deviation coefficient of the target vector, the horizontal deflection angle δ, and the vertical deflection angle θ. That is, the target unit vector is linearly transformed from β2, corresponding to the linear transformation matrix S under basis β1, β2, β3, which can be decomposed into: first, rotating around the Z-axis by an angle δ under basis β1, β2, β3, corresponding to the linear transformation matrix S1; then rotating around the X-axis of the coordinate system obtained from the previous step by an angle θ, corresponding to the linear transformation matrix S2.

[0025] Further, step S31 specifically involves: in the equivalent coordinate system of the regional object space, the state of the target unit vector is determined by the correlation between the equivalent coordinate system of the regional object space, the camera observation coordinate system, the target unit vector, and the vector s2 under basis β1, β2, and β3.

[0026] From the linear algebraic basis coordinates and transformation formulas, we know that in the equivalent coordinate system of the regional object space, in basis α1, α2, α3, the coordinates m of the target unit vector corresponding to the vector s2 under basis β1, β2, β3 are:

[0027] m=Ds2

[0028] For dual cameras, the corresponding target unit vector coordinates m in the two equivalent coordinate systems of the object space of the two regions are as follows:

[0029]

[0030] Further, step S32 specifically includes:

[0031] The equivalent coordinate system of the regional object space and the regional object space coordinate system are related by translation. Therefore, the target unit vector in the equivalent coordinate system corresponds to the target vector x in the regional object space coordinate system. p The solution model is as follows:

[0032]

[0033] In the formula,

[0034] x A x B Let A and B be the coordinates of cameras A and B in the regional object space coordinate system;

[0035]

[0036]

[0037] m A m B Let be the target unit vector coordinates in the equivalent coordinate system of the object space of the region with the optical centers of cameras A and B as the origins;

[0038]

[0039]

[0040] k A k B The magnitude of the target vector (the vector pointing from the optical centers of cameras A and B to the target);

[0041] Therefore, the magnitude of the target vector can be obtained, and the object space coordinates of the target can be further obtained.

[0042] Compared with the prior art, the beneficial effects of the present invention are:

[0043] This invention eliminates the need for expensive professional camera equipment. Utilizing a standard camera with known imaging parameters and observation pose, it achieves target localization based on the principle of spatial forward intersection and coordinate linear transformation. It allows for flexible setting of the observation baseline, making it more suitable for high-precision localization in large spatial scenarios. This method, based on dual-camera spatial forward intersection for target localization, eliminates the need for complex coordinate system establishment. Through a base coordinate linear transformation method, it constructs a target position coordinate calculation model, which can describe the imaging state and is widely applicable to various scenarios, enabling accurate acquisition of the target position in the regional object space coordinate system. Attached Figure Description

[0044] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 Flowchart of a target localization method based on dual-camera spatial forward intersection;

[0046] Figure 2 This application embodiment illustrates the relationship between the regional object space coordinate system and the camera observation coordinate system.

[0047] Figure 3 This application provides a schematic diagram of the target vector pose deviation in its embodiments. Detailed Implementation

[0048] The present invention will be further described below with reference to embodiments:

[0049] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the present invention, and not all of the structures.

[0050] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0051] Example 1

[0052] according to Figure 1 The flowchart shown illustrates target localization calculation based on dual-camera spatial forward intersection. Figure 1As shown, this invention relates to a target localization method based on dual-camera spatial forward intersection. It determines the state information of the vector connecting the optical centers of the two cameras and the target by linearly correlating the regional object space coordinate system, the camera observation coordinate system, and the target image point offset coefficient. The target spatial coordinates are then calculated based on the principle of linear algebraic base coordinate transformation. The specific steps are as follows:

[0053] S1. Construction of the association between the regional object space coordinate system and the camera observation coordinate system

[0054] S11. Establishment of the Regional Object Space Coordinate System: Establish the regional object space coordinate system with the positive directions of east, north, and zenith as the X, Y, and Z axes, respectively (the origin of the coordinate system can be customized or set to the origin of the geographic coordinate system). Use unit vectors originating from the origin and pointing to the positive directions of the X, Y, and Z axes as the bases α1, α2, and α3 of the regional object space coordinate system (for ease of analysis, the regional object space coordinate system will be appropriately translated in the following discussion; the coordinate system whose origin coincides with the camera's optical center is called the equivalent regional object space coordinate system; to avoid too many symbols, the corresponding bases are still denoted as α1, α2, and α3).

[0055] S12. Acquisition of optical center coordinates for dual cameras: Set the camera positions according to the needs of positioning observation, and record the coordinates (x, y) of the optical centers of cameras A and B in the regional object space coordinate system. A x B The camera baseline (the straight-line distance between two cameras) can be flexibly determined based on the needs of visual observation.

[0056] S13. Establishment of the camera observation coordinate system: The camera observation coordinate system is established with the optical centers of cameras A and B as the origin, the rightward direction of the camera (when normally stationary) as the positive X-axis, the camera's line of sight (principal optical axis) as the positive Y-axis, and the upward direction as the positive Z-axis. Unit vectors originating from the camera's optical center and pointing towards the positive X, Y, and Z axes are used as the basis vectors β1, β2, and β3 of the camera observation coordinate system.

[0057] S14. Dual-camera observation pose determination: Determine the camera's observation azimuth angle according to the positioning observation requirements. Set the pitch angle ω, and record the azimuth angles of cameras A and B in the regional object space coordinate system. and pitch angle ω A ω B The dual-camera observation pose is determined by associating the regional object space coordinate system with the camera observation coordinate system, such as... Figure 2 As shown.

[0058] Among them, azimuth angle The angle ω is the angle between the projection of the camera's line of sight (principal optical axis) onto the horizontal plane and the due north direction. Starting from due north as 0°, clockwise rotation (view from zenith downwards) is positive and counterclockwise rotation is negative, with a value range of -360° to +360°. The pitch angle ω is the angle between the camera's line of sight (principal optical axis) and the horizontal plane. Starting from the horizontal direction as 0°, downward tilt is positive and upward tilt is negative, with a value range of -90° to +90°.

[0059] The linear transformation matrix D (under basis α1, α2, α3) is the transition matrix between basis α1, α2, α3 and basis β1, β2, β3, i.e.:

[0060] (β1,β2,β3)=(α1,α2,α3)D

[0061] The three unit column vectors that make up matrix D represent the states of the three coordinate axes (bases β1, β2, β3) of the camera observation coordinate system in the regional object space coordinate system (bases α1, α2, α3). Matrix D is called the camera pose description matrix, representing the relationship between the regional coordinate system and the camera observation coordinate system. It can be decomposed into two steps: (1) Rotate the camera around the Z-axis using bases α1, α2, α3. (1) Corresponding to the linear transformation matrix D1 (under basis α1, α2, α3); (2) Rotation angle ω around the X-axis under basis β1, β2, β3, corresponding to the linear transformation matrix D2 (under basis β1, β2, β3).

[0062] in,

[0063]

[0064] The linear transformation D2 under β1, β2, β3 is a transformation based on the linear transformation D1 under basis α1, α2, α3. Therefore, according to the relationship between linear transformation matrices under different basis in linear algebra, the linear transformation matrix D3 under basis α1, α2, α3 corresponding to D2 is:

[0065] D3 = D1D2D1 -1

[0066] Therefore, the linear transformation matrix D under bases α1, α2, and α3 is:

[0067] D = D3D1 = D1D2D1 -1 D1=D1D2…………(1)

[0068] Right now:

[0069]

[0070] Among them, D1, D2, and D are all orthogonal matrices.

[0071] In this embodiment,

[0072]

[0073]

[0074] S15. Target Image Point Coordinate Acquisition: Based on the determined pose of the dual cameras, use both cameras to take pictures of the target to be located and acquire images (Note: To ensure positioning accuracy, the three points of cameras A, B, and the target must avoid being collinear, i.e., the angle between the line of sight of cameras A and B to the target should be approximately 90°). Establish an image plane coordinate system with the center of the image as the origin, the horizontal direction to the right as the positive x-axis, and the vertical direction downward as the positive y-axis; read the horizontal coordinate u and vertical coordinate v of the target image point in the image.

[0075] S16. Determination of target vector pose deviation coefficient: Obtain camera imaging observation parameters—horizontal field of view (FOV) r Vertical field of view (FOV) c Horizontal pixel count N of the image r Vertical pixel count N c ; and calculate the pose deviation coefficients of the target vector (the vector pointing from the camera optical center to the target) in the camera observation coordinate system (base β1, β2, β3) using the target image point coordinates—horizontal deflection angle δ and vertical deflection angle θ.

[0076] Wherein, the horizontal deflection angle δ is the angle between the projection of the target vector onto the XOY plane in the camera observation coordinate system (base β1, β2, β3) and the camera's line of sight (when looking down at the XOY plane of the camera observation coordinate system, clockwise rotation is positive and counterclockwise rotation is negative, and its value range is -FOV). r / 2~+FOV r / 2), the vertical deflection angle θ is the angle between the target vector and the XOY plane in the camera observation coordinate system (base β1, β2, β3) (starting from 0° on the XOY plane, rotation in the reverse direction of the Z-axis is positive, and rotation in the positive direction of the Z-axis is negative, and its value range is -FOV). c / 2~+FOV c / 2).

[0077] According to the principles of photogrammetry, the field of view (FOV), focal length (f), and number of pixels (N) have the following relationship:

[0078]

[0079] Similarly, the horizontal deflection angle δ and vertical deflection angle θ of the target vector can be calculated based on trigonometric functions from camera imaging parameters such as target image point coordinates, field of view, and number of imaging pixels.

[0080] like Figure 3 As shown, in this embodiment, camera A has: but

[0081]

[0082] and

[0083]

[0084] so

[0085]

[0086] In this embodiment, let And because

[0087]

[0088] so

[0089]

[0090] That is

[0091]

[0092] Similarly, the pose deviation parameters of the target vector in the camera B observation coordinate system can be obtained as follows:

[0093]

[0094] S2. Solving for the target unit vector in the camera observation coordinate system

[0095] The state of the target unit vector (the unit vector pointing from the camera's optical center to the target) in the camera's observation coordinate system (base β1, β2, β3) is determined by the deviation coefficients of the target vector (horizontal deflection angle δ, vertical deflection angle θ). That is, the target unit vector is linearly transformed from β2, corresponding to the linear transformation matrix S (base β1, β2, β3). It can be decomposed into two steps: (1) Rotate the target vector around the Z-axis by an angle δ under base β1, β2, β3, corresponding to the linear transformation matrix S1 (base β1, β2, β3); (2) Rotate the target vector around the X-axis of the coordinate system obtained from the previous step by an angle θ, corresponding to the linear transformation matrix S2.

[0096] Similarly to the camera pose description matrix, we can obtain:

[0097] The linear transformation matrix S under basis β1, β2, β3 is:

[0098] S = S3S1 = S1S2S1 -1 S1=S1S2…………(4)

[0099] Right now:

[0100]

[0101] The three unit column vectors that make up matrix S represent the linear transformation results of the three coordinate axes (basic β1, β2, β3) of the camera observation coordinate system, and its second column vector s2 is the target unit vector. Figure 3 ).

[0102]

[0103] In this embodiment,

[0104]

[0105] S3. Calculation of target position coordinates in the regional object space coordinate system

[0106] S31. Solving for the target unit vector in the equivalent coordinate system of the regional object space: In the equivalent coordinate system of the regional object space, the state of the target unit vector is determined by the relationship between the equivalent coordinate system of the regional object space (basis α1, α2, α3), the camera observation coordinate system (basis β1, β2, β3), and the target unit vector (vector s2 under basis β1, β2, β3).

[0107] From the linear algebraic basis coordinates and transformation formulas, we know that in the equivalent coordinate system (basis α1, α2, α3) of the regional object space, the target unit vector coordinate m corresponding to the vector s2 under basis β1, β2, β3 is:

[0108] m=Ds2…………(6)

[0109] For dual cameras, the corresponding target unit vector coordinates m in the two equivalent coordinate systems of the object space of the two regions are as follows:

[0110]

[0111] S32. Construction of the target position coordinate solution model in the regional object space coordinate system: The equivalent coordinate system of the regional object space is a translation relationship with the regional object space coordinate system. Therefore, the target unit vector in the equivalent coordinate system corresponds to the target vector x in the regional object space coordinate system. p The solution model is as follows:

[0112]

[0113] In the formula,

[0114] x A x B Let A and B be the coordinates of cameras A and B in the regional object space coordinate system;

[0115]

[0116]

[0117] m A m B Let be the target unit vector coordinates in the equivalent coordinate system of the object space of the region with the optical centers of cameras A and B as the origins;

[0118]

[0119]

[0120] k A k B The magnitude of the target vector (the vector pointing from the optical centers of cameras A and B to the target).

[0121] From equation (8), we can obtain: -k A m A +k B m B =x A -x B ,Right now:

[0122]

[0123] Let Q = (-m) A m B The above equation can be simplified to:

[0124]

[0125] Solving the matrix equation yields:

[0126]

[0127] Substituting the modulus back into equation (8), we can obtain the object space coordinates.

[0128] This invention utilizes the azimuth and elevation angle parameters of camera observations to construct a correlation between the regional object space coordinate system and the camera observation coordinate system. Based on the target image point coordinates, it solves for the target vector pose deviation coefficient, obtaining the target unit vector state in the camera observation coordinate system. Then, using a base coordinate linear transformation method, it constructs a target position coordinate calculation model. The model parameters can provide qualitative and quantitative explanations of the imaging principle, achieving accurate acquisition of the target position in the regional object space coordinate system. This effectively solves the positioning accuracy problems caused by the limited length of the dual-camera observation baseline and the avoidance of rigorous derivation of geometric imaging principles in traditional positioning methods. It eliminates the need for preliminary work involving regression calculations based on known points, and positioning can be achieved using ordinary cameras with known imaging parameters and observation poses. It does not require high-cost professional camera equipment, and the observation baseline can be flexibly set, making it suitable for high-precision positioning in scenarios with a larger spatial range.

[0129] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. A target localization method based on dual-camera spatial forward intersection, characterized in that, Includes the following steps: S1. Construction of the association between the regional object space coordinate system and the camera observation coordinate system: S11. Establishment of the regional object space coordinate system; S12. Acquisition of optical center coordinates of dual cameras; S13. Establishment of the camera observation coordinate system; S14. Dual-camera observation pose determination; S15. Obtain the coordinates of the target image point; S16. Determination of target vector pose deviation coefficient; S2. Solving for the target unit vector in the camera observation coordinate system; S3. Calculation of target position coordinates in the regional object space coordinate system; S31. Solving for the target unit vector in the equivalent coordinate system of the regional object space; S32. Construction of the target position coordinate solution model in the regional object space coordinate system; Step S14 specifically involves: determining the camera's observation azimuth angle according to the positioning observation requirements. Pitch angle Set and record the azimuth angles of cameras A and B in the regional object space coordinate system, respectively. and pitch angle , The dual-camera observation pose is determined by the association between the regional object space coordinate system and the camera observation coordinate system; the linear transformation matrix D under basis α1, α2, α3 is the transition matrix between basis α1, α2, α3 and basis β1, β2, β3; Step S2 specifically involves: the state of the target unit vector in the camera observation coordinate system basis β1, β2, β3 is determined by the deviation coefficient of the target vector and the horizontal deflection angle. Vertical deflection angle The target unit vector is determined by a linear transformation of β2, corresponding to the linear transformation matrix S under basis β1, β2, and β3. This transformation matrix S can be decomposed into: a rotation angle around the Z-axis under basis β1, β2, and β3. This corresponds to the linear transformation matrix S1; then rotate the coordinate system around the X-axis of the coordinate system obtained from the previous transformation by an angle. , which corresponds to the linear transformation matrix S2; Step S31 is specifically as follows: In the equivalent coordinate system of the regional object space, the state of the target unit vector is determined by the correlation of the equivalent coordinate system of the regional object space, the camera observation coordinate system, the target unit vector, and the vector s2 under basis β1, β2, and β3. From the linear algebraic basis coordinates and transformation formulas, we know that in the equivalent coordinate system of the regional object space, in the basis α1, α2, α3, the target unit vector coordinates corresponding to the vector s2 under the basis β1, β2, β3 are... m for: ； For dual cameras, the corresponding target unit vector coordinates in the two equivalent coordinate systems of the object space of the two regions m They are respectively: ； Step S32 is as follows: The equivalent coordinate system of the regional object space and the regional object space coordinate system are related by translation. Therefore, the target unit vector in the equivalent coordinate system corresponds to the target vector in the regional object space coordinate system. x p The solution model is as follows: In the formula, Let A and B be the coordinates of cameras A and B in the regional object space coordinate system; ； Let be the target unit vector coordinates in the equivalent coordinate system of the object space of the region with the optical centers of cameras A and B as the origins; ； Let be the target vector, the magnitude of the vector pointing from the optical centers of cameras A and B to the target; Therefore, the magnitude of the target vector can be obtained, and the object space coordinates of the target can be further obtained. In the formula, α1, α2, and α3 are the basis of the regional object space coordinate system, and β1, β2, and β3 are the basis of the camera observation coordinate system.

2. The target localization method based on dual-camera spatial forward intersection according to claim 1, characterized in that, Step S11 is as follows: Establish a regional object space coordinate system with the positive directions of due east, due north, and zenith as the X, Y, and Z axes, respectively, and use the unit vectors starting from the origin and pointing to the positive directions of the X, Y, and Z axes as the basis α1, α2, and α3 of the regional object space coordinate system.

3. The target localization method based on dual-camera spatial forward intersection according to claim 1, characterized in that, Step S12 specifically involves: setting the camera position according to the positioning observation requirements, and recording the coordinates of the optical centers of cameras A and B in the regional object space coordinate system. .

4. The target localization method based on dual-camera spatial forward intersection according to claim 1, characterized in that, Step S13 is as follows: Take the optical centers of cameras A and B as the origin, and take the rightward direction of the camera when it is normally stationary as the positive X-axis, the camera's line of sight as the positive Y-axis, and the upward direction as the positive Z-axis. Take the unit vectors that start from the optical center of the camera and point to the positive X, Y, and Z axes as the basis β1, β2, and β3 of the camera observation coordinate system.

5. The target localization method based on dual-camera spatial forward intersection according to claim 1, characterized in that, Step S15 specifically involves: based on the determined pose of the dual cameras, using the dual cameras to capture images of the target to be located, establishing a plane coordinate system for the images with the center of the image as the origin, the horizontal direction to the right as the positive x-axis, and the vertical direction downward as the positive y-axis; and reading the horizontal coordinates of the target image points in the images. and vertical coordinates .

6. The target localization method based on dual-camera spatial forward intersection according to claim 1, characterized in that, Step S16 specifically involves: acquiring camera imaging observation parameters, including the horizontal field of view (FOV). r Vertical field of view (FOV) c Horizontal pixel count N of the image r and vertical pixel count N c And by using the target image point coordinates, the pose deviation coefficients of the target vector in the camera observation coordinate system, bases β1, β2, and β3, including horizontal and vertical deflection angles, are calculated.

Images