A virtual system-based rotating double-prism target tracking system and method

Abstract

The application provides a rotating double-prism target tracking system and method based on a virtual system, the target tracking system comprising an image acquisition module, an image processing module and a central control module. The steps of the target tracking method comprise: constructing a virtual system based on the vector refraction law, the virtual system mapping the actual pointing direction of the exit light beam of the rotating double-prism system; according to the target image information transmitted by the image processing module, the central control module iteratively estimates the prism rotation angle in the virtual system based on the sine-cosine algorithm; if the estimated error between the exit light beam pointing direction of the virtual rotating double-prism system mapped by the estimated rotation angle of the prism and the target to be tracked is less than the actual error, the first prism and the second prism are driven to rotate to the corresponding estimated rotation angle; and when the actual error between the center of the camera field of view and the center of the target to be tracked is less than the set deviation threshold, the target tracking is completed.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of photoelectric tracking, and particularly relates to a rotating double-prism target tracking system and method based on a virtual system. BACKGROUND

[0002] The rotating double-prism system adjusts the imaging visual axis direction through the coaxial independent rotation of two wedge-shaped prisms. Compared with traditional beam pointing devices such as gimbals and mirrors, the rotating double-prism has the advantages of compact structure, high pointing accuracy, good dynamic characteristics, and high reliability, and has a wide range of applications in space observation, search and rescue, microscopic observation, and other fields.

[0003] In the rotating double-prism system, target tracking refers to the technology of adjusting the prism rotation angle in real time based on the position information of the target, so that the target is always locked at the center of the camera imaging field of view. The existing technology (Li A, Zhong S, Liu X, et al. Double-wedge prism imaging tracking device based on the adaptive boresight adjustment principle[J]. Review of Scientific Instruments, 2019, 90(2): 025107.) proposes an adaptive double-prism visual axis adjustment scheme based on a radial-circumferential decoupling control strategy. This scheme essentially determines the rotation direction and rotation angle of the prism in a trial-and-error manner. This scheme does not require calibration of the initial rotation angle of the prism before rotation, nor does it require prior knowledge of the distance information of the target. However, the execution of each step of this method requires the rotation direction of the prism to be determined based on the error result of the previous step. Although this method can ultimately achieve pointing and tracking control of the target, it requires a large amount of time.

[0004] The patent with application number CN202111001376.0 proposes a rotating double-prism pointing correction method based on a particle swarm algorithm. By collecting the pointing test results of the rotating double-prism experimental prototype, the particle swarm algorithm is used to identify the equivalent error of the experimental prototype based on parameter identification theory, and the physical parameters of the prism are corrected to improve the pointing accuracy of the rotating double-prism system. However, in the target pointing test experiment, the prism rotation angle required to achieve target pointing still relies on the traditional two-step method.

[0005] The patent with the application number CN201310655695.2 proposes a rotating double-prism target tracking control method based on a two-step method, but the method does not consider the errors of the actual parameters of the rotating double-prism system in processing, installation and measurement, which directly affect the target tracking accuracy. In order to realize fast and accurate target pointing tracking control, a new rotating double-prism rotation angle calculation method needs to be found. SUMMARY

[0006] In order to overcome the defects and deficiencies existing in the prior art, the present application proposes a rotating double-prism target tracking system and method based on a virtual system.

[0007] The target tracking system comprises an image acquisition module, an image processing module and a central control module.

[0008] The steps of the target tracking method comprise: constructing a virtual system mapping the actual pointing direction of the exit light beam of the rotating double-prism system based on the vector refraction law, iteratively estimating the prism rotation angle in the virtual system based on the sine-cosine algorithm according to the target image information to be tracked transmitted by the image processing module, driving the first prism and the second prism to rotate to the corresponding estimated rotation angle if the estimated error between the virtual rotating double-prism system exit light beam pointing direction mapped by the estimated rotation angle of the prism and the target to be tracked is less than the actual error, and completing the target tracking when the actual error between the camera field of view center and the target center is less than the set deviation threshold.

[0009] The technical solution adopted by the present application to solve its technical problems is:

[0010] A rotating double-prism target tracking system based on a virtual system, characterized by comprising an image acquisition module, an image processing module and a central control module.

[0011] The image acquisition module is used to acquire images in real time through a camera sensor and send them to the image processing module.

[0012] The image processing module is used to extract the target position information to be tracked in the image and send it to the central control module.

[0013] The central control module comprises an embedded controller, a first prism, a second prism, a first motor and a second motor; the first prism and the second prism are placed in front of the camera; the planes of the prisms are sequentially named S 11 , S 12 , S 21 , S 22 , and a camera coordinate system O t X t Y t Z t, facing the camera lens, horizontal left for X t axis positive direction, vertical upward for Y t axis positive direction, pointing to the prism by the camera lens for Z t axis positive direction;

[0014] The image processing module receives the image as the current frame image, establishes an image coordinate system OXY with the lower left corner of the current frame image as the origin, wherein the horizontal right is the positive direction of the X axis, and the vertical upward is the positive direction of the Y axis, judges whether the current frame image contains the target to be tracked; if not, re-receive the image; if it contains the target to be tracked, frame the rectangular frame in which the target to be tracked is located in the current frame image, and take the center point of the rectangular frame as the center point of the target to be tracked.

[0015] The center control module obtains the estimated rotation angle values of the first prism and the second prism based on the update iteration of the sine-cosine algorithm, analyzes the corresponding estimated prism rotation angle into pulse quantity and drives the first motor and the second motor, thereby driving the first prism and the second prism to rotate to the corresponding angle, so that the center of the camera field of view and the center of the target to be tracked gradually coincide.

[0016] And a virtual system-based rotating double-prism target tracking method based on the above virtual system-based rotating double-prism target tracking system:

[0017] According to the position information of the center of the target to be tracked, the estimated rotation angle values of the first prism and the second prism are obtained in the virtual system based on the update iteration of the sine-cosine algorithm, including the following steps:

[0018] Step S1: constructing a virtual system based on the vector refraction law mapping the actual pointing direction of the exit beam of the rotating double-prism system;

[0019] Step S2: according to the pixel coordinates of the center of the target to be tracked, the estimated rotation angle of the first prism is calculated in the virtual system by the iteration calculation of the sine-cosine algorithm And the estimated rotation angle of the second prism

[0020] Step S3: if the estimated error in the virtual system is less than the actual deviation e between the center of the camera field of view and the target to be tracked, the estimated rotation angles And are sent to the rotating double-prism system; otherwise, the estimated rotation angle values of the first prism and the second prism are continuously iterated and updated in the virtual system based on the sine-cosine algorithm;

[0021] Step S4: The embedded controller resolves the calculated estimated prism rotation angle into pulse number and drives the first motor and the second motor, thereby driving the first prism and the second prism to rotate to the corresponding angle, so that the center of the camera field of view gradually coincides with the center of the target to be tracked.

[0022] Further, in step S1, the specific process of constructing a virtual system mapping the actual pointing of the exit beam of the rotating double-prism system based on the vector refraction law is as follows:

[0023] Step S11: Express the normal vector of each plane of the first prism and the second prism as follows:

[0024]

[0025] Wherein, α1 and α2 are the top angles of the first prism and the second prism, respectively; and are the estimated rotation angles of the first prism and the second prism, respectively; ij is the unit normal vector of the jth plane of the ith prism, i is 1 or 2, j is 1 or 2, n ij is the unit normal vector of the plane S ij ;

[0026] The incident light is incident along the Z t axis direction, and its vector form is expressed as:

[0027] s i =(0,0,1)

[0028] Step S12: According to the vector refraction law, calculate the refraction vector A 11 of the plane S r1 :

[0029]

[0030] Wherein, n1 is the refractive index of the first prism, s i is the incident light vector, and n 11 is the unit normal vector of the plane S 11 ;

[0031] Similarly, the refraction vectors A 12 , A 21 , A 22 of the planes S r2 , S r3 , S r4 are obtained, which are set in the O t X t Y t Z t coordinate system. r4the direction cosine of the estimated azimuth angle of the actual exit beam of the rotating double-prism system in the virtual system is:

[0032]

[0033] the estimated elevation angle of the actual exit beam of the rotating double-prism system in the virtual system is:

[0034]

[0035] the estimated azimuth angle and the estimated elevation angle and the distance between the exit light and the light screen, i.e. the virtual camera field center actually pointed by the exit beam of the rotating double-prism system is calculated

[0036] Further, in step S2, according to the pixel coordinates of the center of the target to be tracked, the estimated rotation angle of the first prism and the estimated rotation angle of the second prism in the virtual system are calculated by iterative calculation of the sine-cosine algorithm

[0037] Step S21: define the position vector W of the sine-cosine algorithm as follows, and randomly initialize the position vector W;

[0038]

[0039] wherein, and are the estimated rotation angle of the first prism and the estimated rotation angle of the second prism of the pth individual, respectively;

[0040] Step S22: send the estimated rotation angle to the virtual system where the exit beam of the rotating double-prism system is actually pointed, to obtain the virtual camera field center pointing direction in the O t X t Y t plane in the camera coordinate system

[0041] Step S23: calculate the estimation error

[0042]

[0043]

[0044]

[0045] wherein, is the virtual camera field center, is the pixel coordinate of the target imaging point in the virtual system, Δx is the horizontal coordinate deviation of the target imaging point in the virtual system from the center of the virtual camera field of view, and Δy is the vertical coordinate deviation of the target imaging point in the virtual system from the center of the virtual camera field of view;

[0046] Step S24: repeating steps S22-S23 for all individuals;

[0047] Step S25: evaluating the global optimal position g t at the tth iteration; the global optimal position g t is the estimated optimal prism rotation angle in the virtual system;

[0048] Step S26: if the estimated error e is smaller than the actual deviation e between the center of the camera field of view of the rotating double-prism system and the target imaging point, sending the estimated optimal prism rotation angle g in the virtual system to the rotating double-prism system;

[0049] Step S27: updating the random parameters r1, r2, r3, and r4 of the sine-cosine algorithm according to the following formula:

[0050]

[0051] r2 = 2 x π x rand

[0052] r3 = 2 x rand

[0053] r4 = rand

[0054] where a is a constant related to the search space range of the algorithm, t is the current iteration number, T is the maximum iteration number, π is the circular constant, and rand is a random number with a value in the range of [0, 1];

[0055] Step S28: updating the position of the individual according to the following formula:

[0056]

[0057] where t is the current iteration number, W t is the position vector at the tth iteration, r1, r2, r3, and r4 are random parameters, sin represents the sine operation, cos represents the cosine operation, and g t is the optimal individual position up to the tth iteration;

[0058] Step S29: determining whether the target has changed; if the target has moved, reinitializing the position vector W and jumping to step S22; otherwise, jumping to step S210;

[0059] Step S210: Repeat steps S22-S29 until the distance between the field center and the target imaging point reaches the set deviation threshold.

[0060] Further, in step S23, the pixel coordinates of the target imaging point in the virtual system are specifically calculated as follows:

[0061]

[0062]

[0063] where (x c1 ,y c1 ) is the camera field center actually pointed by the mapping rotating double-prism system based on the current first prism rotation angle θ1 and the second prism rotation angle θ2, e x is the horizontal coordinate deviation of the actual camera field center from the actual target imaging point, e y is the vertical coordinate deviation of the actual camera field center from the actual target imaging point.

[0064] Compared with the prior art, the present application and the preferred embodiments thereof have at least the following beneficial effects:

[0065] (1) The virtual system is introduced in the rotating double-prism target tracking method, which separates the target tracking system from the prism rotation angle calculation system. Estimating the prism rotation angle in the virtual system does not directly affect the target tracking system, thereby ensuring the stability of the target tracking system.

[0066] (2) The sine-cosine algorithm is introduced into the rotating double-prism target tracking system, which converts the complex inverse problem of the rotating double-prism into an optimization problem of estimating the error . The prism rotation angle is solved by using the cooperation and information sharing mechanism between individuals in the sine-cosine algorithm, which has the advantages of fast calculation and high accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0067] The present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments:

[0068] Figure 1 FIG. 1 is a schematic diagram of the rotating double-prism system of an embodiment of the present application;

[0069] Figure 2 FIG. 2 is a schematic diagram of the rotating double-prism target tracking method based on the virtual system of an embodiment of the present application;

[0070] Figure 3 FIG. 3 is a flowchart of the rotating double-prism target tracking method based on the virtual system of an embodiment of the present application. DETAILED DESCRIPTION

[0071] To make the features and advantages of this patent more clearly understood, the following embodiments are specifically described in detail as follows:

[0072] It should be noted that the following detailed description is illustrative and is intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used in this specification have the same meaning as commonly understood by those skilled in the art to which this application belongs.

[0073] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0074] The rotating dual-prism target tracking system provided by the embodiment of the present invention includes an image acquisition module, an image processing module, and a central control module. The image acquisition module is used to acquire images in real time through a camera sensor and send them to the image processing module; the image processing module is used to extract the position information of the target to be tracked in the image and send it to the central control module; the central control module is mainly composed of an embedded controller, a first prism, a second prism, a first motor, and a second motor. The first prism and the second prism are placed in front of the camera, wherein the first prism and the second prism are wedge prisms with a prism vertex angle α1=α2=14.85° and a refractive index n1=n2=1.515. The planes of the prism are named S from left to right. 11 、S 12 、S 21 、S 22 , its schematic diagram is as follows Figure 1 As shown. With the camera as the origin, establish the camera coordinate system O t X t Y t Z t , facing the camera lens, horizontally to the left is X t The positive direction of the axis, vertically upward is Y t The positive direction of the axis, from the camera lens to the prism is Z t Positive axis direction.

[0075] An image processing module is configured to receive information collected by the image collecting module, take the received image as a current frame image, establish an image coordinate system OXY with the lower left corner of the current frame image as an origin, wherein the horizontal right direction is a positive direction of an X axis, and the vertical upward direction is a positive direction of a Y axis, and determine whether the current frame image contains a target to be tracked.

[0076] A central control module is configured to obtain estimated rotation angle values of the first prism and the second prism based on an updated iteration of a sine-cosine algorithm, analyze the corresponding estimated prism rotation angles into pulse numbers, and drive the first motor and the second motor, so as to drive the first prism and the second prism to rotate to corresponding angles, and gradually make the center of the field of view of the camera coincide with the center of the target to be tracked.

[0077] The principle of the virtual system-based rotating double-prism target tracking method is shown in Figure 2 The corresponding flowchart is shown in Figure 3 The target tracking method specifically includes the following steps:

[0078] Step S1: constructing a virtual system mapping the actual pointing direction of the exit light beam of the rotating double-prism system based on the vector refraction law;

[0079] Step S2: calculating the estimated rotation angle of the first prism and the estimated rotation angle of the second prism in the virtual system through an iterative calculation of a sine-cosine algorithm according to the pixel coordinates of the center of the target to be tracked;

[0080] Step S3: if the estimated error in the virtual system is smaller than the actual deviation e between the center of the field of view of the camera and the target to be tracked, sending the estimated rotation angles to the rotating double-prism system; otherwise, continuing to iteratively update the estimated rotation angles of the first prism and the second prism in the virtual system based on the sine-cosine algorithm;

[0081] Step S4: analyzing the estimated prism rotation angles into pulse numbers by an embedded controller, and driving the first motor and the second motor, so as to drive the first prism and the second prism to rotate to corresponding angles, and gradually make the center of the field of view of the camera coincide with the center of the target to be tracked.

[0082] In the step S1 provided in the embodiment of the application, the specific process of constructing the virtual system mapping the actual pointing direction of the exit light beam of the rotating double-prism system based on the vector refraction law is as follows:

[0083] Step S11: representing the normal vectors of each plane of the first prism and the second prism as:​​​​

[0084]

[0085] wherein a1=a2=14.85° are the vertex angles of the first prism and the second prism respectively; and are the estimated rotation angles of the first prism and the second prism respectively; n ij is the unit normal vector of the jth plane of the ith prism, i is 1 or 2, j is 1 or 2, n ij is the unit normal vector of the plane S ij .

[0086] The incident light is incident along the Z t axis direction, and its vector form is expressed as:

[0087] s i =(0,0,1)

[0088] Step S12: Calculate the refracted vector A 11 of the plane S r1 according to the vector refraction law:

[0089]

[0090] wherein n1=1.515 represents the refractive index of the first prism, s i is the incident light vector, and n 11 is the unit normal vector of the plane S 11 .

[0091] Similarly, the refracted vectors A 12 , A 21 , A 22 of the planes S r2 , S r3 , S r4 can be obtained, and the direction cosines of the outgoing light beam A t in the O t X t Y t Z r4 coordinate system are (K, L, M), the estimated azimuth angle of the actual outgoing light beam of the rotating double-prism system in the virtual system is:

[0092]

[0093] The estimated elevation angle of the actual outgoing light beam of the rotating double-prism system in the virtual system is:

[0094]

[0095] The estimated azimuth angle and the estimated elevation angle are used to calculate the actual outgoing light beam of the rotating double-prism system in the virtual system.​​ and the pitch angle is estimated and the distance D = 670 between the exit light and the screen, the virtual camera field center actually pointed by the exit light beam of the mapping rotating double-prism system can be calculated

[0096] In step S2 provided by the embodiment of the present application, according to the pixel coordinates of the center of the target to be tracked, the estimated rotation angle of the first prism is calculated in the virtual system by iterative calculation of the sine-cosine algorithm and the estimated rotation angle of the second prism The specific process is as follows:

[0097] Step S21: the population size of the sine-cosine algorithm is set to 50, the position vector W of the sine-cosine algorithm is defined as follows, and the position vector W is randomly initialized;

[0098]

[0099] wherein, and are the first prism estimated rotation angle and the second prism estimated rotation angle of the p th individual, respectively, and the value range is [-π, π];

[0100] Step S22: the estimated rotation angle is sent to the virtual system to which the exit light beam of the mapping rotating double-prism system is actually pointed, to obtain the virtual camera field center in the O t X t Y t plane pointing to

[0101] Step S23: the estimation error is calculated

[0102]

[0103]

[0104]

[0105] wherein, is the virtual camera field center, and the value range is is the pixel coordinates of the target imaging point in the virtual system, Δx is the horizontal coordinate deviation of the virtual camera field center from the target imaging point in the virtual system, and Δy is the vertical coordinate deviation of the virtual camera field center from the target imaging point in the virtual system;

[0106] Step S24: steps S22-S23 are repeated for all individuals;

[0107] Step S25: the global optimal position g at the t th iteration is evaluatedt ; if the pth individual has a smaller estimated error, then global optimal position g t i.e. the estimated optimal prism rotation angle in the virtual system;

[0108] Step S26: if the estimated error is smaller than the actual deviation e between the center of the field of view of the rotating double-prism system and the target imaging point, the estimated optimal prism rotation angle in the virtual system is sent to the rotating double-prism system; Step S26: if the estimated error is smaller than the actual deviation e between the center of the field of view of the rotating double-prism system and the target imaging point, the estimated optimal prism rotation angle in the virtual system is sent to the rotating double-prism system; Step S26: if the estimated error is smaller than the actual deviation e between the center of the field of view of the rotating double-prism system and the target imaging point, the estimated optimal prism rotation angle in the virtual system is sent to the rotating double-prism system;

[0109] Step S27: the random parameters r1, r2, r3, r4 of the sine-cosine algorithm are updated according to the following formula:

[0110]

[0111] r2 = 2 x π x rand

[0112] r3 = 2 x rand

[0113] r4 = rand

[0114] wherein a is a constant, set as π, t is the current iteration number, T is the maximum iteration number, set as 200, π is the circular constant, and rand is a random number with a value in [0, 1];

[0115] Step S28: the position of the individual is updated according to the following formula:

[0116]

[0117] wherein t is the current iteration number, W t is the position vector at the tth iteration, r1, r2, r3, r4 are random parameters, sin represents the sine operation, cos represents the cosine operation, g t is the optimal individual position up to the tth iteration;

[0118] Step S29: it is determined whether the target has changed, if the target has moved, the position vector W is reinitialized, and the process jumps to step S22; otherwise, the process jumps to step S210;

[0119] Step S210: steps S22-S29 are repeated until the distance between the center of the field of view and the target imaging point reaches the set deviation threshold.

[0120] More specifically, in step S23 provided in the embodiments of the present application, the pixel coordinates of the target imaging point in the virtual system are specifically calculated as follows:

[0121]

[0122]

[0123] wherein (x c1 ,y c1 ) is the virtual camera field of view center calculated based on the current first prism rotation angle θ1 and the second prism rotation angle θ2 of the rotating double-prism system, e x is the horizontal coordinate deviation of the actual camera field of view center and the actual target imaging point, e y is the vertical coordinate deviation of the actual camera field of view center and the actual target imaging point.

[0124] The above is only the preferred embodiment of the present application, and is not intended to limit the present application in other forms. Any person skilled in the art can use the disclosed technical content to make changes or modifications to equivalent embodiments. However, any simple modification, equivalent change and modification made to the above embodiments without departing from the technical solution of the present application, and in accordance with the technical essence of the present application, shall still fall within the protection scope of the present application.

[0125] The present application is not limited to the above best mode, and any person can derive other various forms of a rotating double-prism target tracking system and method based on a virtual system under the inspiration of the present application. Any equivalent change and modification made in accordance with the scope of the present application shall fall within the scope of the present application.

Claims

1. A virtual system-based rotating double-prism target tracking method, characterized in that: based on a rotating double-prism target tracking system composed of an image acquisition module, an image processing module and a central control module: the image acquisition module is configured to acquire images in real time through a camera sensor and send the images to the image processing module; the image processing module is configured to extract target position information in the images and send the information to the central control module; the image processing module takes the received images as current frame images, establishes an image coordinate system OXY with the lower left corner of the current frame images as the origin, judges whether the current frame images contain a target to be tracked, and if not, re-receives images; if yes, frames a rectangular frame in which the target to be tracked is located in the current frame images, and takes the center of the rectangular frame as the center of the target to be tracked; the central control module obtains estimated rotation angle values of the first prism and the second prism based on the update iteration of the sine-cosine algorithm, analyzes the corresponding estimated prism rotation angle into pulse numbers, drives the first motor and the second motor, thereby driving the first prism and the second prism to rotate to the corresponding angles, and gradually aligns the center of the camera field of view with the center of the target to be tracked; based on the position information of the center of the target to be tracked, the virtual system obtains the estimated rotation angle values of the first prism and the second prism based on the update iteration of the sine-cosine algorithm, including the following steps: step S1: constructing a virtual system that maps the actual pointing direction of the exit light beam of the rotating double-prism system based on the vector refraction law; step S4: the embedded controller analyzes the calculated estimated prism rotation angle into pulse numbers, drives the first motor and the second motor, thereby driving the first prism and the second prism to rotate to the corresponding angles, and gradually aligns the center of the camera field of view with the center of the target to be tracked. 2.The virtual system-based rotating double-prism target tracking method according to claim 1, characterized in that: in step S1, the specific process of constructing the virtual system that maps the actual pointing direction of the exit light beam of the rotating double-prism system based on the vector refraction law is as follows: step S11: representing the normal vectors of each plane of the first prism and the second prism as follows: 3.The virtual system-based rotating double-prism target tracking method according to claim 2, characterized in that: step S21: defining the position vector W of the sine-cosine algorithm as follows and randomly initializing the position vector W: step S24: repeating steps S22-S23 for all individuals; step S27: updating the random parameters r1, r2, r3 and r4 of the sine-cosine algorithm according to the following formula: r2=2×π×rand r3=2×rand r4=rand where ɑ is a constant related to the search space range of the algorithm, t is the current iteration number, T is the maximum iteration number, π is the circular constant, and rand is a random number with a value in [0, 1]; step S28: updating the position of the individual according to the following formula: ​ The center control module comprises an embedded controller, a first prism, a second prism, a first motor and a second motor; the first prism and the second prism are placed in front of the camera; the planes of the prisms are sequentially named S 11 , S 12 , S 21 , S 22 , from left to right, with the camera as the origin, a camera coordinate system O t X t Y t Z t is established, facing the camera lens, the horizontal left is the positive direction of the X t axis, the vertical upward is the positive direction of the Y t axis, and the direction from the camera lens to the prism is the positive direction of the Z t axis. ​ ​ ​ ​ Step S2: According to the pixel coordinates of the center of the target to be tracked, the estimated rotation angle of the first prism is calculated in the virtual system by the sine-cosine algorithm and the estimated rotation angle of the second prism Step S3: If the estimation error in the virtual system The estimated rotation angle is smaller than the actual deviation e between the center of the camera field of view and the target to be tracked. and Send to the rotating dual prism system; conversely, in the virtual system, continue to update the estimated rotation angles of the first prism and the second prism based on the sine-cosine algorithm; ​ ​ ​ ​ wherein a1 and a2 are the apex angles of the first and second prisms, respectively; and are the estimated apex angles of the first and second prisms, respectively; ij is the unit normal vector of the jth plane of the ith prism, i is 1 or 2, j is 1 or 2, n ij is the unit normal vector of the plane S ij ; incident light along the Z t axis direction, which is expressed in vector form as: s i ＝(0,0,1) Step S12: Calculate the refraction vector A 11 of the plane S r1 according to the vector refraction law. where n1 is the first prism refractive index, s i is the incident light vector, n 11 is the unit normal vector of the plane S 11 ; Similarly, the plane S 12 is obtained 21 , the plane S 22 is obtained r2 , the refraction vector A r3 of the plane S r4 is obtained t , the direction cosine of the emergent light beam A t in the coordinate system O t X t Y r4 Z t is (K, L, M), and the estimated azimuth angle of the actual emergent light beam of the mapping rotating double-prism system in the virtual system is Estimating the pitch angle of a real exit beam of a rotating double-prism system mapped in a virtual system is: from the estimated azimuth angle and the estimated elevation angle and the distance between the exit light and the light screen, i.e. the center of the virtual camera field of view that the exit light beam of the calculated mapping rotating biprism system actually points to ​ In step S2, according to the pixel coordinates of the center of the target to be tracked, the estimated rotation angle of the first prism is calculated in the virtual system by a sine-cosine algorithm and the estimated rotation angle of the second prism The specific process is as follows: ​ wherein, and are the first and second prism estimated turn angles, respectively, for the pth individual. Step S22: convert the estimated rotation angle to the virtual system that the exit beam of the mapping rotating biprism system actually points to, to obtain the O t X t Y t the center of the virtual camera field of view in the plane points to Step S23: Calculate the estimation error wherein, is the center of the virtual camera field of view, is the pixel coordinate of the target imaging point in the virtual system, Δx is the horizontal coordinate deviation of the target imaging point in the virtual system from the center of the virtual camera field of view, and Δy is the vertical coordinate deviation of the target imaging point in the virtual system from the center of the virtual camera field of view. ​ Step S25: Evaluate the global optimum position g at the tth iteration t ; if the pth individual has a smaller estimated error, then global optimum position g t i.e. the estimated optimum prism corner in the virtual system; Step S26: if the estimated error the estimated optimal prism rotation angle in the virtual system is sent to the rotating double-prism system if the estimated error is less than the actual deviation e between the center of the camera field of view of the rotating double-prism system and the imaging point of the target. ​ ​ ​ ​ ​ ​ where t is the current iteration number, W t is the position vector at the tth iteration, r1, r2, r3, r4 are random parameters, sin denotes a sine operation, cos denotes a cosine operation, g t is the optimal individual position up to the tth iteration; Step S29: judging whether the target has changed, if the target has moved, reinitializing the position vector W, and jumping to step S22; otherwise, jumping to step S210; Step S210: repeating steps S22-S29 until the distance between the field center and the target imaging point reaches the set deviation threshold.

4. The method of claim 3, wherein: In step S23, the specific calculation process of the pixel coordinates of the target imaging point in the virtual system is as follows: wherein (x c1 ,y c1 ) is the center of the camera field of view calculated based on the current first prism rotation angle θ1 and the second prism rotation angle θ2 of the rotating double-prism system, e x is the horizontal coordinate deviation of the actual center of the camera field of view from the actual target imaging point, e y is the vertical coordinate deviation of the actual center of the camera field of view from the actual target imaging point.

Images