A rotating double-prism target tracking control method

Abstract

The application relates to a rotating double-prism target tracking control method. By analyzing the propagation path of a virtual pointing axis and a target visual axis in double Risley prisms, combining the visual axis adjustment law of the rotating double prisms, a control method for adjusting the visual axis pointing based on a virtual axis angle error is provided. The core lies in introducing a coupling proportional controller, realizing real-time stable tracking of the target, and avoiding the problem of low tracking efficiency caused by directly using approximate equivalent information of the image. Example results show that in dynamic target tracking, the method provided by the application can track moving targets with a wide range of speeds. Compared with existing tracking control methods, the application significantly improves the tracking accuracy and stability of the rotating double-prism system for moving targets.

Description

Technical Field

[0001] This invention relates to the field of photoelectric tracking, and more specifically to a rotating double prism target tracking control method. Background Technology

[0002] In the field of photoelectric tracking technology, existing technologies typically employ gimbals or fast-reflecting mirrors as the actuators for target tracking. Gimbals can achieve large-angle deflections, making them suitable for tracking targets moving over a wide range. However, their complex structure, large size, and high inertia result in slow dynamic response speeds, making it difficult to meet the precise tracking requirements of high-speed moving targets, and they also consume a lot of power. Fast-reflecting mirrors offer advantages such as fast response speeds and high tracking accuracy, enabling fine adjustments of micro-arcs. However, their deflection angles are relatively small, typically only suitable for tracking small-range targets, and unsuitable for handling the needs of large-range moving targets. Therefore, both gimbals and fast-reflecting mirrors in existing technologies have limitations and cannot simultaneously meet the requirements of large-range, high-precision, and fast-response target tracking.

[0003] Compared to gimbals and quick-reflecting mirror mechanisms, rotating biprisms offer several advantages: compact structure, high precision, vibration insensitivity, and low cost, making them suitable as target tracking mechanisms. The azimuth and deflection angles of the emitted beam, refracted by the prisms, are coordinated and controlled by two coaxially connected circular wedge prisms. When applying rotating biprisms to photoelectric tracking, the camera detector serves as the sole source of target information, and its control objective is the miss distance (pixel error of the target in the image). Achieving stable and accurate tracking requires maintaining a stable and sufficiently small miss distance.

[0004] Patent application CN201310655695.2 proposes a two-step rotating biprism target tracking control method. However, this method requires repeated optimization iterations after each frame of feedback, which directly affects the real-time performance of target tracking control. To achieve fast and accurate target pointing and tracking control, a more efficient tracking control method is needed.

[0005] The existing technology (Yuan L, Li J, Huang Y, et al. High-precision closed-loop tracking of moving targets based on rotational double prisms[J]. Optical Engineering, 2021, 60(11): 114107-114107.) proposes a real-time region selection closed-loop tracking method (RTSS), which uses the input error value to feed back the detector and the output prism adjustment value to continuously track moving targets in different regions of the field of view for a period of time. However, the experimental setting uses a relatively low target speed, and the algorithm is quite complex due to the control switching based on different regions. Jitter occurs during high-speed switching, which limits the application effect of the rotational double prism.

[0006] Existing technology (Li J, Yuan L, Xia H, et al. Rotation matrix error-decoupling methods for Risley prism closed-loop tracking[J]. Precision Engineering-Journal of the International Societies for Precision Engineering and Nanotechnology, 2022, 76: 66-74.) proposes a rotation matrix error decoupling method (RMED), which claims to eliminate the coupling relationship between tracking errors in the x and y directions and prism rotation in a rotating biprism tracking device by removing them from the image. Based on the decoupled image tracking error, rotations in the same and opposite directions are applied simultaneously to the two prisms to adjust the line of sight, achieving closed-loop tracking. However, this method is essentially still a coupling control, only performing approximate equivalent calculations based on image errors, ignoring the accurate correlation between image errors and deflection angles, thus limiting its tracking performance.

[0007] Existing target tracking methods based on rotating biprisms only approximate the pointing error using image information and directly apply it to target feedback, thus limiting the improvement of their tracking performance. To overcome the shortcomings of existing technologies, we propose a novel rotating biprism target tracking control method based on virtual axis angle error feedback to improve the tracking accuracy of rotating biprisms for high-speed moving line-of-sight. Summary of the Invention

[0008] The purpose of this invention is to provide a rotating biprism target tracking control method, which can improve the line-of-sight tracking accuracy of the rotating biprism for high-speed moving targets.

[0009] The objective of this invention is achieved through the following technical solution:

[0010] A rotating biprism target tracking control method is disclosed. This method, based on a rotating biprism target tracking system, includes the following steps: A CMOS camera receives image acquisition information and determines whether the current frame image contains the target to be tracked. If the target is not contained, the tracking system is placed in a scanning search mode. If the target is contained, the centroid pixel position of the target is extracted. Based on this pixel position information and the law of vector refraction, a virtual path of the target's visual axis in the rotating biprism system is constructed. Furthermore, a path pointing to the visual axis corresponding to the image center is constructed, and the virtual axis angle error between the two visual axes is obtained and described in two dimensions: azimuth angle error and deflection angle error. An embedded drive controller, combined with the visual axis adjustment law of the rotating biprism, introduces a coupling proportional controller into the control loop. After calculating the next rotation angle increment of the two prisms, the first and second prisms are driven to rotate to the corresponding rotation angle, ensuring that the center of the target to be tracked is always maintained at the center of the camera's field of view, thus achieving target tracking control based on a rotating biprism.

[0011] The target tracking system based on rotating double prisms includes a CMOS camera for real-time acquisition and extraction of the position information of the target to be tracked in the image; an embedded drive controller for receiving image feedback information; a first motor and a second motor for driving prisms ∏1 and Π2 to rotate; and an embedded drive controller for controlling the first motor and the second motor.

[0012] Both prisms Π1 and Π2 can rotate around Z. W The optical axis rotates independently. Prisms Π1 and Π2 are arranged in series along the system's optical axis, with the refractive surfaces configured as flat-wedge-wedge-flat. The rotation angles are defined as θ1 and θ2, and the angular velocities are defined as ω1 and ω2. A CMOS camera Cartesian coordinate system O is established. C X C Y C Z C and the world rectangular coordinate system O W X W Y W Z W The origin O is the coordinate system. C and O W They are located at the optical center of the CMOS camera lens and the center of the incident plane of the prism ∏1, respectively.

[0013] The method specifically includes the following steps:

[0014] Step S1: Based on the pixel coordinates (Δp) of the center of the target to be tracked x ,Δp y ), calculate the incident vector on the CMOS camera side. And the outgoing target's line-of-sight vector is calculated based on the vector refraction law. Similarly, the incident vector from the camera side is obtained using the pixel coordinates (0,0) of the image center. Then calculate the pointing vector of the emitted line of sight.

[0015]

[0016] Where: n0, n1, and n2 are the refractive indices of air, prism 1, and prism 2, respectively, and n a n b ∈[n0,n1,n2];

[0017] Let be the unit normal vector of the incident plane of prism 1.

[0018] Let be the unit normal vector of the exit surface of prism 1.

[0019] Let be the unit normal vector of the incident plane of prism 2.

[0020] Let be the unit normal vector of the exit surface of prism 2.

[0021] Where: θ1 and θ2 are the prism rotation angles; α1 and α2 are the prism apex angles;

[0022] Step S2: Calculate the errors of the two outgoing line-of-sight vectors, expressed in two dimensions: the azimuth angle error projected onto the Z-axis normal plane and the deflection angle error of the two line-of-sight vectors relative to the Z-axis.

[0023]

[0024]

[0025]

[0026] Step S3: If the center of the target to be tracked does not coincide with the center of the image, then there is an error due to the non-coincidence between the two outgoing line vectors; calculate the azimuth error and deflection error between the two lines, and calculate the prism rotation angle increment based on the beam adjustment law; when the two prisms rotate in the same direction at the same speed, only the azimuth angle of the outgoing light is changed, while the deflection angle of the outgoing light remains unchanged; while when the two prisms rotate in opposite directions at the same speed, only the deflection angle of the outgoing light is changed, while the azimuth angle remains unchanged.

[0027] Step S4: Introduce a coupled proportional controller to independently calculate the rotation angles of the two prisms. When the deflection angle of the target's line-of-sight vector needs to be reduced to coincide with the pointing line-of-sight vector, the angle between the principal sections of the two prisms should be increased. Therefore, the two prisms should rotate in the direction that increases the angle, ensuring that the rotation directions of the two prisms are opposite and the rotation angles are the same, Δρ. Conversely, when the angle between the principal sections of the two prisms is decreased, the two prisms should rotate in the direction that decreases the angle, ensuring that the rotation directions of the two prisms are opposite and the rotation angles are the same, Δρ.

[0028] Step S5: When the azimuth angle of the target's line-of-sight vector needs to be reduced to coincide with the pointing line-of-sight vector, the two prisms should ensure that their rotation directions and rotation angles are the same. Both prisms rotate in the direction that decreases the azimuth angle of the target's line of sight vector; conversely, both prisms rotate in the direction that increases the azimuth angle of the target's line of sight vector.

[0029] Step S6: In order to adjust the convergence effect of the line-of-sight error, multiply the deflection angle error by a proportional coefficient k1 and the azimuth angle error by a proportional coefficient k2; combine the required rotation angles of the prism in the two dimensions to obtain the independent adjustment angles of the two prisms respectively.

[0030] or

[0031] Where k1 and k2 are constants. Different values ​​of gain k1 and k2 will affect the convergence of target tracking error. In the specific implementation process, k1 and k2 are continuously adjusted according to the tracking test results, and finally an optimal value is determined.

[0032] Step S7: The embedded drive controller drives the first motor and the second motor with the two prism angle adjustment commands Δθ1 and Δθ2 calculated based on the above process, thereby driving the first prism and the second prism to rotate to the corresponding angle, so that the center of the CMOS camera field of view always coincides with the center of the target to be tracked.

[0033] The CMOS camera continuously detects the position of the target center pixel in the image. The embedded drive controller continuously calculates the rotation angle of the two prism drive motors based on the camera feedback information and continuously drives the prisms to rotate to the corresponding angle position. When the target center coincides with the image center, the two outgoing line-of-sight vectors will also coincide, and there will be no line-of-sight error. The rotation angle of the two prisms calculated based on the line-of-sight error is 0, and the prisms no longer need to be adjusted, thus achieving target tracking.

[0034] The beneficial effects of this invention are as follows:

[0035] In the rotating biprism target tracking method, a virtual line-of-sight error is introduced instead of a simple image error. The virtual line-of-sight error directly reflects the essence of the target tracking error. Combined with the unique line-of-sight adjustment law of the rotating biprism system, a coupled proportional controller is introduced to achieve high-precision and stable target tracking.

[0036] Real-time tracking of high-speed moving targets has been achieved, and the proposed method can stably track moving targets with different speeds and trajectories. Compared with existing tracking methods, the proposed method has improved tracking performance, which has been verified. Attached Figure Description

[0037] The present invention will now be described in further detail with reference to the accompanying drawings and specific implementation methods.

[0038] Figure 1 This is a schematic diagram of the rotating double prism system of the present invention performing pointing tracking;

[0039] Figure 2 This is a control block diagram of the rotating double prism target tracking control method of the present invention;

[0040] Figure 3 The present invention describes the movement trajectory using horizontal, vertical, and diamond-shaped trajectories;

[0041] Figure 4 This invention relates to the tracking and control process of the rotating double prism target tracking and control method.

[0042] Figure 5 The tracking error curve of this invention shows that the target movement speed of the horizontal and vertical segmented trajectory is 2.13 mm / s.

[0043] Figure 6 The tracking error curve of this invention shows that the target moving speed is 3.01 mm / s based on the diamond segmented trajectory.

[0044] Figure 7 The tracking error curve of this invention shows that the horizontal and vertical segmented trajectory tracks the target's moving speed of 7.90 mm / s.

[0045] Figure 8 The tracking error curve of this invention shows that the target moving speed of the diamond segmented trajectory is 11.17 mm / s.

[0046] Figure 9 The tracking error curve of this invention shows that the horizontal and vertical segmented trajectory tracks the target's moving speed of 23.7 mm / s.

[0047] Figure 10 The tracking error curve of this invention shows that the target moving speed of the diamond segmented trajectory is 33.52 mm / s. Detailed Implementation

[0048] The present invention will now be described in further detail with reference to the accompanying drawings.

[0049] like Figures 1 to 10 As shown, in order to achieve the technical effect of "improving the line-of-sight tracking accuracy of a rotating biprism for high-speed moving targets", the steps and functions of a rotating biprism target tracking control method are explained in detail below.

[0050] A rotating biprism target tracking control method is disclosed. This method, based on a rotating biprism target tracking system, includes the following steps: A CMOS camera receives image acquisition information and determines whether the current frame image contains the target to be tracked. If the target is not contained, the tracking system is placed in a scanning search mode. If the target is contained, the centroid pixel position of the target is extracted. Based on this pixel position information and the law of vector refraction, a virtual path of the target's visual axis in the rotating biprism system is constructed. Furthermore, a path pointing to the visual axis corresponding to the image center is constructed, and the virtual axis angle error between the two visual axes is obtained and described in two dimensions: azimuth angle error and deflection angle error. An embedded drive controller, combined with the visual axis adjustment law of the rotating biprism, introduces a coupling proportional controller into the control loop. After calculating the next rotation angle increment of the two prisms, the first and second prisms are driven to rotate to the corresponding rotation angle, ensuring that the center of the target to be tracked is always maintained at the center of the camera's field of view, thus achieving target tracking control based on a rotating biprism.

[0051] The target tracking system based on rotating double prisms includes a CMOS camera for real-time acquisition and extraction of the position information of the target to be tracked in the image; an embedded drive controller for receiving image feedback information; a first motor and a second motor for driving prisms ∏1 and Π2 to rotate; and an embedded drive controller for controlling the first motor and the second motor.

[0052] Both prisms Π1 and Π2 can rotate around Z. W The axis rotates independently, with prism Π1 being closer to the CMOS camera and prism Ώ2 being closer to the target; prisms Ώ1 and Ώ2 are arranged in series along the optical axis of the system, and the configuration of the refractive surfaces is flat-wedge-wedge-flat; the rotation angles are defined as θ1 and θ2, and the angular velocities are defined as ω1 and ω2;

[0053] like Figure 1 As shown, a Cartesian coordinate system O for the CMOS camera was established. C X C Y C Z C and the world rectangular coordinate system O W X W Y W Z W The origin O is the coordinate system. C and O WThese are located at the optical center of the CMOS camera lens and the center of the incident plane of the prism ∏1, respectively; if the CMOS camera is properly mounted in the rotating double prism assembly, then the camera's line of sight Z... C and the system optical axis Z W The axes theoretically coincide;

[0054] The principle of the rotating biprism target tracking control method based on virtual axis angle error is as follows: Figure 2 As shown, the corresponding flowchart is as follows: Figure 4 As shown;

[0055] The method specifically includes the following steps:

[0056] Step S1: Based on the pixel coordinates (Δp) of the center of the target to be tracked x ,Δp y ), calculate the incident vector on the CMOS camera side. And the outgoing target's line-of-sight vector is calculated based on the vector refraction law. Similarly, the incident vector from the camera side is obtained using the pixel coordinates (0,0) of the image center. Then calculate the pointing vector of the emitted line of sight.

[0057]

[0058] Where: n0, n1, and n2 are the refractive indices of air, prism 1, and prism 2, respectively, and n a n b ∈[n0,n1,n2];

[0059] Let be the unit normal vector of the incident plane of prism 1.

[0060] Let be the unit normal vector of the exit surface of prism 1.

[0061] Let be the unit normal vector of the incident plane of prism 2.

[0062] Let be the unit normal vector of the exit surface of prism 2.

[0063] Where: θ1 and θ2 are the prism rotation angles; α1 and α2 are the prism apex angles;

[0064] Step S2: Calculate the errors of the two outgoing line-of-sight vectors, expressed in two dimensions: the azimuth angle error projected onto the Z-axis normal plane and the deflection angle error of the two line-of-sight vectors relative to the Z-axis.

[0065]

[0066] Step S3: If the center of the tracked target does not coincide with the center of the image, then there is an error due to the non-coincidence between the two outgoing line vectors; calculate the azimuth error and deflection error between the two lines, and calculate the prism rotation angle increment based on the beam adjustment law; when the two prisms rotate in the same direction at the same speed, only the azimuth angle of the outgoing light is changed, while the deflection angle of the outgoing light remains unchanged; while when the two prisms rotate in opposite directions at the same speed, only the deflection angle of the outgoing light is changed, while the azimuth angle remains unchanged.

[0067] Step S4: Introduce a coupled proportional controller to independently calculate the rotation angles of the two prisms. When the deflection angle of the target's line-of-sight vector needs to be reduced to coincide with the pointing line-of-sight vector, the angle between the principal sections of the two prisms should be increased. Therefore, the two prisms should rotate in the direction that increases the angle, ensuring that the rotation directions of the two prisms are opposite and the rotation angles are the same, Δρ. Conversely, when the angle between the principal sections of the two prisms is decreased, the two prisms should rotate in the direction that decreases the angle, ensuring that the rotation directions of the two prisms are opposite and the rotation angles are the same, Δρ.

[0068] Step S5: When the azimuth angle of the target's line-of-sight vector needs to be reduced to coincide with the pointing line-of-sight vector, the two prisms should ensure that their rotation directions and rotation angles are the same. Both prisms rotate in the direction that decreases the azimuth angle of the target's line of sight vector; conversely, both prisms rotate in the direction that increases the azimuth angle of the target's line of sight vector.

[0069] Step S6: In order to adjust the convergence effect of the line-of-sight error, multiply the deflection angle error by a proportional coefficient k1 and the azimuth angle error by a proportional coefficient k2; combine the required rotation angles of the prism in the two dimensions to obtain the independent adjustment angles of the two prisms respectively.

[0070] or

[0071] Where k1 and k2 are constants. Different values ​​of gain k1 and k2 will affect the convergence of target tracking error. In the specific implementation process, k1 and k2 are continuously adjusted according to the tracking test results to finally determine an optimal value; in this example, k1 = 0.8 and k2 = 1.2.

[0072] Step S7: The embedded drive controller drives the first motor and the second motor with the two prism angle adjustment commands Δθ1 and Δθ2 calculated based on the above process, thereby driving the first prism and the second prism to rotate to the corresponding angle, so that the center of the CMOS camera field of view always coincides with the center of the target to be tracked.

[0073] The CMOS camera continuously detects the position of the center pixel of the target in the image. The embedded drive controller continuously calculates the rotation angle of the two prism drive motors based on the camera feedback information and continuously drives the prism to rotate to the corresponding angle position. When the target center coincides with the image center, the two outgoing line-of-sight vectors will also coincide, and there will be no line-of-sight error. The rotation angle of the two prisms calculated based on the line-of-sight error is 0, and the prisms no longer need to be adjusted, thus achieving target tracking.

[0074] The following is a verification of dynamic targets. When implementing dynamic target tracking and positioning, the actual simulated target moves along a certain trajectory at a certain speed. The rotating dual-prism target tracking system automatically controls the rotation of the two prisms to change the camera's viewing axis based on imaging feedback, moving the center of the CMOS camera's field of view to the center of the target.

[0075] The parameters involved in the embodiments are listed in Table 1. When tracking moving targets, the target is set to move at six different speeds: 2.13 mm / s, 7.90 mm / s, 23.7 mm / s, 3.01 mm / s, 11.17 mm / s, and 33.52 mm / s, which correspond to maximum relative angular velocities of 0.59° / s, 2.20° / s, 6.59° / s, 0.84° / s, 3.10° / s, and 9.24° / s, respectively.

[0076] Along two sets of multi-segment trajectories respectively: such as Figure 3 As shown, horizontal, vertical, and diamond-shaped trajectories are used for movement. The tracking error is evaluated using the pixel distance between the target center and the image center, defined as... Where Δp x and Δp y This represents the pixel error between the target imaging point and the image center in different directions; it utilizes the root mean square error of the entire dynamic tracking process. To evaluate the performance of the tracking system in dynamic target tracking; such as Figures 5 to 10 The figures shown are the tracking results of this method.

[0077] Table 1 Parameters of the Rotating Biprism Target Tracking System

[0078]

[0079] By utilizing the current prism angle state and target image feedback, closed-loop control eliminates the angular error between the pointing and target viewing axis vectors, stabilizing the moving target as close as possible to the image center. First, a target feature extraction algorithm is used to obtain the target's miss distance in the image. Then, the vector refraction law is used to determine the deflection and azimuth errors between the pointing and target axes, and the independent adjustment amounts of the two prisms are calculated based on the viewing axis adjustment rules. Subsequently, a coupled proportional controller is introduced into the prism drive control loop to calculate and execute the relative rotation angle of each prism in real time, quickly eliminating target tracking errors.

[0080] The results demonstrate that this method significantly improves the line-of-sight tracking accuracy of moving targets using a rotating biprism. For example, under the tracking control strategy VAAD based on virtual axis angle error feedback, the total root mean square (RMS) tracking errors for targets with relative angular velocities of 0.59° / s, 2.2° / s, 6.59° / s, 0.84° / s, 3.10° / s, and 9.24° / s are 1.18, 1.76, 6.43, 9.23, 23.19, and 32.78 pixels, respectively. Compared with existing RTSS and RMED methods, this method improves both the maximum error and RMS error metrics, demonstrating enhanced and validated tracking performance.

Claims

1. A rotating biprism target tracking control method, the method being based on a rotating biprism target tracking system, characterized in that: The method includes the following steps: The CMOS camera receives image acquisition information, determines whether the current frame image contains the target to be tracked, and if it does not contain the target, the tracking system is placed in scan search mode; if it contains the target, the centroid pixel position of the target is extracted, and target tracking control is performed, specifically including: Step S1: Based on the centroid pixel coordinates (Δp) of the target to be tracked x ,Δp y ), calculate the incident vector on the CMOS camera side. And the outgoing target's line-of-sight vector was calculated based on the vector refraction law. Similarly, the incident vector on the camera side is obtained using the pixel coordinates (0,0) of the image center. Calculate the outgoing line-of-sight vector Step S2: Calculate the errors of the two outgoing line-of-sight vectors, expressed in two dimensions: the azimuth angle error projected onto the Z-axis normal plane and the deflection angle error of the two line-of-sight vectors relative to the Z-axis. Step S3: If the center of the tracked target does not coincide with the center of the image, then there is an error between the two outgoing line vectors; calculate the virtual axis angle error between the two lines, namely the azimuth angle error and the deflection angle error, and calculate the prism rotation angle increment based on the beam adjustment law; Step S4: Introduce a coupling proportional controller to independently calculate the rotation angles of the two prisms; Step S5: When the azimuth angle of the target's line-of-sight vector needs to be reduced to coincide with the pointing line-of-sight vector, it should be ensured that the rotation directions and rotation angles of the two prisms are the same. Both prisms rotate in the direction that decreases the azimuth angle of the target's line of sight vector; conversely, both prisms rotate in the direction that increases the azimuth angle of the target's line of sight vector. Step S6: Adjust the convergence effect of the line-of-sight error by multiplying the deflection angle error by a proportional coefficient k1 and the azimuth angle error by a proportional coefficient k2; combine the required rotation angles of the prism in the two dimensions to obtain the independent adjustment angles of the two prisms. Step S7: The embedded drive controller drives the first motor and the second motor based on the calculated angle adjustment commands Δθ1 and Δθ2 of the two prisms, thereby rotating prisms Π1 and Π2 to the corresponding angles, so that the center of the CMOS camera's field of view always coincides with the center of the target to be tracked.

2. The rotating double prism target tracking control method according to claim 1, characterized in that: The target tracking system based on a rotating double prism includes a CMOS camera for real-time acquisition and extraction of the position information of the target to be tracked in images; An embedded driver controller that receives image feedback information; And a first motor and a second motor that drive prisms Π1 and Π2 to rotate; the first motor and the second motor are controlled by an embedded drive controller.

3. The rotating double prism target tracking control method according to claim 2, characterized in that: Both prisms Π1 and Π2 can rotate around Z. W The optical axis rotates independently. Prisms Π1 and Π2 are arranged in series along the system's optical axis, with the refractive surfaces configured as flat-wedge-wedge-flat. The rotation angles are defined as θ1 and θ2, and the angular velocities are defined as ω1 and ω2. A CMOS camera Cartesian coordinate system O is established. C X C Y C Z C and the world rectangular coordinate system O W X W Y W Z W The origin O is the coordinate system. C and O W They are located at the optical center of the CMOS camera lens and the center of the incident plane of prism Π1, respectively.

4. The rotating double prism target tracking control method according to claim 1, characterized in that: In step S1: Where: n0, n1, and n2 are the refractive indices of air, prism Π1, and prism Π2, respectively, and n a n b ∈[n0,n1,n2]; Let be the unit normal vector of the incident plane of prism Π1. Let be the unit normal vector of the exit surface of prism Π1. Let be the unit normal vector of the incident plane of prism Π2. Let be the unit normal vector of the exit surface of prism Π2. Where: θ1 and θ2 are rotation angles; α1 and α2 are prism apex angles.

5. The rotating double prism target tracking control method according to claim 4, characterized in that: In step S2:

6. The rotating double prism target tracking control method according to claim 4, characterized in that: In step S3: when the two prisms rotate at the same speed and in the same direction, only the azimuth angle of the emitted light is changed, while the deflection angle of the emitted light remains unchanged; while when the two prisms rotate at the same speed in opposite directions, only the deflection angle of the emitted light is changed, while the azimuth angle remains unchanged.

7. The rotating double prism target tracking control method according to claim 4, characterized in that: In step S4: when the deflection angle of the target line-of-sight vector needs to be reduced to coincide with the pointing line-of-sight vector, the angle between the principal sections of the two prisms should be increased. Therefore, the two prisms should rotate in the direction that increases the angle, and ensure that the rotation directions of the two prisms are opposite and the rotation angles are the same, Δρ. Conversely, when the angle between the principal sections of the two prisms is reduced, the two prisms should rotate in the direction that decreases the angle, and ensure that the rotation directions of the two prisms are opposite and the rotation angles are the same, Δρ.

8. The rotating double prism target tracking control method according to claim 4, characterized in that: In step S6: k1 and k2 are constants. Different values ​​of k1 and k2 will affect the convergence of the target tracking error. In the specific implementation process, k1 and k2 are continuously adjusted according to the tracking test results, and finally an optimal value is determined.

9. The rotating double prism target tracking control method according to claim 4, characterized in that: In step S7: the CMOS camera continuously detects the position of the target center pixel in the image, and the embedded drive controller continuously calculates the rotation angle of the two prism drive motors based on the camera feedback information, and continuously drives the prisms to rotate to the corresponding angle position. When the target center coincides with the image center, the two outgoing line-of-sight vectors will also coincide, so there is no line-of-sight error. The rotation angle of the two prisms calculated based on the line-of-sight error is 0, and the prisms no longer need to be adjusted, thus finally achieving target tracking.

Images