A multi-dimensional detection device and method for space targets

Abstract

This application discloses a multi-dimensional detection device and method for space targets, relating to the field of photoelectric imaging detection technology. The device includes a tracking turntable subsystem, a telescope subsystem, a multi-spectral polarization imaging subsystem, and an image processing and display subsystem. The multi-spectral polarization imaging subsystem and the image processing and display subsystem are connected. The telescope subsystem and the multi-spectral polarization imaging subsystem are integrated on the tracking turntable subsystem. The tracking turntable subsystem rotates within a set angle range, thereby rotating the telescope subsystem and the multi-spectral polarization imaging subsystem. The telescope subsystem acquires multi-field-of-view optical paths and incident these paths onto the multi-spectral polarization imaging subsystem. The multi-spectral polarization imaging subsystem determines polarization images at different field-of-view angles based on the multi-field-of-view optical paths. The image processing and display subsystem stitches and fuses the polarization images at different field-of-view angles based on image processing algorithms to obtain a wide-area detection image. This application improves the spatial coverage and resolution of observations.

Description

Technical Field

[0001] This application relates to the field of photoelectric imaging detection technology, and in particular to a multi-dimensional detection device and method for space targets. Background Technology

[0002] With the increasing frequency of space launches, the number of space targets is rapidly increasing, and the risk of on-orbit satellites being threatened by space debris impacts is drastically rising, making space target detection extremely important. Ground-based space target detection systems have become one of the important means of space target detection due to their advantages such as low cost, short construction period, and mature technology.

[0003] Traditional ground-based detection bands include visible light and infrared bands, which have different detection advantages. Currently, there are various existing integrated visible and infrared imaging optical systems that can effectively reduce interference factors in the light propagation environment. However, ground-based multi-dimensional detection systems for space targets still have the problems of large size and small field of view. A single detector cannot simultaneously meet the requirements of the space exploration field for a large field of view and high resolution. Summary of the Invention

[0004] The purpose of this application is to provide a multi-dimensional detection device and method for space targets, which can acquire detailed space target features with a large field of view and high resolution, thereby improving the spatial coverage and resolution of observations.

[0005] To achieve the above objectives, this application provides the following solution:

[0006] Firstly, this application provides a multi-dimensional detection device for space targets, comprising:

[0007] The system includes a tracking turntable subsystem, a telescope subsystem, a multi-band polarization imaging subsystem, and an image processing and display subsystem; the multi-band polarization imaging subsystem and the image processing and display subsystem are connected; the telescope subsystem and the multi-band polarization imaging subsystem are integrated on the tracking turntable subsystem.

[0008] The tracking turntable subsystem is used to achieve rotation within a set angle range, thereby driving the telescope subsystem and the multi-band polarization imaging subsystem to rotate.

[0009] The telescope subsystem is used to acquire multiple field-of-view optical paths and incident the multiple field-of-view optical paths onto the multi-spectral polarization imaging subsystem; the multiple field-of-view optical paths include the optical paths reflected or emitted by spatial targets under different field-of-view angles.

[0010] The multi-band polarization imaging subsystem is used to determine polarization images at different field angles based on the multi-field-of-view optical path.

[0011] The image processing and display subsystem is used to stitch and fuse polarized images under different field of view angles based on image processing algorithms to obtain a wide-area detection image and display the wide-area detection image.

[0012] Secondly, this application provides a method for multi-dimensional detection of space targets, wherein the method is used in any of the aforementioned multi-dimensional detection devices for space targets, and the method includes:

[0013] The polarization images under different field of view angles are acquired; the polarization images under different field of view angles are determined by the multi-spectral polarization imaging subsystem based on the multi-field-view angle optical path; the multi-field-view angle optical path includes the optical path of reflection or emission from the spatial target under different field of view angles.

[0014] Based on image processing algorithms, polarization images under different field of view are stitched and fused to obtain a wide-area detection image, which is then displayed.

[0015] According to the specific embodiments provided in this application, this application has the following technical effects:

[0016] This application provides a multi-dimensional detection device and method for space targets. A telescope subsystem and a multi-spectral polarization imaging subsystem are integrated on a tracking turntable subsystem. The tracking turntable subsystem rotates within a set angle range, thereby driving the telescope subsystem and the multi-spectral polarization imaging subsystem to rotate. This allows for coverage of a larger space where the target is located, enabling more comprehensive observation of the trajectory, position, and attitude changes of the space target. This achieves wider-range detection and identification, solving the problem of the small field of view in traditional single-detector imaging systems. This enables collaborative detection of space targets from different angles. By combining target information obtained from overlapping different fields of view, multiple polarization images from different field of view angles are stitched and fused based on image processing algorithms. This enhances image details, obtains more detailed target features, and outputs a complete and high-resolution image of the space target scene, i.e., a wide-area detection image, thereby improving the spatial coverage and resolution of the observation. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the functional modules of a multi-dimensional detection device for space targets in one embodiment of this application.

[0019] Figure 2This is a schematic diagram of the functional modules of a multi-spectral polarization imaging subsystem in a space target multi-dimensional detection device provided in an embodiment of this application.

[0020] Figure 3 This is a schematic diagram of the first adaptive optics module in a multi-spectral polarization imaging subsystem of a space target multi-dimensional detection device provided in an embodiment of this application.

[0021] Figure 4 This is a model diagram of a multi-dimensional detection device for space targets provided in an embodiment of this application.

[0022] Figure 5 A flowchart of a multi-dimensional detection method for space targets provided in an embodiment of this application.

[0023] Reference numerals: 1-First polarization modulation unit, 2-Second polarization modulation unit, 3-Third polarization modulation unit, 4-First filter unit, 5-Second filter unit, 6-Third filter unit, 7-First imaging lens unit, 8-Second imaging lens unit, 9-Third imaging lens unit, 10-First imaging detection unit, 11-Second imaging detection unit, 12-Third imaging detection unit, 13-First image acquisition unit, 14-Second image acquisition unit, 15-Second image acquisition unit. Detailed Implementation

[0024] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0025] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0026] In one exemplary embodiment, such as Figure 1 , Figure 2 , Figure 3 and Figure 4 As shown, a multi-dimensional detection device for space targets is provided, including: a tracking turntable subsystem, a telescope subsystem, a multi-band polarization imaging subsystem, and an image processing and display subsystem; the multi-band polarization imaging subsystem and the image processing and display subsystem are connected; the telescope subsystem and the multi-band polarization imaging subsystem are integrated on the tracking turntable subsystem.

[0027] The tracking turntable subsystem is used to achieve rotation within a set angle range, thereby driving the rotation of the telescope subsystem and the multispectral polarization imaging subsystem. Specifically, the tracking turntable subsystem can achieve rotation from 0° to 180°.

[0028] The telescope subsystem is used to acquire multi-field-of-view optical paths and incident the multi-field-of-view optical paths onto the multi-spectral polarization imaging subsystem; the multi-field-of-view optical paths include the optical paths reflected or emitted by space targets at different field-of-view angles.

[0029] The multi-band polarization imaging subsystem is used to determine polarization images at different field angles based on multi-field-of-view optical paths.

[0030] The image processing and display subsystem is used to stitch and fuse polarized images from different field angles based on image processing algorithms to obtain and display wide-area detection images.

[0031] Image stitching technology has received considerable attention in recent years. Current image stitching techniques are mostly based on traditional photoelectric detection methods. However, due to complex backgrounds, interference from background stars, relatively weak target signals, poor beam penetration, and difficulty in distinguishing faint targets in actual detection, traditional ground-based detection systems have limited target detection and identification capabilities and low accuracy. Therefore, wide-area, high-resolution detection remains a requirement for ground-based space target detection and represents a development trend in space target detection. This embodiment acquires multiple polarization images from different field-of-view angles and uses image processing algorithms to stitch and fuse these images to generate a wide-area detection image, which is of great significance for space target detection and aerospace research.

[0032] By capturing images simultaneously from multiple detectors in different directions, with some overlap, and then performing image registration and fusion, a new, wide-angle, high-resolution, and seamless image is created. Image stitching technology can produce panoramic images with high resolution and a wide field of view.

[0033] Polarization imaging technology digitizes the measured polarization information of a space target field, expanding the target information from light intensity and position to polarization degree and polarization angle, thus broadening the information dimension. Due to the difference in polarization characteristics between man-made targets and the deep space background, polarization imaging can highlight the target contour information and enhance image details. It is also beneficial for identifying faint targets in strong light backgrounds. Combining polarization imaging technology with image stitching technology in ground-based photoelectric detection systems can effectively reduce interference factors in the light propagation environment and achieve wide-area, high-resolution space target detection.

[0034] SIFT (Scale-Invariant Feature Transform) is a robust feature point detection and matching algorithm widely used in remote sensing image registration. However, traditional SIFT algorithms suffer from high computational cost and low matching efficiency when processing large-scale remote sensing images. While solutions combining multi-scale wavelet transform and SIFT exist to improve registration speed, further improvements in the efficiency of remote sensing image registration are still needed, driven by the ongoing developments in the aerospace field.

[0035] Particle Swarm Optimization (PSO) is a swarm intelligence-based optimization algorithm with advantages such as strong global search capability and fast convergence speed. Introducing PSO into multi-scale wavelet transform and SIFT algorithms can further improve the efficiency and accuracy of feature point matching.

[0036] To address the above issues, this application proposes a ground-based multi-dimensional detection system and wide-area detection method for space targets, combining multi-spectral polarization imaging technology, adaptive adjustment technology, and image stitching technology. Based on adaptive adjustment technology to correct atmospheric wavefront interference, a large field-of-view target measurement is achieved through polarization degree image field-of-view stitching. Target information obtained by overlapping different fields of view is combined, and image processing algorithms are used to stitch and reconstruct visible light / shortwave infrared polarization images from different fields of view. During the stitching process, a fast remote sensing image registration algorithm based on particle swarm optimization combined with wavelet transform and SIFT algorithm is optimized. By calculating the correspondence of the same feature point in the scene across different polarization images, the projection relationship between the images is calculated. Multiple images are then projected onto the same plane, and finally, a complete wide-area polarization degree image is output through fusion, obtaining a high-resolution polarization image of the scene from which detailed target features with a large field of view and high resolution are extracted.

[0037] In another exemplary embodiment of this application, the image processing and display subsystem includes an information processing unit and a display unit; the information processing unit is connected to the multi-band polarization imaging subsystem and the display unit, respectively.

[0038] The information processing unit is used to: sharpen polarized images under different field of view angles based on the median filtering algorithm to obtain images to be stitched under different field of view angles; stitch the images to be stitched under different field of view angles based on wavelet transform algorithm, scale-invariant feature transform algorithm and particle swarm optimization algorithm to obtain the stitched image; fuse the overlapping areas of adjacent images in the stitched image based on the wavelet transform algorithm to obtain a wide-area detection image; the display unit is used to display the wide-area detection image.

[0039] In another exemplary embodiment of this application, the telescope subsystem includes: a first front telescope unit, a second front telescope unit, and a third front telescope unit arranged in an equilateral triangle.

[0040] The first front telescope unit is used to incident the first optical path onto the multi-spectral polarization imaging subsystem; the first optical path is the optical path reflected or emitted by the spatial target at the first field of view.

[0041] The second front telescope unit is used to incident the second optical path onto the multi-spectral polarization imaging subsystem; the second optical path is the optical path reflected or emitted by the spatial target at the second field of view.

[0042] The third front telescope unit is used to incident the third optical path onto the multi-spectral polarization imaging subsystem; the third optical path is the optical path reflected or emitted by the spatial target at the third field of view.

[0043] The first, second, and third optical paths are parallel.

[0044] In another exemplary embodiment of this application, such as Figure 2 As shown, the multi-band polarization imaging subsystem includes: a spectral polarization state modulation module, a correction module, and an image acquisition module; the spectral polarization state modulation module is located on the output optical path of the telescope subsystem; the correction module is located on the output optical path of the spectral polarization state modulation module; and the image acquisition module is located on the output optical path of the correction module.

[0045] The spectral polarization state modulation module is used to generate polarization state modulated output light at different field angles based on the multi-field-of-view optical path.

[0046] The correction module is used to correct the polarization state of the modulated output light under different field of view angles, so as to obtain the corrected output light under different field of view angles.

[0047] The image acquisition module is used to image the corrected outgoing light at different field of view angles to obtain polarization images at different field of view angles.

[0048] In another exemplary embodiment of this application, the spectral polarization state modulation module includes: a first polarization state modulation module, a second polarization state modulation module, and a third polarization state modulation module; the optical axis of the first polarization state modulation module is disposed on the output optical path of the first front telescope unit, the optical axis of the second polarization state modulation module is disposed on the output optical path of the second front telescope unit, and the optical axis of the third polarization state modulation module is disposed on the output optical path of the third front telescope unit.

[0049] The first polarization modulation module is used to generate polarization-modulated outgoing light at a first field of view according to the first optical path.

[0050] The second polarization modulation module is used to generate polarization-modulated outgoing light at the second field of view according to the second optical path.

[0051] The third polarization modulation module is used to generate polarization-modulated outgoing light at the third field of view according to the third optical path.

[0052] In another exemplary embodiment of this application, the first polarization modulation module includes: a first polarization modulation unit and a first filtering unit; the first polarization modulation unit is disposed on the output optical path of the first front telescope unit; the first filtering unit is disposed on the output optical path of the first polarization modulation unit.

[0053] The first polarization modulation unit is used to generate linearly polarized light at a first field of view according to the first optical path. The linearly polarized light includes: 0° linearly polarized light, 45° linearly polarized light, 90° linearly polarized light, and 135° linearly polarized light.

[0054] The first filter unit is used to filter the linearly polarized light under the first field of view to obtain the polarized modulated outgoing light under the first field of view.

[0055] Furthermore, the optical axes of the first front telescope unit, the first polarization modulation unit, and the first filter unit are on the same straight line; the optical axes of the second front telescope unit, the second polarization modulation unit, and the second filter unit are on the same straight line; the optical axes of the third front telescope unit, the third polarization modulation unit, and the third filter unit are on the same straight line; and the optical axes of the first polarization modulation module, the second polarization modulation module, and the third polarization modulation module are parallel.

[0056] Furthermore, the first, second, and third polarization modulation units employ high-speed, high-positioning-precision rotating wheels to rapidly drive the polarizers to rotate synchronously.

[0057] The second polarization modulation module includes a second polarization modulation unit and a second filter unit; the third polarization modulation module includes a third polarization modulation unit and a third filter unit; the second polarization modulation module and the third polarization modulation module have the same structure as the first polarization modulation module, and will not be described further here.

[0058] Since the polarization-modulated output light may contain atmospheric interference information, after obtaining the polarization-modulated output light, it is necessary to remove the atmospheric interference information in the polarization-modulated output light, as follows.

[0059] In another exemplary embodiment of this application, the correction module includes: a first adaptive optics module, a second adaptive optics module, and a third adaptive optics module; the first adaptive optics module is disposed on the output optical path of the first polarization state modulation module; the second adaptive optics module is disposed on the output optical path of the second polarization state modulation module; and the third adaptive optics module is disposed on the output optical path of the third polarization state modulation module.

[0060] The first adaptive optics module is used to remove atmospheric interference information from the polarization-modulated outgoing light at the first field of view, so as to obtain the corrected outgoing light at the first field of view.

[0061] The second adaptive optics module is used to remove atmospheric interference information from the polarization-modulated outgoing light at the second field of view, so as to obtain the corrected outgoing light at the second field of view.

[0062] The third adaptive optics module is used to remove atmospheric interference information from the polarization-modulated output light at the third field of view, so as to obtain the corrected output light at the third field of view.

[0063] In another exemplary embodiment of this application, such as Figure 3 As shown, the first adaptive optics module includes: a first wavefront correction unit, a first beam splitter unit, a first wavefront sensing unit, and a first wavefront control unit; the first wavefront sensing unit and the first wavefront control unit are connected.

[0064] The first wavefront correction unit includes: a deformable mirror and a deformable mirror driver; the deformable mirror driver is connected to the deformable mirror and the first wavefront control unit respectively; the deformable mirror is disposed on the output optical path of the first polarization state modulation module; the first beam splitter is disposed on the output optical path of the deformable mirror; and the first wavefront sensing unit is disposed on the output optical path of the first beam splitter.

[0065] The deformable mirror is used to deform the polarization-modulated outgoing light at the first field of view to generate deformed outgoing light at the first field of view.

[0066] The first beam splitting unit is used to split the deformed outgoing light at the first field of view into two beams, which enter the first wavefront sensing unit and the image acquisition module respectively.

[0067] The first wavefront sensing unit is used to measure the deformed outgoing light under the first field of view in real time to obtain the first wavefront distortion.

[0068] The first wavefront control unit generates a control signal based on the first wavefront distortion variable and sends it to the deformable mirror actuator. The first wavefront distortion variable is the optical wave distortion generated by the deformed outgoing light under the first field of view due to atmospheric interference. The first wavefront control unit determines whether the first wavefront distortion variable is within a set distortion range. If the first wavefront distortion variable is within the set distortion range, no control signal is generated, i.e., the deformable mirror actuator is not controlled. If the first wavefront distortion variable exceeds the set distortion range, a control signal is generated to control the deformable mirror actuator.

[0069] The deformable mirror actuator is used to change the shape of the deformable mirror according to the control signal to generate corrected outgoing light at the first field of view.

[0070] The first beam-splitting unit can be a beam-splitting mirror. The main purpose of the beam-splitting mirror is to separate the optical paths, ensure that the optical paths of the wavefront sensor and the imaging system do not interfere with each other, and avoid blocking the imaging optical path.

[0071] The second adaptive optics module includes: a second wavefront correction unit, a second beam splitter unit, a second wavefront sensing unit, and a second wavefront control unit; the third adaptive optics module includes: a third wavefront correction unit, a third beam splitter unit, a third wavefront sensing unit, and a third wavefront control unit. The second and third adaptive optics modules have the same structure as the first adaptive optics module, and will not be described further here.

[0072] The modulated outgoing light from the first optical path enters the first adaptive optics module. The first wavefront correction unit reflects the target beam, which then passes through the beam splitting unit and enters the first wavefront sensing unit. The first wavefront sensing unit measures the wavefront distortion of the target beam in real time and transmits the wavefront distortion signal to the first wavefront control unit, generating a control signal for the first wavefront correction unit. Finally, the first wavefront correction unit corrects the originally distorted beam into a plane wave. The entire system achieves closed-loop negative feedback, completes the correction of aberrations, and improves the system resolution.

[0073] The light from the second and third optical paths is directed into the second and third adaptive optics modules, respectively, to correct aberrations using the same steps.

[0074] The first, second, and third wavefront sensing units all use Hartmann-Shack sensors. The first, second, and third wavefront correction units change the beam direction by mounting multiple piezoelectric ceramic actuators on the deformable mirror.

[0075] In another exemplary embodiment of this application, the image acquisition module includes: a first image acquisition module, a second image acquisition module, and a third image acquisition module connected to the image processing and display subsystem; the first image acquisition module is disposed on the output optical path of the first adaptive optics module; the second image acquisition module is disposed on the output optical path of the second adaptive optics module; and the third image acquisition module is disposed on the output optical path connected to the third adaptive optics module.

[0076] The first image acquisition module is used to image the corrected outgoing light under the first field of view to obtain a polarization image under the first field of view.

[0077] The second image acquisition module is used to image the corrected outgoing light under the second field of view to obtain a polarization image under the second field of view.

[0078] The third image acquisition module is used to image the corrected outgoing light at the third field of view to obtain a polarization image at the third field of view.

[0079] The first, second, and third image acquisition modules, in order to acquire real-time image streams, use the wavefront distortion thresholds set in the first, second, and third wavefront control units to filter for clear images. The process is as follows:

[0080] First, each wavefront sensing unit measures the wavefront distortion in real time at a high frequency (500 Hz). Each wavefront control unit drives the deformable mirror to compensate for the distortion and calculates the corrected wavefront residual (RMS value). Each image acquisition module continuously captures images at a slightly lower frequency (e.g., 50 Hz), and each frame is marked with an exposure start and end timestamp (accuracy ≤ 1 μs).

[0081] Secondly, when the wavefront residual RMS value is below the threshold for multiple consecutive frames (10 frames @ 500 Hz = 20 ms), it is determined to be a stable period. The start timestamp (T_start) and end timestamp (T_end) of the stable period are recorded.

[0082] Furthermore, images completely contained within [T_start, T_end] are selected from the massive image stream for post-processing, i.e., the frames with the best correction effect.

[0083] The first image acquisition module includes a first imaging lens unit, a first imaging detection unit, and a first image acquisition unit connected in sequence; the second image acquisition module includes a second imaging lens unit, a second imaging detection unit, and a second image acquisition unit connected in sequence; the third image acquisition module includes a third imaging lens unit, a third imaging detection unit, and a third image acquisition unit connected in sequence.

[0084] The optical axes of the first imaging lens unit, the first imaging detection unit, and the first image acquisition unit are on the same straight line; the optical axes of the second imaging lens unit, the second imaging detection unit, and the second image acquisition unit are on the same straight line; the optical axes of the third imaging lens unit, the third imaging detection unit, and the third image acquisition unit are on the same straight line; and the optical axes of the first image acquisition module, the second image acquisition module, and the third image acquisition module are parallel.

[0085] The first, second, and third imaging detection units are electrically connected to the first, second, and third image acquisition units, respectively; the first, second, and third image acquisition modules are electrically connected to the information processing unit, and the information processing unit is electrically connected to the display unit.

[0086] In addition, the first, second, and third image acquisition modules are also connected to the display unit.

[0087] The multi-dimensional detection device for space targets also includes a control subsystem.

[0088] Furthermore, the telescope subsystem is connected to the multi-band polarization imaging subsystem, which in turn is connected to the image processing and display subsystem. The control subsystem is connected to the multi-band polarization imaging subsystem, the image processing and display subsystem, and the tracking turntable subsystem, respectively.

[0089] The first, second, and third imaging detection units complete polarization imaging in the visible light and near-infrared 0.4~1.7µm band, while the first, second, and third image acquisition units acquire images of the visible light and near-infrared 0.4~1.7µm band at the same time and in the same field of view.

[0090] Furthermore, based on the obtained target information image, the information processing unit obtains polarization degree and polarization angle images of different fields of view, and stitches and reconstructs the polarization degree images of different fields of view using image processing algorithms to output a complete wide-area detection image; the display unit displays the linearly polarized image and the wide-area detection image based on the polarization degree.

[0091] The tracking turntable system can rotate from 0° to 180° to acquire multiple images of the target from different perspectives.

[0092] The control subsystem provides regulation for the multi-band polarization imaging subsystem and the image processing and display subsystem.

[0093] Traditional single-detector imaging systems have a small field of view. Detecting multiple detectors at an angle can expand the field of view, enabling monitoring and identification over a wider area. Dividing the target's reflected light into three paths is equivalent to three detectors coordinating detection of the target from different angles. Two detectors combined have insufficient field of view coverage and reduced fault tolerance. If one system fails, the wide-area detection advantage is lost. Four-detector collaborative detection is costly and has high maintenance costs, and requires high precision in both hardware and software. Considering all factors, a three-path, three-detector approach is adopted to detect the target. Each path of light will produce different degrees of error due to atmospheric wavefront interference, so all three paths of light need to be corrected.

[0094] The three light beams have different detection angles towards the target, and all three detectors can image the target in different regions of the field of view of each detector. The three light paths have parallel optical axes, and the three detectors are distributed in an equilateral triangle. In the actual detection process, the detectors are aligned with the target and take pictures through a tracking turntable subsystem.

[0095] All three channels use the same modules and hardware, and have the same function: to correct aberrations and obtain clear polarization images. The difference lies in the different detection angles of the three detectors towards the target, resulting in different backgrounds and different interference with light. Each channel uses the same modules and methods to correct different aberrations, but the specific parameters are different.

[0096] Furthermore, the working principle of the multi-dimensional detection device for space targets is as follows.

[0097] S1. The control system provides power to the polarization modulation unit, the filtering unit, the imaging detection unit, and the image processing and display subsystem. The polarization modulation unit, filtering unit, imaging detection unit, imaging lens unit, and image acquisition unit described below all include the first, second, and third polarization modulation units, filtering units, imaging detection units, imaging lens units, and image acquisition units.

[0098] S2. The target-reflected light, after being affected by atmospheric turbulence, enters the multi-spectral polarization imaging subsystem through the first, second, and third telescope subsystems. Visible / near-infrared light (0.4–1.7 μm) passes through the polarization modulation unit, where a high-speed, high-precision rotating wheel rapidly rotates the polarizer, enabling linear polarization modulation. The visible / near-infrared light (0.4–1.7 μm) passes through the filter unit, where a high-speed, high-precision filter rotating wheel rotates the filter, enabling multispectral observation with 0.44 μm, 0.694 μm, and 1.49 μm filters. Switching to the panchromatic channel allows for panchromatic polarization detection. The modulated outgoing light carries polarization and atmospheric interference information.

[0099] S3. Taking the first adaptive optics module as an example, the modulated outgoing light enters the first wavefront sensing unit through the first wavefront correction unit and the first beam splitting unit.

[0100] The control system supplies power to the first wavefront correction unit, the first beam splitting unit, the first wavefront sensing unit, and the first wavefront control unit. The first wavefront sensing unit uses a Hartmann-Shack sensor, which segments and samples the wavefront of the target reflected light through a lens array. The wavefront tilt within each sub-aperture range will cause the focused spot of the unit lens to drift laterally. The drift of the spot center in two directions relative to the reference position marked by a parallel line is measured, thereby calculating the average slope of the wavefront in two directions within each sub-aperture range. This enables real-time measurement of the wavefront distortion of the target beam and transmits the wavefront distortion signal to the first wavefront control unit.

[0101] The first wavefront control unit obtains the wavefront slope from the sensor, calculates the wavefront phase using an algorithm, and converts it into a control signal for the first wavefront correction unit. This signal is then fed back to the deformable mirror actuator to change the shape of each aperture module of the deformable mirror to compensate for the distorted wavefront, correcting the originally distorted beam into a plane wave. The entire system achieves closed-loop negative feedback, completing the correction of aberrations. The outgoing light carries polarization information and corrects for atmospheric interference.

[0102] The steps for the second and third adaptive optics modules are the same as those for the first adaptive optics module.

[0103] S4. The emitted light, after being corrected by the correction module, is imaged onto the imaging detection unit by the imaging lens unit, and then onto the display unit after passing through the image acquisition unit to obtain four polarization linear polarization components from different perspectives.

[0104] S5. The image processing and display subsystem includes an information processing unit and a display unit. The information processing unit obtains target information images with different viewing angles and polarization angles, i.e., polarization images under different field of view angles, based on the above subsystem.

[0105] S6. The reflected light from the target object needs to pass through optical components such as optical lenses, which will affect the clarity of the polarization image to a certain extent. The polarization images obtained from different sensors are preprocessed separately by a median filtering algorithm based on variance.

[0106] S7. Based on the preprocessed polarization image, obtain spatial targets from different viewpoints. , , , information, This represents the irradiance of the corresponding beam, which is obtained using the Stokes component calculation formula. , , , The component image is used to calculate the degree of polarization. ) and polarization angle ( )image.

[0107] S8. A fast registration algorithm for remote sensing images based on particle swarm optimization combined with wavelet transform and SIFT algorithm.

[0108] S9. Project multiple images onto the same plane, and use wavelet transform to perform pixel-level fusion processing on the overlapping areas of adjacent images to eliminate stitching seams and obtain a smooth and seamless panoramic image.

[0109] This application proposes a multi-dimensional detection device and method for space targets by combining multi-spectral polarization imaging technology, adaptive adjustment technology, and image stitching technology, which has the following beneficial effects.

[0110] (1) This application can obtain intensity detection and polarization information of visible light and short-wave infrared. Based on the polarization advantage, it can realize multi-dimensional spatial target detection, obtain high-resolution polarization images of the scene, enhance image details, and obtain more detailed target features.

[0111] (2) By combining target information obtained from overlapping different fields of view, and stitching and reconstructing visible light / shortwave infrared polarization images from different fields of view using image processing algorithms, the observation field of view can be expanded, enabling large-field spatial target detection and outputting a complete wide-area polarization detection image that covers a larger space where the target is located. This allows for a more comprehensive observation of the trajectory, position, and attitude changes of spatial targets. The fast remote sensing image registration algorithm based on particle swarm optimization combined with wavelet transform and SIFT algorithms can improve the efficiency and accuracy of the algorithm, thereby increasing the image registration speed and further enhancing the system's detection efficiency and accuracy for spatial targets.

[0112] This application combines multi-spectral polarization imaging, adaptive adjustment, and image stitching techniques to generate wide-area detection images, which have broad application value in space target observation and can improve the spatial coverage, resolution, and accuracy of observations. This method has significant positive benefits for space target monitoring, early warning, and trajectory prediction.

[0113] Two embodiments are provided to illustrate the processes of image acquisition and image processing.

[0114] Example 1.

[0115] This embodiment addresses the problems of traditional photoelectric detection methods being susceptible to environmental interference, complex actual detection backgrounds, relatively weak space target signals, weak beam penetration, atmospheric interference affecting imaging quality, and difficulty in distinguishing faint targets. It provides a multi-dimensional space target detection device based on adaptive adjustment technology and multi-spectral polarization imaging technology, enabling the acquisition of polarization images from different perspectives.

[0116] In this embodiment.

[0117] (1) The first, second and third polarization modulation units all use Thorlabs wire grid polarizer WP50L-UB.

[0118] (2) The first, second and third filter units all use the filter turntable FW102CWNEB driven by the stepper motor of Thorlabs. The filters selected are Thorlabs FBH440-10, FBH694-10 and FBH1490-12 filters.

[0119] (3) The first, second and third imaging lens units all use Lingyun shortwave infrared lens M5018-VSW.

[0120] (4) The first, second and third imaging detection units all use the Lingyun shortwave infrared detection module Cobra2000-U31280-130VT1-00.

[0121] In this embodiment, the control subsystem comprises a spectral polarization modulation module, a correction module, an image acquisition module, and a power supply for the image processing and display subsystem.

[0122] The reflected light from the target, after being disturbed by atmospheric turbulence, enters the multi-spectral polarization imaging subsystem through the telescope subsystem. A high-speed, high-precision rotating wheel rapidly drives the polarizer to rotate. The light from the first optical path passes through the first polarization modulation unit, the light from the second optical path passes through the second polarization modulation unit, and the light from the third optical path passes through the third polarization modulation unit, achieving polarization modulation including 0° linear polarization, 45° linear polarization, 90° linear polarization, and 135° linear polarization.

[0123] A high-speed, high-positioning-precision filter rotating wheel drives the filter to rotate. Each filter unit has four channels, including three spectral channels and one all-pass channel. The visible light filter unit includes filters with center wavelengths of 0.44µm, 0.694µm, and 1.49µm. The electronically controlled high-speed, high-positioning-precision filter rotating wheel can be moved to different filter positions to achieve multispectral observation. Switching to the panchromatic channel enables panchromatic polarization detection. The modulated outgoing light carries polarization information and atmospheric interference information.

[0124] The light corrected by the adaptive optics module is imaged by the first, second, and third imaging lens units into the first, second, and third imaging detection units, respectively. By rotating the polarizer, the polarization directions are obtained at 0°, 45°, 90°, and 135°. The image acquisition unit can obtain images with different spectral intensities, namely polarized images at 0°, 45°, 90°, and 135°, thus completing polarization imaging in the visible and near-infrared 0.4~1.7µm band. The first, second, and third image acquisition units respectively acquire images of the visible and near-infrared 0.4~1.7µm band under different fields of view at the same time, and display them on the display unit to obtain four linear polarization components.

[0125] This embodiment can acquire intensity detection and polarization information in multiple bands and dimensions of visible light and short-wave infrared, enabling multi-dimensional spatial target detection, obtaining high-resolution polarization images of the scene, and acquiring more detailed target features.

[0126] Example 2.

[0127] This embodiment addresses the challenge of simultaneously achieving high resolution and small size in traditional photoelectric detection systems. Based on an image processing and display subsystem, it processes the acquired target information image using a variance-based median filtering algorithm. By obtaining polarization degree and polarization angle images from different fields of view, and optimizing the image using a particle swarm optimization algorithm combined with wavelet transform and SIFT algorithms, the polarization degree and polarization angle images of adjacent detectors at different field of view angles are stitched together pairwise to generate a complete and seamless imaging result for the same field of view. The resulting complete wide-area high-resolution detection image is then displayed by the display unit.

[0128] The specific method in this embodiment is as follows: Figure 5 As shown.

[0129] Step 1: Input the image to be registered.

[0130] The image processing and display subsystem includes an information processing unit and a display unit. The information processing unit processes target information images obtained from different viewpoints and polarization angles.

[0131] Step 2: Preprocessing of the images to be registered.

[0132] The reflected light from the target object needs to pass through optical components such as optical lenses, which will affect the clarity of the polarization image to a certain extent. The polarization images obtained from different sensors are preprocessed separately by a variance-based median filtering algorithm.

[0133] The sampling window iterates through the target polarization image, with a window size of 5×5 and 25 pixels per group. The center point coordinates are denoted as . grayscale value Its average value can be expressed as: .

[0134] The variance is: .

[0135] When grayscale value satisfy If so, then the pixel value is retained.

[0136] When grayscale value satisfy When using the neighborhood substitution method, the pixel value is replaced with the maximum gray value of the pixel in the 3×3 neighborhood. Based on the median filtering algorithm, the median of the useful signal value is taken as the new median in the 5×5 neighborhood. This can effectively filter out noise and also better protect the edge details of the target image.

[0137] Step 3: Obtaining polarization information.

[0138] Based on the preprocessed polarization image, spatial targets are obtained from different viewpoints. , , , The information is obtained through the Stokes component calculation formula. , , , The component image is used to calculate the degree of polarization. ) and polarization angle ( )image.

[0139] Let the Stokes vector represent the various polarization states of a light wave. The four parameters of the Stokes vector are: .

[0140] It is the total irradiance of the 0° polarization component plus the total irradiance of the 90° polarization component, representing the total irradiance of the beam; It is the total irradiance of the 0° polarization component minus the total irradiance of the 90° polarization component; It is the total irradiance of the 45° polarization component minus the total irradiance of the 135° polarization component; It is the total irradiance of the right-hand polarization component minus the total irradiance of the left-hand polarization component.

[0141] Degree of polarization (DOP) and angle of polarization (AOP) are two indicators for studying the polarization state of light. Changes in polarization state and angle of polarization can represent the polarization characteristics of polarized light.

[0142] .

[0143] .

[0144] Step 4: Wavelet decomposition preprocessing.

[0145] The input image is decomposed into a low-frequency component (LL) and three high-frequency components (LH, HL, HH) by performing layer-by-layer wavelet decomposition in the horizontal and vertical directions based on two-dimensional discrete wavelet transform (DWT). The low-frequency information is preserved, resulting in a low-frequency image containing a large amount of information.

[0146] For image matrix First, perform wavelet decomposition on the image in the horizontal direction to generate low-frequency... and high-frequency subband .

[0147] .

[0148] .

[0149] These two formulas represent wavelet decomposition of an image in the horizontal direction. First, each row of the original image is extracted and decomposed using wavelet decomposition to obtain high and low frequency information in the horizontal direction.

[0150] The direction is the direction of the line. The direction is the column direction, and it is fixed. Direction, right Direction filtering, in the formula The x-coordinate of the image is fixed in this formula, meaning it only operates on pixels in a specific vertical direction. n is a variable representing the number of pixels in the input image traversed along the vertical direction (y-axis). For each fixed x-coordinate... , Will take all All values ​​in the direction.

[0151] and This represents low-pass and high-pass filters, containing low-frequency and high-frequency filtering coefficients, used to extract low-frequency approximation information and high-frequency detail information from image rows.

[0152] Indicates the direction of the filter along the row ( Displacement (direction).

[0153] It is a filter function that calculates the pixel values ​​of the output image based on the neighboring pixels in the vertical direction of the input image.

[0154] It represents the low-frequency horizontal component and indicates smoothing information.

[0155] It represents the horizontal high-frequency component and the horizontal edge information.

[0156] The image has already been decomposed into high-frequency components using wavelet decomposition along the row direction. and low frequency This is a one-dimensional transformation. Wavelet decomposition is applied to each column of these two sub-bands to extract each column and perform wavelet decomposition to obtain high and low frequency information in the vertical direction.

[0157] right and Each column is decomposed using one-dimensional wavelet decomposition, producing four sub-bands. , , , .

[0158] .

[0159] .

[0160] .

[0161] .

[0162] in: This represents the low-frequency subband, which contains the main structural information of the image; These are low-frequency horizontal components and high-frequency vertical components, representing the vertical edge information of the image; These are high-frequency horizontal components and low-frequency vertical components, representing the horizontal edge information of the image; This is a high-frequency subband that represents the diagonal information of the image.

[0163] The four formulas above represent wavelet decomposition of two sub-bands in the column direction (perpendicular). The direction is the direction of the line. The direction is the column direction, and it is fixed. Direction, right Direction filtering, in the formula The vertical coordinate of the image is fixed in this formula, meaning that the operation is performed only on pixels in a certain horizontal direction of the image. `x` is a variable representing all pixels of the input image traversed horizontally. For each fixed `x`... , Will take all All values ​​in the direction.

[0164] Through multi-level wavelet decomposition, it is possible to... Subbands are further decomposed, generating new ones. , , and Subband decomposition can continue at multiple levels to form a multi-scale image decomposition with a pyramid structure, resulting in a wavelet pyramid of the image.

[0165] In this embodiment, a three-level wavelet pyramid is obtained through decomposition. The SIFT algorithm is used to extract features from the low-frequency component images of the wavelet pyramid image, transforming the registration problem of high-resolution images into the registration problem of low-frequency approximate component images. Coarse registration of the image is achieved in a large-scale (low-resolution) space. Then, using the obtained registration result as the initial value, the phase correlation method is used to perform translation compensation on the coarsely registered image to achieve fine registration of the image and improve the registration efficiency.

[0166] Step 5: SIFT algorithm feature extraction.

[0167] Based on feature point monitoring technology, polarization degree or polarization angle images from different viewpoints are registered. First, the polarization degree images of adjacent detectors are matched and stitched together. The polarization degree images of adjacent detectors are denoted as... and The SIFT (Scale Invariant Feature Transform) algorithm is used to extract features from the LL subband of the low-frequency image. Feature points are extracted from polarization degree images or polarization angle images obtained and processed by different viewpoints, i.e., different detectors, and image feature points are extracted in the scaling space.

[0168] Spatial transformation utilizes a two-dimensional Gaussian distribution function. Images are transformed to form a sequence of images in a multi-scale space, and then scale-invariant feature points are found in these image sequences. The two-dimensional Gaussian distribution function takes the form shown below.

[0169] .

[0170] It is the variance of a two-dimensional standard normal distribution.

[0171] The image is blurred and downsampled using a Gaussian function to obtain a Gaussian pyramid, and a Gaussian scale space is constructed. During the scaling process: .

[0172] Representing a two-dimensional image, This represents convolution, and the formula represents a point on the image. Spatial convolution transformation based on a two-dimensional Gaussian distribution function is performed until the scale factor is variance. In Gaussian scale space, according to the scale factor Construct a Gaussian pyramid from smallest to largest value A difference Gaussian pyramid is constructed using the differences between adjacent Gaussian scale spaces.

[0173] .

[0174] in, That is, the scale factor representing different scales of space; Represents different scales of space within the Gaussian pyramid; This represents the difference pyramid.

[0175] scale factor Changes will have an impact on scale space. As the size decreases, the Gaussian filter (two-dimensional Gaussian distribution function) has a weaker smoothing effect on images, but it preserves image details and edge information. Increasing the size of the Gaussian filter (two-dimensional Gaussian distribution function) enhances its smoothing effect on the image, removing small details and noise, and blurring the image. Changing the scale factor... By obtaining the value of , we can obtain the features of the image at different scales. This represents different scale factors, according to different scale factors The values ​​are arranged from smallest to largest to construct a Gaussian pyramid containing spaces of different scales. .

[0176] For Gaussian pyramids, they provide representations of images at different scales, but do not provide information on variations across different scales. The difference pyramid is a Gaussian pyramid generated by subtracting adjacent layers from each group except the top and bottom layers (the next layer minus the previous layer). Each difference image reflects the detailed information of the image at that scale, representing information that can be represented at the current scale in the Gaussian pyramid but cannot be represented at a smoother scale at the next higher level. It can effectively extract the location of stable key points in the scale space of the image.

[0177] Keypoints are composed of local extrema in the difference Gaussian pyramid space. Each pixel is compared with all its neighbors to see if it is larger or smaller than its neighbors in both the image domain and scale domain. The middle detection point is compared with its eight neighbors at the same scale and 9×2 points corresponding to the adjacent scales above and below, totaling 26 points, to ensure that extrema are detected in both the scale space and the two-dimensional image space. After detecting the extrema in the discrete space, a quadratic curve is fitted to the obtained local extrema to more accurately pinpoint the location and scale value of the feature points. The fitting method is as follows.

[0178] Suppose there is a point in a certain difference Gaussian pyramid space. By performing scale space analysis on this point Expand the series and take the squared terms, then find the first derivative to determine the extrema.

[0179] .

[0180] .

[0181] .

[0182] In the above formula, D is When the series expands, the contents of the parentheses are hidden, and it is represented by the following formula.

[0183] , .

[0184] For scale coordinates, , These represent the corresponding extreme positions and extreme values.

[0185] Set threshold The feature points satisfy These are selected as candidate feature points. A threshold is then set. The value is 0.03. After determining the feature points, the principal direction of the eigenvalues ​​is calculated. The gradient values ​​in each direction around the feature point and the product of the contribution (weight value) of each direction to the center point are calculated and summed in eight directions. The direction of the maximum value is taken as the principal direction. Similar matching points are searched for and paired in the three polarization images. Four pairs of matching points are randomly selected to perform linear estimation of the projection matrix M.

[0186] .

[0187] Establish a distance estimation function for feature point pairs: .

[0188] in, , For a pair of matching feature points, M To represent the projection matrix of the feature point coordinate transformation between images, m 0、 m 1、 m 2、 m 3、 m 4、 m 5、 m 6、 m 7 is the projection matrix M The elements in.

[0189] If the calculated distance is less than a certain threshold, the matching point pair is retained; otherwise, it is discarded. Low-contrast key points and unstable edge response points are removed to enhance matching stability.

[0190] Step 6: Construct feature point descriptors.

[0191] Assigning orientations to keypoints. Histograms are used to statistically analyze the gradient directions of pixels in the vicinity of each keypoint; the peak value represents the keypoint orientation. The coordinate axes are rotated to align with the keypoint orientation. A window centered on the keypoint is then selected and divided into 4×4 regions. Gradient histograms for eight directions are calculated for each region. The accumulated value of each gradient yields a seed point. There are 16 seed points, each composed of eight numbers, resulting in a 4×4×8=128-dimensional descriptor. This descriptor exhibits scaling, rotation, and affine invariance. Feature point matching is achieved by comparing these descriptors.

[0192] Step 7: Optimize feature matching based on particle swarm optimization algorithm.

[0193] The feature point matching results are used as the initial solution of the particle swarm algorithm. Based on the particle swarm algorithm (PSO), the particles are encoded and each particle is represented as a vector containing the translation and rotation parameters of the image to be registered. A fitness function is defined to evaluate the registration effect of each particle, and the degree of feature point matching is used as the evaluation index.

[0194] A certain number of particles are randomly generated, and their positions and velocities are initialized. Based on the current positions and velocities of the particles, the PSO algorithm is used to update the positions and velocities of the particles.

[0195] Calculate the fitness value of each particle, select the best particle based on the fitness value, and record its position as the current best solution to find the best matching result.

[0196] Determine whether the termination condition is met, such as reaching the maximum number of iterations or the fitness value reaching the threshold, and output the registration parameters corresponding to the optimal solution, namely the translation and rotation parameters of the image.

[0197] Step 8: Image registration.

[0198] Perspective transformation and resampling are performed on the image to achieve coarse registration. The phase correlation method is then used to perform translation compensation on the coarsely registered image to achieve fine registration.

[0199] First, coarse registration is performed using the extracted feature point pairs to estimate the preliminary transformation matrix. Assume the feature points in the two low-frequency images correspond to... , The transformation relationship between the two is as follows.

[0200] .

[0201] .

[0202] .

[0203] The transformation matrix can be a translation, rotation, scaling, or affine transformation. , This indicates scaling and rotation in the horizontal direction; , This indicates scaling and rotation in the vertical direction; , These represent the translation amounts in the horizontal and vertical directions, respectively.

[0204] Using all feature point pairs, fit the transformation matrix using the least squares method. This yields a system of equations.

[0205] .

[0206] , is a column vector containing all feature points. , is a matrix containing all feature points.

[0207] , which is the column vector of transformation matrix parameters.

[0208] Based on the least squares method, we can obtain: The superscript T denotes transpose. The resulting six matrix parameters can be used to... and Image coarse registration.

[0209] .

[0210] This represents the image after transformation; Represents the coordinates after the inverse transformation, and gives The corresponding pixel position in the image It will undergo initial registration and transformation to align with the image. The image is obtained after coarse alignment.

[0211] Optimization of the transformation matrix based on the phase correlation method This method achieves fine registration in high-resolution space, keeping the image scale and rotation parameters unchanged and only considering translation transformation. Based on Fourier transform, it estimates the image translation by calculating phase information in the frequency domain. Two images are subjected to Fourier transform, and their cross-spectrum in the frequency domain is calculated. The inverse Fourier transform of the cross-spectrum is then calculated to obtain the cross-correlation function in the spatial domain. The location of its maximum value represents the translation between the two images; the maximum value is... The translation transformation matrix for fine matching is: .

[0212] The final fine registration transformation matrix is: .

[0213] .

[0214] In high-resolution space, This indicates the amount of translation of the image in the horizontal and vertical directions. This represents the image after transformation; Represents the coordinates after the inverse transformation, and gives... The corresponding pixel position in the image It will undergo fine registration and transformation to align with the image. The aligned image is obtained from the above.

[0215] The third image was stitched together using the same steps.

[0216] Step 9: Image fusion processing.

[0217] Wavelet transform is used to perform pixel-level fusion processing on the overlapping areas of adjacent images, eliminating stitching seams and obtaining a smooth and seamless panoramic image.

[0218] The original image to be fused is subjected to wavelet transform and then decomposed into different feature domains (high-frequency components and low-frequency components) in different frequency bands. Combining the characteristics of the high-frequency components and low-frequency components, image fusion is performed separately to form a new wavelet pyramid structure. Then, inverse wavelet transform is performed on it to synthesize the fused image. The corresponding pixels in each sub-band image of each layer and each direction are fused to generate fused sub-band images. Finally, inverse wavelet transform is performed on the fused sub-band image sequence to reconstruct the fused image.

[0219] The obtained image is enhanced by wavelet transform algorithm. Based on the multi-resolution analysis principle of wavelets, the image is subjected to multi-level two-dimensional discrete wavelet transform, which decomposes the image into low-frequency sub-band of image approximate signal and high-frequency sub-band of image detail signal. Nonlinear image enhancement is performed on the low-frequency sub-band, and wavelet denoising is performed on the high-frequency part to reduce the impact of noise on the image and enhance the image contrast. Finally, the enhanced image is obtained by wavelet reconstruction, and the display unit displays the processed image, outputting a complete wide-area detection image.

[0220] This embodiment combines target information obtained by overlapping different fields of view, and uses image processing algorithms to stitch together and reconstruct visible light / shortwave infrared polarization images from different fields of view. This enables the detection of spatial targets in a large field of view, outputs a complete wide-area polarization detection image, improves image registration speed, and enhances the recognition effect of spatial targets against a space background.

[0221] Based on the same inventive concept, this application also provides a method for multi-dimensional space target detection to implement the aforementioned multi-dimensional space target detection device. The method includes: acquiring polarization images at different field-of-view angles; the polarization images at different field-of-view angles are determined by a multi-spectral polarization imaging subsystem based on multi-field-of-view optical paths; the multi-field-of-view optical paths include the optical paths reflected or emitted by the space target at different field-of-view angles. The polarization images at different field-of-view angles are stitched and fused using an image processing algorithm to obtain and display a wide-area detection image.

[0222] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0223] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A multi-dimensional detection device for space targets, characterized in that, The multi-dimensional detection device for space targets includes: a tracking turntable subsystem, a telescope subsystem, a multi-band polarization imaging subsystem, and an image processing and display subsystem; the multi-band polarization imaging subsystem and the image processing and display subsystem are connected; the telescope subsystem and the multi-band polarization imaging subsystem are integrated on the tracking turntable subsystem; The tracking turntable subsystem is used to achieve rotation within a set angle range, thereby driving the telescope subsystem and the multi-band polarization imaging subsystem to rotate. The telescope subsystem is used to acquire multi-field-of-view optical paths and incident the multi-field-of-view optical paths onto the multi-spectral polarization imaging subsystem; the multi-field-of-view optical paths include optical paths reflected or emitted by spatial targets under different field-of-view angles; The multi-band polarization imaging subsystem is used to determine polarization images at different field angles based on the multi-field-of-view optical path; The image processing and display subsystem is used to stitch and fuse polarization images under different field of view angles based on image processing algorithms to obtain a wide-area detection image and display the wide-area detection image; The image processing and display subsystem includes an information processing unit and a display unit; the information processing unit is connected to both the multi-band polarization imaging subsystem and the display unit. The information processing unit is used for: The median filtering algorithm is used to sharpen polarized images under different field of view to obtain images to be stitched under different field of view. Images to be stitched together under different field of view angles are stitched together using wavelet transform, scale-invariant feature transform, and particle swarm optimization algorithms to obtain stitched images. The wide-area detection image is obtained by fusing overlapping regions of adjacent images in the stitched image based on the wavelet transform algorithm. The display unit is used to display the wide-area detection image; The specific processing procedure of the information processing unit is as follows: Polarization images are acquired from different viewpoints; the polarization images include: polarization degree images and polarization angle images; The variance-based median filtering algorithm preprocesses polarization images obtained from different sensors to obtain images to be stitched under different field of view angles. The wavelet transform algorithm is used to decompose the image to be stitched into low-frequency and high-frequency components. Feature points and descriptors are extracted using the scale-invariant feature transform algorithm on low-frequency components; The feature point matching results are used as the initial solution of the particle swarm optimization algorithm, and the particles are encoded based on the particle swarm optimization algorithm. Each particle represents a set of registration parameters; the registration parameters are the translation and rotation parameters of the image. Calculate the fitness function, update the particle velocity and position, and update the optimal position and global optimal position of the particles based on the fitness. Determine whether the termination condition is met, output the best matching parameters, apply image transformation, and achieve image registration; Wavelet transform is used to fuse the overlapping areas of adjacent images, eliminating stitching seams and obtaining a smooth and seamless panoramic image. The obtained panoramic image is enhanced by an algorithm to generate a wide-area detection image; The telescope subsystem includes: a first front telescope unit, a second front telescope unit, and a third front telescope unit arranged in an equilateral triangle. The first front telescope unit is used to incident a first optical path onto the multi-spectral polarization imaging subsystem; the first optical path is the optical path reflected or emitted by a spatial target at a first field of view. The second front telescope unit is used to incident the second optical path onto the multi-spectral polarization imaging subsystem; the second optical path is the optical path reflected or emitted by the spatial target under the second field of view. The third front telescope unit is used to incident the third optical path onto the multi-band polarization imaging subsystem; the third optical path is the optical path reflected or emitted by the spatial target under the third field of view. The first optical path, the second optical path, and the third optical path are parallel.

2. The multi-dimensional detection device for space targets according to claim 1, characterized in that, The multi-band polarization imaging subsystem includes: a spectral polarization state modulation module, a correction module, and an image acquisition module; the spectral polarization state modulation module is disposed on the output optical path of the telescope subsystem; the correction module is disposed on the output optical path of the spectral polarization state modulation module; and the image acquisition module is disposed on the output optical path of the correction module. The spectral polarization state modulation module is used to generate polarization state modulated output light at different field angles according to the multi-field-of-view optical path; The correction module is used to correct the polarization state modulated output light under different field of view angles to obtain the corrected output light under different field of view angles; The image acquisition module is used to image the corrected outgoing light at different field of view angles to obtain polarization images at different field of view angles.

3. The multi-dimensional detection device for space targets according to claim 2, characterized in that, The spectral polarization state modulation module includes: a first polarization state modulation module, a second polarization state modulation module, and a third polarization state modulation module; the optical axis of the first polarization state modulation module is set on the output optical path of the first front telescope unit, the optical axis of the second polarization state modulation module is set on the output optical path of the second front telescope unit, and the optical axis of the third polarization state modulation module is set on the output optical path of the third front telescope unit. The first polarization state modulation module is used to generate polarization state modulated outgoing light at a first field of view according to the first optical path; The second polarization modulation module is used to generate polarization-modulated outgoing light at a second field of view according to the second optical path; The third polarization modulation module is used to generate polarization-modulated outgoing light at a third field of view according to the third optical path.

4. The multi-dimensional detection device for space targets according to claim 3, characterized in that, The first polarization modulation module includes: a first polarization modulation unit and a first filtering unit; the first polarization modulation unit is disposed on the output optical path of the first front-mounted telescope unit; the first filtering unit is disposed on the output optical path of the first polarization modulation unit; The first polarization modulation unit is used to generate linearly polarized light at a first field of view according to the first optical path; The first filtering unit is used to filter the linearly polarized light under the first field of view to obtain the polarized state modulated outgoing light under the first field of view.

5. The multi-dimensional detection device for space targets according to claim 3, characterized in that, The correction module includes: a first adaptive optics module, a second adaptive optics module, and a third adaptive optics module; the first adaptive optics module is disposed on the output optical path of the first polarization state modulation module; the second adaptive optics module is disposed on the output optical path of the second polarization state modulation module; and the third adaptive optics module is disposed on the output optical path of the third polarization state modulation module. The first adaptive optics module is used to remove atmospheric interference information from the polarization-modulated output light under the first field of view to obtain the corrected output light under the first field of view; The second adaptive optics module is used to remove atmospheric interference information from the polarization-modulated outgoing light under the second field of view, so as to obtain the corrected outgoing light under the second field of view; The third adaptive optics module is used to remove atmospheric interference information from the polarization-modulated outgoing light under the third field of view, so as to obtain the corrected outgoing light under the third field of view.

6. The multi-dimensional detection device for space targets according to claim 5, characterized in that, The first adaptive optics module includes: a first beam splitter, a first wavefront correction unit, a first wavefront sensing unit, and a first wavefront control unit; the first wavefront sensing unit and the first wavefront control unit are connected. The first wavefront correction unit includes: a deformable mirror and a deformable mirror driver; the deformable mirror driver is connected to the deformable mirror and the first wavefront control unit respectively; the deformable mirror is disposed in the output optical path of the first polarization state modulation module; the first beam splitting unit is disposed in the output optical path of the deformable mirror, and the first wavefront sensing unit is disposed in the output optical path of the first beam splitting unit; The deformable mirror is used to deform the polarization-modulated outgoing light under the first field of view to generate deformed outgoing light under the first field of view. The first beam splitting unit is used to split the deformed outgoing light under the first field of view into two beams, which enter the first wavefront sensing unit and the image acquisition module respectively. The first wavefront sensing unit is used to: measure the deformed outgoing light under the first field of view in real time to obtain the first wavefront distortion; the first wavefront distortion is the light wave distortion generated by the deformed outgoing light under the first field of view under the influence of atmospheric interference; The first wavefront control unit is used to generate a control signal based on the first wavefront distortion and send it to the deformable mirror driver; The deformable mirror driver is used to change the shape of the deformable mirror according to the control signal to generate corrected outgoing light at a first field of view.

7. The multi-dimensional detection device for space targets according to claim 5, characterized in that, The image acquisition module includes: a first image acquisition module, a second image acquisition module, and a third image acquisition module connected to the image processing and display subsystem; the first image acquisition module is disposed on the output optical path of the first adaptive optics module; the second image acquisition module is disposed on the output optical path of the second adaptive optics module; and the third image acquisition module is disposed on the output optical path connected to the third adaptive optics module. The first image acquisition module is used to image the corrected outgoing light under the first field of view to obtain a polarization image under the first field of view; The second image acquisition module is used to image the corrected outgoing light under the second field of view to obtain a polarization image under the second field of view; The third image acquisition module is used to image the corrected outgoing light under the third field of view to obtain a polarization image under the third field of view.

8. A method for multi-dimensional detection of space targets, characterized in that, The space target multidimensional detection method is used in the space target multidimensional detection device according to any one of claims 1-7; The multi-dimensional detection method for space targets includes: The polarization images under different field of view angles are acquired; the polarization images under different field of view angles are determined by the multi-spectral polarization imaging subsystem based on the multi-field-view angle optical path; the multi-field-view angle optical path includes the optical path reflected or emitted by the spatial target under different field of view angles; Based on image processing algorithms, polarization images under different field of view are stitched and fused to obtain a wide-area detection image, which is then displayed.

Images