Foundation high-resolution celestial body imaging method based on lucky imaging

Abstract

The invention discloses a ground-based high-resolution celestial body imaging method based on lucky imaging, relates to the field of astronomical observation, and is technically characterized by aiming at the problems of limited application scene and the like of a lucky imaging method based on preceding-stage adaptive optical correction. By reducing the number of the required pupil planes, the image signal-to-noise ratio of the beacon light is improved, the requirement on the guide star brightness is reduced, and the observable astronomical target range is expanded. Furthermore, an initial value of an iterative algorithm is replaced by a phase solved by a historical frame, so that the convergence speed of the iterative algorithm is improved. According to the iterative algorithm, the residual wavefront with relatively low aberration can be obtained through single-step iteration, so that the time bandwidth is improved while the correction precision of the system is ensured, and meanwhile, the lucky frame proportion of a post-stage lucky imaging system is improved. The method has the advantages that the preceding-stage adaptive optical system has higher correction efficiency, the complexity of an off-pupil image acquisition system is reduced, and compared with a traditional method, the method has excellent image quality improvement capability under various intensity turbulent flow fields, and is suitable for obtaining high-resolution astronomical images under the conditions of limited guide star intensity and strong turbulent flow fields.

Description

Technical Field

[0001] The invention belongs to the field of astronomical observation, and in particular relates to a ground-based high-resolution celestial body imaging method based on lucky imaging. Background Art

[0002] In the process of ground-based astronomical observation, increasing the aperture of the imaging system is the main means to improve its imaging spatial resolution. However, due to the influence of atmospheric turbulence, the imaging resolution of the system decreases significantly with the increase of the aperture and cannot reach the theoretical diffraction limit. The adaptive optical method that uses the optomechanical system to correct the wavefront distorted by atmospheric turbulence in real time can effectively improve the imaging resolution of the system. However, the beam corrected by the adaptive optical system still has residual distortion to varying degrees, so the aperture advantage of the system cannot be fully utilized. In order to solve the problem of image quality degradation caused by the imperfect correction of the adaptive optical system, people have proposed the idea of ​​further fusion processing of the image carrying residual distortion, thereby developing the lucky imaging technology based on the correction of the previous adaptive optical system.

[0003] Lucky imaging technology evaluates and screens the image quality of a series of short-exposure images taken in time sequence, and fuses the image frames with less distortion to obtain higher quality imaging effects. As an adaptive optical system for pre-correction, its correction efficiency determines the operating frequency and image quality improvement effect of the subsequent lucky imaging system. The adaptive optical system based on nonlinear wavefront curvature sensing technology has significant advantages in terms of wavefront sensing and the dynamic range of distortion correction. However, its iterative wavefront sensing algorithm has problems such as high time consumption and easy to fall into local optimal solutions, which affects the time bandwidth of the entire adaptive optical system and the working efficiency of the subsequent lucky imaging method.

[0004] The high time consumption of the iterative algorithm of the nonlinear wavefront curvature sensing technology can be effectively alleviated by using the wavefront calculation method of single-step iteration, and the risk of falling into the local optimal solution is also reduced. However, the limited sensing accuracy of the single-step iteration significantly increases the number of frames required for the subsequent fusion of lucky frames. When the residual wavefront distortion is too large, the lucky imaging method will fail. In addition, the four off-pupil surfaces required by the nonlinear wavefront curvature sensing technology impose strict requirements on the brightness of the guide star, resulting in a significant reduction in the number of available natural guide stars, reducing the number of celestial bodies that can be observed based on natural guide stars.

[0005] Therefore, the lucky imaging technology based on the adaptive optical system based on the traditional nonlinear wavefront curvature sensor as the pre-correction has the following disadvantages: 1. The time-consuming wavefront iteration calculation reduces the closed-loop time bandwidth of the system; 2. The limited wavefront sensing accuracy of the single-step iteration increases the risk of the lucky imaging method being effectively applied; 3. The four-pupillary structure weakens the scene adaptability of the system. Therefore, it is necessary to propose a ground-based celestial body imaging method that meets a wider range of application scenarios and needs and is more efficient. Summary of the invention

[0006] The present invention proposes a ground-based high-resolution celestial body imaging method based on lucky imaging. In order to improve the adaptability of the method to weak guide star intensity scenes, the traditional four-pupillary plane structure is replaced by a double-pupillary plane structure. Correspondingly, the iterative algorithm is modified to be based on a dual-plane form, and the initial value of the iteration is replaced by the wavefront phase solved in the previous frame from the random matrix, and the inter-frame correlation of the turbulent time series change is used to improve the solution efficiency of the dual-plane iterative algorithm. In order to improve the time bandwidth of the front-stage adaptive optical system, the convergence condition of the wavefront iterative algorithm is modified to a single-step iteration. The rear-stage lucky imaging system obtains a high-resolution fused image after image quality evaluation, image denoising and image superposition steps according to a specific lucky frame ratio. Compared with the existing method, the optical path structure involved in the present invention is simple, and the wavefront iterative algorithm can improve the wavefront sensing accuracy while maintaining the speed advantage, and can also be applied in certain weak guide star scenes and strong turbulent fields.

[0007] The specific technical solution of the present invention is:

[0008] A ground-based high-resolution celestial body imaging method based on lucky imaging, characterized in that the method comprises the following steps:

[0009] S1. Use the beacon light out-of-pupil image acquisition module in the previous adaptive optical system to collect two out-of-pupil intensity images as information sampling of the distorted wavefront of the current frame;

[0010] S2. Input the two off-pupil intensity matrices and the wavefront matrix solved in the previous frame into the dual-plane GS (Gerchberg-Saxton) algorithm, and obtain the phase distribution at the pupil plane of the current frame through a single-step iteration;

[0011] S3. The wavefront phase is analyzed by the control system, and a corresponding correction voltage signal is output to the wavefront correction module, and finally the imaging module perceives the corrected image;

[0012] S4. Then, subsequent lucky imaging analysis is performed according to the input lucky frame ratio. First, the time-series image data obtained by the imaging module is stored in the memory, and image quality evaluation analysis such as peak signal-to-noise ratio and mean square error is performed on all frames. Then, the lucky frames with less turbulent distortion are screened according to the input lucky frame ratio, and then certain image noise reduction processing is performed on them. The grayscale peak points of each lucky frame image are located, and then image registration and cropping are performed on all lucky frames with each peak point as the center, and finally all lucky frames are superimposed to obtain a high-resolution image.

[0013] The beacon light out-of-pupil image acquisition module in step S1 can be a high-photon utilization optical path based on an electrically controlled zoom lens, or can be a real-time acquisition optical path based on a branching system.

[0014] The iterative plane sequence of the dual-plane GS algorithm in step S2 is: pupil plane, short-distance pupil plane, long-distance pupil plane and pupil plane.

[0015] The wavefront correction module in step S3 may be a wavefront shaping device such as a deformable mirror or a spatial light modulator.

[0016] The lucky frame ratio in step S4 is usually determined by the turbulence field intensity. As the turbulence field intensity increases, the lucky frame ratio decreases accordingly.

[0017] The lucky imaging analysis in step S4 can be implemented by MATLAB, Python, etc., or by using software such as AutoStakkert.

[0018] The beneficial effect of the present invention is that in the face of the actual use needs of actual high-resolution celestial imaging systems, a ground-based high-resolution celestial imaging method based on lucky imaging is proposed. The dual-plane nonlinear wavefront curvature sensing technology involved in this method has lower requirements on the guide star strength than the traditional four-plane solution, and its pupil image acquisition optical path is simpler. Compared with the traditional algorithm, the wavefront calculation efficiency of this method has lower residual errors under single-step constraints, which can effectively improve the lucky frame ratio, so that the system can still achieve high-resolution image fusion under strong turbulent fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of this application, and do not constitute a limitation of the embodiments of the present invention. In the drawings:

[0020] Figure 1 It is a flow chart of a ground-based high-resolution celestial body imaging method based on lucky imaging in an embodiment of the present invention;

[0021] Figure 2 This is a data processing flow chart of the lucky imaging method in an embodiment of the present invention;

[0022] Figure 3 A conceptual schematic diagram of the optical-mechanical structure of the system in an embodiment of the present invention;

[0023] Figure 3 Middle: dotted lines indicate signal transmission between corresponding devices;

[0024] Figure 3 In: ① refers to the wavefront sensing module, and in this embodiment specifically refers to the beacon light out-of-pupil image acquisition optical path;

[0025] Figure 3 In: ② refers to the wavefront correction module, which can be a wavefront shaping device such as a deformable mirror or a spatial light modulator;

[0026] Figure 3 In: ③ refers to image sensors such as CCD or CMOS;

[0027] Figure 4 Schematic diagram of point spread function comparison of the imaging system under different intensities of turbulence fields and different combinations of adaptive optics modes and lucky imaging modes in an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments and drawings. The exemplary embodiments of the present invention and their description are only used to explain the present invention and are not intended to limit the present invention.

[0029] A ground-based high-resolution celestial body imaging method based on lucky imaging, characterized in that the method comprises the following steps;

[0030] S1. Use the beacon light out-of-pupil image acquisition module in the previous adaptive optical system to collect two out-of-pupil intensity images as information sampling of the distorted wavefront of the current frame;

[0031] S2. Input the two off-pupil intensity matrices and the wavefront matrix solved in the previous frame into the dual-plane GS (Gerchberg-Saxton) algorithm, and obtain the phase distribution at the pupil plane of the current frame through a single-step iteration;

[0032] S3. The wavefront phase is analyzed by the control system, and a corresponding correction voltage signal is output to the wavefront correction module, and finally the imaging module perceives the corrected image;

[0033] S4. Then, subsequent lucky imaging analysis is performed according to the input lucky frame ratio. First, the time-series image data obtained by the imaging module is stored in the memory, and image quality evaluation analysis such as peak signal-to-noise ratio and mean square error is performed on all frames. Then, the lucky frames with less turbulent distortion are screened according to the input lucky frame ratio, and then certain image noise reduction processing is performed on them. The grayscale peak points of each lucky frame image are located, and then image registration and cropping are performed on all lucky frames with each peak point as the center, and finally all lucky frames are superimposed to obtain a high-resolution image.

[0034] The beacon light out-of-pupil image acquisition module in step S1 can be a high-photon utilization optical path based on an electrically controlled zoom lens, or can be a real-time acquisition optical path based on a branching system.

[0035] The iterative plane sequence of the dual-plane GS algorithm in step S2 is: pupil plane, short-distance pupil plane, long-distance pupil plane and pupil plane.

[0036] The wavefront correction module in step S3 may be a wavefront shaping device such as a deformable mirror or a spatial light modulator.

[0037] The lucky frame ratio in step S4 is usually determined by the turbulence field intensity. As the turbulence field intensity increases, the lucky frame ratio decreases accordingly.

[0038] The lucky imaging analysis in step S4 can be implemented by MATLAB, Python, etc., or by using software such as AutoStakkert.

[0039] Embodiment 1: The workflow of a ground-based high-resolution celestial body imaging method based on lucky imaging is as follows:

[0040] according to Figure 1 As described in step S1 of the embodiment, the two beacon light off-pupil images required by the dual-plane nonlinear wavefront sensing method are obtained by using the off-pupil image acquisition optical path. According to the scene requirements, a high photon utilization optical path based on an electrically controlled zoom lens or a real-time acquisition optical path based on a branching system can be selected.

[0041] according to Figure 1 As described in step S2 in the above, the two collected off-pupil intensity matrices and the wavefront matrix solved in the previous frame are input into the dual-plane GS algorithm, and the iteration termination condition is set to a single-step iteration, thereby finally realizing high-speed wavefront sensing.

[0042] according to Figure 1 As described in step S3 of FIG. 1 , the wavefront phase obtained by iterative calculation is input into the control system and analyzed, and then the corresponding correction voltage signal is output to the wavefront correction module. The distorted light beam incident on a wavefront shaping device such as a deformable mirror or a spatial light modulator is compensated for the corresponding phase, and finally the imaging module perceives the corrected image.

[0043] according to Figure 1 Step S4 and Figure 2 As described, subsequent lucky imaging analysis is performed according to the input lucky frame ratio. First, the time-series image data obtained by the imaging module is stored in the memory, and the total number of frames required is determined according to the lucky frame ratio. Image quality evaluation analysis is performed on all frames, such as peak signal-to-noise ratio or mean variance, etc., and lucky frames with less turbulent distortion are screened according to the input lucky frame ratio, and then certain image noise reduction processing is performed on all lucky frames. Subsequently, the grayscale peak point of each lucky frame image is determined and used as the image fusion positioning point, and then image registration and cropping are performed on all lucky frames with the positioning point as the center, and finally all lucky frames are superimposed to obtain a high-resolution image. The lucky frame probability formula containing the target diffraction limit information is:

[0044] P≈5.6exp[-0.1557(D / r0) 2 ],D / r0≥3.5

[0045] Where D is the diameter of the telescope's primary mirror and r0 is the atmospheric coherence length.

[0046] The hardware and software devices used in the method involved in the present invention are: Windows 11 operating system, AMD Ryzen 75800H with Radeon Graphics CPU, MATLAB 2023a simulation platform, and AutoStakkert software.

[0047] To further verify the actual performance of the ground-based high-resolution celestial imaging method based on lucky imaging, we compared the point spread function of the imaging system under turbulent fields of different intensities. Different combinations of "adaptive optics mode" and "lucky imaging mode" were used for comparison. The comparison results are shown in Figure 2. Figure 4 As shown in Figure 2. According to the convolution theory of imaging, the uniformity of the point spread function of the system directly determines the actual resolution of the imaging system. Figure 4 , the lucky imaging mode is turned off, and the single-step iterative correction effect of the traditional two-plane and four-plane methods is poor when the turbulence intensity is greater than 16, and the lucky imaging mode does not contribute much to the improvement of imaging resolution. However, the single-step correction mode based on the historical frame two-plane method has a significant correction effect in all turbulence intensity scenarios. And after the lucky imaging mode is turned on, the imaging resolution is further improved.

[0048] In summary, the present invention relates to a ground-based high-resolution celestial imaging method based on lucky imaging. In the application scenario of observing celestial targets by the lucky imaging method based on the front-stage adaptive optics correction, only two off-pupil beacon light intensity maps are required to obtain a low residual aberration image through single-step iteration, and the imaging resolution is further improved by lucky imaging technology processing. The dual-plane based iterative algorithm proposed in the present invention effectively reduces the complexity of the system optical path structure, and also reduces the system's requirements for the beacon light intensity, improves the overall working time bandwidth of the cascade system, and significantly increases the number of celestial targets observable based on natural guide stars. In actual application scenarios, the traditional method has limited image quality improvement capabilities in strong turbulence scenes, while the method involved in the present invention has stronger robustness in turbulent environments.

Claims

1. A ground-based high-resolution celestial body imaging method based on lucky imaging, characterized in that The method comprises the following steps: S1. Use the beacon light out-of-pupil image acquisition module in the previous adaptive optical system to collect two out-of-pupil intensity images as information sampling of the distorted wavefront of the current frame; S2. Input the two off-pupil intensity matrices and the wavefront matrix solved in the previous frame into the dual-plane GS (Gerchberg-Saxton) algorithm, and obtain the phase distribution at the pupil plane of the current frame through a single-step iteration; S3. The wavefront phase is analyzed by the control system, and a corresponding correction voltage signal is output to the wavefront correction module, and finally the imaging module perceives the corrected image; S4. Then, subsequent lucky imaging analysis is performed according to the input lucky frame ratio. First, the time-series image data obtained by the imaging module is stored in the memory, and image quality evaluation analysis such as peak signal-to-noise ratio and mean square error is performed on all frames. Then, the lucky frames with less turbulent distortion are screened according to the input lucky frame ratio, and then certain image noise reduction processing is performed on them. The grayscale peak points of each lucky frame image are located, and then image registration and cropping are performed on all lucky frames with each peak point as the center, and finally all lucky frames are superimposed to obtain a high-resolution image.

2. The method of ground-based high-resolution celestial body imaging based on lucky imaging according to claim 1, characterized in that: The beacon light out-of-pupil image acquisition module in step S1 can be a high-photon utilization optical path based on an electrically controlled zoom lens, or can be a real-time acquisition optical path based on a branching system.

3. The method for ground-based high-resolution celestial body imaging based on lucky imaging according to claim 1, characterized in that: The iterative plane sequence of the dual-plane GS algorithm in step S2 is: pupil plane, short-distance pupil plane, long-distance pupil plane and pupil plane.

4. The method of ground-based high-resolution celestial body imaging based on lucky imaging according to claim 1, characterized in that: The wavefront correction module in step S3 may be a wavefront shaping device such as a deformable mirror or a spatial light modulator.

5. The method of ground-based high-resolution celestial body imaging based on lucky imaging according to claim 1, characterized in that: The lucky frame ratio in step S4 is usually determined by the turbulence field intensity. As the turbulence field intensity increases, the lucky frame ratio decreases accordingly.

6. The method of ground-based high-resolution celestial body imaging based on lucky imaging according to claim 1, characterized in that: The lucky imaging analysis in step S4 can be implemented by MATLAB, Python, etc., or by using software such as AutoStakkert.

Images