Multi-scene high-precision foundation optical imaging processing method based on focal length inverse solution and degradation processing

Abstract

The invention discloses a multi-scene high-precision foundation optical imaging processing method based on focal length inverse solution and degradation processing, and belongs to the technical field of high-precision space target observation and astronomical research. The method comprises the following steps: collecting actual parameters under different orbit conditions to build a simulation space environment, adding targets according to actual requirements, simulating motion laws of space targets under different orbit conditions, and exporting a short-exposure image based on orbit dynamics and changes of an observation environment; performing superposition operation on the exported short-exposure images according to a time sequence to simulate a long-exposure image; constructing a degradation model, and performing degradation processing on the short-exposure image and the long-exposure image; and based on a peer-to-peer relationship between a sensor measurement view field and an actual observation view field and a camera imaging principle, reversely solving a focal length, and calculating angle measurement information of the space target in the degraded image. The method not only can obtain a high-quality ground-based image, but also can be used for generating a space-based image.

Description

Technical Field

[0001] The present invention belongs to the technical field of high-precision space target observation and astronomy, and particularly relates to a multi-scene high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing. Background Art

[0002] With the continuous development of the fields of space science and astronomical observation, the demand for high-precision space target observation is increasing day by day. As an important observation tool, ground-based optical imaging systems are widely used in astronomical observation, satellite tracking, space target positioning and other fields. However, in practical applications, ground-based optical imaging systems need to capture real images. Due to complex environments and demanding geographical conditions, this method faces challenges such as high simulation costs, strict observation conditions, inability to obtain the true value of the target position, and low image diversity. Currently, the mainstream methods of space target simulation mainly fall into two categories: one is represented by Satellite Tool Kit (STK), which generates images by building a scene; the other is represented by Satsim, which adds newly rendered space targets to existing space images.

[0003] When the first type of observation method represented by STK performs ground-based image simulation, it is necessary to build a simulation scene and introduce additional optoelectronic modules to set parameter information. However, this method has extremely high video memory occupancy and it is difficult to obtain high-quality images under continuous processes. If additional optoelectronic modules are not introduced, although the original 3D window can directly export ground-based images, they will lack parameter information such as focal length, and the exported images cannot be used to determine the observation angle, and the image data is difficult to be used for subsequent orbit determination research. In addition, many simulation engines represented by STK highly rely on fixed focal length parameters and idealized optical conditions when generating ground-based images, and it is difficult to achieve a good balance between parameter measurability and high-quality images.

[0004] The second type of observation method represented by Satsim is to add newly rendered space targets to existing space images. Different from the first type of method that highly relies on scene building, this method realizes simulation under various observation conditions by dynamically adding information such as the trajectory, position, and speed of the target, and is particularly suitable for real-time tracking and observation of high-precision space targets. Although many simulators represented by Satsim can flexibly simulate target observations in different environments, in the processing of important parameters such as the focal length and field of view angle of the optical imaging system, the simulation accuracy is difficult to guarantee. In addition, this type of method cannot directly export the trajectory information of space targets. Although it can provide better image quality, the operation is complex, and the loss of parameter information is a pain point that is difficult to solve. Summary of the Invention

[0005] To solve the above technical problems, the present invention provides a multi-scenario high-precision ground-based optical imaging processing method based on inverse focal length solution and degradation processing. Through the deep integration of multi-scenario modeling, inverse focal length parameter solution, and degradation models, the accuracy and authenticity of ground-based optical imaging simulation are significantly improved. Without the need to introduce additional plugins, it overcomes the limitation that the STK software cannot directly set the focal length, significantly reduces the computational resource occupancy of the optoelectronic module, and has high practical value.

[0006] To achieve the above object, the present invention proposes the following technical solutions:

[0007] A multi-scenario high-precision ground-based optical imaging processing method based on inverse focal length solution and degradation processing, the method comprising:

[0008] Step 1, build diverse simulation scenarios: Collect actual parameters under different orbital conditions to build a simulation space environment, add space targets according to actual needs, simulate the motion laws of space targets under different orbital conditions, and based on the changes in orbital dynamics and observation environments, export short-exposure images;

[0009] Step 2, sequence image superposition and degradation model processing: Perform superposition operations on the exported short-exposure images in chronological order to simulate long-exposure images; construct a degradation model and perform degradation processing on the exported short-exposure images and the obtained long-exposure images;

[0010] Step 3, inverse solve the focal length and solve parameters: Based on the equivalence relationship between the sensor measurement field of view and the actual observation field of view and the camera imaging principle, inverse solve the focal length, and calculate the angular measurement information of the space target in the degraded short-exposure images and long-exposure images according to the position coordinates of the space target in the image.

[0011] On the other hand, the present invention provides a multi-scenario high-precision ground-based optical imaging processing device based on inverse focal length solution and degradation processing, comprising:

[0012] A preprocessing module for collecting actual parameters under different orbital conditions to build a simulation space environment, adding space targets according to actual needs, simulating the motion laws of space targets under different orbital conditions, and based on the changes in orbital dynamics and observation environments, exporting short-exposure images;

[0013] A degradation module for performing superposition operations on the exported short-exposure images in chronological order to simulate long-exposure images; constructing a degradation model and performing degradation processing on the exported short-exposure images and the obtained long-exposure images;

[0014] A calculation module, configured to inversely solve the focal length based on the equivalence relationship between the sensor-measured field of view and the actual observed field of view and the camera imaging principle, and calculate the angular measurement information of the space target in the short-exposure image and the long-exposure image after the degradation processing according to the position coordinates of the space target in the image.

[0015] In a third aspect, the present invention provides an electronic device, including: one or more processors; a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the foregoing multi-scenario high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing.

[0016] In a fourth aspect, the present invention provides a computer-readable storage medium, on which executable instructions are stored, and when the instructions are executed by a processor, the processor can be caused to implement the foregoing multi-scenario high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing.

[0017] The beneficial effects of the present invention are as follows:

[0018] Without introducing additional plugins and minimizing computing resources to the greatest extent, an independent high-quality ground-based observation image acquisition and target parameter calculation process based on STK is created; the scene construction parameters are derived from real data, deeply integrating the degradation model, considering the interference of equipment, environment, etc. on the imaging process, which is real and reliable; this method covers two processing methods under long and short exposures, and can not only obtain high-quality ground-based images, but also be used to generate space-based images. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a flowchart of a multi-scenario high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing of the present invention;

[0020] Figure 2 is a schematic diagram of the principle of focal length inverse solution and angular measurement calculation process;

[0021] Figure 3 is a comparison diagram of the effects before and after the improvement of the STK-based method, where (a) is the directly exported effect diagram, (b) is the effect diagram after superposition, and (c) is the effect diagram after degradation processing. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] The present invention will be further described below with reference to the drawings and embodiments.

[0023] As Figure 1As shown, the present invention proposes a multi-scene high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing, which is divided into three steps: building a variety of simulation scenes, superimposing sequential images and fusing image degradation models, and inversely solving focal length parameters and combining imaging principles to obtain image angle measurement information. In the process of generating images, the above three steps are repeated to obtain multi-scene high-precision and high-quality ground-based optical images in a simulated environment. Specifically, the ground-based optical image uses the earth as a reference benchmark, and the conversion from the image coordinate system to the space coordinate system can be achieved through the internal and external parameters of the camera. The most common method for setting camera parameters in STK is to introduce a photoelectric module, but the introduction of a photoelectric module in STK not only seriously occupies computer memory resources, but also makes it difficult to complete the simulation of complex scenes, and the image quality is low; in addition, the STK software itself generates scenes based on ideal optical conditions, and can only export short-exposure images. Therefore, in order to make the obtained image both authentic and diverse, the present invention adopts the processing operation means of superimposing sequential images and fusing degradation models, so that the obtained image is more authentic and reliable. The method specifically includes:

[0024] Step 1. Build a variety of simulation scenarios: collect actual parameters under different orbits (such as Starlink, Beidou, etc.) to build a simulated space environment, add targets according to actual needs, and simulate the motion laws of space targets under different orbital conditions. Combine the different orbital characteristics of low orbit, medium orbit and high orbit, design a variety of typical target motion modes, and export short exposure images based on changes in orbital dynamics and observation environment;

[0025] In the process of scene generation, the present invention adopts the joint debugging method of Matlab and STK to construct four optional scenes to simulate different orbit satellite systems: Scene one focuses on low-orbit satellites, and simulates complex low-orbit satellite group behavior by setting up multiple high-density constellations to enhance the complexity of space motion; Scene two focuses on medium-orbit satellites, which are suitable for positioning and navigation applications in view of their smaller number and larger coverage; Scene three concentrates on high-orbit satellites, simulating geostationary orbits and relay missions to ensure wide coverage and communication capabilities; Scene four combines low-orbit, medium-orbit and high-orbit satellites to comprehensively study the interaction and collaborative work between satellites in different orbits, highlighting the collaborative mechanism of satellites in different orbits in the space environment.

[0026] To further enhance the complexity, dynamics and authenticity of the simulation environment, the present invention introduces slight random changes in orbital altitude and inclination with a standard deviation of 10 km and 1 degree to simulate orbital changes caused by factors such as fuel consumption and external disturbances, providing an innovative simulation platform for the study of multi-orbit satellite systems.

[0027]

[0028] : It represents the orbital altitude of the th satellite at the moment. : It represents the change in orbital altitude of the th satellite at the moment. Similarly, : It represents the orbital inclination of the th satellite at the moment. : It represents the change in orbital inclination of the th satellite at the moment.

[0029] Based on the STK simulation engine, after completing the above basic parameter settings, denote the vertical half - angle of the rectangular field of view as , the horizontal half - angle of the rectangular field of view as , and the angle of the three - dimensional window observation view as . This step establishes the equivalence relationship between the sensor detection field of view and the actual observation field of view on the premise of ensuring that the shape of the sensor field of view is adjusted to a square, that is, the above angles satisfy the following relationship:

[0030] ,

[0031] Different scenarios can be selected or corresponding parameters can be modified according to different research tasks. The Matlab co - debugging STK method adopted in this step has good universality, can reduce the maintenance cost while improving the modification efficiency, and export multi - scenario short - exposure images that meet the requirements.

[0032] Step 2: Sequence image superposition and degradation model processing: Superpose the sequence images under short exposure in chronological order to simulate long - exposure images; construct a degradation model to perform degradation processing on the exported short - exposure images and the obtained long - exposure images, including noise simulation, uneven brightness, blurring effects, etc.

[0033] In practice, long - exposure images can capture information throughout the entire time period and the shape information of the captured targets is more obvious. However, STK can only export short - exposure images and cannot directly export long - exposure image sequences. To explore the imaging characteristics under different modes, combined with the short - exposure images exported in Step 1, the present invention adopts the method of sequence image superposition to synthesize the exported short - exposure image sequence to obtain long - exposure images

[0034] ,

[0035] where, I i(x, y) are the pixel coordinates of the short-exposure image. This step performs pixel clipping on the superimposed result to ensure that all pixel values are within the range of [0, 255]. To enhance the authenticity of the image, a degradation model is constructed to degrade the exported short-exposure image and the superimposed long-exposure image, such as adding noise and uneven brightness processing, so as to more accurately simulate the actual imaging process. Among them, the degradation processing methods of the two are similar:

[0036] ,

[0037] where, is the pixel point of the image without degradation processing, is the pixel point of the image after degradation processing, is the consideration of the noise mixing component by the degradation model, is the consideration of the uneven brightness of shooting by the degradation model. Combining the characteristics of the sensor and the image propagation process, a noise mixing model is constructed:

[0038] ,

[0039] where, represent Gaussian noise, Poisson noise, and salt-and-pepper noise respectively, represents the weight factor corresponding to each item. To simulate uneven brightness, a distance-based attenuation function needs to be introduced to control the change of brightness:

[0040] ,

[0041] where, is the sampling point coordinate, is the width to control the gradual change of brightness. The farther the sampling point is, the more significant the attenuation. The feathering effect realizes the maximum brightness in the central area of the image and gradually decreases the brightness at the edge by smoothly transitioning the boundary. In addition, this step uses the method of superimposing multiple sampling points to generate brightness changes of different shapes, thereby achieving the expected ideal effect.

[0042] After completing the above two operation steps, a degraded image very close to the real scene is obtained. However, if the target monitoring task is to be completed, the angular information of the moving target needs to be solved to complete the subsequent positioning operation.

[0043] To simulate uniform defocus, the present invention uses Gaussian blur; at the same time, it combines convolution with the point spread function to reproduce the real observation blur effect.

[0044] Such as Figure 2As shown in the figure, in step 3, back-calculate the focal length and solve the parameters: Based on the equivalence relationship between the sensor measurement field of view and the actual observation field of view, as well as the camera imaging principle, back-calculate the focal length, and calculate the angular measurement information of the space target in the short-exposure image and the long-exposure image after the degradation process according to the position coordinates of the space target in the image.

[0045] Based on step 1, establish the equivalence relationship between the sensor detection field of view and the observation field of view. Assume that the size of the derived image is pixels, then the coordinates of the origin of the image coordinate system in the pixel plane coordinate system are .

[0046] Based on step 2, the long-exposure image can be obtained on the basis of the short-exposure image by using the sequence image superposition and fusion degradation process. To complete the subsequent orbit determination work, the present invention combines the camera imaging principle and the focal length back-solving method to solve the azimuth angle and elevation angle of the space target satellite in the time series.

[0047] Based on the camera imaging principle, it is known that:

[0048] f = s,

[0049] In the formula, f is the camera focal length, and s is the image distance. According to the equivalence relationship between the sensor measurement field of view and the actual observation field of view, the expression of the image distance s converted to the unit pixel in the pixel plane coordinate system is:

[0050] ,

[0051] In the formula, is the vertical half angle of the rectangular field of view set in step 1, and thus the camera focal length f can be back-calculated. Let the centroid coordinates of the observed satellite on the degraded image at any time be ( , ), is the correction term (the azimuth angle in the XYZ plane, usually referring to the angle from the positive X axis, rotating clockwise around the Z axis to the direction of the target object), then the expression of the azimuth angle of the image space target is:

[0052] ,

[0053] Using the complementary relationship between the elevation angle of the image space target and the angle with the vertical direction, the expression of the elevation angle is derived as:

[0054] ,

[0055] In the formula, is the angle between the line connecting the camera pinhole and the space target and the vertical direction, is the elevation angle of the space target. According to mathematical derivation, it can be obtained that The expression is:

[0056] = ,

[0057] A series of actual pitch angles and azimuth angles of the space target with respect to the camera coordinate system over time are obtained through the sensor. Denote the pitch angle of the space target at a certain moment as , and the azimuth angle as . Comparing with the pitch angle and azimuth angle of the image space target obtained by the above-mentioned inverse solution based on the focal length, the error is controlled within 0.5 degrees.

[0058] It is not difficult to measure the distance between the space target and the sensor through known technical means. Denote the distance between the space target and the sensor measured at a certain moment as . Combining the pitch angle and azimuth angle of the space target at this moment obtained by the above-mentioned method of using the camera imaging principle and inverse focal length solution, the rectangular coordinates of the space target at this moment in the camera coordinate system can be obtained as:

[0059] ,

[0060] The rectangular coordinates of the space target with respect to the earth coordinate system are obtained as:

[0061] ,

[0062] are the rectangular coordinates of the space target at this moment in the earth coordinate system. In the formula, R and T are the external parameters of the camera, representing the rotation matrix of the camera coordinate system with respect to the earth coordinate system and the translation matrix of the camera coordinate system with respect to the earth coordinate system respectively. Thus, the coordinates of the space target at this moment in the earth coordinate system can be obtained.

[0063] As Figure 3 shows, after building a diversified simulation scenario, the short-exposure image shown in (a) of Figure 3 can be obtained. Using the principle of sequence image superposition, the short-exposure image can be processed into the long-exposure image shown in (b) of Figure 3 . Then, a degradation model is constructed, and after degrading the obtained long-exposure image, Figure 3 Figure 3 can be obtained.The effect shown in (c) is close to the image obtained using the real scene. To study the subsequent orbit determination work, it is necessary to analyze the data contained in the image. Based on the equivalence relationship between the sensor measurement field of view and the actual observation field of view and the camera imaging principle, the angular measurement information of the space target in the image can be calculated through parameter inverse solution. Thus, a series of processes from scene construction to the superposition and degradation processing of the exported sequence images, and then to the derivation of the angular measurement information of the space target are completed. For the obtained angular measurement information, the subsequent orbit determination work can be studied using the orbit mechanics model.

[0064] On the other hand, the present invention provides a multi-scene high-precision ground-based optical imaging processing device based on focal length inverse solution and degradation processing. Each module included therein can implement each step of the foregoing method. Specifically, it includes:

[0065] A preprocessing module for collecting actual parameters under different orbit conditions to build a simulation space environment, adding space targets according to actual needs, simulating the motion laws of space targets under different orbit conditions, and exporting short-exposure images based on the changes in orbit dynamics and the observation environment.

[0066] A degradation module for performing a superposition operation on the exported short-exposure images in chronological order to simulate long-exposure images; constructing a degradation model to perform degradation processing on the exported short-exposure images and the obtained long-exposure images.

[0067] A calculation module for inversely solving the focal length based on the equivalence relationship between the sensor measurement field of view and the actual observation field of view and the camera imaging principle, and calculating the angular measurement information of the space target in the degraded short-exposure images and long-exposure images according to the position coordinates of the space target in the image.

[0068] In a third aspect, the present invention provides an electronic device, including: one or more processors; a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the foregoing multi-scene high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing.

[0069] In a fourth aspect, the present invention provides a computer-readable storage medium, on which executable instructions are stored. When the instructions are executed by a processor, the processor can implement the foregoing multi-scene high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing.

[0070] The specific embodiments described above further elaborate on the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A multi-scene high-precision ground-based optical imaging processing method based on focal length inversion and degradation processing, characterized in that: The method comprises: Step 1. Build a variety of simulation scenarios: collect actual parameters under different orbital conditions to build a simulated space environment, add space targets according to actual needs, simulate the motion laws of space targets under different orbital conditions, and export short-exposure images based on changes in orbital dynamics and observation environment; Step 2: Sequence image superposition and degradation model processing: superimpose the exported short exposure images in chronological order to simulate the long exposure images; construct a degradation model, and perform degradation processing on the exported short exposure images and the obtained long exposure images; Step 3, reverse solve the focal length and solve the parameters: Based on the equivalence between the sensor measurement field of view and the actual observation field of view and the camera imaging principle, reverse solve the focal length, and calculate the angular measurement information of the space target in the degraded short exposure image and the long exposure image according to the position coordinates of the space target on the image.

2. According to claim 1, a multi-scene high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing is characterized in that: The simulation of the motion laws of space targets under different orbital conditions in step 1 includes constructing four scenarios to simulate satellite systems under different orbital conditions: low-orbit satellites, medium-orbit satellites, high-orbit satellites, and a fusion scenario of low-orbit, medium-orbit and high-orbit satellites.

3. The multi-scene high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing according to claim 2 is characterized in that: In step 1, deriving the short exposure image based on the changes in orbital dynamics and observation environment includes: Based on the STK simulation engine, considering the random changes in orbital height and inclination, the vertical half angle of the rectangular field of view is , the horizontal half angle of the rectangular field of view is , the angle of the three-dimensional window observation angle is , under the premise of adjusting the sensor field of view to a square, establish an equivalent relationship between the sensor detection field of view and the actual observation field of view: ， Select different scenes or modify corresponding parameters according to different research tasks to export multi-scene short-exposure images that meet the requirements.

4. The multi-scene high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing according to claim 1 is characterized in that: In step 2, the exported short exposure images are stacked in time sequence to simulate long exposure images, including: The short exposure image sequence is exported by using the method of sequence image superposition. Synthesize to get a long exposure image : ， Among them, I i (x , y) is the pixel coordinate of the short exposure image.

5. The multi-scene high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing according to claim 1 is characterized in that: In step 2, the degradation model is constructed, and the degradation processing of the derived short exposure image and the obtained long exposure image includes: Construct a degradation model, add noise and perform brightness and darkness imbalance processing on the exported short exposure image and the obtained long exposure image: ， In the formula, is the pixel point of the image that has not been degraded. is the image pixel after degradation processing, is the degradation model's consideration of the noise mixture component, It is the degradation model that takes into account the imbalance of light and dark in shooting; among them: ， ， In the formula, Represent Gaussian noise, Poisson noise, and salt and pepper noise, respectively. Represents the weight factor of each item, represents the sampling point coordinates, Used to control the width of the gradient; Gaussian blur combined with point spread function convolution is used to achieve the observation blur effect.

6. The multi-scene high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing according to claim 1 is characterized in that: In step 3, based on the equivalence between the sensor measured field of view and the actual observed field of view and the camera imaging principle, the inverse focal length solution includes: Based on the camera imaging principle, f = s, f is the focal length of the camera, and s is the image distance; According to the equivalence relationship between the sensor measurement field of view and the actual observation field of view, the image distance The expression converted to the unit pixel in the pixel plane coordinate system is: ， The camera focal length can be obtained by inverse analysis .

7. The multi-scene high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing according to claim 1 is characterized in that: Calculating the angle measurement information of the space target in the degraded short exposure image and the long exposure image according to the position coordinates of the space target on the image in step 3 includes: Suppose the centroid coordinates of the observation satellite on the degraded image at any time are ( , ), is the correction term, t is the time variable, then the expression of the image space target azimuth is: ， The pitch angle is derived by using the complementary relationship between the pitch angle of the target in image space and the vertical angle. The expression is: ， In the formula, The vertical angle between the camera pinhole and the space target is: = , Thus, a series of actual pitch angle and azimuth angle information of the space target with respect to the camera coordinate system in time sequence is obtained.

8. A multi-scene high-precision ground-based optical imaging processing device based on focal length inverse solution and degradation processing, characterized in that: include: The preprocessing module is used to collect actual parameters under different orbital conditions to build a simulated space environment, add space targets according to actual needs, simulate the motion laws of space targets under different orbital conditions, and export short-exposure images based on changes in orbital dynamics and observation environment; A degradation module is used to perform a superposition operation on the exported short exposure images in time sequence to simulate the long exposure images; a degradation model is constructed to perform degradation processing on the exported short exposure images and the obtained long exposure images; The calculation module is used to inversely solve the focal length based on the equivalence relationship between the sensor measured field of view and the actual observed field of view and the camera imaging principle, and calculate the angle measurement information of the space target in the degraded short exposure image and the long exposure image according to the position coordinates of the space target on the image.

9. An electronic device, characterized in that: include: one or more processors; A memory for storing one or more programs; Among them, when one or more programs are executed by the one or more processors, the one or more processors implement a multi-scene high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: Executable instructions are stored thereon, which, when executed by a processor, enable the processor to implement a multi-scene high-precision ground-based optical imaging processing method based on focal length inverse solution and degradation processing as described in any one of claims 1-7.

Images