A method for constructing a physical feature library of space targets

Abstract

The application discloses a kind of space target physical feature library construction methods, comprising the following steps: determining space target type and total type quantity;Determine the physical feature category of each space target;For each physical feature, respectively determine its key feature item, construct feature information library;Obtain the key feature data of known space target type space target or simulation generation target, input feature information library, form physical feature library;Acquire the key feature data of the space target to be classified, match the key feature data of the space target to be classified with the key feature data in the physical feature library, determine the space target type of the space target to be classified;The space target type of the space target to be classified and its key feature data are input into the physical feature library, and the update of the physical feature library is completed.The application constructs the feature set of general common type, realizes that target physical feature can be intelligently classified and the physical feature attribute of space target category is improved.

Description

Technical Field

[0001] This invention belongs to the field of research on the physical characteristics of space targets, and in particular relates to a method for constructing a physical characteristic database of space targets. Background Technology

[0002] The physical characteristics of space targets include photometric features, infrared features, structural features, and radar RCS features. Space targets are diverse, and their characteristics vary greatly. Based on these different characteristics, space targets can be classified and categorized, such as remote sensing space targets, reconnaissance space targets, communication space targets, and combat space targets. Different types of space targets have significantly different physical characteristics. By constructing a space target physical characteristic database, after acquiring some of the physical characteristics of space targets, they can be matched and categorized according to these characteristics. Simultaneously, the physical characteristics of detected targets should also be stored in the space target characteristic database in a certain way and updated regularly. Traditional space target databases generally include physical and logical databases. Because space targets include various spacecraft and space debris, their numbers are vast. Constructing a physical characteristic database requires specialized database software for maintenance. Such physical characteristic databases do not integrate and summarize the common characteristics of similar types of space targets to construct a set of universal, common type features. To achieve this function, traditional space target databases would need to develop specialized feature summarization software for reclassification. Therefore, it is necessary to establish a method for constructing a physical feature library of space targets that can comprehensively summarize the physical characteristics of different types of space targets. Summary of the Invention

[0003] The technical problem solved by this invention is to overcome the shortcomings of existing methods and provide a method for constructing a physical feature library of space targets. This method integrates and summarizes the common characteristics of space targets of the same type, constructs a feature set of general common types, and ultimately realizes the intelligent classification and improvement of the physical feature attributes of space targets based on their physical features.

[0004] The objective of this invention is achieved through the following technical solutions:

[0005] This invention discloses a method for constructing a physical feature library of space targets, comprising:

[0006] Determine the types of space targets and the total number of types;

[0007] The types of physical characteristics for each type of space target are determined, including photometric, infrared, orbital, and external shape / structure.

[0008] For each type of physical feature, its key feature terms are determined, and a feature information database is constructed.

[0009] Obtain key feature data of space targets of known space target types or simulated generated targets, and input them into the feature information database to form a physical feature database;

[0010] The key feature data of the spatial target to be classified is obtained, and the key feature data of the spatial target to be classified is matched with the existing key feature data in the physical feature library. The spatial target type of the spatial target to be classified is determined according to the matching result.

[0011] Input the spatial target type and key feature data of the spatial target to be classified into the physical feature database to complete the update of the physical feature database.

[0012] In the above method for constructing a physical feature database of space targets, the key feature information of each type of physical feature is determined, and a feature information database is constructed. The specific steps are as follows:

[0013] The key features of the luminance are determined to be shape, size, material, distance between the target and the observation point, angle between the observation point, the sun, and the target, and luminance value. These features are then stored in a tree structure to form a luminance feature table.

[0014] The key infrared features are identified as material, size, temperature, and infrared radiation brightness. These features are then arranged in order and stored in a tree structure to form an infrared feature table.

[0015] The key feature of the track is determined to be the number of 6 track elements, and the number of 6 track elements is stored to form a track feature table;

[0016] The key features of the external structure are identified as target characteristics, key component information, and a binary rendering image of the spatial target. Target characteristics include target shape, material, and size; key component information includes component type and component size; the binary rendering image of the spatial target includes the distance between the target and the observation point during imaging, the latitude and longitude angle of the observation point with the target as the center of the celestial sphere, and the binary image itself; the key features of the external structure are stored to form an external structure feature table.

[0017] The tree-shaped photometric feature table, infrared feature table, orbital feature table, and external structure feature table are stored in the input database to form a feature information database.

[0018] In the above method for constructing a physical feature library of space targets, key feature data of space targets of known types or simulated targets are obtained and input into the feature information library to form a physical feature library. The specific steps are as follows:

[0019] (1) Obtain key characteristic data of photometry by relying on simulation methods or measured data. The specific method is as follows:

[0020] Measurable key feature data are determined based on historical experience data. Measurable key feature data includes target shape, target area S, target-observation point distance R, and the angle between observation point, sun, and target. β i ;

[0021] Based on measurable key feature data, calculate the spatial target surface element. ds Capture the total solar energy dE within the wavelength range of sunlight;

[0022] Calculate the reflectance distribution of the surface element of the space target based on the total solar energy dE;

[0023] Based on measurable key feature data, calculate the luminosity value Em produced by the sunlight reflected by the target at the observation point;

[0024] (2) Obtain key infrared characteristic data by means of simulation or measured data. The specific methods are as follows: measure the target material, size, temperature, angle between the observation point and the sun and the target, the distance between the target and the observation point, and the infrared radiation brightness value; in the simulation, determine the target material, size, and temperature values ​​according to the required target type and historical experience data; calculate the infrared radiation brightness of the space target itself.

[0025] (3) Obtain key characteristic data of the target by relying on simulation methods or measured data, including: the value of the number of track 6;

[0026] (4) Obtain key feature data of the external structure by relying on simulation methods or measured data. The specific method is as follows:

[0027] Key feature data of known space target types are obtained through measured data, including target shape, material, size, key component type, key component size, distance between the target and the observation point during imaging, latitude and longitude angle of the observation point with the target as the center of the celestial sphere, and values ​​of the binarized image; a binarized image is generated through simulation.

[0028] (5) Input the key feature data of photometry, key feature data of infrared, key feature data of target, and key feature data of shape and structure into the feature information database to form a physical feature database.

[0029] In the above method for constructing the physical feature database of space targets, the total solar energy dE is calculated using the following formula:

[0030]

[0031] in, The wavelength of electromagnetic radiation, β i The angle between the observation point, the sun, and the target.

[0032] In the above method for constructing a physical feature library of space targets, the formula for calculating the reflectance distribution of surface elements of the space target is:

[0033]

[0034] in, For the bidirectional reflection distribution function of different materials, For (θ) r ,φ r ) represents the reflected brightness in the direction, (θ) r ,φ r ) represents the two included angles between the position vector of the observation point and the normal direction of the surface of the space target in the coordinate system of the space target surface.

[0035] In the above method for constructing the physical feature database of space targets, the luminosity value Em generated by the sunlight reflected by the target at the observation point is calculated using the following formula:

[0036]

[0037] Where D is the aperture of the observation equipment at the observation point.

[0038] In the above method for constructing a physical feature database of space targets, the formula for calculating the infrared radiation brightness of the space target itself is as follows:

[0039] ,

[0040] in, c 1 represents the first radiation coefficient. c 2 is the second radiation coefficient. ε λ The emissivity spectrum of the space target surface is calculated based on the target material and the angle between the observation point, the sun, and the target. λ is the infrared radiation wavelength, and T is the absolute temperature.

[0041] In the above-mentioned method for constructing a physical feature library of spatial targets, the specific method for generating a binarized image through simulation is as follows:

[0042] An orthogonal projection imaging mode is adopted, a free camera is added, and the lens is aimed at the spatial target model to ensure that the z-axis direction of the two is consistent, so that the model is located at the center of the image plane and is fully visible in the camera view in the initial state;

[0043] A free parallel light source with no attenuation is added behind the camera's initial position, and the light color is white.

[0044] Adjust the relative positions of the camera, light source, and model. Changes in relative positions will cause changes in viewpoint and lighting. Bind the light source and camera to keep their relative positions constant, keep the model stationary, and move the camera on a sphere with the model's center as the center.

[0045] The visible part of the model body is output onto the image plane using the segmentation map to generate a binarized image.

[0046] In the above-mentioned method for constructing a physical feature database of spatial targets, the specific method for obtaining key feature data of the spatial targets to be classified, matching the key feature data of the spatial targets to be classified with existing key feature data in the physical feature database, and determining the spatial target type of the spatial targets to be classified based on the matching results is as follows:

[0047] (1) Obtain key feature data of N sets of spatial targets to be classified;

[0048] (2) Based on a set of key feature data of the spatial target to be classified, determine the type of physical feature to which the key feature data of the spatial target to be classified belongs;

[0049] (3) Using the correlation analysis method, a set of key feature data of the spatial target to be classified is matched with the key feature data in the physical feature library to pre-determine the spatial target type to which the set of key feature data belongs;

[0050] (4) Repeat step (3) until all N sets of key feature data have been traversed; obtain the spatial target type to which the corresponding N sets of spatial targets to be classified belong;

[0051] (5) Using direct voting, Bayesian inference or DS evidence theory to judge, combined with the total number of spatial target types, the spatial target type of N groups of spatial targets to be classified is judged to determine the spatial target type of the spatial targets to be classified, where N is a positive integer greater than 1.

[0052] In the above-mentioned method for constructing a physical feature database of spatial targets, the types of spatial targets are remote sensing, communication, navigation, reconnaissance, or attack and defense.

[0053] Compared with the prior art, the present invention has the following advantages:

[0054] (1) This invention adopts the method of integrating and summarizing the common characteristics of the same type of space targets, and constructs a feature set of general common types for space target type classification, thereby realizing the quantification of the physical characteristics of space targets.

[0055] (2) The present invention uses a tree structure to construct a physical feature library of space targets, establishes the correspondence between the physical feature quantities of space target observation and the observation conditions, and improves the classification efficiency of space targets. Attached Figure Description

[0056] Figure 1 This is a flowchart of the method for constructing a physical feature library of space targets according to the present invention;

[0057] Figure 2 This is a schematic diagram illustrating the establishment of a simulated feature library for the photometric characteristics of spatial targets provided in an embodiment of the present invention;

[0058] Figure 3 This is a schematic diagram illustrating the establishment of a simulated feature library of infrared characteristics of space targets provided in an embodiment of the present invention;

[0059] Figure 4 This is a schematic diagram of the simulation process for external structural characteristics provided in an embodiment of the present invention;

[0060] Figure 5 This is a schematic diagram of the external shape and structural characteristic library data structure provided in an embodiment of the present invention;

[0061] Figure 6 This is a schematic diagram of the overall framework of the space target physical feature library provided in the embodiments of the present invention. Detailed Implementation

[0062] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0063] like Figure 1 The present invention discloses a method for constructing a physical feature library of space targets, comprising:

[0064] Step (1): Determine the types of space targets and the total number of types; space target types include remote sensing, communication, navigation, reconnaissance, and attack / defense targets.

[0065] Step (II) Determine the types of physical characteristics for each type of space target. These physical characteristics include photometric, infrared, orbital, and external shape.

[0066] Step (3): For each type of physical feature, determine its key feature items and construct a feature information database. The specific steps are as follows:

[0067] The key features for determining luminosity are shape, size, material, distance between the target and the observation point, angle between the observation point, the sun, and the target, and luminosity value. These features are arranged in order and stored in a tree structure to form a luminosity feature table.

[0068] The key infrared characteristics are identified as material, size, temperature, and infrared radiation intensity. These are then arranged in order and stored in a tree structure to form an infrared characteristic table.

[0069] The key feature of the track is determined to be the number of the 6 track elements, and the number of the 6 track elements is stored to form a track feature table;

[0070] The key features of the external structure are identified as target characteristics, key component information, and a binary rendering image of the spatial target. Target characteristics include target shape, material, and size; key component information includes component type and component size; the binary rendering image of the spatial target includes the distance between the target and the observation point during imaging, the latitude and longitude angle of the observation point with the target as the center of the celestial sphere, and the binary image; the key features of the external structure are stored to form an external structure feature table.

[0071] The tree-shaped photometric feature table, infrared feature table, orbital feature table, and external structure feature table are stored in the input database to form a feature information database.

[0072] Step (iv): Obtain key feature data of space targets of known space target types or simulated targets, input them into the feature information database to form a physical feature database. The specific steps are as follows:

[0073] (1) Obtain key characteristic data of photometry by relying on simulation methods or measured data. The specific method is as follows:

[0074] Measurable key feature data are determined based on historical experience data. Measurable key feature data includes target shape, target area S, target-observation point distance R, and the angle between observation point, sun, and target. β i ;

[0075] Based on measurable key feature data, calculate the spatial target surface element. ds The total solar energy dE captured within the wavelength range; the formula for calculating the total solar energy dE is:

[0076]

[0077] Where λ is the wavelength of visible light. Represents the area of ​​tiny surface cells on the surface of a space target. β i The angle between the observation point, the sun, and the target.

[0078] Based on the total solar energy dE, the reflectance distribution of the surface element of the space target is calculated using the following formula:

[0079]

[0080] in, For the bidirectional reflection distribution function of different materials, For (θ) r ,φ r ) represents the reflected brightness in the direction, (θ) r ,φ r ) represents the two included angles between the position vector of the observation point and the normal direction of the surface of the space target in the coordinate system of the space target surface.

[0081] Based on measurable key feature data, the luminosity value Em produced by the sunlight reflected from the target at the observation point is calculated using the following formula:

[0082]

[0083] In the formula, R is the distance from the observation platform to the space target, and D is the aperture of the observation equipment at the observation point. This represents the area of ​​a tiny surface element on the surface of a space target.

[0084] (2) Obtain key infrared characteristic data by means of simulation or measured data. The specific methods are as follows: measure the target material, size, temperature, angle between the observation point and the sun and the target, the distance between the target and the observation point, and the infrared radiation brightness value; in the simulation, determine the target material, size, and temperature values ​​according to the required target type and historical experience data.

[0085] The formula for calculating the infrared radiation brightness of a space target is:

[0086] ,

[0087] in, c 1 represents the first radiation coefficient. c 2 is the second radiation coefficient. ε λ The emissivity spectrum of the space target surface can be calculated based on the target material and the angle between the observation point, the sun, and the target. λ is the infrared radiation wavelength, and T is the absolute temperature.

[0088] (3) Obtain key characteristic data of the target by relying on simulation methods or measured data, including: the value of the number of track 6;

[0089] (4) Obtain key feature data of the external structure by relying on simulation methods or measured data. The specific method is as follows:

[0090] Key feature data of known space target types were obtained through measured data, including target shape, material, size, key component type, key component size, distance between the target and the observation point during imaging, latitude and longitude angle between the observation point and the target as the center of the celestial sphere, and the value of the binarized image.

[0091] The binarized image is generated through simulation. The specific method is as follows:

[0092] An orthogonal projection imaging mode is adopted, a free camera is added, and the lens is aimed at the spatial target model to ensure that the z-axis direction of the two is consistent, so that the model is located at the center of the image plane and is fully visible in the camera view in the initial state;

[0093] A free parallel light source with no attenuation is added behind the camera's initial position, and the light color is white.

[0094] Adjust the relative positions of the camera, light source, and model. Changes in relative positions will cause changes in viewpoint and lighting. Bind the light source and camera to keep their relative positions constant, keep the model stationary, and move the camera on a sphere with the model's center as the center.

[0095] The visible part of the model body is output onto the image plane using the segmentation map to generate a binarized image.

[0096] (5) Input the key feature data of photometry, key feature data of infrared, key feature data of target, and key feature data of shape and structure into the feature information database to form a physical feature database.

[0097] Step (5): Obtain the key feature data of the spatial target to be classified, match the key feature data of the spatial target to be classified with the existing key feature data in the physical feature database, and determine the spatial target type of the spatial target to be classified based on the matching results. The specific method is as follows:

[0098] (1) Obtain key feature data of N sets of spatial targets to be classified through measurement methods;

[0099] (2) Based on a set of key feature data of the spatial target to be classified, determine the type of physical feature to which the key feature data of the spatial target to be classified belongs;

[0100] (3) Using the correlation analysis method, a set of key feature data of the spatial target to be classified is matched with the key feature data in the physical feature library to pre-determine the spatial target type to which the set of key feature data belongs;

[0101] (4) Repeat step (3) until all N sets of key feature data have been traversed; obtain the spatial target type to which the corresponding N sets of spatial targets to be classified belong;

[0102] (5) Using direct voting, Bayesian inference or DS evidence theory to determine the spatial target type of the N groups of spatial targets to be classified, and in combination with the total number of types of spatial targets, the spatial target type of the spatial targets to be classified is determined.

[0103] Step (VI): Input the spatial target type and key feature data of the spatial target to be classified into the physical feature database to complete the update of the physical feature database.

[0104] Example 1

[0105] This embodiment provides a method for constructing a physical feature library of space targets, which includes the following steps:

[0106] Step 1: Deconstruct and analyze the physical characteristics of the space target, construct a preliminary database using simulation methods, and determine the database storage structure. The physical characteristics of the space target include four parts: photometry, infrared signature, orbital characteristics, and external shape. The simulation database for each of these four parts is established in four separate steps.

[0107] Step 2: Based on the classification of spatial target physical features in Step 1, classify the captured spatial target physical features and assign them to various spatial target categories.

[0108] The physical disassembly characteristics of space targets include four parts: photometric, infrared, orbital, and external structure. A simulation database for each of these four parts is established in four separate steps:

[0109] Step S11: Establishment of a simulation database for the photometric characteristics of space targets. Photometric information is related to factors such as the size, shape, distance, and orientation of space targets. The detected space targets are generally very far away, making them typical point source targets. The surface shapes of space targets are complex, and can be roughly divided into three categories: cuboids, cylinders, and spheres. According to the variables summarized above, photometric information is related to five aspects: the size, shape, distance, solar angle, and surface material of the space target. The establishment process consists of the following four steps:

[0110] (1) Calculate the target surface element in space ds The formula for calculating the total solar energy captured in the (500-800nm) wavelength range is as follows:

[0111]

[0112] in, β i Connect the sun to the target in space and the surface element of the target in space. ds The angle between the normals (rad).

[0113] (2) Based on the total solar energy captured by the space target surface source in the 500nm~800nm ​​band calculated in (1), the space target surface element can be calculated. ds The reflected brightness distribution is as follows:

[0114]

[0115] in, f r ( β i , θ r , φ r () is the bidirectional reflection distribution function. L r ( β i , θ r , φ r ) as ( θ r , φ r θ represents the reflected brightness in the direction. r ,φ r ) represents the two included angles between the position vector of the observation point and the normal direction of the surface of the space target in the coordinate system of the space target surface.

[0116] (3) Based on the reflected brightness of the spatial target surface element in the specified direction calculated in (2), let the azimuth angle of the target be ( θ r , φ r If the sunlight reflected by the target at this moment is at the target detection sensor in the observation space, then the luminosity produced is:

[0117]

[0118] Wherein, R is the distance from the observed space target to the target space target.

[0119] (4) Follow the steps in (1)-(3), as follows Figure 2 As shown, photometric information simulations were performed on three spatial target shapes: sphere, square, and cylinder. Based on their respective characteristics, a tree-shaped photometric information storage structure was constructed.

[0120] Step S12: Establishment of the space target orbital characteristic database. The orbital characteristics of space targets do not require simulation; the cataloged orbital data (number of orbital elements) of space targets can be directly obtained from publicly available materials and added to the space target physical database.

[0121] Step S13: Establishment of the Infrared Characteristics Database for Space Targets. The infrared radiation characteristics of a target are related to five factors: target surface temperature, size, material, solar angle direction, and distance. After determining the factors affecting the infrared characteristics of space targets, the following analysis process is followed: using the controlled variable method, a database describing the infrared characteristics of space targets under different characteristics is constructed for different variables. This mainly consists of the following four steps:

[0122] (1) Define the infrared radiation brightness of the space target itself as:

[0123] ,

[0124] in, c 1 represents the first radiation coefficient. c 2 is the second radiation coefficient. ε λ Let λ be the emissivity spectrum of the space target's surface, λ be the wavelength, and T be the absolute temperature [K]. Therefore, the surface temperature of the space target determines its infrared radiation characteristics. The factors affecting the surface temperature of the space target can be decomposed into several parts, including direct radiative heating of the space target surface by the sun, radiative heating of the space target by solar radiation reflected from the Earth, radiative heating of the space target by Earth / atmospheric thermal radiation, and heat conduction generated by internal heat-generating components of the space target.

[0125] (2) Calculate the temperature rise caused by direct solar radiation. From Earth's perspective, the Sun occupies only 0.5 spherical degrees. Therefore, on the order of the Earth-Sun distance, the Sun can be treated as a point source. Moreover, the volume of the target in space is relatively small, so sunlight can be considered as parallel light, and its radiative heat flux density is a solar constant S = 1353 W / m². 2 Any surface unit on the outer surface of a space target The direct solar radiation heat flow received by the surface is:

[0126]

[0127] in, The surface absorptivity to solar radiation. Let be the angle between the direction of solar incidence and the direction of the surface element normal. When the value is less than 0, it indicates that the sun is not shining on the surface of the unit. . The solar radiation transfer coefficient is the coefficient of a surface unit. When a space target is within the Earth's shadow, its surface does not receive solar heat. .

[0128] (3) Calculate the temperature rise caused by solar radiation reflected from the Earth. Due to the complex distribution of surface features and the variation of reflectivity with time and season, simplification is needed when describing the reflection of solar radiation from the Earth's surface. Assuming the Earth is a diffuse reflector, its reflection of solar radiation obeys Lambert's law, its reflectance spectrum is the same as the solar spectrum, and its reflectivity is taken as the global average reflectivity of 0.35, then any surface unit on the surface of a space target... The received heat flux from Earth's albedo radiation is:

[0129]

[0130] in, The absorption rate of sunlight by the surface of a space target; This is the Earth's albedo coefficient. When a space target is within the Earth's shadow, When the space target is in the sunlit area: , It is the angle between the sunlight and the line connecting the space target and the Earth's center.

[0131] (4) Calculate the temperature rise caused by solar radiation reflected from the Earth. The Earth / atmosphere system absorbs solar radiation and simultaneously radiates energy into space. The energy radiated into space and the energy absorbed by the solar radiation reach equilibrium, and the Earth / atmosphere can be approximated as a uniformly radiating thermal equilibrium body. In practical engineering thermal design, the Earth is generally assumed to be a diffuse emitter. Any surface unit on the surface of a space target The infrared radiation heat flux received from the entire Earth is:

[0132]

[0133] in, Let be the infrared emissivity of the space target's surface (assuming the space target's surface is a diffuser, and the absorptivity equals the emissivity). The Earth's reflectivity to solar radiation; This is the Earth's infrared angular coefficient.

[0134] (5) Calculate the heat balance equation according to steps (1)-(5), and thus calculate the surface temperature T of the space target. Ignoring the volume effect of the space target structure, the surface of the space target is treated as a thin-walled structure for heat transfer analysis. The heat balance equation satisfied by the surface element is:

[0135]

[0136] in, As the internal heat source of the surface unit, Surface area of ​​surface unit Let j be the area of ​​surface element j. The thickness of the surface unit, c, the specific heat of the surface unit. Represents the surface unit density. Thermal emissivity of the surface unit Let T be the emissivity coefficient of the space target surface, and T be the temperature of the space target surface. Let be the radiative exchange coefficient of surface element j over surface element i. For time; when calculating the temperature of the solar panel, the heat energy dissipated in the solar panel is treated as an internal heat source, taking into account the photoelectric conversion efficiency of the solar panel.

[0137] (6) Using the temperature calculated in step (5) above, substitute it into the thermal radiation calculation in step (1) to obtain the infrared radiation value, and simultaneously perform target simulation, such as... Figure 3 As shown, a database is established that relates the infrared radiation characteristics of a target to five factors: target surface temperature, size, material, solar angle direction, and distance.

[0138] Step S14: Generating visible light simulation images of space targets is a necessary prerequisite for establishing 3D models of space targets and forms the basis for building a space target library. This involves creating 3D models of space targets required for the library using either CAD software or 3D models provided by relevant software. The spatial target shape and structural feature library is primarily stored in image format, supplemented by necessary detection and recognition results. Subsequently, the target library is established based on the 3D models. The specific generation of simulation images for the feature library is as follows... Figure 4 As shown, the specific steps are as follows:

[0139] (1) Camera parameter settings. In space-based visible light imaging detection of space targets, the target is generally far from the observation camera, and the observation distance is much larger than the target size. For simplicity, an orthogonal projection imaging mode is used in the simulation. A free camera is added, with the lens aligned with the space target model, ensuring that the z-axis directions of both are consistent, so that the model is located at the center of the image plane and is fully visible in the initial state of the camera view.

[0140] (2) Light source parameter settings. Considering that sunlight in the space environment is approximately parallel light, a free parallel light source with no attenuation is added behind the initial position of the camera, and the light color is white.

[0141] (3) Relative Position Adjustment. Changes in the relative positions of the camera, light source, and model cause changes in viewpoint and illumination. To obtain a full-viewpoint simulation image of the model, the light source and camera are bound together, keeping their relative positions constant. The model remains stationary, and the camera is moved on a sphere centered on the model. To clearly describe the changes in viewpoint position, the latitude and longitude representation method from geography is adopted, and the initial position of the camera is defined as having both latitude and longitude of zero.

[0142] (4) Rendering Output. Set the output image resolution and image attributes. Use the segmentation map to output all visible parts of the model onto the image plane, with gradients at the edges. After binarization, this becomes the binarized image corresponding to the color image, saving the target's contour information. Simultaneously, construct a spatial target structure feature database based on steps (1)-(3) above. The data structure of the spatial target shape structure feature database is as follows: Figure 5 As shown.

[0143] After classifying the physical characteristics of space targets and assigning these characteristics to various space target categories, the following steps are required for feature classification of space targets acquired in orbit, ultimately forming the framework of the space target physical feature library, as follows: Figure 6 As shown:

[0144] Step S21: First, perform physical feature classification to determine the type of spatial target feature to which the collected information belongs.

[0145] Step S22: Match the spatial target characteristic acquisition conditions with the data in the physical feature library, find the spatial target simulation environment that is closest to the corresponding physical feature attribute value, and determine the spatial target type corresponding to each group of physical features based on the matching results, until all N groups of key feature data of the targets to be classified have been traversed; obtain the spatial target type to which the corresponding N groups of spatial targets to be classified belong;

[0146] Step S23: Using direct voting, Bayesian inference, or DS evidence theory to determine the spatial target type of the N groups of spatial targets to be classified, and combining the total number of spatial target types, determine the spatial target type of the spatial targets to be classified.

[0147] Step 3: Input the spatial target type and key feature data of the spatial target to be classified into the physical feature database to complete the update of the physical feature database.

[0148] Through the above steps, the construction of the physical feature library of space targets is achieved.

[0149] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.

Claims

1. A method for constructing a library of physical characteristics of space objects, characterized in that, The application relates to a space target classification method and device. Determine the type of space target and the total number of types; Determine the type of physical characteristics of each type of space target, which includes luminosity, infrared, orbit, and shape structure; Determine the key feature items of each type of physical characteristics, and construct a feature information library; Obtain the key feature data of known space targets or simulation generated targets, input the feature information library, and form a physical feature library; Obtain the key feature data of the space target to be classified, match the key feature data of the space target to be classified with the key feature data in the physical feature library, and determine the type of the space target to be classified according to the matching result; Input the type of the space target to be classified and the key feature data of the space target to be classified into the physical feature library, and complete the update of the physical feature library; Determine the key feature information of each type of physical characteristics, and construct a feature information library, and the specific steps are as follows: Determine the key feature items of the luminosity as shape, size, material, target-observation point distance, observation point-sun-target angle, and luminosity value, store the key feature items in a tree form according to the arrangement order, and form a luminosity feature table; Determine the key feature items of the infrared as material, size, and temperature infrared radiation brightness, store the key feature items in a tree form according to the arrangement order, and form an infrared feature table; Determine the key feature items of the orbit as 6 orbital elements, and store the 6 orbital elements to form an orbit feature table; Determine the key feature items of the shape structure as target characteristics, key component information, and space target binary rendering image, wherein the target characteristics include target shape, material, and size, the key component information includes component type and component size, and the space target binary rendering image includes target-observation point distance, observation point-observation point angle in the target as a celestial center, and a binary image; store the key feature items of the shape structure to form a shape structure feature table; Store the tree-shaped luminosity feature table, infrared feature table, orbit feature table, and shape structure feature table into an input library to form a feature information library; Obtain the key feature data of known space targets or simulation generated targets, input the feature information library, and form a physical feature library, and the specific steps are as follows: (1) Obtain the key feature data of the luminosity by relying on simulation means or measured data, and the specific method is as follows: Determining measurable key feature data from historical empirical data, the measurable key feature data including target shape, target area S, target to observation point distance R, observation point-sun-target angle β i ; Computing a space target facet based on measurable key feature data According to the total solar energy dE, the space target surface element reflection brightness distribution is calculated; Capturing the total solar energy dE in a band range of the solar light; According to the measurable key feature data, the luminosity value Em generated by the target reflected sunlight at the observation point is calculated; (2) Obtain the key feature data of the infrared by relying on simulation means or measured data, and the specific method is as follows: the target material, size, temperature, observation point-sun-target angle, target-observation point distance, and infrared radiation brightness value are measured; in simulation, the values of the target material, size, and temperature are determined according to historical experience data according to the required target type; the space target itself infrared radiation brightness is calculated; (3) Obtain the key feature data of the target by relying on simulation means or measured data, including the value of 6 orbital elements; (4) Obtain the key feature data of the shape structure by relying on simulation means or measured data, and the specific method is as follows: ​ The key feature data of the known space target type of the space target obtained by the measured data includes target shape, material, size, key component type, key component size, target distance from the observation point at the time of imaging, the latitude and longitude angle of the observation point with the target as the celestial center, and the value of the binary image; and the binary image is generated by simulation means; (5) inputting the key feature data of the luminosity, the key feature data of the infrared, the key feature data of the target, and the key feature data of the external shape structure into the feature information library to form a physical feature library.

2. The method of claim 1, wherein: The total solar energy dE is calculated by the following formula: wherein is the wavelength of visible light, β i is the observer-sun-target angle.

3. The method of claim 1, wherein: The formula for calculating the reflection luminance distribution of the space target surface element is: wherein is the bidirectional reflectance distribution function for different materials, is the bidirectional reflectance distribution function for different materials, r ,φ r ) are two angles in the spatial target surface coordinate system between the observation point position vector and the normal direction of the spatial target surface. r ,φ r ) are two angles in the spatial target surface coordinate system between the observation point position vector and the normal direction of the spatial target surface.

4. The method of claim 1, wherein: The luminosity value Em of the target reflected sunlight at the observation point is calculated by the following formula: Wherein, D is the aperture of the observation equipment at the observation point.

5. The method of claim 1, wherein: The formula for calculating the infrared radiation luminance of the space target itself is: ， wherein, c 1 is a first radiation coefficient, c 2 is a second radiation coefficient, The binary image is generated by simulation means, and the specific method is as follows: λ The emissivity spectrum of the space target surface is calculated according to the target material and the observation point-sun-target angle, λ is the infrared radiation wavelength, and T is the absolute temperature.

6. The method of claim 1, wherein: An orthogonal projection imaging mode is adopted, a free camera is added, the lens is aligned with the space target model, the z-axis directions of the two are ensured to be consistent, the model is located at the center of the image plane and is fully visible in the initial state of the camera view; A free parallel light source is added behind the initial position of the camera, without attenuation, and the light color is white; The relative positions of the camera, the light source, and the model are adjusted, the change of the relative positions causes the change of the view point and the light, the light source and the camera are bound to keep the relative positions unchanged, and the model is fixed, and the camera is moved on the spherical surface with the center of the model as the spherical center; The visible part of the model body is output to the image plane by using the segmentation graph to generate a binary image. The key feature data of the space target to be classified is obtained, the key feature data of the space target to be classified is matched with the key feature data in the physical feature library, and the space target type of the space target to be classified is determined according to the matching result, and the specific method is as follows:

7. The method of claim 1, wherein: (1) obtaining N groups of key feature data of space targets to be classified; (2) judging the physical feature category to which the key feature data of the space target to be classified belongs according to a group of key feature data of the space target to be classified; (3) matching a group of key feature data of the space target to be classified with the key feature data in the physical feature library by using a correlation analysis method, and pre-judging the space target type to which the space target to be classified corresponding to the group of key feature data belongs; (4) repeating step (3) until all N groups of key feature data are traversed; obtaining the space target types to which N groups of space targets to be classified belong; (5) judging the space target types to which N groups of space targets to be classified belong by using a direct voting voting, Bayesian inference or D-S evidence theory judgment mode, and combining the total number of the space target types, to determine the space target type to which the space target to be classified belongs, wherein N is a positive integer greater than 1. The space target type is remote sensing, communication, navigation, detection, or attack and defense.

8. The method of claim 1, wherein: ​

Images