A high-precision space target positioning and processing system

Abstract

The present invention relates to the field of space target positioning, specifically a high-precision space target positioning and processing system, which integrates advanced image processing, pattern recognition, multi-sensor data fusion and high-precision positioning algorithms to achieve efficient recognition and continuous tracking of space targets and aircraft. The system uses a variety of sensors (such as cameras, radars, infrared sensors, laser radars, etc.) to obtain target data, and accurately identifies and continuously tracks targets through background modeling, foreground extraction, target detection, feature extraction and pattern matching.

Description

Technical Field

[0001] The present invention relates to the field related to space target positioning, and in particular to a high-precision space target positioning and processing system. Background Art

[0002] With the increasing number of space exploration and space activities, the precise positioning and processing of space targets has become a key technology. The existing space target positioning system has certain deficiencies in accuracy, real-time performance and processing capabilities, and it is difficult to meet the increasingly complex space mission requirements. Therefore, a space target positioning and processing system with high accuracy, high real-time performance and high processing capability is urgently needed. Summary of the invention

[0003] The purpose of the present invention is to provide a high-precision space target positioning and processing system to solve the problems raised in the above background technology.

[0004] To achieve the above object, the present invention provides the following technical solution: a high-precision space target positioning and processing system, comprising the following modules:

[0005] Multi-sensor fusion module: It is composed of multiple sensors, including optical sensors, radar sensors and infrared sensors. The sensors provide spatial target information and improve the accuracy and reliability of target positioning through data fusion technology;

[0006] High-precision positioning algorithm module: Using the Kalman filter algorithm, the system processes sensor data in real time to achieve high-precision spatial target positioning;

[0007] Efficient data processing architecture: Distributed computing and parallel processing technologies are used to process large amounts of sensor data. At the same time, the efficient data processing architecture has data compression and storage functions to ensure the real-time and reliability of data processing;

[0008] Target recognition and tracking module: Use image processing and pattern recognition technology to identify and track space targets. Through the cooperation with the high-precision positioning algorithm module, continuous tracking and precise positioning of the target are achieved.

[0009] Preferably, the optical sensor in the multi-sensor fusion module provides the shape, brightness, color and texture information of the target; the radar sensor measures the distance, speed and orientation information of the target; the infrared sensor identifies the temperature distribution and thermal characteristics of the target, wherein the specific working steps of the data fusion technology are as follows:

[0010] Step 1: Data collection: Each sensor simultaneously collects multi-source data of the space target;

[0011] Step 2: Data preprocessing: preprocess the collected multi-source data, including data calibration, noise filtering and data alignment steps;

[0012] Step 3: Data fusion: According to actual needs, the pre-processed data is fused using a fusion algorithm. The data fusion process selects data layer fusion, feature layer fusion or decision layer fusion according to the specific application scenario.

[0013] Step 4: State estimation: Based on the fused data, the state of the target is estimated, including the target's position, velocity, acceleration, and motion trajectory information. The results of the state estimation are used for subsequent target recognition and tracking.

[0014] Preferably, the specific working steps of the high-precision positioning algorithm module are as follows:

[0015] Step S1, system model construction: In the space target positioning scenario, assuming that a target moving in space is tracked, using Kalman filtering, we first need to define the state and measurement model of the system. The state of the space target is represented by a six-dimensional vector, including the position, velocity and acceleration of the target. Set the target initial state estimate According to the prior knowledge or the initial measurement data of the sensor, the initial covariance matrix P0 is set, and the process noise covariance matrix Q is set. k-1 and the measurement noise covariance matrix R k ;

[0016] Step S2, prediction step: predict the current state according to the state transition equation Predict the covariance matrix of the current state

[0017] Step S3, update step: calculate Kalman gain K k ; Update state estimates based on measurement data Update the state covariance matrix P k ;

[0018] Step S4, repeat: repeat the prediction and update steps for each time step to achieve real-time estimation and tracking of the target state.

[0019] Preferably, the specific contents of the efficient data processing architecture are as follows:

[0020] A. Distributed computing architecture

[0021] Master node: responsible for task scheduling, node coordination and result aggregation. The master node receives the raw data stream from the sensor, decomposes it into multiple subtasks, and assigns them to the working nodes.

[0022] Worker node: responsible for specific data processing tasks. Each worker node receives subtasks from the master node, performs data parsing, filtering, and compression operations, and returns the processing results to the master node;

[0023] Task scheduling: The master node uses a dynamic task scheduling algorithm to dynamically allocate tasks according to the load of each working node to avoid overloading of certain nodes, thereby achieving load balancing and maximizing system throughput;

[0024] Fault-tolerance mechanism: When a working node fails, the master node will reallocate the tasks on that node to other nodes that are working normally to ensure that data processing tasks are not interrupted;

[0025] B. Parallel processing technology

[0026] Data sharding: The system first shards the received sensor data, breaking down large-scale data into multiple small-scale data blocks. Each data block is processed in parallel by different working nodes;

[0027] Multi-threaded parallel processing: The system uses pipeline processing technology. The data processing process is broken down into multiple stages;

[0028] C. Data compression

[0029] The system uses compression ratio optimization technology to dynamically adjust compression parameters to achieve the best compression effect. Specifically, when the network bandwidth is sufficient, the compression ratio is reduced to reduce the computational overhead during decompression; when the network bandwidth is limited, the compression ratio is increased to reduce the amount of data transmission.

[0030] D. Data storage

[0031] The system adopts a distributed storage system. Sensor data is distributed and stored on multiple storage nodes. Each storage node is responsible for storing a part of the data. Distributed storage improves storage capacity and improves data reliability and availability through data redundancy and copy mechanisms.

[0032] E. Data redundancy and copy mechanism

[0033] To ensure data reliability and availability, the distributed storage system adopts data redundancy and replication mechanisms. Specifically: Data redundancy: When storing sensor data, multiple redundant data will be generated to prevent sensor data loss; Replication mechanism: In the distributed storage system, each sensor data will have multiple copies, stored on different storage nodes;

[0034] F. Real-time guarantee

[0035] The system provides real-time monitoring and alarm functions, which can monitor sensor data in real time and trigger alarms when the data is abnormal. Specifically: when the sensor data exceeds the preset threshold, the system will immediately trigger an alarm and notify relevant personnel to take measures; when the change trend of sensor data is abnormal, the system will trigger an alarm to warn of potential risks in advance; anomaly detection: through machine learning or statistical analysis methods, the system automatically detects abnormal behavior in sensor data and triggers an alarm.

[0036] Preferably, the specific working steps of the target identification and tracking module are as follows:

[0037] a. The target recognition and tracking module first receives data from multiple sensors of the multi-sensor fusion module, and fuses the data of different sensors through multi-sensor data fusion technology to obtain more comprehensive and accurate target information;

[0038] b. For the image data collected by the camera, image preprocessing is first performed to improve the accuracy of subsequent target recognition and tracking. The preprocessing steps include: removing noise in the image through filtering algorithms; converting color images into grayscale images to reduce computational complexity; enhancing image contrast and improving image quality through methods such as histogram equalization; and correcting image distortion caused by the camera lens to ensure image accuracy;

[0039] c. Use background modeling technology to build a background model of the scene. Then use foreground extraction technology to separate the foreground target from the background. Then, in the extracted foreground area, the system uses the target detection algorithm to detect the specific target object;

[0040] d. On the basis of detecting the specific target object, further extract the target features for subsequent target recognition, that is, classify the features extracted from the target object through the pattern recognition algorithm to identify the category of the target;

[0041] e. Target tracking: In continuous video frames, the target in the current frame is matched with its position in the previous frames through the target matching algorithm, so as to achieve continuous tracking of the target. The target matching algorithm includes:

[0042] Feature-based matching: By extracting the features of the target, specifically SIFT, SURF, and ORB, the features of the same target are matched in consecutive frames to achieve tracking;

[0043] Color-based matching: Tracking is achieved by calculating the color histogram of the target and matching areas of similar color in consecutive frames;

[0044] Contour-based matching: Tracking is achieved by calculating the contour features of the target and matching the areas with the same contour in consecutive frames;

[0045] f. Collaborative work of high-precision positioning algorithm module: To achieve accurate positioning of the target, the target recognition and tracking module works in collaboration with the high-precision positioning algorithm module. The high-precision positioning algorithm module provides high-precision target location information by integrating multiple positioning information, specifically GPS, UWB, and RTK; that is, during the target tracking process, the current position of the target is corrected through the location information provided by the high-precision positioning algorithm module. When the target enters an area with poor GPS signals, it automatically switches to UWB or RTK positioning information to ensure the accuracy of the target position; by working in collaboration with the high-precision positioning algorithm module, the target recognition and tracking module can achieve continuous tracking and precise positioning of the target, ensuring that the target's location information is always accurate and reliable during the entire movement process;

[0046] g. Anomaly detection and processing:

[0047] During the target tracking process, if the target is blocked, out of the field of view, or the sensor data is abnormal, the target loss detection algorithm is used to detect whether the target is lost. The specific target loss detection algorithm includes: motion trajectory prediction: by predicting the target's motion trajectory, determine whether the target deviates from the expected position; feature matching failure: by detecting the success rate of target feature matching, determine whether the target is lost;

[0048] When the target is lost, the target re-identification algorithm finds the target again in the new field of view or sensor data by comparing the target's features. The target re-identification algorithm includes: feature-based re-identification: re-matching and identifying the target in the new field of view or sensor data by extracting the target's features. The target re-identification algorithm includes: feature-based re-identification: re-matching and identifying the target in the new field of view or sensor data by extracting the target's features; deep learning-based re-identification: automatically learning and extracting high-level features of the target through a deep neural network, and then using these features for re-identification;

[0049] The target is re-identified, and the system attempts to restore the target's motion trajectory through the target trajectory recovery algorithm. Specifically, the interpolation algorithm is used to fill in the trajectory data during the period when the target is lost. Specifically, linear interpolation or spline interpolation is used to estimate the target's motion trajectory during the period when the target is lost based on the position information before and after the target is lost.

[0050] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention can realize high-precision identification of the target by combining a variety of image processing and pattern recognition technologies. At the same time, by using advanced target tracking algorithms, the target can be tracked continuously and stably to ensure that the position information of the target is always accurate and reliable throughout the movement process; the present invention realizes all-round monitoring of the target by fusing data from a variety of sensors (such as cameras, radars, infrared sensors, lidars, etc.). Multi-sensor data fusion can not only improve the robustness and anti-interference ability of the system, but also maintain stable identification and tracking of the target under different environmental conditions (such as daytime, night, rainy and snowy weather, etc.). For example, when the camera vision is limited, radar and lidar data can provide additional distance and speed information to ensure uninterrupted tracking; the present invention realizes high-precision positioning of the target by working in collaboration with the high-precision positioning algorithm module, integrating multiple positioning technologies such as GPS, UWB, RTK, etc. This high-precision positioning information can ensure that the location information of the target in open areas and indoor environments is always accurate and reliable, meeting the application requirements in various complex scenarios; the present invention can maintain continuous tracking of the target even when the target is blocked, out of the field of view, or the sensor data is abnormal through target loss detection and target re-identification mechanism. For example, when the target is temporarily blocked, the system can quickly re-identify and locate the target through methods such as feature matching and trajectory prediction to ensure uninterrupted tracking; the present invention adopts a modular design, and each functional module (such as image processing module, target tracking module, positioning module, etc.) can be independently developed and upgraded to ensure the scalability of the system. By adding or reducing computing nodes and storage nodes, the system can flexibly respond to application scenarios of different scales and meet the needs of different users. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] See also Figure 1 The present invention provides a technical solution: a high-precision space target positioning and processing system, comprising the following modules:

[0054] Multi-sensor fusion module: It is composed of multiple sensors, including optical sensors, radar sensors and infrared sensors. The sensors provide spatial target information and improve the accuracy and reliability of target positioning through data fusion technology;

[0055] High-precision positioning algorithm module: Using the Kalman filter algorithm, the system processes sensor data in real time to achieve high-precision spatial target positioning;

[0056] Efficient data processing architecture: Distributed computing and parallel processing technologies are used to process large amounts of sensor data. At the same time, the efficient data processing architecture has data compression and storage functions to ensure the real-time and reliability of data processing;

[0057] Target recognition and tracking module: Use image processing and pattern recognition technology to identify and track space targets. Through the cooperation with the high-precision positioning algorithm module, continuous tracking and precise positioning of the target are achieved.

[0058] Furthermore, in the multi-sensor fusion module, the optical sensor provides the shape, brightness, color and texture information of the target; the radar sensor measures the distance, speed and direction information of the target; the infrared sensor identifies the temperature distribution and thermal characteristics of the target, wherein the specific working steps of the data fusion technology are as follows:

[0059] Step 1: Data collection: Each sensor simultaneously collects multi-source data of the space target;

[0060] Step 2: Data preprocessing: preprocess the collected multi-source data, including data calibration, noise filtering and data alignment steps;

[0061] Step 3: Data fusion: According to actual needs, the pre-processed data is fused using a fusion algorithm. The data fusion process selects data layer fusion, feature layer fusion or decision layer fusion according to the specific application scenario.

[0062] Step 4: State estimation: Based on the fused data, the state of the target is estimated, including the target's position, velocity, acceleration, and motion trajectory information. The results of the state estimation are used for subsequent target recognition and tracking.

[0063] Furthermore, the specific working steps of the high-precision positioning algorithm module are as follows:

[0064] Step S1, system model construction: In the space target positioning scenario, assuming that a target moving in space is tracked, using Kalman filtering, we first need to define the state and measurement model of the system. The state of the space target is represented by a six-dimensional vector, including the position, velocity and acceleration of the target. Set the target initial state estimate According to the prior knowledge or the initial measurement data of the sensor, the initial covariance matrix P0 is set, and the process noise covariance matrix Q is set. k-1 and the measurement noise covariance matrix R k ;

[0065] Step S2, prediction step: predict the current state according to the state transition equation Predict the covariance matrix of the current state

[0066] Step S3, update step: calculate Kalman gain K k ; Update state estimates based on measurement data Update the state covariance matrix P k ;

[0067] Step S4, repeat: repeat the prediction and update steps for each time step to achieve real-time estimation and tracking of the target state.

[0068] Furthermore, the specific contents of the efficient data processing architecture are as follows:

[0069] A. Distributed computing architecture

[0070] Master node: responsible for task scheduling, node coordination and result aggregation. The master node receives the raw data stream from the sensor, decomposes it into multiple subtasks, and assigns them to the working nodes.

[0071] Worker node: responsible for specific data processing tasks. Each worker node receives subtasks from the master node, performs data parsing, filtering, and compression operations, and returns the processing results to the master node;

[0072] Task scheduling: The master node uses a dynamic task scheduling algorithm to dynamically allocate tasks according to the load of each working node to avoid overloading of certain nodes, thereby achieving load balancing and maximizing system throughput;

[0073] Fault-tolerance mechanism: When a working node fails, the master node will reallocate the tasks on that node to other nodes that are working normally to ensure that data processing tasks are not interrupted;

[0074] B. Parallel processing technology

[0075] Data sharding: The system first shards the received sensor data, breaking down large-scale data into multiple small-scale data blocks. Each data block is processed in parallel by different working nodes;

[0076] Multi-threaded parallel processing: The system uses pipeline processing technology. The data processing process is broken down into multiple stages;

[0077] C. Data compression

[0078] The system uses compression ratio optimization technology to dynamically adjust compression parameters to achieve the best compression effect. Specifically, when the network bandwidth is sufficient, the compression ratio is reduced to reduce the computational overhead during decompression; when the network bandwidth is limited, the compression ratio is increased to reduce the amount of data transmission.

[0079] D. Data storage

[0080] The system adopts a distributed storage system. Sensor data is distributed and stored on multiple storage nodes. Each storage node is responsible for storing a part of the data. Distributed storage improves storage capacity and improves data reliability and availability through data redundancy and copy mechanisms.

[0081] E. Data redundancy and replication mechanism

[0082] To ensure data reliability and availability, the distributed storage system adopts data redundancy and replication mechanisms. Specifically: Data redundancy: When storing sensor data, multiple redundant data will be generated to prevent sensor data loss; Replication mechanism: In the distributed storage system, each sensor data will have multiple copies, stored on different storage nodes;

[0083] F. Real-time guarantee

[0084] The system provides real-time monitoring and alarm functions, which can monitor sensor data in real time and trigger alarms when the data is abnormal. Specifically: when the sensor data exceeds the preset threshold, the system will immediately trigger an alarm and notify relevant personnel to take measures; when the change trend of sensor data is abnormal, the system will trigger an alarm to warn of potential risks in advance; anomaly detection: through machine learning or statistical analysis methods, the system automatically detects abnormal behavior in sensor data and triggers an alarm.

[0085] Furthermore, the specific working steps of the target recognition and tracking module are as follows:

[0086] a. The target recognition and tracking module first receives data from multiple sensors of the multi-sensor fusion module, and fuses the data of different sensors through multi-sensor data fusion technology to obtain more comprehensive and accurate target information;

[0087] b. For the image data collected by the camera, image preprocessing is first performed to improve the accuracy of subsequent target recognition and tracking. The preprocessing steps include: removing noise in the image through filtering algorithms; converting color images into grayscale images to reduce computational complexity; enhancing image contrast and improving image quality through methods such as histogram equalization; and correcting image distortion caused by the camera lens to ensure image accuracy;

[0088] c. Use background modeling technology to build a background model of the scene. Then use foreground extraction technology to separate the foreground target from the background. Then, in the extracted foreground area, the system uses the target detection algorithm to detect the specific target object;

[0089] d. On the basis of detecting the specific target object, further extract the target features for subsequent target recognition, that is, classify the features extracted from the target object through the pattern recognition algorithm to identify the category of the target;

[0090] e. Target tracking: In continuous video frames, the target in the current frame is matched with its position in the previous frames through the target matching algorithm, so as to achieve continuous tracking of the target. The target matching algorithm includes:

[0091] Feature-based matching: By extracting the features of the target, specifically SIFT, SURF, and ORB, the features of the same target are matched in consecutive frames to achieve tracking;

[0092] Color-based matching: Tracking is achieved by calculating the color histogram of the target and matching areas of similar color in consecutive frames;

[0093] Contour-based matching: Tracking is achieved by calculating the contour features of the target and matching the areas with the same contour in consecutive frames;

[0094] f. Collaborative work of high-precision positioning algorithm module: To achieve accurate positioning of the target, the target recognition and tracking module works in collaboration with the high-precision positioning algorithm module. The high-precision positioning algorithm module provides high-precision target location information by integrating multiple positioning information, specifically GPS, UWB, and RTK; that is, during the target tracking process, the current position of the target is corrected through the location information provided by the high-precision positioning algorithm module. When the target enters an area with poor GPS signals, it automatically switches to UWB or RTK positioning information to ensure the accuracy of the target position; by working in collaboration with the high-precision positioning algorithm module, the target recognition and tracking module can achieve continuous tracking and precise positioning of the target, ensuring that the target's location information is always accurate and reliable during the entire movement process;

[0095] g. Anomaly detection and processing:

[0096] During the target tracking process, if the target is blocked, out of the field of view, or the sensor data is abnormal, the target loss detection algorithm is used to detect whether the target is lost. The specific target loss detection algorithm includes: motion trajectory prediction: by predicting the target's motion trajectory, determine whether the target deviates from the expected position; feature matching failure: by detecting the success rate of target feature matching, determine whether the target is lost;

[0097] When the target is lost, the target re-identification algorithm finds the target again in the new field of view or sensor data by comparing the target's features. The target re-identification algorithm includes: feature-based re-identification: re-matching and identifying the target in the new field of view or sensor data by extracting the target's features. The target re-identification algorithm includes: feature-based re-identification: re-matching and identifying the target in the new field of view or sensor data by extracting the target's features; deep learning-based re-identification: automatically learning and extracting high-level features of the target through a deep neural network, and then using these features for re-identification;

[0098] The target is re-identified, and the system attempts to restore the target's motion trajectory through the target trajectory recovery algorithm. Specifically, the interpolation algorithm is used to fill in the trajectory data during the period when the target is lost. Specifically, linear interpolation or spline interpolation is used to estimate the target's motion trajectory during the period when the target is lost based on the position information before and after the target is lost.

[0099] The present invention mainly relates to target recognition and tracking technology in the field of space exploration and aviation, and aims to achieve efficient recognition and continuous tracking of space targets and aircraft through advanced image processing, pattern recognition, multi-sensor data fusion and high-precision positioning algorithms. The system integrates data from multiple sensors (such as cameras, radars, infrared sensors, lidar, etc.) and uses background modeling, foreground extraction, target detection, feature extraction and pattern recognition to accurately identify and track space targets and aircraft.

[0100] A key innovation of the present invention is the multi-sensor data fusion technology, which can fuse different types of sensor data to provide more comprehensive and accurate target information. In addition, the system achieves accurate positioning and continuous tracking of the target through the collaborative work of high-precision positioning algorithm modules, and can maintain high-precision positioning through UWB, RTK and other technologies even in areas with poor GPS signals.

[0101] In order to cope with complex situations such as the target being obscured, leaving the field of view or sensor data being abnormal, the present invention designs an anomaly detection and processing mechanism, including functions such as target loss detection, target re-identification and trajectory recovery, to ensure that the system can maintain stable tracking of the target in various complex scenarios.

[0102] The present invention also improves the processing efficiency and scalability of the system through parallel computing, hardware acceleration and modular design, ensuring its high efficiency and stability in space exploration and aviation applications.

[0103] In summary, the present invention provides an advanced target recognition and tracking solution suitable for space exploration and aviation fields, with the characteristics of high precision, high reliability and high scalability, providing important support for the technological development and application of related fields.

[0104] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A high-precision space target positioning and processing system, characterized in that: Includes the following modules: Multi-sensor fusion module: It is composed of multiple sensors, including optical sensors, radar sensors and infrared sensors. The sensors provide spatial target information and improve the accuracy and reliability of target positioning through data fusion technology; High-precision positioning algorithm module: Using the Kalman filter algorithm, the system processes sensor data in real time to achieve high-precision spatial target positioning; Efficient data processing architecture: Distributed computing and parallel processing technologies are used to process large amounts of sensor data. At the same time, the efficient data processing architecture has data compression and storage functions to ensure the real-time and reliability of data processing; Target recognition and tracking module: Use image processing and pattern recognition technology to identify and track space targets. Through the cooperation with the high-precision positioning algorithm module, continuous tracking and precise positioning of the target are achieved.

2. A high-precision space target positioning and processing system according to claim 1, characterized in that: The optical sensor in the multi-sensor fusion module provides the shape, brightness, color and texture information of the target; the radar sensor measures the distance, speed and direction of the target; the infrared sensor identifies the temperature distribution and thermal characteristics of the target. The specific working steps of the data fusion technology are as follows: Step 1: Data collection: Each sensor simultaneously collects multi-source data of the space target; Step 2: Data preprocessing: preprocess the collected multi-source data, including data calibration, noise filtering and data alignment steps; Step 3: Data fusion: According to actual needs, the pre-processed data is fused using a fusion algorithm. The data fusion process selects data layer fusion, feature layer fusion or decision layer fusion according to the specific application scenario. Step 4: State estimation: Based on the fused data, the state of the target is estimated, including the target's position, velocity, acceleration, and motion trajectory information. The results of the state estimation are used for subsequent target recognition and tracking.

3. A high-precision space target positioning and processing system according to claim 1, characterized in that: The specific working steps of the high-precision positioning algorithm module are as follows: Step S1, system model construction: In the space target positioning scenario, assuming that a target moving in space is tracked, using Kalman filtering, we first need to define the state and measurement model of the system. The state of the space target is represented by a six-dimensional vector, including the position, velocity and acceleration of the target. Set the target initial state estimate According to the prior knowledge or the initial measurement data of the sensor, the initial covariance matrix P0 is set, and the process noise covariance matrix Q is set. k-1 and the measurement noise covariance matrix R k ; Step S2, prediction step: predict the current state according to the state transition equation Predict the covariance matrix of the current state Step S3, update step: calculate Kalman gain K k ; Update state estimates based on measurement data Update the state covariance matrix P k ; Step S4, repeat: repeat the prediction and update steps for each time step to achieve real-time estimation and tracking of the target state.

4. A high-precision space target positioning and processing system according to claim 1, characterized in that: The specific contents of the efficient data processing architecture are as follows: A. Distributed computing architecture Master node: responsible for task scheduling, node coordination and result aggregation. The master node receives the raw data stream from the sensor, decomposes it into multiple subtasks, and assigns them to the working nodes. Worker node: responsible for specific data processing tasks. Each worker node receives subtasks from the master node, performs data parsing, filtering, and compression operations, and returns the processing results to the master node; Task scheduling: The master node uses a dynamic task scheduling algorithm to dynamically allocate tasks according to the load of each working node to avoid overloading of certain nodes, thereby achieving load balancing and maximizing system throughput; Fault-tolerance mechanism: When a working node fails, the master node will reallocate the tasks on that node to other nodes that are working normally to ensure that data processing tasks are not interrupted; B. Parallel processing technology Data sharding: The system first shards the received sensor data, breaking down large-scale data into multiple small-scale data blocks. Each data block is processed in parallel by different working nodes; Multi-threaded parallel processing: The system uses pipeline processing technology. The data processing process is broken down into multiple stages; C. Data compression The system uses compression ratio optimization technology to dynamically adjust compression parameters to achieve the best compression effect. Specifically, when the network bandwidth is sufficient, the compression ratio is reduced to reduce the computational overhead during decompression; when the network bandwidth is limited, the compression ratio is increased to reduce the amount of data transmission. D. Data storage The system adopts a distributed storage system. Sensor data is distributed and stored on multiple storage nodes. Each storage node is responsible for storing a part of the data. Distributed storage improves storage capacity and improves data reliability and availability through data redundancy and copy mechanisms. E. Data redundancy and replication mechanism To ensure data reliability and availability, the distributed storage system adopts data redundancy and replication mechanisms. Specifically: Data redundancy: When storing sensor data, multiple redundant data will be generated to prevent sensor data loss; Replication mechanism: In the distributed storage system, each sensor data will have multiple copies, stored on different storage nodes; F. Real-time guarantee The system provides real-time monitoring and alarm functions, which can monitor sensor data in real time and trigger alarms when the data is abnormal. Specifically: when the sensor data exceeds the preset threshold, the system will immediately trigger an alarm and notify relevant personnel to take measures; when the change trend of sensor data is abnormal, the system will trigger an alarm to warn of potential risks in advance; anomaly detection: through machine learning or statistical analysis methods, the system automatically detects abnormal behavior in sensor data and triggers an alarm.

5. The high-precision space target positioning and processing system according to claim 1, characterized in that: The specific working steps of the target recognition and tracking module are as follows: a. The target recognition and tracking module first receives data from multiple sensors of the multi-sensor fusion module, and fuses the data of different sensors through multi-sensor data fusion technology to obtain more comprehensive and accurate target information; b. For the image data collected by the camera, image preprocessing is first performed to improve the accuracy of subsequent target recognition and tracking. The preprocessing steps include: removing noise in the image through filtering algorithms; converting color images into grayscale images to reduce computational complexity; enhancing image contrast and improving image quality through methods such as histogram equalization; and correcting image distortion caused by the camera lens to ensure image accuracy; c. Use background modeling technology to build a background model of the scene. Then use foreground extraction technology to separate the foreground target from the background. Then, in the extracted foreground area, the system uses the target detection algorithm to detect the specific target object; d. On the basis of detecting the specific target object, further extract the target features for subsequent target recognition, that is, classify the features extracted from the target object through the pattern recognition algorithm to identify the category of the target; e. Target tracking: In continuous video frames, the target in the current frame is matched with its position in the previous frames through the target matching algorithm, so as to achieve continuous tracking of the target. The target matching algorithm includes: Feature-based matching: By extracting the features of the target, specifically SIFT, SURF, and ORB, the features of the same target are matched in consecutive frames to achieve tracking; Color-based matching: Tracking is achieved by calculating the color histogram of the target and matching areas of similar color in consecutive frames; Contour-based matching: Tracking is achieved by calculating the contour features of the target and matching the areas with the same contour in consecutive frames; f. Collaborative work of high-precision positioning algorithm module: To achieve accurate positioning of the target, the target recognition and tracking module works in collaboration with the high-precision positioning algorithm module. The high-precision positioning algorithm module provides high-precision target location information by integrating multiple positioning information, specifically GPS, UWB, and RTK; that is, during the target tracking process, the current position of the target is corrected through the location information provided by the high-precision positioning algorithm module. When the target enters an area with poor GPS signals, it automatically switches to UWB or RTK positioning information to ensure the accuracy of the target position; by working in collaboration with the high-precision positioning algorithm module, the target recognition and tracking module can achieve continuous tracking and precise positioning of the target, ensuring that the target's location information is always accurate and reliable during the entire movement process; g. Anomaly detection and processing: During the target tracking process, if the target is blocked, out of the field of view, or the sensor data is abnormal, the target loss detection algorithm is used to detect whether the target is lost. The specific target loss detection algorithm includes: motion trajectory prediction: by predicting the target's motion trajectory, determine whether the target deviates from the expected position; feature matching failure: by detecting the success rate of target feature matching, determine whether the target is lost; When the target is lost, the target re-identification algorithm finds the target again in the new field of view or sensor data by comparing the target's features. The target re-identification algorithm includes: feature-based re-identification: re-matching and identifying the target in the new field of view or sensor data by extracting the target's features. The target re-identification algorithm includes: feature-based re-identification: re-matching and identifying the target in the new field of view or sensor data by extracting the target's features; deep learning-based re-identification: automatically learning and extracting high-level features of the target through a deep neural network, and then using these features for re-identification; The target is re-identified, and the system attempts to restore the target's motion trajectory through the target trajectory recovery algorithm. Specifically, the interpolation algorithm is used to fill in the trajectory data during the period when the target is lost. Specifically, linear interpolation or spline interpolation is used to estimate the target's motion trajectory during the period when the target is lost based on the position information before and after the target is lost.

Images