Joint space sensing and communication system and method based on computer vision

Abstract

The invention discloses a combined space sensing and communication system and method based on computer vision, and belongs to the field of outer space combined space sensing and communication. The system comprises an optical camera, a computer vision unit, an information processing unit and a communication component. The optical camera is used for collecting data of an outer space scene covering a preset area. And the computer vision unit preprocesses the collected image and analyzes the image to obtain target information. And the information processing unit performs beam former selection and pilot design according to the target information data. The communication assembly is used for completing the communication process of the outer space. According to the invention, the computer vision technology CV is combined with communication processing, accurate positioning of a target position is realized based on the CV technology, and joint perception and communication are expanded to the outer space for satellite detection of surrounding environment and identification of a target user, and are used for tracking the target user. Therefore, high spectrum efficiency and low resource cost of space sensing and communication are realized.

Description

Technical Field

[0001] The present invention relates to a joint space sensing and communication system and method based on open-loop computer vision applicable to outer space scenes, belonging to the field of outer space joint space sensing and communication. Background Art

[0002] Satellite communications can be combined with traditional terrestrial wireless communications to achieve seamless global coverage. For example, three geostationary orbit satellites can provide continuous, fixed services to customers and ground stations. Recently, the number of non-geostationary orbit satellites that can operate at speeds of up to 25,000 km / h has also increased significantly, causing outer space to become increasingly crowded and dangerous. As a result, it is difficult to communicate effectively with target users in such a complex and changeable outer space. On the other hand, since traditional RF technology is limited in transmitting large amounts of data, satellite communications have recently adopted terahertz bands and free-space optics to meet the growing high-speed transmission needs of inter-satellite and satellite-to-ground links in advanced broadband Internet mega-constellations. Therefore, this high-bandwidth, high-frequency satellite link has strict requirements for high directivity.

[0003] In this highly dynamic outer space scenario, frequency communication technology is an inevitable and huge challenge. To meet the above challenges, inspired by the concept of joint perception and communication proposed for traditional ground communications, extending the concept of joint perception and communication to outer space can help satellites detect the surrounding environment, identify target users and track them, thereby achieving high spectrum efficiency and low resource cost. However, the existing joint space communication technology is not suitable for space-based communication systems due to two main reasons. First, passive radars, which are usually used to detect the environment, cannot be used in outer space environments due to the difficulty in detecting signal fluctuations. Second, active radar detection technology brings additional problems such as long delay time and security issues. In traditional joint space station systems, closed-loop procedures detect the current situation and provide feedback information for communication. However, due to the considerable distance and relative speed between satellites, this process will greatly delay the outer space scene. In addition, security issues cannot be avoided because the increase in the number of satellites makes the space crowded and vulnerable to attackers eavesdropping on the active radar pilot signal.

[0004] Inspired by the Joint Sensing and Communication (JSC) concept proposed for traditional terrestrial communications, extending the JSC concept to outer space can help satellites detect the surrounding environment, identify target users and track them, achieving high spectrum efficiency and low resource cost.

[0005] Existing joint space communication technology is not directly applicable to space-based communication systems, mainly due to two key factors. First, traditional passive radar technology, although commonly used for environmental detection, is not suitable for direct application because it is difficult to cope with signal fluctuations in outer space. Second, although active radar detection technology has its advantages, the delay problems and potential security risks it introduces, such as being vulnerable to attackers eavesdropping on active radar pilot signals, have become obstacles to its application in outer space.

[0006] In the traditional joint space station system, a closed-loop procedure is used to detect the current situation and provide feedback information for communication. However, due to the considerable distance and relative speed between satellites, this process will greatly delay the outer space scene. In addition, as the number of satellites increases, the space congestion increases, and security issues become increasingly prominent. Active radar pilot signals are more vulnerable to eavesdropping by attackers. Summary of the invention

[0007] The main purpose of the present invention is to provide a joint space sensing and communication system and method based on computer vision, and to construct an open-loop joint space sensing and communication JSSC method that integrates computer vision CV (Computer Vision) decision-making method. Combining CV technology with communication processing, accurately positioning the target position based on CV technology, and extending joint perception and communication to outer space for satellite detection of the surrounding environment, identification of target users, and tracking of target users, thereby achieving high spectrum efficiency and low resource cost of space sensing and communication.

[0008] The purpose of the present invention is achieved through the following technical solutions.

[0009] The present invention discloses a computer vision-based joint space sensing and communication system, which includes an optical camera, a computer vision unit, an information processing unit and a communication component.

[0010] The optical camera is used to collect data of outer space scenes covering a preset area to achieve space monitoring and situational awareness of outer space scenes. The optical camera inputs the collected satellite image data into the computer vision unit.

[0011] The computer vision unit performs preprocessing on the collected images, including denoising, normalization and image resizing. The computer vision unit uses computer vision algorithms to analyze the images and obtain target information. The specific implementation method is: using integral image technology to quickly calculate the pixel sum of any area in the image, significantly improving the efficiency of image processing. The most representative feature points are selected from the image through the feature selection algorithm, and the essential attributes of the target object are accurately reflected through the most representative feature points. The detection cascade method is adopted to combine multiple simple detectors into a cascade detector. Through step-by-step filtering, the target in the image is accurately detected and located to obtain target information. The computer vision unit sends the obtained target information data to the information processing unit.

[0012] The information processing unit performs beamformer selection and pilot design based on the target information data, and significantly reduces the beamforming training time of each communication consistency block through the angle information in the target information input by the computer vision unit. In the implementation of beamforming training, the information processing unit constructs a model called "beam prediction network". The beam prediction network takes the angle information in the target information data as input, and predicts the optimal beamforming parameters through analysis. The beamforming parameters include the direction, width and gain of the beam. Through the target information input by the computer vision unit, the beam prediction network determines the beamforming parameters, significantly reducing the beamforming training time of each communication consistency block. At the same time, combined with the beam prediction network topology information, the information processing unit can also optimize the power and signal transmission parameters to improve the communication performance of the joint space sensing and communication system.

[0013] The beam prediction network is a multi-layer perceptron MLP structure, in which each layer of perceptron maps input data to output data through a series of weight matrices W and bias terms b.

[0014] 1. Input layer: The input layer receives target information from the computer vision unit, especially angle information. The target information is represented in the form of vectors or matrices, such as [angle1,angle2,...,angleN].

[0015] 2. Hidden layer: The hidden layer transforms the input data through nonlinear activation functions ReLU, Sigmoid and Tanh.

[0016] The output of each hidden layer is the input to the next hidden layer or output layer.

[0017] h i =f(W i ·h i-1 +b i )

[0018] Among them, hi is the output of the i-th hidden layer, w i is the weight matrix of this layer, b i is the bias term and f is the activation function.

[0019] 3. Output layer: The output layer generates beamforming parameters such as beam direction, width, and gain. These parameters may be expressed in the form of a vector: [beam_direction, beam_width, beam_gain]. The calculation of the output layer is similar to that of the hidden layer, and the activation function is:

[0020] output=W output ·h last +b output

[0021] Where output is the beamforming parameter vector, W output is the weight matrix of the output layer, b output is the bias term, h last is the output of the last hidden layer.

[0022] Training the beam prediction network involves optimizing the weights and bias terms of the beam prediction network using labeled training data, i.e., known target information and corresponding beamforming parameters. New target information is input into the beam prediction network, and the output is calculated by forward propagation of the beam prediction network. The beam prediction network is trained by minimizing the mean square error between the predicted output and the true output. The gradient descent optimization algorithm used in the training process updates the weights and bias terms to gradually reduce the value of the loss function. When the beam prediction network is trained, the trained beam prediction network is used to predict the beamforming parameters corresponding to the new target information.

[0023] The communication component is used for communication in outer space, using radio frequency, laser or a combination of the two. According to the power and signal transmission parameters determined by the information processing unit, the communication component is configured for data transmission, and the data is transmitted to the target user or ground station through the communication component to achieve efficient communication.

[0024] The RF beamforming range covers high frequency (20-30 MHz) to S band (3.4 GHz).

[0025] The present invention discloses a computer vision-based joint space sensing and communication method, which is implemented based on the computer vision-based joint space sensing and communication system. The computer vision-based joint space sensing and communication method comprises the following steps:

[0026] Step 1: Use a high-resolution optical camera to collect data on outer space scenes covering a preset area to achieve spatial monitoring and situational awareness of outer space scenes. Adjust the shooting parameters according to the required outer space scene range requirements to ensure that the optical camera works normally, and the shooting parameters include exposure time and focal length.

[0027] Step 2: The collected outer space scene data is transferred to the computer vision unit;

[0028] Step 3: The computer vision unit preprocesses the collected images. The preprocessing steps include denoising, normalization, and image resizing.

[0029] Among them, denoising is to remove the random phase noise caused by transmission in the image according to the current orbital parameters of the satellite; normalization is to adjust the brightness and contrast of the desired pixels in the image according to the attributes of the pixels with the highest brightness and contrast in the current image, so that they are all within the predetermined numerical range; image resizing is to adjust the image quantization bit width according to the system's preset input requirements.

[0030] Step 4: The computer vision unit uses computer vision algorithms to analyze the image and obtain target information. The specific implementation method is: use integral image technology to quickly calculate the pixel sum of any area in the image, significantly improving the efficiency of image processing. The most representative feature points are selected from the image through the feature selection algorithm, and the essential attributes of the target object are accurately reflected through the most representative feature points. The detection cascade method is adopted to combine multiple simple detectors into a cascade detector. Through step-by-step filtering, the target in the image is accurately detected and located to obtain target information. The computer vision unit sends the obtained target information data to the information processing unit.

[0031] Step 5: The information processing unit decodes and interprets the target information data input by the computer vision unit to understand the meaning of the target information data and further begins to determine the specific communication requirements and parameters of the open-loop joint space sensing and communication JSSC method.

[0032] Among them, in the process of determining communication needs, the information processing unit will comprehensively consider various factors to make a series of decisions, including selecting a beamformer to optimize the direction and strength of signal transmission, and designing a pilot signal to ensure the realization of joint space sensing and communication.

[0033] The multiple factors include the location and moving speed of the target, the complexity of the environment, and the real-time requirements of the communication.

[0034] Step 6: Configure the communication component for data transmission according to the communication requirements and parameters determined by the information processing unit. The communication component uses radio frequency, laser or a combination of the two to select the appropriate communication method according to the specific scenario and requirements. During the configuration process, the communication component is ensured to work properly by dynamically adjusting parameters such as the satellite orbit altitude, satellite cruise trajectory and optical camera exposure factor. The communication component transmits data to the target user or ground station to achieve efficient communication.

[0035] Beneficial effects:

[0036] 1. The present invention discloses a computer vision-based joint spatial sensing and communication system and method, which performs information analysis based on computer vision images, uses integral image technology to quickly calculate the pixel sum of any area in the image, and significantly improves the efficiency of image processing. The most representative feature points are selected from the image through a feature selection algorithm, and the essential attributes of the target object are accurately reflected through the most representative feature points. The detection cascade method is adopted to combine multiple simple detectors into a cascade detector, and the target in the image is accurately detected and located by step-by-step filtering to obtain target information.

[0037] 2. The present invention discloses a computer vision-based joint space sensing and communication system and method, which performs beamforming training based on the input of the computer vision module, and constructs a beam prediction network model through the target information input by the computer vision unit. The model can quickly and accurately determine the beamforming parameters, thereby significantly reducing the beamforming training time of each communication consistency block. At the same time, combined with the network topology information, the information processing unit can also optimize the power and signal transmission parameters to improve the communication performance of the joint space sensing and communication system.

[0038] 3. The present invention discloses a computer vision-based joint space sensing and communication system and method, which combines CV technology with communication processing, realizes precise positioning of the target position based on CV technology, and extends joint perception and communication to outer space for satellite detection of the surrounding environment, identification of target users, and tracking of target users, thereby achieving high spectrum efficiency and low resource cost of space sensing and communication. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1The present invention discloses a flow chart of a joint spatial sensing and communication method based on computer vision. DETAILED DESCRIPTION

[0041] The present invention is further described and described in detail below with reference to the accompanying drawings and embodiments.

[0042] like Figure 1 As shown, this embodiment discloses a computer vision-based joint spatial sensing and communication system and method, including an optical camera, a computer vision unit, an information processing unit and a communication component.

[0043] The optical camera is used to collect data of outer space scenes covering a preset area to achieve space monitoring and situational awareness of outer space scenes. The optical camera inputs the collected satellite image data into the computer vision unit.

[0044] The computer vision unit performs preprocessing on the collected images, including denoising, normalization and image resizing. The computer vision unit uses computer vision algorithms to analyze the images and obtain target information. The specific implementation method is: using integral image technology to quickly calculate the pixel sum of any area in the image, significantly improving the efficiency of image processing. The most representative feature points are selected from the image through the feature selection algorithm, and the essential attributes of the target object are accurately reflected through the most representative feature points. The detection cascade method is adopted to combine multiple simple detectors into a cascade detector. Through step-by-step filtering, the target in the image is accurately detected and located to obtain target information. The computer vision unit sends the obtained target information data to the information processing unit.

[0045] The information processing unit selects and pre-designs the beamformer based on the target information data, and significantly reduces the beamforming training time of each communication consistency block through the angle information in the target information input by the computer vision unit. In the implementation of beamforming training, the information processing unit first constructs a model called "beam prediction network". This model uses the angle information in the target information data as the main input, and predicts the optimal beamforming parameters, such as the direction, width and gain of the beam through analysis. Through the target information input by the computer vision unit, the beam prediction network can quickly and accurately determine the beamforming parameters, thereby significantly reducing the beamforming training time of each communication consistency block. At the same time, combined with the network topology information, the information processing unit can also optimize the power and signal transmission parameters to improve the communication performance of the joint space sensing and communication system.

[0046] The model of the “beam prediction network” is a multi-layer perceptron (MLP) structure, in which each layer of perceptron maps input data to output data through a series of weight matrices (W) and bias terms (b).

[0047] 4. Input layer: The input layer receives target information from the computer vision unit, especially angle information. This information may be represented in the form of vectors or matrices, such as [angle1,angle2,...,angleN].

[0048] 5. Hidden layer: The hidden layer transforms the input data through nonlinear activation functions such as ReLU, Sigmoid, and Tanh. The output of each hidden layer is the input of the next hidden layer or output layer. Mathematically, this can be expressed as:

[0049] h i =f(W i ·h i-1 +b i )

[0050] Among them, h i is the output of the i-th hidden layer, w i is the weight matrix of this layer, b i is the bias term and f is the activation function.

[0051] 6. Output layer: The output layer generates beamforming parameters such as beam direction, width, and gain. These parameters may be represented in the form of a vector: [beam_direction, beam_width, beam_gain]. The calculation of the output layer is similar to that of the hidden layer, and the activation function is:

[0052] output=W output ·h last +b output

[0053] Where output is the beamforming parameter vector, W output is the weight matrix of the output layer, b output is the bias term, h last is the output of the last hidden layer.

[0054] Training a beam prediction network involves optimizing the network's weights and bias terms using labeled training data (i.e., known target information and corresponding beamforming parameters). This is typically done by minimizing the mean squared error between the predicted output and the true output. The gradient descent optimization algorithm used during training updates the weights and bias terms to gradually reduce the value of the loss function. Once the network is trained, it can be used to predict beamforming parameters for new target information. This simply requires feeding the new target information into the network and computing the output through the network's forward propagation.

[0055] The communication component uses radio frequency, laser or a combination of the two to communicate. According to the power and signal transmission parameters determined by the information processing unit, the communication component is configured for data transmission, and the data is transmitted to the target user or ground station through the communication component to achieve an efficient communication process.

[0056] The RF beamforming range covers high frequency (20-30 MHz) to S band (3.4 GHz).

[0057] like Figure 1 As shown, the computer vision-based joint spatial sensing and communication method disclosed in this embodiment is implemented based on the computer vision-based joint spatial sensing and communication system. The computer vision-based joint spatial sensing and communication method comprises the following steps:

[0058] Step 1: Use a high-resolution optical camera to collect data on outer space scenes covering a preset area to achieve spatial monitoring and situational awareness of outer space scenes. Adjust the shooting parameters according to the required outer space scene range requirements to ensure that the optical camera works normally, and the shooting parameters include exposure time and focal length.

[0059] Step 2: The collected outer space scene data is transferred to the computer vision unit;

[0060] Step 3: The computer vision unit preprocesses the collected images. The preprocessing steps include denoising, normalization, and image resizing.

[0061] Among them, denoising is to remove the random phase noise caused by transmission in the image according to the current orbital parameters of the satellite; normalization is to adjust the brightness and contrast of the desired pixels in the image according to the attributes of the pixels with the highest brightness and contrast in the current image, so that they are all within the predetermined numerical range; image resizing is to adjust the image quantization bit width according to the system's preset input requirements.

[0062] Step 4: The computer vision unit uses computer vision algorithms to analyze the image and obtain target information. The specific implementation method is: use integral image technology to quickly calculate the pixel sum of any area in the image, significantly improving the efficiency of image processing. The most representative feature points are selected from the image through the feature selection algorithm, and the essential attributes of the target object are accurately reflected through the most representative feature points. The detection cascade method is adopted to combine multiple simple detectors into a cascade detector. Through step-by-step filtering, the target in the image is accurately detected and located to obtain target information. The computer vision unit sends the obtained target information data to the information processing unit.

[0063] Step 5: The information processing unit decodes and interprets the target information data input by the computer vision unit to understand the meaning of the target information data and further begins to determine the specific communication requirements and parameters of the open-loop joint space sensing and communication JSSC method.

[0064] Among them, in the process of determining communication needs, the information processing unit will comprehensively consider various factors to make a series of decisions, including selecting a beamformer to optimize the direction and strength of signal transmission, and designing a pilot signal to ensure the realization of joint space sensing and communication.

[0065] The multiple factors include the location and moving speed of the target, the complexity of the environment, and the real-time requirements of the communication.

[0066] Step 6: According to the communication requirements and parameters determined by the information processing unit, configure the communication component for data transmission. The communication component uses radio frequency, laser or a combination of the two to select the appropriate communication method according to the specific scenario and requirements. During the configuration process, the normal operation of the communication component is ensured by dynamically adjusting parameters such as the satellite orbit altitude, satellite cruise trajectory, and optical camera exposure factor. The communication component transmits data to the target user or ground station to achieve an efficient communication process.

[0067] The main content of the joint spatial sensing and communication system based on open-loop computer vision described above, through the system structure provided by the embodiment of the present invention, can detect the environment and users by using the observed image or video data, without waiting for feedback signals, and has high concealment, low power consumption and low latency characteristics. Through the description of the above implementation methods, those skilled in the art can clearly understand the structure of the joint spatial sensing and communication system based on open-loop computer vision.

Claims

1. A computer vision-based joint spatial sensing and communication system, characterized in that: Includes optical camera, computer vision unit, information processing unit and communication components; The optical camera is used to collect data of outer space scenes covering a preset area to achieve space monitoring and situational awareness of outer space scenes; the optical camera inputs the collected satellite image data into the computer vision unit; The computer vision unit performs preprocessing on the collected images, including denoising, normalization and image resizing. The computer vision unit uses computer vision algorithms to analyze the images and obtain target information. The specific implementation method is: using integral image technology to quickly calculate the pixel sum of any area in the image, significantly improving the efficiency of image processing; The most representative feature points are selected from the image through the feature selection algorithm, and the essential attributes of the target object are accurately reflected through the most representative feature points; the detection cascade method is adopted to combine multiple simple detectors into a cascade detector, and the target in the image is accurately detected and located through step-by-step filtering to obtain target information; The computer vision unit sends the acquired target information data to the information processing unit; The information processing unit selects and pre-designs a beamformer according to the target information data, and significantly reduces the beamforming training time of each communication consistency block through the angle information in the target information input by the computer vision unit; in the implementation of beamforming training, the information processing unit constructs a model called "beam prediction network"; the beam prediction network takes the angle information in the target information data as input, and predicts the optimal beamforming parameters through analysis, and the beamforming parameters include the direction, width and gain of the beam; the beam prediction network determines the beamforming parameters through the target information input by the computer vision unit, significantly reducing the beamforming training time of each communication consistency block; at the same time, combined with the topology information of the beam prediction network, the information processing unit can also optimize the power and signal transmission parameters to improve the communication performance of the joint space sensing and communication system; Training the beam prediction network involves optimizing weights and bias items of the beam prediction network using labeled training data, wherein the labeled training data used is known target information and corresponding beamforming parameters; inputting new target information into the beam prediction network, and calculating the output through forward propagation of the beam prediction network, and training the beam prediction network by minimizing the mean square error between the predicted output and the true output; the gradient descent optimization algorithm used in the training process updates the weights and bias items to gradually reduce the value of the loss function; when the beam prediction network is trained, using the trained beam prediction network to predict the beamforming parameters corresponding to the new target information; The communication component is used for communication in outer space, using radio frequency, laser or a combination of the two. According to the power and signal transmission parameters determined by the information processing unit, the communication component is configured for data transmission, and the data is transmitted to the target user or ground station through the communication component to achieve efficient communication.

2. A computer vision-based joint spatial sensing and communication system as claimed in claim 1, characterized in that: The beam prediction network is a multi-layer perceptron MLP structure, in which each layer of perceptron maps input data to output data through a series of weight matrices W and bias terms b; Input layer: The input layer receives target information from the computer vision unit, especially angle information; The target information is represented in the form of vector or matrix; Hidden layer: The hidden layer transforms the input data through nonlinear activation functions such as ReLU, Sigmoid, and Tanh; the output of each hidden layer is the input of the next hidden layer or output layer; h i =f(W i ·h i-1 +b i ) Among them, h i is the output of the i-th hidden layer, w i is the weight matrix of this layer, b i is the bias term and f is the activation function. Output layer: The output layer generates beamforming parameters, which are expressed in the form of a vector: [beam_direction, beam_width, beam_gain]; the calculation of the output layer is similar to that of the hidden layer, and the activation function is: output=W output ·h last +b output Where output is the beamforming parameter vector, W output is the weight matrix of the output layer, b output is the bias term, h last is the output of the last hidden layer.

3. A computer vision-based joint spatial sensing and communication system as claimed in claim 2, characterized in that: The RF beamforming range covers high frequency 20-30 MHz to S-band 3.4 GHz.

4. A computer vision-based joint spatial sensing and communication method, implemented based on a computer vision-based joint spatial sensing and communication system as claimed in claim 1, 2 or 3, characterized in that: The following steps are included: Step 1: Use a high-resolution optical camera to collect data of outer space scenes covering a preset area to achieve space monitoring and situation awareness of outer space scenes; adjust shooting parameters according to the required outer space scene range requirements to ensure the normal operation of the optical camera, and the shooting parameters include exposure time and focal length; Step 2: The collected outer space scene data is transferred to the computer vision unit; Step 3: The computer vision unit preprocesses the collected images. The preprocessing steps include denoising, normalization, and image resizing. Step 4: The computer vision unit uses computer vision algorithms to analyze the image and obtain target information. The specific implementation method is: using integral image technology to quickly calculate the pixel sum of any area in the image, significantly improving the efficiency of image processing; The most representative feature points are selected from the image through the feature selection algorithm, and the essential attributes of the target object are accurately reflected through the most representative feature points; the detection cascade method is adopted to combine multiple simple detectors into a cascade detector, and the target in the image is accurately detected and located through step-by-step filtering to obtain target information; the computer vision unit sends the obtained target information data to the information processing unit; Step 5: The information processing unit decodes and interprets the target information data input by the computer vision unit to understand the meaning of the target information data, and further begins to determine the specific communication requirements and parameters of the open-loop joint space sensing and communication JSSC method; In the process of determining communication requirements, the information processing unit will make a series of decisions based on a variety of factors, including selecting a beamformer to optimize the direction and strength of signal transmission, designing a pilot signal to ensure the realization of joint space sensing and communication; The multiple factors include the location and movement speed of the target, the complexity of the environment, and the real-time requirements of the communication; Step 6: According to the communication requirements and parameters determined by the information processing unit, configure the communication components for data transmission; the communication components use radio frequency, laser or a combination of the two communication methods, and select the appropriate communication method according to the specific scenario and requirements; during the configuration process, dynamically adjust the satellite orbit altitude, satellite cruise trajectory, optical camera exposure factor and other parameters to ensure the normal operation of the communication components; transmit data to the target user or ground station through the communication components to achieve efficient communication.

5. A computer vision-based joint spatial sensing and communication method as claimed in claim 4, characterized in that: In step 3, Denoising is to remove the random phase noise caused by transmission in the image based on the current orbital parameters of the satellite; Normalization is to adjust the brightness and contrast of the desired pixels in the image according to the attributes of the pixels with the highest brightness and contrast in the current image, so that they are both within the predetermined value range; image resizing is to adjust the image quantization bit width according to the system's preset input requirements.

Images