Visual axis stable video generation method and device based on dynamic fuzzy integral

Abstract

The invention provides a visual axis stable video generation method and device based on dynamic fuzzy integration, and the method comprises the steps: collecting inertial pointing data of a servo system in an inertial stability model, and reading a gray value of an initial scene image as a reference frame; performing interpolation processing on the acquired azimuth angle, pitch angle and roll angle data to generate a high-density pointing angle sequence, and calculating azimuth stability precision and pitch stability precision; reversely rotating the reference frame to generate imaging slices, superposing all corresponding imaging slices pixel by pixel in single-frame exposure time, and normalizing a result; and overlapping character information including a field angle, stability precision and an imaging parameter on each frame of image to generate a visual axis stabilization effect video file. In the embodiment of the invention, the visual axis stability precision effect video is generated according to the pointing data and the imaging sensor parameters, the intuition of stability precision effect display in a design stage can be improved, and the dependence of a servo system on an optoelectronic equipment image link in a physical test stage is reduced.

Description

Technical Field

[0001] This invention belongs to the field of motion control technology for optoelectronic devices, and specifically relates to a method and apparatus for generating line-of-sight stabilized video based on dynamic fuzzy integral. Background Technology

[0002] The line-of-sight stabilization accuracy of optoelectronic equipment refers to the stability of the optical axis under conditions such as vehicle-mounted, airborne, and shipborne environments. It reflects the ability to isolate carrier disturbances and is a core indicator of the servo system for optoelectronic equipment on moving platforms. Currently, there are two main methods for evaluating stability performance during the servo system design phase: The first method involves servo system simulation or testing to obtain gyroscope data or geographic pointing angle data from the optoelectronic device. Based on this data, stability accuracy and jitter peak values ​​are calculated, and an angle curve is plotted. The stabilization effect is then evaluated based on these metrics. This method is often used in the servo system design phase. Its drawback is that it requires highly experienced professionals to accurately understand the stabilization effect based on numerical data, and it is not as intuitive and easy to understand as video analysis.

[0003] The second method involves testing with physical optoelectronic equipment to obtain inertial pointing angle data and real imaging video. Frame-by-frame image processing is then performed on the video to extract the pointing angle sequence of a fixed reference target. Based on this angle sequence, stability accuracy and jitter peak are calculated, and angle curves are plotted. The stabilization effect is evaluated based on stability accuracy, jitter peak, angle curves, and the actual video. The disadvantages of this method are the high manpower and material costs of setting up the testing environment, high requirements for weather conditions and a fixed reference target, and the need for extensive software for data storage, video acquisition, data analysis, and video analysis.

[0004] Therefore, how to provide a method and device for generating line-of-sight stabilized video based on dynamic fuzzy integral, which generates a video of line-of-sight stabilization accuracy based on pointing data and imaging sensor parameters, restores intra-frame blurring, improves the intuitiveness of displaying stabilization accuracy in the design stage, and reduces the dependence of the servo system on the image link of optoelectronic equipment in the physical testing stage has become an urgent technical problem to be solved. Summary of the Invention

[0005] This invention provides a method and apparatus for generating line-of-sight stabilized video based on dynamic fuzzy integral. It generates a video of line-of-sight stabilization accuracy based on pointing data and imaging sensor parameters, restores intra-frame blurring, improves the intuitiveness of displaying stabilization accuracy in the design stage, and reduces the dependence of the servo system on the image link of the optoelectronic device in the physical testing stage.

[0006] In one embodiment of the present invention, a method for generating line-of-sight stabilized video based on dynamic fuzzy integral is provided, the method comprising: S101. Data Acquisition: During the simulation or actual test of the servo system, the inertial pointing data of the servo system under the inertial stability model is acquired. The inertial pointing data includes the azimuth angle, pitch angle and roll angle of the line of sight. The data update frequency is not less than 1kHz and the total data duration is not less than 20 seconds. S102, Parameter setting: Set the exposure time, frame rate, image resolution and field of view parameters of the imaging sensor, and read the grayscale value of the initial scene image as the reference frame; S103. Data processing: Based on the exposure time, interpolate the collected azimuth, pitch and roll angle data respectively to generate a high-density pointing angle sequence, and calculate the azimuth stability accuracy and pitch stability accuracy. S104. Generate image: Using the time interval of the interpolated pointing angle sequence as the unit, rotate the reference frame in reverse according to the roll angle corresponding to each unit time to generate an imaging slice; stack all imaging slices corresponding to a single frame exposure time pixel by pixel, and normalize the stacking result to obtain a single frame image. S105. Generate video: Superimpose character information containing field of view, stabilization accuracy and imaging parameters onto each frame image generated in step S104, and combine all single frame images in chronological order to generate a video file of the view stabilization effect.

[0007] Furthermore, the interpolation process employs cubic spline interpolation, piecewise linear interpolation, or radial basis function interpolation methods, and the interpolated data time interval satisfies the following condition: the exposure time is more than 20 times the interpolated data time interval.

[0008] Furthermore, the inertial pointing data originates from the data output by the servo system simulation software, or from the data measured by the gyroscope or inertial measurement unit (IMU) in the physical optoelectronic device.

[0009] Further, the generated image includes: Establish a timeline, align the interpolated angle data sequence with the imaging time, and determine the angle data sequence segment corresponding to the exposure time of each frame image; Based on the reference frame and the roll angle corresponding to each angle data time unit, the grayscale matrix of the imaging slice corresponding to that time unit is calculated through coordinate transformation. The cumulative grayscale matrix is ​​obtained by summing the grayscale matrices of the imaging slices corresponding to all time units within the exposure time of a frame of an image pixel by pixel. The values ​​of each element in the cumulative grayscale matrix are normalized to a preset image grayscale range to obtain the image data of the frame.

[0010] Furthermore, the superimposed character information includes at least: horizontal field of view, vertical field of view, azimuth stabilization accuracy, pitch stabilization accuracy, frame rate, exposure time, and current frame number.

[0011] Furthermore, a fixed reference marker is superimposed at the center of each generated image frame.

[0012] The assessment results of the subsystem health status assessment Weighting based on degradation degree And the weight of subsystem performance evaluation.

[0013] In another embodiment of the present invention, a line-of-sight stabilized video generation device based on dynamic fuzzy integral is provided, based on the line-of-sight stabilized video generation method based on any one of the above claims, the device comprising: The data acquisition module is used to acquire inertial pointing data of the servo system under the inertial stability model during servo system simulation or actual measurement. The data includes azimuth angle, pitch angle and roll angle. The parameter configuration module is used to set the exposure time, frame rate, image resolution, and field of view parameters of the imaging sensor, and to load the initial scene reference frame image; The data processing module is used to interpolate the collected pointing angle data to generate a high-density sequence and calculate the azimuth and pitch stability accuracy. The image synthesis module is used to rotate the reference frame time-by-time according to the interpolated roll angle sequence to generate imaging slices, and to superimpose and normalize all slices within a single frame exposure time to synthesize a single frame image. The video generation module is used to overlay parameter characters onto the synthesized single-frame image and combine them in sequence to generate a video file.

[0014] Furthermore, the data processing module executes a cubic spline interpolation algorithm, and ensures that the data time interval after interpolation satisfies that the exposure time is more than 20 times the interval.

[0015] Furthermore, the operations performed by the image synthesis module include coordinate rotation, pixel translation, matrix summation, and normalization.

[0016] In another embodiment of the present invention, a computer-readable storage medium is provided having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the above-described method for generating a dual-axis-stabilized video based on dynamic fuzzy integral.

[0017] The beneficial effects of this invention are as follows: As can be seen from the above scheme, the embodiments of the present invention provide a method and apparatus for generating line-of-sight stabilized video based on dynamic fuzzy integral. The method involves acquiring inertial pointing data of the servo system under an inertial stabilization model. The inertial pointing data includes the azimuth, pitch, and roll angles of the line of sight. The data update frequency is not less than 1 kHz, and the total data duration is not less than 20 seconds. The method sets the exposure time, frame rate, image resolution, and field of view parameters of the imaging sensor, and reads the grayscale value of the initial scene image as the reference frame. The acquired azimuth, pitch, and roll angle data are interpolated to generate a high-density pointing angle sequence, and the azimuth stabilization accuracy and pitch stabilization accuracy are calculated. Using the time interval of the interpolated pointing angle sequence as the unit, the reference frame is rotated in reverse according to the roll angle corresponding to each unit of time to generate an imaging slice. All imaging slices corresponding to a single frame exposure time are superimposed pixel-by-pixel, and the superposition result is normalized to obtain a single-frame image. Character information containing the field of view, stabilization accuracy, and imaging parameters is superimposed on each of the generated single-frame images, and all single-frame images are combined in chronological order to generate a line-of-sight stabilized video file. In this embodiment of the invention, a video of the visual axis stabilization accuracy effect is generated based on pointing data and imaging sensor parameters, restoring intra-frame blurring, improving the intuitiveness of the stabilization accuracy effect display during the design phase, and reducing the servo system's dependence on the optoelectronic device image link during the physical testing phase. Attached Figure Description

[0018] Figure 1 This is a flowchart of a line-of-sight stabilized video generation method based on dynamic fuzzy integral according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the initial video scene and imaging coordinate system of a line-of-sight stabilization video generation method based on dynamic fuzzy integral according to an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0020] Existing methods for evaluating stabilization effects can only rely on data and curves during the design phase, lacking the intuitive visual impact of photoelectric imaging. While real-world video recordings are available during the physical testing phase, the testing environment is often complex. This invention, under the condition that imaging sensor parameters such as exposure time, frame rate, and image resolution are known, only requires the simulated or measured inertial pointing angle. Through data interpolation, dynamic fuzzy integration, pixel coordinate transformation, slice overlay, normalization, and character overlay, image frames are generated. Finally, the frame sequence is combined to form a video of the visual axis stabilization effect. This effectively improves the intuitiveness of stabilization effect evaluation during the servo system design phase or simplifies the physical testing environment.

[0021] like Figures 1 to 2 As shown, Figure 1 This is a flowchart of a line-of-sight stabilized video generation method based on dynamic fuzzy integral according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the initial video scene and imaging coordinate system of a line-of-sight stabilization video generation method based on dynamic fuzzy integral according to an embodiment of the present invention.

[0022] Figure 1 A line-stabilized video generation method based on dynamic fuzzy integral includes: S101. Data Acquisition: During the simulation or actual test of the servo system, the inertial pointing data of the servo system under the inertial stability model is acquired. The inertial pointing data includes the azimuth angle, pitch angle and roll angle of the line of sight. The data update frequency is not less than 1kHz and the total data duration is not less than 20 seconds. S102, Parameter setting: Set the exposure time, frame rate, image resolution and field of view parameters of the imaging sensor, and read the grayscale value of the initial scene image as the reference frame; S103. Data processing: Based on the exposure time, interpolate the collected azimuth, pitch and roll angle data respectively to generate a high-density pointing angle sequence, and calculate the azimuth stability accuracy and pitch stability accuracy. S104. Generate image: Using the time interval of the interpolated pointing angle sequence as the unit, rotate the reference frame in reverse according to the roll angle corresponding to each unit time to generate an imaging slice; stack all imaging slices corresponding to a single frame exposure time pixel by pixel, and normalize the stacking result to obtain a single frame image. S105. Generate video: Superimpose character information containing field of view, stabilization accuracy and imaging parameters onto each frame image generated in step S104, and combine all single frame images in chronological order to generate a video file of the view stabilization effect.

[0023] In this embodiment of the invention, servo system simulation software or physical servo control software acquires inertial pointing data of the servo system under an inertial stability model in real time during simulation or actual measurement. Data acquisition is performed according to the original update frequency calculated from the gyroscope or attitude data, with an update frequency ≥1kHz and a total data duration ≥20s. Imaging exposure time, imaging frame rate, image resolution, and field of view are set. The grayscale value of the initial scene of the required video is read as the reference frame, and a sensor coordinate system is established. Based on the pointing angle update cycle and exposure time, interpolation is performed on three sets of data: azimuth angle, pitch angle, and roll angle, generating a high-density pointing angle sequence. This ensures that after interpolation, the exposure time is more than 20 times the pointing angle time interval. Then, the azimuth stability accuracy, pitch stability accuracy, and vertical field of view are calculated. Using the interpolated pointing angle data time interval as a unit, the reference frame is rotated according to the roll angle in the sensor coordinate system to obtain an imaging slice within each time unit. All imaging slices within a single frame exposure time are added pixel by pixel, and after normalization, a single-frame image with dynamic fuzzy integral is obtained. The field of view, azimuth stabilization accuracy, pitch stabilization accuracy, frame rate, exposure time, frame number, and other characters are superimposed on the corners of each frame image. Then, all the single frame images are combined according to the required video format to obtain a stabilized video file.

[0024] In this embodiment of the invention, the gyro integral or line-of-sight attitude angle is collected as the line-of-sight pointing angle. Cubic spline interpolation is used to interpolate the line-of-sight pointing angle to ensure that the imaging exposure time is more than 20 times the time interval of the pointing angle. In the imaging coordinate system, the target scene is rotated in the opposite direction by the line-of-sight roll angle to obtain an imaging slice for each time unit. All imaging slices within a single frame exposure time are added pixel by pixel and normalized to obtain a single frame image. The single frame image is superimposed with characters such as field of view, azimuth stabilization accuracy, pitch stabilization accuracy, frame rate, exposure time, and frame number to form an effect video.

[0025] Compared with existing technologies, this invention can generate a video of the visual axis stabilization accuracy effect based on pointing data and imaging sensor parameters during the servo system design or physical testing phases. It can restore intra-frame blurring phenomena, significantly improve the intuitiveness of the stabilization accuracy effect display during the design phase, and greatly reduce the servo system's dependence on the optoelectronic device image link during the physical testing phase.

[0026] In one embodiment of the present invention, the interpolation process employs cubic spline interpolation, piecewise linear interpolation, or radial basis function interpolation, and the interpolated data time interval satisfies the following condition: the exposure time is more than 20 times the interpolated data time interval.

[0027] In another embodiment of the present invention, the inertial pointing data is derived from data output by servo system simulation software, or from data measured by a gyroscope or inertial measurement unit (IMU) in a physical optoelectronic device.

[0028] The servo control software, either from servo system simulation software or the physical servo control software, collects and saves the inertial pointing data of the line of sight during stable accuracy simulation or actual measurement. Data sources include, but are not limited to, gyro integrals and attitude angles measured by the IMU. The acquisition frequency is the original data update frequency, and the data content includes azimuth data. Pitch data and rolling data The data update cycle is The length of all three sets of angle data is [missing information]. .

[0029] In another embodiment of the present invention, the generated image includes: Establish a timeline, align the interpolated angle data sequence with the imaging time, and determine the angle data sequence segment corresponding to the exposure time of each frame image; Based on the reference frame and the roll angle corresponding to each angle data time unit, the grayscale matrix of the imaging slice corresponding to that time unit is calculated through coordinate transformation. The cumulative grayscale matrix is ​​obtained by summing the grayscale matrices of the imaging slices corresponding to all time units within the exposure time of a frame of an image pixel by pixel. The values ​​of each element in the cumulative grayscale matrix are normalized to a preset image grayscale range to obtain the image data of the frame.

[0030] In another embodiment of the present invention, the superimposed character information includes at least: horizontal field of view, vertical field of view, azimuth stabilization accuracy, pitch stabilization accuracy, frame rate, exposure time, and current frame number.

[0031] In another embodiment of the present invention, a fixed reference mark is superimposed at the center position of each generated image frame.

[0032] The assessment results of the subsystem health status assessment Weighting based on degradation degree And the weight of subsystem performance evaluation.

[0033] In this embodiment of the invention, imaging parameters are set, including exposure time. Imaging frame rate Image horizontal resolution Image vertical resolution Horizontal field of view The grayscale values ​​of the initial scene of the required video are read as the reference frame to obtain the grayscale data of the reference frame in the sensor coordinate system. This data is... Line × Column matrix, denoted as ,in Using pixel coordinates, , All are integers, with ranges of 1 and 2. , .like Figure 2 As shown, a sensor coordinate system is established with the image center as the origin 0, the right as x+, and the top as y+.

[0034] Data processing: right , and Perform interpolation separately to obtain the interpolated values. , and Interpolation methods include, but are not limited to, cubic spline interpolation, piecewise linear interpolation, and radial basis function interpolation. After interpolation, , and Data time interval and data length for: (Equation 1) In the formula, min() is a minimum value function that returns the minimum value among the input values; ceil() is a floor function that returns the smallest integer greater than or equal to the input value.

[0035] , and Subtract the value of the first data point from each to obtain the azimuth angle of the line of sight. Pitch angle and roll angle .

[0036] (Equation 2) beg , The standard deviation yields the azimuth stability accuracy. and pitch stability accuracy .

[0037] (Equation 3) In the formula , They are respectively , The average value.

[0038] Based on image horizontal resolution Image vertical resolution Horizontal field of view Calculate the vertical field of view .

[0039] (Equation 4) According to azimuth Pitch angle and roll angle Data length Data interval time and imaging frame rate Calculate the total video duration With total frames .

[0040] (Equation 5) Generate an image, establish a timeline, taking the first angle data as the starting point, align the angle data with the imaging time to obtain the angle data range corresponding to the exposure time of each frame, denoted as . , The frame number, This is the angle data sequence number corresponding to the start time of exposure for this frame of the image. This is the angle data sequence number corresponding to the end of the exposure of this frame. This represents the number of imaging slices in this frame. The grayscale data of the reference frame After pixel translation and rotation, the grayscale data matrix of a single imaging slice is obtained. .

[0041] (Equation 6) In the formula: The frame number has a value range of 1. ; The slice number within the frame, with a value range of... ; , They are respectively and The number of pixels that need to be translated in the direction; round() is a rounding function that returns the integer closest to the input value; ceil() is a floor function that returns the smallest integer greater than or equal to the input value. This is the grayscale data matrix of the translated image; This is the pixel coordinate system after rotating the image around the origin by an angle.

[0042] The grayscale data matrices of all imaging slices within each frame are summed and normalized to obtain a single-frame image with dynamic blur integration. .

[0043] (Equation 7) In the formula, It is the sum of the grayscale data matrices of all imaging slices within this frame. for The maximum value among the matrix elements.

[0044] Overlay the field of view at the corners of each frame. Azimuth stability accuracy and pitch stability accuracy Frame rate Exposure time Frame number The characters are superimposed on a reference cross at the center of the image. Then, all the single-frame images are combined according to the required video format to obtain a stable video file. The video file format includes, but is not limited to, mp4, wmv, avi and other formats.

[0045] In another embodiment of the present invention, a line-of-sight stabilized video generation device based on dynamic fuzzy integral is provided, based on the line-of-sight stabilized video generation method based on any one of the above claims, the device comprising: The data acquisition module is used to acquire inertial pointing data of the servo system under the inertial stability model during servo system simulation or actual measurement. The data includes azimuth angle, pitch angle and roll angle. The parameter configuration module is used to set the exposure time, frame rate, image resolution, and field of view parameters of the imaging sensor, and to load the initial scene reference frame image; The data processing module is used to interpolate the collected pointing angle data to generate a high-density sequence and calculate the azimuth and pitch stability accuracy. The image synthesis module is used to rotate the reference frame time-by-time according to the interpolated roll angle sequence to generate imaging slices, and to superimpose and normalize all slices within a single frame exposure time to synthesize a single frame image. The video generation module is used to overlay parameter characters onto the synthesized single-frame image and combine them in sequence to generate a video file.

[0046] In another embodiment of the present invention, the interpolation algorithm executed by the data processing module is cubic spline interpolation, and ensures that the data time interval after interpolation satisfies that the exposure time is more than 20 times the interval.

[0047] In another embodiment of the present invention, the operations performed by the image synthesis module include coordinate rotation, pixel translation, matrix summation, and normalization.

[0048] In another embodiment of the present invention, a computer-readable storage medium is provided having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the above-described method for generating a dual-axis-stabilized video based on dynamic fuzzy integral.

[0049] This invention provides a method and apparatus for generating line-of-sight stabilized video based on dynamic fuzzy integral. The method involves acquiring inertial pointing data from a servo system operating under an inertial stabilization model. This inertial pointing data includes the azimuth, pitch, and roll angles of the line of sight. The data update frequency is no less than 1 kHz, and the total data duration is no less than 20 seconds. The method sets the exposure time, frame rate, image resolution, and field of view parameters of the imaging sensor, and reads the grayscale value of the initial scene image as a reference frame. The acquired azimuth, pitch, and roll angle data are interpolated to generate a high-density pointing angle sequence, and the azimuth and pitch stabilization accuracy is calculated. Using the time interval of the interpolated pointing angle sequence as a unit, the reference frame is rotated in reverse according to the roll angle corresponding to each unit of time to generate an imaging slice. All imaging slices corresponding to a single frame exposure time are superimposed pixel-by-pixel, and the superposition result is normalized to obtain a single-frame image. Character information containing the field of view, stabilization accuracy, and imaging parameters is superimposed on each generated frame image, and all single-frame images are combined in chronological order to generate a video file with a line-of-sight stabilized effect.

[0050] In this embodiment of the invention, a video of the visual axis stabilization accuracy effect is generated based on pointing data and imaging sensor parameters, restoring intra-frame blurring, improving the intuitiveness of the stabilization accuracy effect display during the design phase, and reducing the servo system's dependence on the optoelectronic device image link during the physical testing phase.

[0051] The above are preferred embodiments of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for generating line-of-sight stabilized video based on dynamic fuzzy integral, characterized in that, The method includes: S101. Data Acquisition: During the simulation or actual test of the servo system, the inertial pointing data of the servo system under the inertial stability model is acquired. The inertial pointing data includes the azimuth angle, pitch angle and roll angle of the line of sight. The data update frequency is not less than 1kHz and the total data duration is not less than 20 seconds. S102, Parameter setting: Set the exposure time, frame rate, image resolution and field of view parameters of the imaging sensor, and read the grayscale value of the initial scene image as the reference frame; S103. Data processing: Based on the exposure time, interpolate the collected azimuth, pitch and roll angle data respectively to generate a high-density pointing angle sequence, and calculate the azimuth stability accuracy and pitch stability accuracy. S104. Generate image: Using the time interval of the interpolated pointing angle sequence as the unit, rotate the reference frame in reverse according to the roll angle corresponding to each unit time to generate an imaging slice; stack all imaging slices corresponding to a single frame exposure time pixel by pixel, and normalize the stacking result to obtain a single frame image. S105. Generate video: Superimpose character information containing field of view, stabilization accuracy and imaging parameters onto each frame image generated in step S104, and combine all single frame images in chronological order to generate a video file of the view stabilization effect.

2. The method for generating line-of-sight stabilized video based on dynamic fuzzy integral according to claim 1, characterized in that, The interpolation process employs cubic spline interpolation, piecewise linear interpolation, or radial basis function interpolation methods. The interpolated data time interval satisfies the following condition: the exposure time is more than 20 times the interpolated data time interval.

3. The method for generating line-of-sight stabilized video based on dynamic fuzzy integral according to claim 1, characterized in that, The inertial pointing data comes from the data output by the servo system simulation software, or from the data measured by the gyroscope or inertial measurement unit (IMU) in the physical optoelectronic device.

4. The method for generating line-of-sight stabilized video based on dynamic fuzzy integral according to claim 1, characterized in that, The generated image includes: Establish a timeline, align the interpolated angle data sequence with the imaging time, and determine the angle data sequence segment corresponding to the exposure time of each frame image; Based on the reference frame and the roll angle corresponding to each angle data time unit, the grayscale matrix of the imaging slice corresponding to that time unit is calculated through coordinate transformation. The cumulative grayscale matrix is ​​obtained by summing the grayscale matrices of the imaging slices corresponding to all time units within the exposure time of a frame of an image pixel by pixel. The values ​​of each element in the cumulative grayscale matrix are normalized to a preset image grayscale range to obtain the image data of the frame.

5. The method for generating line-of-sight stabilized video based on dynamic fuzzy integral as described in claim 1, characterized in that, The superimposed character information includes at least: horizontal field of view, vertical field of view, azimuth stabilization accuracy, pitch stabilization accuracy, frame rate, exposure time, and current frame number.

6. The method for generating line-of-sight stabilized video based on dynamic fuzzy integral as described in claim 1, characterized in that, A fixed reference marker is superimposed at the center of each generated image frame. The assessment result G of the subsystem health status assessment i The degradation degree is assessed by weight Z. i And the weight of subsystem performance evaluation.

7. A line-of-sight stabilized video generation device based on dynamic fuzzy integral, comprising a line-of-sight stabilized video generation method based on dynamic fuzzy integral as described in claims 1 to 6, characterized in that, The device includes: The data acquisition module is used to acquire inertial pointing data of the servo system under the inertial stability model during servo system simulation or actual measurement. The data includes azimuth angle, pitch angle and roll angle. The parameter configuration module is used to set the exposure time, frame rate, image resolution, and field of view parameters of the imaging sensor, and to load the initial scene reference frame image; The data processing module is used to interpolate the collected pointing angle data to generate a high-density sequence and calculate the azimuth and pitch stability accuracy. The image synthesis module is used to rotate the reference frame time-by-time according to the interpolated roll angle sequence to generate imaging slices, and to superimpose and normalize all slices within a single frame exposure time to synthesize a single frame image. The video generation module is used to overlay parameter characters onto the synthesized single-frame image and combine them in sequence to generate a video file.

8. The line-of-sight stabilized video generation device based on dynamic fuzzy integral according to claim 7, characterized in that, The data processing module executes a cubic spline interpolation algorithm, ensuring that the data time interval after interpolation is at least 20 times the exposure time.

9. The method for generating line-of-sight stabilized video based on dynamic fuzzy integral according to claim 7, characterized in that, The operations performed by the image synthesis module include coordinate rotation, pixel translation, matrix summation, and normalization.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the line-of-sight stabilization video generation method according to any one of claims 1 to 6.

Images