A blue-green laser cross-domain communication alignment method, device, system and equipment

Abstract

This invention relates to the field of optical communication, specifically to a method, apparatus, system, and device for cross-domain blue-green laser communication alignment. After a first blue-green laser communication device begins emitting a laser signal, the coordinates of the current scanning target point are acquired. Based on the coordinates, an alignment control system is controlled to scan the current scanning target point, obtaining a scanned image of the target point. The laser spot in the scanned image is detected; if a laser spot is detected, its intensity is determined. If the intensity of the laser spot exceeds a preset capture threshold, it is determined whether the intensity of a preset number of consecutive historical scanning target points exceeds the preset capture threshold. If the intensity of a preset number of consecutive historical scanning target points exceeds the preset capture threshold, the coordinates of the next scanning target point are determined to align and track the laser signal. This achieves accurate identification of the laser spot position and improves alignment accuracy.

Description

Technical Field

[0001] This invention relates to the field of optical communication, specifically to a blue-green laser cross-domain communication alignment method, apparatus, system, and device. Background Technology

[0002] With the increasing demand for integrated underwater and aerial communication, high-speed information transmission technologies across the water-air interface are gradually becoming an important research direction for marine information networks, unmanned system operations, and underwater monitoring and control. While traditional underwater acoustic communication has certain advantages in long-distance transmission, it is limited by factors such as low sound speed, narrow bandwidth, and large latency, making it difficult to meet the requirements for real-time performance and high data rates. Furthermore, radio waves attenuate significantly in water, enabling only short-range communication in extremely shallow water, and cannot support stable links between underwater equipment and aerial platforms.

[0003] Blue-green lasers (wavelengths of approximately 450nm-550nm) exhibit low absorption and scattering losses in water, while also occupying an optical "window" region in the atmosphere. Combining the characteristics of low underwater attenuation and high transmission efficiency in the air, they have become an important carrier for cross-water-air interface wireless communication. However, cross-domain communication using blue-green lasers still faces several challenges in engineering applications. Factors such as water surface fluctuations, bubble disturbances, water flow gradients, and platform attitude changes lead to difficulties in establishing optical communication links and poor stability. Furthermore, due to interference from the marine environment and mechanical vibrations, traditional beam alignment methods based on fixed scanning or manual adjustment suffer from slow response and low accuracy, making it difficult to meet real-time communication requirements. Summary of the Invention

[0004] The purpose of this invention is to provide a blue-green laser cross-domain communication alignment method, device, system and equipment, which solves the problems in the prior art.

[0005] This invention is achieved through the following technical solution:

[0006] In a first aspect, embodiments of the present invention provide a blue-green laser cross-domain communication alignment method, applied to a blue-green laser cross-domain communication alignment system. The system includes an underwater data acquisition node and a shore-based data aggregation node. The underwater data acquisition node includes a first blue-green laser communication device for emitting laser signals. The shore-based data aggregation node includes an alignment control system. The method includes:

[0007] After the first blue-green laser communication device starts emitting laser signals, the coordinates of the current scanning target point that is scanning the laser signals are obtained.

[0008] The alignment control system is controlled to scan the current target point based on the coordinate values, thereby obtaining a scanned image of the current target point;

[0009] The laser spot in the scanned image is detected. If the laser spot is detected, the light intensity of the laser spot is determined.

[0010] If the light intensity of the laser spot exceeds the preset capture threshold, then it is determined whether the light intensity of a preset number of historical scan target points that are consecutive to the current scan target point all exceed the preset capture threshold.

[0011] If the light intensity corresponding to a consecutive preset number of historical scan target points all exceeds the preset capture threshold, then the coordinates of the next scan target point of the current scan target point are determined to align and track the laser signal.

[0012] Preferably, obtaining the coordinates of the current scanning target point for scanning the laser signal includes:

[0013] If the current scan target point is the first scan target point, then the coordinate value of the current scan target point is the initial value;

[0014] If the current scanning target point is not the first scanning target point, then the coordinate value of the current scanning target point is obtained based on the coordinate values ​​and light intensity of the scanning target points in the two most recent scans of the current scanning target point.

[0015] Preferably, obtaining the coordinates of the current scanning target point based on the coordinates and light intensity of the target point in the two most recent scans includes:

[0016] The most recent light intensity increment is obtained based on the light intensity difference between the two previous scans of the current scan target point and a preset gain coefficient.

[0017] The current step size of the current scan target point is determined based on the most recent light intensity increment and the step size of the previous scan target point.

[0018] The current orientation angle of the current scanning target point is determined based on the preset cellular spiral scanning path and the orientation angle of the previous scanning target point.

[0019] The coordinates of the current scanning target point are obtained based on the current direction angle and the current step size.

[0020] Preferably, the step of detecting the laser spot in the scanned image, and determining the intensity of the laser spot if the laser spot is detected, includes:

[0021] The scanned image is subjected to median filtering and mean filtering to obtain a denoised image;

[0022] After converting the denoised image from the BGR color space to the HSV color space, threshold segmentation is performed on the denoised image based on a preset blue-green laser HSV threshold range to generate a binary mask image.

[0023] Contour extraction is performed on the binary mask image to obtain multiple closed contours;

[0024] The multiple closed contours are filtered according to the preset laser spot contour to determine whether a laser spot exists.

[0025] If a laser spot exists, determine the center coordinates of the laser spot;

[0026] The intensity of the laser spot is determined based on the center coordinates.

[0027] Preferably, the contour extraction of the binary mask image to obtain multiple closed contours includes:

[0028] The binary mask image is convolved to calculate the brightness gradient of each pixel in the horizontal and vertical directions.

[0029] Based on the brightness gradient of each pixel in the horizontal and vertical directions, a non-maximum suppression operation is performed on each pixel to retain the local maximum points in the gradient direction.

[0030] Based on the local maximum value of each pixel, each pixel is marked as a strong edge pixel, a weak edge pixel, or a non-edge pixel;

[0031] Weak edge pixels connected to strong edge pixels are identified as final edge pixels, thus obtaining the edge image;

[0032] The starting point is determined from each edge image. According to the preset connectivity rules, the edge pixels adjacent to each starting point are recursively searched and connected until the tracking returns to the starting point, resulting in multiple closed contours.

[0033] Preferably, the alignment control system includes a two-dimensional turntable, a turntable controller, and an industrial camera. The step of determining the coordinates of the next scanning target point of the current scanning target point to align and track the laser signal includes:

[0034] The current light intensity increment is obtained based on the difference between the light intensity of the current scan target point and the light intensity of the previous scan target point.

[0035] Adaptive state estimation is performed based on the light intensity increment and the preset state vector to obtain the estimated position of the laser spot at the current moment. The state vector includes the estimated position of the laser spot in the coordinate system of the industrial camera and the speed of the laser spot movement.

[0036] The position error is obtained by comparing the estimated position of the light spot with the desired position.

[0037] The position error is input into the PID controller, and the angle control quantity of the two-dimensional turntable in the azimuth and pitch directions is calculated through proportional, integral and derivative operations.

[0038] The two-dimensional turntable is controlled according to the angle control value to continuously align and track the laser signal until the obtained position error is less than the preset convergence threshold.

[0039] Preferably, the step of adaptively estimating the state based on the light intensity increment and a preset state vector to obtain the estimated position of the light spot at the current moment includes:

[0040] Based on the current light intensity increment, the current process noise covariance matrix at the current moment is obtained. With the current observation noise covariance matrix The current process noise covariance matrix With the current observation noise covariance matrix satisfy:

[0041] ;

[0042] in, , These are the initial state noise variance and observation noise variance, respectively. for Light intensity at any given moment for Light intensity at any given moment For normalized gain, , This is the noise adjustment factor;

[0043] The noise covariance matrix of the previous process at the previous time step. The previous state vector estimate Compared with the previous state vector estimate covariance matrix Input the prediction step of the unscented Kalman filter to obtain the mean of the current predicted state at the current time. With the mean of the current predicted state The current covariance matrix Wherein, the estimated value of the previous state vector Including the estimated position of the previous light spot in the coordinate system of the industrial camera at the previous moment. and the speed of the previous light spot ;

[0044] According to the preset observation function The noise covariance matrix of the previous observation To obtain the mean of the measurement prediction at the current moment. Measurement and prediction of covariance Cross covariance and Kalman gain ;

[0045] Average the current predicted state The current covariance matrix The Kalman gain The covariance of the measurement prediction and measurement predicted mean This yields the updated estimate of the state vector at the current time. With the updated estimate covariance matrix ;

[0046] From the updated estimate Extract the estimated position of the light spot in the camera coordinate system at the current moment. .

[0047] Secondly, embodiments of the present invention provide a blue-green laser cross-domain communication alignment device, applied to a blue-green laser cross-domain communication alignment system. The system includes an underwater data acquisition node and a shore-based data aggregation node. The underwater data acquisition node includes a first blue-green laser communication device for emitting laser signals. The shore-based data aggregation node includes an alignment control system. The device includes:

[0048] The acquisition module is used to acquire the coordinates of the current scanning target point that is scanning the laser signal after the first blue-green laser communication device starts to emit a laser signal;

[0049] The scanning module is used to control the alignment control system to scan the current scanning target point according to the coordinate value, so as to obtain a scanned image of the current scanning target point;

[0050] The detection module is used to detect laser spots in the scanned image. If a laser spot is detected, the intensity of the laser spot is determined.

[0051] The judgment module is used to determine whether the light intensity of a preset number of historical scanning target points that are consecutive to the current scanning target point exceeds the preset capture threshold if the light intensity of the laser spot exceeds the preset capture threshold.

[0052] The alignment and tracking module is used to determine the coordinates of the next scanning target point of the current scanning target point and perform alignment and tracking of the laser signal if the light intensity corresponding to a consecutive preset number of historical scanning target points exceeds the preset capture threshold.

[0053] Thirdly, embodiments of the present invention provide a blue-green laser cross-domain communication alignment system, comprising:

[0054] An underwater data acquisition node, the underwater data acquisition node including a first blue-green laser communication device, the first blue-green laser communication device being used to transmit laser signals;

[0055] The shore-based data aggregation node includes an alignment control system and a second blue-green laser communication device. The alignment control system includes a two-dimensional turntable, a turntable controller, and an industrial camera. The second blue-green laser communication device and the industrial camera are fixed to the two-dimensional turntable by clamps and are installed as a whole in the shore-side support structure.

[0056] The control terminal is connected to both the underwater data acquisition node and the onshore data aggregation node, and is used to execute the method of the first aspect.

[0057] Fourthly, embodiments of the present invention provide an electronic device, including: at least one processor, at least one memory, and computer program instructions stored in the memory, which, when executed by the processor, implement the method of the first aspect described above.

[0058] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0059] This invention achieves rapid acquisition and high-precision alignment of the laser link through precise angle control of a two-dimensional turntable and real-time feedback from an industrial camera. A spot detection algorithm based on HSV color space segmentation and OpenCV image processing is employed, enabling rapid extraction of the spot centroid under complex background light interference, improving spot positioning accuracy and providing reliable input for subsequent control stages. The proposed adaptive step-size cellular spiral scanning algorithm dynamically adjusts the step size according to the spot signal intensity and search progress, achieving efficient beam acquisition. To address spot jitter caused by underwater disturbances and platform attitude changes, a composite control algorithm combining an improved unscented Kalman filter based on adaptive noise adjustment and PID control is used to achieve robust estimation and dynamic compensation of the spot position, effectively improving link stability. This invention integrates the laser communication unit, two-dimensional turntable, turntable controller, and camera into a single, compact system that is easy to calibrate and adaptable to various optical communication platforms, enabling rapid deployment and field application. This invention enables rapid acquisition, high-precision alignment, and stable maintenance of cross-medium blue-green laser communication links. It is applicable to cross-domain communication scenarios such as unmanned underwater vehicles (UUVs), unmanned aerial vehicles (UAVs), and sea surface buoys. It features high bandwidth, low latency, and good anti-interference performance, and has significant engineering application value and promotion potential. Attached Figure Description

[0060] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort. In the drawings:

[0061] Figure 1 A flowchart illustrating the blue-green laser cross-domain communication alignment method provided by this invention;

[0062] Figure 2 This is a schematic diagram illustrating the center coordinate detection effect of the laser spot provided by the present invention.

[0063] Figure 3 This is a schematic diagram of alignment and tracking simulation provided by the present invention;

[0064] Figure 4 A schematic diagram of the blue-green laser cross-domain communication alignment device provided by the present invention;

[0065] Figure 5 This is a schematic diagram of the blue-green laser cross-domain communication alignment system provided by the present invention;

[0066] Figure 6 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0067] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0068] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0069] It should be noted that all actions involving the acquisition of signals, information, or data in this invention are carried out in compliance with the relevant data protection laws and regulations of the locality and with authorization from the owner of the relevant device.

[0070] Example 1

[0071] Please see Figure 1 This invention provides a blue-green laser cross-domain communication alignment method, applied to a blue-green laser cross-domain communication alignment system. The system includes an underwater data acquisition node and a shore-based data aggregation node. The underwater data acquisition node includes a first blue-green laser communication device for emitting laser signals. The shore-based data aggregation node includes an alignment control system. The method includes:

[0072] S1. After the first blue-green laser communication device starts emitting laser signals, the coordinates of the current scanning target point for scanning the laser signals are obtained;

[0073] Specifically, this step is the initial action of the scanning alignment process, and its function is to provide the receiving end with an initial spatial search reference before the communication link is established. The first blue-green laser communication device refers to the laser device deployed at the underwater node for emitting a communication beam; the laser signal emitted by it is the physical object being scanned and captured. The coordinates of the current scan target point are parameters used to describe the expected pointing position of the receiving end's line of sight, usually expressed in the form of azimuth and pitch angles. These coordinates are generated by the scanning control algorithm based on a preset search strategy. By first acquiring these coordinates, the system clarifies the specific spatial target of this scanning cycle, providing a basis for subsequent pointing control and image acquisition, thereby initiating a complete scan-detection cycle.

[0074] S2. Based on the coordinate values, control the alignment system to scan the current scanning target point to obtain a scanned image of the current scanning target point;

[0075] Specifically, this step transforms the spatial coordinates obtained in the previous step into concrete mechanical movements and data acquisition actions. The alignment control system is a hardware combination integrating a 2D turntable, controller, and imaging device. Its function is to execute precise spatial pointing and obtain visual feedback. Control is performed based on the coordinate values; that is, the turntable controller drives the 2D turntable to rotate, ensuring its line of sight is precisely aligned with the spatial orientation determined by S1. Subsequently, the imaging device fixed on the turntable acquires a frame of the scene image under this stable pointing state, thus obtaining the scanned image of the current scanning target point. This process realizes the conversion from digital commands to physical space exploration. The scanned image serves as direct visual evidence to determine whether a laser signal exists within the field of view, providing a data foundation for subsequent spot recognition.

[0076] S3. Detect the laser spot in the scanned image. If the laser spot is detected, determine the light intensity of the laser spot.

[0077] Specifically, this step involves feature extraction and preliminary interpretation of the acquired visual information. A laser spot is a bright spot formed by a laser beam on the target surface of an imaging sensor, representing the presence of a laser signal. Detecting it is a typical image recognition process, usually involving separating regions from complex background information that match the color, shape, and brightness characteristics of the laser. If such regions are successfully located using image processing methods, a laser spot is detected. Subsequently, the light intensity of the spot needs to be determined. Light intensity is a quantitative indicator representing the energy intensity of the spot region, generally obtained by calculating the sum or average of the pixel grayscale values ​​within the spot area. This step completes the transformation from the raw image to key decision data; the spot detection result and its light intensity value provide the core basis for subsequent judgments on whether to proceed to the fine alignment stage.

[0078] S4. If the light intensity of the laser spot exceeds the preset capture threshold, determine whether the light intensity of a preset number of historical scan target points that are consecutive to the current scan target point all exceed the preset capture threshold.

[0079] Specifically, this step introduces a historical signal-based verification mechanism to improve the reliability of capture judgment. The preset capture threshold is a pre-defined threshold value used to distinguish valid laser signals from background noise or transient interference. A light intensity exceeding this threshold indicates the presence of a potentially valid signal at the current scan target point. The preset quantity is an integer parameter, such as 3 or 5, which defines the number of scan cycles required for continuous verification. The logic of this step is that not only is the signal at the current scan target point valid, but also the signals from several consecutive historical scan points preceding it are valid. This design effectively filters out occasional false detections caused by water surface fluctuations, transient occlusion, or random noise, ensuring that the system only triggers subsequent, more precise, and time-consuming tracking processes when it continuously and stably receives a laser signal, thereby improving the robustness of system decisions.

[0080] S5. If the light intensity corresponding to a consecutive preset number of historical scan target points all exceeds the preset capture threshold, then determine the coordinate value of the next scan target point of the current scan target point to align and track the laser signal.

[0081] Specifically, this step is the decision-making and transition point for the system to switch from global search mode to local tracking mode. When the persistence condition described in S4 is met, the system determines that it has successfully captured and initially locked the laser signal, and the scanning phase ends. At this point, the coordinates of the next scanning target point are no longer generated by the scanning path planning algorithm covering the unknown area, but by a dedicated alignment tracking algorithm. Alignment tracking is a control mode designed to maintain precise alignment between the optical axis and the beam. Its core is to dynamically predict the movement trend of the light spot based on its real-time position and calculate the turntable adjustment command required to stabilize the light spot at the center of the field of view. Therefore, the coordinates of the next scanning target point determined here are essentially the first tracking control command calculated by the tracking algorithm based on the current and historical light spot position information. The execution of this step lays the foundation for establishing a stable, closed-loop communication link for the system.

[0082] In some embodiments, S1, obtaining the coordinates of the current scanning target point for scanning the laser signal, includes:

[0083] S11. If the current scanning target point is the first scanning target point, then the coordinate value of the current scanning target point is the initial value;

[0084] Specifically, this step defines the initialization conditions of the scan sequence, ensuring that the search process has a definite starting point. The current scan target point is an element in a sequence of spatial locations visited sequentially over time during the scan. The first scan target point specifically refers to the starting position of the sequence, and its coordinates are not dynamically calculated by the algorithm but are set to a predefined system initial value. This initial value is typically set based on the system's mechanical zero point, the expected signal occurrence area, or historical communication records, providing a fixed and known reference origin for the entire scan process. In this way, the system avoids coordinate uncertainty when starting the search task, making the scan path predictable and repeatable, laying the foundation for subsequent orderly and efficient coverage of the target search space.

[0085] S12. If the current scanning target point is not the first scanning target point, then the coordinate value of the current scanning target point is obtained based on the coordinate values ​​and light intensity of the scanning target points in the two most recent scans of the current scanning target point.

[0086] Specifically, this step constitutes the core feedback mechanism of dynamic programming of the scanning path, realizing the transformation from fixed-pattern search to adaptive search. After the system has completed at least one scan, the determination of the current scanning target point depends on historical scanning data, specifically the coordinates of the target point and the corresponding light intensity measurements from the previous two cycles (i.e., the two most recent scans). The coordinate values ​​provide sequential information about the spatial location, while the light intensity reflects the probability and strength of the signal at that location. The scanning control algorithm comprehensively analyzes this data. For example, if the light intensity shows an increasing trend, it may reduce the step size and perform a fine search in the same direction; if the light intensity is weak or there is no signal, it may increase the step size or change the direction to quickly explore new areas. Through this dynamic decision-making based on recent feedback, the system can autonomously converge the scanning trajectory towards areas with stronger signals, thereby improving acquisition efficiency, avoiding ineffective dwelling in areas without signals, and achieving optimized allocation of search resources and processes.

[0087] In some embodiments, S12, obtaining the coordinates of the current scanning target point based on the coordinates and light intensity of the target point in the two most recent scans, includes:

[0088] The most recent light intensity increment is obtained based on the light intensity difference between the two previous scans of the current scan target point and a preset gain coefficient.

[0089] The current step size of the current scan target point is determined based on the most recent light intensity increment and the step size of the previous scan target point.

[0090] The current orientation angle of the current scanning target point is determined based on the preset cellular spiral scanning path and the orientation angle of the previous scanning target point.

[0091] The coordinates of the current scanning target point are obtained based on the current direction angle and the current step size.

[0092] Specifically, the core of this step lies in dynamically adjusting the search behavior based on historical scan feedback. Specifically, the system first calculates the light intensity difference between the two previous scans of the current target point and multiplies this difference by a preset gain coefficient to obtain a quantified recent light intensity increment; this increment reflects the recent trend in signal strength. Subsequently, the system performs a comprehensive calculation with this recent light intensity increment and the scan step size set for the previous target point. The calculation rule ensures that when the light intensity increment indicates a stronger signal, the current step size is appropriately reduced relative to the previous scan step size to allow for a more refined local search in potential target areas; conversely, the step size is increased to quickly cover unknown areas. While determining the movement step size, the system sequentially selects the next azimuth angle of the previous target point's azimuth angle as the current azimuth angle, based on six fixed directions defined by the preset cellular spiral scan path. This ensures that the scan path continuously expands outward in a spiral shape, avoiding path repetition and achieving efficient coverage. Finally, combining the calculated current step size and current azimuth angle, the precise coordinates of the current target point in two-dimensional space can be calculated using a coordinate recursion formula. This method incorporates real-time signal quality feedback into the scanning path planning, enabling the search process to smoothly switch between coarse and fine search modes autonomously based on environmental feedback. This improves the convergence speed and reliability of the initial acquisition phase in complex and ever-changing cross-domain channel environments.

[0093] Furthermore, in cross-domain laser communication, due to platform micro-motion, refraction disturbance, pointing error, etc., the optical axes of the transmitting and receiving ends often deviate. In order to achieve beam capture within a limited time, a scan-detection-feedback mechanism is often adopted. This embodiment uses an adaptive step-size cellular spiral scanning algorithm, which has advantages such as high coverage efficiency, fast convergence, smooth path, and real-time controllability. Its specific processing flow is as follows:

[0094] ① Initialization parameters: Set the coordinates of the scan start point Scan step size Capture threshold .

[0095] ② Image acquisition / light intensity detection: Calculate the light intensity index of the acquired image. The light intensity is calculated using ROI integration, specifically:

[0096]

[0097] in, For pixel coordinates, For pixel grayscale values, Set it to a rectangular area within a certain range (e.g., 20x20 pixels).

[0098] ③ Calculate the light intensity increment: .in For normalized gain, For the first The light intensity captured in the second scan. For the first The light intensity captured in the second scan. This represents the increase in light intensity.

[0099] ④ Adaptive step size adjustment: The step size adjustment rule is as follows ,in The step size contraction factor, This is the sensitivity adjustment coefficient. For the first Step size of each scan No. The step size of each scan. It can be known that when... When it increases, Then the update step size Then it becomes smaller, enabling fine scanning. Conversely, when When decreasing, update step size If the value increases, a coarse search or rollback will be performed.

[0100] ⑤ Displacement Update: Calculate new coordinates Advance along the six directions of the cellular network. The recursive formula for the scan point location is:

[0101]

[0102] Among them, the direction angle Rotate sequentially in the six directions of the honeycomb, moving one step at a time. .

[0103] ⑥ Light intensity determination and capture determination: If continuous The second sampling satisfies If the target is successfully acquired, the alignment and tracking phase can begin. The threshold is set for capture. If capture is unsuccessful, return to step ② to continue scanning.

[0104] In some embodiments, S3 involves detecting a laser spot in the scanned image; if a laser spot is detected, the intensity of the laser spot is determined, including:

[0105] S31. Perform median filtering and mean filtering on the scanned image to obtain a denoised image;

[0106] Specifically, this step involves preprocessing the original scanned image to suppress noise interference and improve image quality, laying the foundation for subsequent spot recognition. The scanned image, acquired by an imaging sensor, typically contains random noise introduced by photoelectric conversion, water scattering, and ambient lighting. Median filtering, a non-linear filtering technique, replaces the gray value of each pixel in the image with the median of the gray values ​​of all pixels in its neighborhood. This process effectively removes salt-and-pepper noise while preserving the sharpness of image edges. Mean filtering, a linear filtering method, replaces the original pixel value with the average gray value of its neighborhood, thus smoothing the image and suppressing Gaussian noise. These two filtering methods are applied sequentially to the scanned image: median filtering first removes prominent noise points, followed by mean filtering to smooth the overall image, ultimately resulting in a denoised image with significantly reduced noise levels. This preprocessing enhances the signal-to-noise ratio in the image, allowing subsequent color segmentation and contour extraction operations to be performed on a clearer and more stable data basis, improving the robustness of the entire detection process.

[0107] S32. After converting the denoised image from the BGR color space to the HSV color space, threshold segmentation is performed on the denoised image based on the preset blue-green laser HSV threshold range to generate a binary mask image.

[0108] Specifically, the core of this step is to use color information to separate potential laser spot regions from a complex background. BGR is the color space typically output by imaging devices; its components (blue, green, and red) are closely related to the physical characteristics of the sensor but are sensitive to changes in light intensity. The HSV color space, on the other hand, decouples color information into three independent components: hue, saturation, and lightness. Hue directly corresponds to the type of color and has better invariance to changes in light intensity. Converting the denoised image from BGR to HSV space is to isolate color recognition from the influence of brightness. The preset blue-green laser HSV threshold range is a set of numerical intervals pre-calibrated based on the typical spectral characteristics of blue-green lasers, used to define the possible color range of the laser spot in the HSV space. Thresholding is performed based on this threshold range, that is, the HSV value of each pixel in the image is judged. Pixels falling within the preset range are classified as targets (usually set to white), and the remaining pixels are classified as background (usually set to black), thus generating a binary mask image. In this image, the white area represents the candidate spot area that matches the color characteristics of blue-green laser. This process effectively filters out background interference that differs significantly from the laser color, achieving preliminary coarse localization of the target.

[0109] S33. Extract contours from the binary mask image to obtain multiple closed contours;

[0110] Specifically, this step aims to identify and locate the complete geometric boundaries of all candidate targets from a binary mask image. Although the binary mask image separates the foreground and background, foreground targets typically exist as connected white regions. Contour extraction aims to find the precise boundaries of these connected regions. The contour extraction algorithm uses edge tracking technology to connect adjacent foreground pixels in the binary image, forming closed curves representing region boundaries—i.e., closed contours. Each closed contour corresponds to an independent, continuous candidate target region in the image. This process transforms the image from a pixel-level region representation to a higher-level geometric contour representation, with each contour containing a series of ordered boundary point coordinates. Outputting multiple closed contours means the system may have detected several candidate regions that match color features, providing processing objects for subsequent fine-tuning based on shape features.

[0111] S34. Filter the multiple closed contours according to the preset laser spot contour to determine whether there is a laser spot.

[0112] Specifically, this step involves a secondary screening of candidate targets based on shape features to eliminate false targets that match the color but not the shape, ultimately confirming the existence of the laser spot. The preset laser spot contour features are a set of predefined geometric criteria based on laser beam characteristics (such as approximate circularity and ellipticity range) and imaging system characteristics, such as the area range of the contour, the ratio of perimeter to area (circularity), and the aspect ratio of the circumscribed rectangle. The system calculates these geometric features for each closed contour obtained in step S33 and compares them with the preset criteria. Only contours whose geometric features also match the expected contours are determined to be true laser spot contours. By traversing and screening multiple closed contours, erroneous detections caused by water surface reflection, floating objects, or other interference sources with similar colors can be effectively eliminated. If, after screening, at least one contour satisfies all shape criteria, it is determined that a laser spot exists; otherwise, it is determined that no valid laser spot exists in the current field of view. This step, through dual verification of color and shape, greatly improves the accuracy of spot detection.

[0113] S35. If a laser spot exists, determine the center coordinates of the laser spot;

[0114] Specifically, this step involves precisely locating the light spot after confirming its existence. The center coordinates are two-dimensional coordinates representing the precise position of the light spot in the image pixel coordinate system, typically in pixels. For the laser spot outline confirmed after S34 filtering, the system uses geometric calculation methods to determine its center. Common methods include calculating the centroid of all pixels in the outline, or calculating the center of its smallest bounding rectangle. For approximately circular light spots, calculating the center of its smallest bounding circle is a common and effective method. Determining these center coordinates transforms the visual existence of the light spot into quantifiable and precise spatial location information. This location information is a crucial input necessary for subsequent calculations of light intensity and alignment tracking control; its positioning accuracy directly affects the performance of the entire alignment system.

[0115] S36. Determine the light intensity of the laser spot based on the center coordinates.

[0116] Specifically, this step quantifies the energy intensity of the located laser spot. Light intensity is a physical quantity characterizing the energy level of a spot region. In image processing, it is usually approximated by calculating the sum or average of the gray values ​​of pixels within the local area where the spot is located. In practice, the system defines a fixed-size rectangular region as the Region of Interest (ROI) based on the center coordinates of the spot determined in S35. This region should be large enough to cover the main energy distribution of the spot. Then, the sum of the gray values ​​of all pixels within this ROI is calculated, and this sum is used as the light intensity value of the current laser spot. The light intensity obtained in this way comprehensively reflects the brightness and size of the spot and is a more stable energy metric than a single pixel gray value. This light intensity value serves as a crucial feedback signal, used in scanning capture judgment and subsequent adaptive algorithms in tracking control. For example, it is used to determine if the signal is strong enough to trigger a state switch, or to adjust filter parameters to cope with light intensity fluctuations.

[0117] Furthermore, one of the key technologies for realizing the capture, aiming, and tracking system is the rapid and accurate detection of the laser spot and the localization of its center position. This embodiment employs a spot center detection algorithm based on the fusion of HSV (Hue, Saturation, Value) segmentation and OpenCV image processing to detect and locate the laser spot. The main steps of this algorithm include:

[0118] ① Read video frames and remove noise: Access the camera, read video frames, perform image preprocessing, and remove image noise through median filtering and mean filtering.

[0119] Median filtering replaces each pixel value with the median of its neighborhood, effectively removing salt-and-pepper noise.

[0120] ;

[0121] in, For the original image in coordinates The pixel grayscale value at that location; This represents the relative coordinate offset within the current pixel's neighborhood window; The neighborhood radius determines the size of the filtering window; This indicates that the median of the grayscale values ​​of all pixels in the set is retrieved. For the image after median filtering The pixel value at that location.

[0122] Mean filtering replaces each pixel value with the mean value of its neighborhood, smoothing the image and reducing noise.

[0123] ;

[0124] in, It represents the grayscale value of a pixel within the neighborhood. The radius of the filtering window; The image after mean filtering The pixel value at that location.

[0125] ②HSV Conversion: Convert the image from the BGR color space to the HSV color space for easier color segmentation. Define the HSV color ranges for the blue and green laser points and create a mask where only the areas of laser points that meet the criteria are white, and the rest are black. The RGB to HSV conversion formula is as follows:

[0126] ;

[0127] in, , They represent the blue, green, and red components, respectively. These represent Hue, Saturation, and Value, respectively. By taking the maximum value of the three color channel components R (red), G (green), and B (blue), the HSV color space has good robustness to changes in illumination and is suitable for processing color information in underwater environments.

[0128] ③ Thresholding Segmentation: Laser spot segmentation is performed based on a predefined HSV threshold range. The image is divided into foreground and background by comparing image pixel values ​​with the predefined threshold range. The thresholding segmentation formula is as follows:

[0129] ;

[0130] in, , These are the lower and upper limits of the set Hue value threshold, respectively. For blue-green lasers, the Hue value range is usually set to (35, 130). These are the pixel values ​​after thresholding.

[0131] In some embodiments, S33, contour extraction is performed on the binary mask image to obtain multiple closed contours, including:

[0132] The binary mask image is convolved to calculate the brightness gradient of each pixel in the horizontal and vertical directions.

[0133] Based on the brightness gradient of each pixel in the horizontal and vertical directions, a non-maximum suppression operation is performed on each pixel to retain the local maximum points in the gradient direction.

[0134] Based on the local maximum value of each pixel, each pixel is marked as a strong edge pixel, a weak edge pixel, or a non-edge pixel;

[0135] Weak edge pixels connected to strong edge pixels are identified as final edge pixels, thus obtaining the edge image;

[0136] The starting point is determined from each edge image. According to the preset connectivity rules, the edge pixels adjacent to each starting point are recursively searched and connected until the tracking returns to the starting point, resulting in multiple closed contours.

[0137] Furthermore, based on threshold segmentation ③, this embodiment performs the following processing:

[0138] ④ Canny Edge Detection: In OpenCV's edge detection algorithm, a Gaussian filter is first used to smooth the grayscale image, and the Sobel operator is used to calculate the gradient values ​​of the grayscale image in the horizontal and vertical directions. Then, non-maximum suppression is performed. For each pixel, the gradient values ​​of its two neighboring pixels in the gradient direction are compared, and only pixels with the largest gradient values ​​in the gradient direction are retained, which helps to eliminate the blurring effect on the edges.

[0139]

[0140] in, Represents pixel coordinates. , These represent the gradient components of the image in the horizontal and vertical directions, respectively. This represents the gradient intensity at a pixel. This indicates the gradient direction of the pixel. Then, accurate edge detection is achieved through a double-threshold hysteresis connection, which divides pixels into strong edges, weak edges, and non-edges, and connects weak edges with their surrounding strong edges to form complete edges.

[0141] ⑤ Contour Extraction: The contour information of the light spot is extracted by drawing the edge detection results. Specifically, the `findContours` function in OpenCV is used to find the contour (i.e., the contour of the laser point) in the mask. The `findContours` function divides the foreground pixels in a binary image into several closed Jordan curves with 4 or 8 connectivity and records their nested tree. The task of light spot center detection and localization requires obtaining the precise boundary and shape features of the light spot. The light spot in the image represents the outer contour to be extracted. Using this function can avoid the problem of multi-layered nested connected components caused by the halo effect of light spots in water.

[0142] ⑥ Laser spot center positioning: Calculate the center coordinates of the laser spot using the contour geometric features obtained in step ⑤, such as... Figure 2 The image shows a schematic diagram illustrating the detection effect of the laser spot's center coordinates. First, a circular similarity metric is constructed based on the contour geometric features:

[0143] ;

[0144] in, The similarity between the outline and the circle; , The area and perimeter are calculated using OpenCV's cv2.contourArea and cv2.arcLength functions, respectively.

[0145] Since the light spot will have some distortion, after filtering out the contours that are too different from the circle, the cv2.minEnclosingCircle function is used to calculate the position of the center point of the light spot. This function is mainly used to calculate the minimum circumcircle of the contour. Its advantage is that the calculation speed is fast. The obtained light intensity centroid Center is the center coordinate of the laser point, which further improves the real-time performance and accuracy of the algorithm.

[0146] In some embodiments, the alignment control system includes a two-dimensional turntable, a turntable controller, and an industrial camera. Step S5 involves determining the coordinates of the next scanning target point from the current scanning target point to align and track the laser signal, including:

[0147] S51. Based on the difference between the light intensity corresponding to the current scanning target point and the light intensity of the previous scanning target point, obtain the current light intensity increment;

[0148] S52. Based on the light intensity increment and the preset state vector, perform adaptive state estimation to obtain the estimated position of the laser spot at the current moment. The state vector includes the estimated position of the laser spot in the coordinate system of the industrial camera and the speed of the laser spot movement.

[0149] S53. The position error is obtained based on the difference between the estimated position and the desired position of the light spot;

[0150] S54. Input the position error into the PID controller, and calculate the angle control quantity of the two-dimensional turntable in the azimuth and pitch directions through proportional, integral and derivative operations.

[0151] S55. Control the two-dimensional turntable to continuously align and track the laser signal according to the angle control amount until the obtained position error is less than the preset convergence threshold.

[0152] In some implementations, S52, adaptive state estimation is performed based on the light intensity increment and a preset state vector to obtain the estimated position of the light spot at the current moment, including:

[0153] Based on the current light intensity increment, the current process noise covariance matrix at the current moment is obtained. With the current observation noise covariance matrix The current process noise covariance matrix With the current observation noise covariance matrix satisfy:

[0154] ;

[0155] in, , These are the initial state noise variance and observation noise variance, respectively. for Light intensity at any given moment for Light intensity at any given moment For normalized gain, , This is the noise adjustment factor;

[0156] The noise covariance matrix of the previous process at the previous time step. The previous state vector estimate Compared with the previous state vector estimate covariance matrix Input the prediction step of the unscented Kalman filter to obtain the mean of the current predicted state at the current time. With the mean of the current predicted state The current covariance matrix Wherein, the estimated value of the previous state vector Including the estimated position of the previous light spot in the coordinate system of the industrial camera at the previous moment. and the speed of the previous light spot ;

[0157] According to the preset observation function The noise covariance matrix of the previous observation To obtain the mean of the measurement prediction at the current moment. Measurement and prediction of covariance Cross covariance and Kalman gain ;

[0158] Average the current predicted state The current covariance matrix The Kalman gain The covariance of the measurement prediction and measurement predicted mean This yields the updated estimate of the state vector at the current time. With the updated estimate covariance matrix ;

[0159] From the updated estimate Extract the estimated position of the light spot in the camera coordinate system at the current moment. .

[0160] Furthermore, the alignment and tracking algorithm in this embodiment is a composite control algorithm consisting of two parts: an improved unscented Kalman filter based on adaptive noise adjustment (ANA-UKF) and PID control. Spot measurement noise is often non-Gaussian, especially in underwater environments (scattering, bubbles, planktonic objects). ANA-UKF is used to address the problem of excessive spot positioning errors caused by underwater factors such as water flow disturbances, resulting in spot image distortion and jitter. PID control is used for rapid response of the gimbal control, thereby achieving the establishment and stable maintenance of the laser link in a short time.

[0161] To track underwater light spots, the state vector is first established:

[0162] ;

[0163] in, Let be the position of the light spot in the camera coordinate system. It is the speed of the light spot movement, used for prediction.

[0164] Establish the state equations:

[0165] ;

[0166] in, For spot prediction models, Process noise represents underwater disturbances (refraction, scattering, etc.). The sampling time interval, for The variance of state noise at any given time.

[0167] Establish the measurement equation:

[0168] ;

[0169] in, The observation function maps the state to camera pixel coordinates. To observe the noise, for The variance of observation noise at any given time.

[0170] Figure 3 The simulation diagram for alignment and tracking is shown below. The entire alignment and tracking process is as follows:

[0171] ① Obtain the current image frame and calculate the light intensity: Obtain the light intensity using the ROI integration calculation method described above. .

[0172] ②UKF Sigma point generation: For dimensional state vector Generate a set of Sigma points (sampling points) using the following formula:

[0173] ;

[0174] in, for The covariance matrix at time t, parameters It is a scaling parameter used to reduce the overall prediction error. These are the tuning parameters for UKF.

[0175] ③Sigma point propagation: calculation One-step prediction of a Sigma point set: , for The i-th Sigma point at time i;

[0176] Predicted state mean: ;

[0177] Predicting covariance: ;

[0178] in, for Always The mean of the predicted state at time t, Weighted by mean, For covariance weights, for The variance of state noise at any given time.

[0179] ④ Measurement and prediction: using observation functions The mean of the measurement predictions is obtained: ;

[0180] Covariance of measurement predictions: ;

[0181] Cross covariance: ;

[0182] Kalman gain: ;

[0183] in, for The variance of observation noise at time step, , , , They are the predicted ones The time-measured mean, the covariance of the measurement prediction, the cross covariance, and the Kalman gain.

[0184] ⑤ Adaptive noise adjustment: To address the issues of large fluctuations in underwater light spot signals and non-Gaussian noise, adaptive covariance adjustment is introduced, updating the noise level based on light intensity fluctuations. , :

[0185] ;

[0186] in , These are the initial state noise variance and observation noise variance, respectively. for Integral light intensity at ROI at any given time for Integral light intensity at ROI at any given time For normalized gain, , The noise adjustment coefficient is obtained. , In the next moment It is used for state prediction and measurement prediction, thereby forming an adaptive closed-loop update mechanism for time-varying noise covariance.

[0187] ⑥ State Update: Calculate the state update Update with covariance:

[0188] ;

[0189] in, for Always The mean of the predicted state at time t, For Kalman gain, for The measured value at time, for Always The measured predicted mean at time, for Always The predicted state covariance at time t. for Always The covariance of the time measurement prediction.

[0190] ⑦ PID Control: The PID controller uses the UKF output to estimate the spot position error.

[0191] ;

[0192] in, The desired position of the light spot on the camera's imaging plane (e.g., the center of the field of view). For UKF The estimation of the spot position at time step (i.e., the UKF state vector estimation obtained in step 6) In ), The position error of the light spot is the input to the PID controller.

[0193] PID output angle command:

[0194] ;

[0195] Among them, Indicates proportional gain. Indicates integral gain. Represents differential gain. , After PID calculation, the gimbal is in , Directional control quantity.

[0196] ⑧ Drive the two-dimensional turntable to rotate and adjust the optical axis.

[0197] ⑨ In the next frame loop, continue iterative processing until the 2-norm of the spot center deviation converges, i.e., it is less than the set threshold. .

[0198] Example 2

[0199] Please see Figure 4This invention provides a blue-green laser cross-domain communication alignment device, applied to a blue-green laser cross-domain communication alignment system. The system includes an underwater data acquisition node and a shore-based data aggregation node. The underwater data acquisition node includes a first blue-green laser communication device for emitting laser signals. The shore-based data aggregation node includes an alignment control system. The device includes:

[0200] The acquisition module 401 is used to acquire the coordinates of the current scanning target point that is scanning the laser signal after the first blue-green laser communication device starts to emit a laser signal;

[0201] The scanning module 402 is used to control the alignment control system to scan the current scanning target point according to the coordinate value, so as to obtain a scanned image of the current scanning target point;

[0202] The detection module 403 is used to detect the laser spot in the scanned image. If the laser spot is detected, the intensity of the laser spot is determined.

[0203] The judgment module 404 is used to determine whether the light intensity of a preset number of historical scanning target points that are consecutive to the current scanning target point exceeds the preset capture threshold if the light intensity of the laser spot exceeds the preset capture threshold.

[0204] The alignment and tracking module 405 is used to align and track the laser signal if the light intensity corresponding to a consecutive preset number of historical scan target points exceeds the preset capture threshold, and then determine the coordinate value of the next scan target point of the current scan target point.

[0205] It should be noted that each module and unit in the blue-green laser cross-domain communication alignment device in this embodiment corresponds one-to-one with each step in the blue-green laser cross-domain communication alignment method in the aforementioned embodiment. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned blue-green laser cross-domain communication alignment method, and will not be repeated here.

[0206] Example 3

[0207] Please see Figure 5 This invention provides a blue-green laser cross-domain communication alignment system, comprising:

[0208] An underwater data acquisition node, the underwater data acquisition node including a first blue-green laser communication device, the first blue-green laser communication device being used to transmit laser signals;

[0209] The shore-based data aggregation node includes an alignment control system and a second blue-green laser communication device. The alignment control system includes a two-dimensional turntable, a turntable controller, and an industrial camera. The second blue-green laser communication device and the industrial camera are fixed to the two-dimensional turntable by clamps and are installed as a whole in the shore-side support structure.

[0210] The control terminal is connected to the underwater data acquisition node and the onshore data aggregation node, and is used to execute the method described in Embodiment 1.

[0211] Example 4

[0212] Please see Figure 6 This embodiment provides an electronic device, including at least one processor 601 and a memory 602. Optionally, the device further includes a communication component 603. The processor 601, memory 602, and communication component 603 are connected via a bus 604.

[0213] In a specific implementation, at least one processor 601 executes computer execution instructions stored in memory 602, causing at least one processor 601 to perform the above-described method.

[0214] The specific implementation process of processor 601 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0215] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0216] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0217] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0218] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0219] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0220] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0221] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0222] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0223] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0224] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0225] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0226] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0227] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A cross-domain communication alignment method using blue-green lasers, characterized in that, An alignment system for cross-domain communication using blue-green lasers is provided. The system includes an underwater data acquisition node and a shore-based data aggregation node. The underwater data acquisition node includes a first blue-green laser communication device for emitting laser signals. The shore-based data aggregation node includes an alignment control system. The method includes: After the first blue-green laser communication device starts emitting laser signals, the coordinates of the current scanning target point that is scanning the laser signals are obtained. The alignment control system is controlled to scan the current target point based on the coordinate values, thereby obtaining a scanned image of the current target point; The laser spot in the scanned image is detected. If the laser spot is detected, the light intensity of the laser spot is determined. If the light intensity of the laser spot exceeds the preset capture threshold, then it is determined whether the light intensity of a preset number of historical scan target points that are consecutive to the current scan target point all exceed the preset capture threshold. If the light intensity corresponding to a consecutive preset number of historical scan target points all exceeds the preset capture threshold, then the coordinate value of the next scan target point of the current scan target point is determined to align and track the laser signal. The alignment control system includes a two-dimensional turntable, a turntable controller, and an industrial camera. The step of determining the coordinates of the next scanning target point from the current scanning target point to align and track the laser signal includes: The current light intensity increment is obtained based on the difference between the light intensity of the current scan target point and the light intensity of the previous scan target point. Adaptive state estimation is performed based on the light intensity increment and the state vector of the previous moment to obtain the estimated position of the laser spot at the current moment. The state vector includes the estimated position of the laser spot in the coordinate system of the industrial camera and the speed of the laser spot movement. The position error is obtained by comparing the estimated position of the light spot with the desired position. The position error is input into the PID controller, and the angle control quantity of the two-dimensional turntable in the azimuth and pitch directions is calculated through proportional, integral and derivative operations. The two-dimensional turntable is controlled according to the angle control value to continuously align and track the laser signal until the obtained position error is less than the preset convergence threshold.

2. The method according to claim 1, characterized in that, The step of obtaining the coordinates of the current scanning target point for scanning the laser signal includes: If the current scan target point is the first scan target point, then the coordinate value of the current scan target point is the initial value; If the current scanning target point is not the first scanning target point, then the coordinate value of the current scanning target point is obtained based on the coordinate values ​​and light intensity of the scanning target points in the two most recent scans of the current scanning target point.

3. The method according to claim 2, characterized in that, The step of obtaining the coordinates of the current scanning target point based on the coordinates and light intensity of the target point in the two most recent scans includes: The most recent light intensity increment is obtained based on the light intensity difference between the two previous scans of the current scan target point and a preset gain coefficient. The current step size of the current scan target point is determined based on the most recent light intensity increment and the step size of the previous scan target point. The current orientation angle of the current scanning target point is determined based on the preset cellular spiral scanning path and the orientation angle of the previous scanning target point. The coordinates of the current scanning target point are obtained based on the current direction angle and the current step size.

4. The method according to claim 1, characterized in that, The step of detecting the laser spot in the scanned image, and determining the intensity of the laser spot if a laser spot is detected, includes: The scanned image is subjected to median filtering and mean filtering to obtain a denoised image; After converting the denoised image from the BGR color space to the HSV color space, threshold segmentation is performed on the denoised image based on a preset blue-green laser HSV threshold range to generate a binary mask image. Contour extraction is performed on the binary mask image to obtain multiple closed contours; The multiple closed contours are filtered according to the preset laser spot contour to determine whether a laser spot exists. If a laser spot exists, determine the center coordinates of the laser spot; The intensity of the laser spot is determined based on the center coordinates.

5. The method according to claim 4, characterized in that, The contour extraction of the binary mask image yields multiple closed contours, including: The binary mask image is convolved to calculate the brightness gradient of each pixel in the horizontal and vertical directions. Based on the brightness gradient of each pixel in the horizontal and vertical directions, a non-maximum suppression operation is performed on each pixel to retain the local maximum points in the gradient direction. Based on the local maximum value of each pixel, each pixel is marked as a strong edge pixel, a weak edge pixel, or a non-edge pixel; Weak edge pixels connected to strong edge pixels are identified as final edge pixels, thus obtaining the edge image; The starting point is determined from each edge image. According to the preset connectivity rules, the edge pixels adjacent to each starting point are recursively searched and connected until the tracking returns to the starting point, resulting in multiple closed contours.

6. The method according to claim 1, characterized in that, The step of adaptively estimating the state based on the light intensity increment and a preset state vector to obtain the estimated position of the light spot at the current moment includes: Based on the current light intensity increment, the current process noise covariance matrix at the current moment is obtained. With the current observation noise covariance matrix The current process noise covariance matrix With the current observation noise covariance matrix satisfy: ； in, , These are the initial state noise variance and observation noise variance, respectively. for Light intensity at any given moment for Light intensity at any given moment For normalized gain, , This is the noise adjustment factor; The noise covariance matrix of the previous process at the previous time step. The previous state vector estimate Compared with the previous state vector estimate covariance matrix Input the prediction step of the unscented Kalman filter to obtain the mean of the current predicted state at the current time. With the mean of the current predicted state The current covariance matrix Wherein, the estimated value of the previous state vector Including the estimated position of the previous light spot in the coordinate system of the industrial camera at the previous moment. and the speed of the previous light spot ; According to the preset observation function The noise covariance matrix of the previous observation To obtain the mean of the measurement prediction at the current moment. Measurement and prediction of covariance Cross covariance and Kalman gain ; Based on the current predicted state average The current covariance matrix The Kalman gain The covariance of the measurement prediction and measurement predicted mean This yields the updated estimate of the state vector at the current time. With the updated estimate covariance matrix ; From the updated estimate Extract the estimated position of the light spot in the camera coordinate system at the current moment. .

7. A blue-green laser cross-domain communication alignment device, characterized in that, An alignment system for cross-domain communication using blue-green lasers is provided. The system includes an underwater data acquisition node and a shore-based data aggregation node. The underwater data acquisition node includes a first blue-green laser communication device for emitting laser signals. The shore-based data aggregation node includes an alignment control system. The device comprises: The acquisition module is used to acquire the coordinates of the current scanning target point that is scanning the laser signal after the first blue-green laser communication device starts to emit a laser signal; The scanning module is used to control the alignment control system to scan the current scanning target point according to the coordinate value, so as to obtain a scanned image of the current scanning target point; The detection module is used to detect laser spots in the scanned image. If a laser spot is detected, the intensity of the laser spot is determined. The judgment module is used to determine whether the light intensity of a preset number of historical scanning target points that are consecutive to the current scanning target point exceeds the preset capture threshold if the light intensity of the laser spot exceeds the preset capture threshold. The alignment and tracking module is used to align and track the laser signal if the light intensity corresponding to a preset number of consecutive historical scan target points exceeds the preset capture threshold, and then determine the coordinate value of the next scan target point of the current scan target point. The alignment and tracking module is also used for: The current light intensity increment is obtained based on the difference between the light intensity of the current scan target point and the light intensity of the previous scan target point. Adaptive state estimation is performed based on the light intensity increment and the state vector of the previous moment to obtain the estimated position of the laser spot at the current moment. The state vector includes the estimated position of the laser spot in the coordinate system of the industrial camera and the speed of the laser spot movement. The position error is obtained by comparing the estimated position of the light spot with the desired position. The position error is input into the PID controller, and the angle control quantity of the two-dimensional turntable in the azimuth and pitch directions is calculated through proportional, integral and derivative operations. The two-dimensional turntable is controlled according to the angle control value to continuously align and track the laser signal until the obtained position error is less than the preset convergence threshold.

8. A blue-green laser cross-domain communication alignment system, characterized in that, include: An underwater data acquisition node, the underwater data acquisition node including a first blue-green laser communication device, the first blue-green laser communication device being used to transmit laser signals; The shore-based data aggregation node includes an alignment control system and a second blue-green laser communication device. The alignment control system includes a two-dimensional turntable, a turntable controller, and an industrial camera. The second blue-green laser communication device and the industrial camera are fixed to the two-dimensional turntable by clamps and are installed as a whole in the shore-side support structure. The control terminal is connected to the underwater data acquisition node and the onshore data aggregation node, respectively, and is used to execute the method described in any one of claims 1-6.

9. An electronic device, characterized in that, include: At least one processor, at least one memory, and computer program instructions stored in the memory, which, when executed by the processor, implement the method as described in any one of claims 1-6.

Images