An underwater three-dimensional scanning imaging system based on a streak tube lidar

Abstract

The present invention relates to the field of surveying and mapping technology, and discloses an underwater three-dimensional scanning imaging system based on a streak tube lidar, which includes an environmental perception and initialization module. According to the known environmental conditions, it calculates and determines the initial refractive index, sets the working parameters of the lidar, including the wavelength and pulse frequency, and configures an automatic calibration function, setting the calibration period and trigger conditions; a laser emission control module that emits laser pulses at a predetermined time interval and records the exact time of each emission. The present invention can effectively address the problem of light attenuation in water and the possible small offsets of the internal components of the system after long-term operation. Even in a complex and variable underwater environment, it can maintain high-precision distance measurement and three-dimensional modeling capabilities. Through the automatic calibration and dynamic adjustment mechanism, it ensures the consistency and reliability of the system during long-term operation.

Description

Technical Field

[0001] The present invention relates to the field of surveying and mapping, and more specifically, it relates to an underwater three-dimensional scanning imaging system based on a streak tube lidar. Background Art

[0002] With the increasing demand for ocean exploration and resource development, the precise surveying and monitoring of the underwater environment have become a key area in scientific research, environmental protection, and industrial applications. Although traditional sonar technology is widely used in underwater detection, its resolution is limited and it is difficult to obtain high-precision three-dimensional images. As an emerging remote sensing technology, lidar has achieved remarkable success in land surveying due to its high resolution and fast data acquisition ability.

[0003] Introducing lidar technology into the underwater environment faces many challenges, especially the attenuation of light in water. After long-term operation, there may be slight offsets in the internal components of the lidar system, resulting in inaccurate measurement results. Summary of the Invention

[0004] The present invention provides an underwater three-dimensional scanning imaging system based on a streak tube lidar to solve the technical problems in the related art.

[0005] The present invention provides an underwater three-dimensional scanning imaging system based on a streak tube lidar, including:

[0006] Environmental perception and initialization module: Determine the initial refractive index according to environmental conditions, set the wavelength and pulse frequency of the lidar at the same time, and configure the calibration period and trigger conditions of the automatic calibration function;

[0007] Laser emission control module: Emit laser pulses at a predetermined pulse frequency and record their emission times;

[0008] Reflected signal receiving module: Receive the reflected signals from the surface of the target object and record their arrival times;

[0009] Original distance calculation module: Calculate the original distance estimate value based on the emission time, arrival time, and the speed of light;

[0010] Refraction correction module: Correct the distance deviation caused by refraction. For the interface from air to water, determine the incident angle and refraction angle, and calculate the actual ranging value;

[0011] Multipath reflection suppression module: Evaluate whether each point cloud data point is generated by multipath reflection, remove the false echoes caused by multipath reflection, and retain the true reflected signals;

[0012] Sensor Drift Correction Module: Regularly check the status of the sensor, compare the current readings with the reference benchmark, and if the deviation exceeds the preset threshold, start the correction program;

[0013] Data Integration and Optimization Module: Combine the corrected and purified data points to form the final 3D point cloud image;

[0014] User Interface and Output Module: Provide an operation interface, display the finally generated 3D map or model, and support export and sharing functions.

[0015] Furthermore, the initial refractive index in the environmental perception and initialization module is obtained by using a CTD probe to measure the temperature , salinity and pressure ;

[0016] Based on the obtained environmental conditions, use an empirical formula to determine the initial refractive index:

[0017] ;

[0018] where is the refractive index under reference conditions, represents the laser wavelength.

[0019] Furthermore, in the refraction correction module, the incident angle is determined according to the installation angle of the lidar system and the relative position of the target object;

[0020] The angle of the incident angle is calculated through the known sensor attitude information and target position;

[0021] ;

[0022] where and are the distance differences in the vertical and horizontal directions respectively;

[0023] Use Snell's law to calculate the refraction angle of light when it enters water from air ;

[0024] ;

[0025] Solve this equation to obtain :

[0026] ;

[0027] where is the refractive index of light in air, is the refractive index of light in water, obtained from the environmental perception and initialization module.

[0028] Further, calculate the actual ranging value The specific steps are as follows:

[0029] Distance traveled in air:

[0030] ;

[0031] Distance traveled in water:

[0032] ;

[0033] Actual ranging value Total distance considering refraction effect:

[0034] ;

[0035] Where represents the actual ranging value considering refraction effect, represents the distance traveled in air, represents the distance traveled in water.

[0036] Further, the processing steps of the multipath reflection suppression module are as follows:

[0037] Preliminary screening: Filter out data points that obviously do not conform to physical laws based on a time window;

[0038] Feature extraction: Extract key features from the received reflected signals;

[0039] Time-domain analysis:

[0040] Peak detection: Identify multiple peaks in the reflected signal. The first peak corresponds to the direct reflection, and subsequent peaks are caused by multipath reflections;

[0041] Delay difference: Calculate the time delay between different peaks ;

[0042] ;

[0043] Where , represents the arrival times of the first and second peaks;

[0044] Intensity ratio: Compare the relative intensities of different peaks , if one peak is much weaker than the other, it is a false echo;

[0045] ;

[0046] Where , represents the reflected signal intensities of the first and second peaks;

[0047] Frequency domain analysis:

[0048] Fourier transform: Perform a fast Fourier transform (FFT) on the reflected signal and convert it to the frequency domain for analysis;

[0049] Spectrum characteristics: Distinguish real reflections and false echoes by analyzing spectrum characteristics;

[0050] Design filters: Design digital filters according to known target characteristics and environmental conditions. The filters include low-pass filters, high-pass filters, and band-pass filters;

[0051] Apply filters: Apply the designed filters to the reflected signal to remove unnecessary noise and false echoes and retain the most likely real reflected signal.

[0052] Furthermore, the implementation steps of the sensor drift correction module are as follows:

[0053] Implement self-check: The automatic calibration module starts the self-check program according to the set calibration period or trigger condition, and obtains the status data of each key component of the lidar system, including the position of the mirror, the laser emission angle, and the temperature sensor reading;

[0054] Compare with reference data: Load reference benchmark data from pre-stored standard values or the latest calibration information sent by the ground station, compare the current status data with the reference benchmark, and identify any changes or offsets beyond the allowable range;

[0055] By calculating the relative error to quantify:

[0056] ;

[0057] where is the current reading, is the reference benchmark;

[0058] Dynamic adjustment: If any offset exceeds the preset threshold, immediately start the correction program and use a driving device of a precision motor or piezoelectric ceramic to finely adjust the position of the mirror to compensate for mechanical drift caused by temperature changes or other factors;

[0059] Assume the angle to be adjusted is :

[0060] ;

[0061] where is the detected angle deviation, is the proportionality coefficient used to control the adjustment amplitude;

[0062] Compensate for the offset on the optical path by changing the direction or tilt angle of the laser emitter;

[0063] Assume the angle to be adjusted is :

[0064] ;

[0065] where is the detected angle deviation, is the proportionality coefficient;

[0066] Update the model parameters: Recalculate the refractive index in water based on the latest calibration results and update the refractive index model;

[0067] If the environmental conditions change, the refractive index needs to be adjusted:

[0068] ;

[0069] where is the initial refractive index, , are the empirical coefficients of the effects of temperature and salinity respectively, 、 are the temperature and salinity under reference conditions respectively;

[0070] All parameters dependent on the sensor state are updated in a timely manner;

[0071] Verification and recording: After completing the dynamic adjustment, perform a self-check again and compare it with the reference benchmark to confirm that all offsets are within the acceptable range, and save the time, adjustment parameters and results of each calibration.

[0072] Furthermore, the execution steps of the data integration and optimization module are as follows:

[0073] Merge all valid point cloud data points after refraction correction and multipath reflection suppression to form a complete three-dimensional point cloud data set;

[0074] Ensure that each point contains its coordinates (x, y, z) and intensity information , where (x, y, z) represents the spatial coordinates of the point cloud data point, represents the reflection signal intensity;

[0075] If data from multiple sensors or different time periods is used, time synchronization is required.

[0076] Furthermore, the data integration and optimization module also performs the following steps:

[0077] Identify and remove isolated points, outliers, and noise points through algorithms, train the model to recognize the normal point cloud distribution pattern, and mark the points deviating from these patterns as noise. Reduce the irregularities in the point cloud through a smoothing algorithm, perform weighted averaging by combining spatial distance and intensity difference, and smooth the surface while preserving the edge features;

[0078] Extract the geometric features in the point cloud and calculate the normal vector of each point :

[0079] ;

[0080] wherein is the current point, is its neighboring point;

[0081] Evaluate the curvature change around each point in the point cloud, segment the point cloud according to geometric features or semantic information, and separate independent objects or regions;

[0082] Use Delaunay triangulation and Poisson reconstruction to convert the point cloud data into a triangular mesh.

[0083] Furthermore, the calculation formula of the smoothing algorithm is as follows:

[0084] ;

[0085] ;

[0086] wherein represents the position of the smoothed point, wherein is the position of the original point, is the weight function, based on distance attenuation, represents the number of neighboring points participating in the smoothing calculation.

[0087] The present invention also provides a storage medium storing non-transitory computer-readable instructions for performing the steps corresponding to one or more modules in the foregoing underwater three-dimensional scanning imaging system based on a streak tube lidar.

[0088] The beneficial effects of the present invention are as follows:

[0089] The present invention can effectively address the problem of light attenuation in water and the possible small offsets of the internal components of the system after long-term operation. Even in a complex and changeable underwater environment, it can maintain high-precision distance measurement and three-dimensional modeling capabilities. Through an automatic calibration and dynamic adjustment mechanism, it ensures the consistency and reliability of the system during long-term operation, reduces noise and errors, improves the quality of the point cloud data, and makes the generated three-dimensional image clearer and more accurate. Brief Description of the Drawings

[0090] Figure 1 It is a structural block diagram of an underwater three-dimensional scanning imaging system based on a streak tube lidar proposed by the present invention.

[0091] In the figure: 101, environmental perception and initialization module; 102, laser emission control module; 103, reflected signal receiving module; 104, original distance calculation module; 105, refraction correction module; 106, multipath reflection suppression module; 107, sensor drift correction module; 108, data integration and optimization module; 109, user interface and output module. Specific embodiments

[0092] Now, the subject matter described herein will be discussed with reference to example embodiments. It should be understood that discussing these embodiments is only to enable those skilled in the art to better understand and thus implement the subject matter described herein, and the functions and arrangements of the elements discussed can be changed without departing from the scope of protection of the content of this specification. Each example can omit, substitute, or add various processes or components as needed. Additionally, the features described relative to some examples can also be combined in other examples.

[0093] An underwater three-dimensional scanning imaging method based on a streak tube lidar includes the following steps:

[0094] S100, initialization settings: According to known environmental conditions, calculate and determine the initial refractive index, set the operating parameters of the lidar, including wavelength and pulse frequency, configure the automatic calibration module, and set the calibration period and trigger conditions (such as once per hour or when a significant change is detected);

[0095] In an embodiment of the present invention, the specific steps are as follows:

[0096] Determine the initial refractive index :

[0097] Environmental condition monitoring: Use auxiliary sensors (such as CTD probes) to obtain the temperature , salinity and pressure ;

[0098] Look up or calculate the refractive index: According to the obtained environmental conditions, use empirical formulas to determine the initial refractive index ;

[0099] In an embodiment of the present invention, the seawater refractive index formula used is based on the Cox-Munk model or other verified formulas, for example:

[0100] ;

[0101] where is the refractive index under reference conditions, , , p are the empirical coefficients of the influence of temperature, salinity, and pressure on the refractive index respectively, , , are the temperature, salinity, and pressure under reference conditions, where represents the refractive index, which is a dimensionless physical quantity representing the ratio of the propagation speed of light in a certain medium to its propagation speed in a vacuum.

[0102] Select the wavelength : Select the laser wavelength according to the application requirements and water body characteristics. For underwater applications, wavelengths in the blue-green light region are usually selected because these wavelengths have better penetration in water.

[0103] Set the pulse frequency: Set an appropriate pulse emission frequency according to the size of the target area and the desired point cloud density , which determines the number of laser pulses emitted per second.

[0104] Set the calibration period: Define the time interval for automatic calibration, such as once per hour. This time interval can be adjusted according to the duration of the task and the accuracy requirements.

[0105] Trigger condition configuration: Set the trigger conditions to initiate additional immediate calibration, such as when significant changes are detected (such as rapid temperature changes, obvious deviation of sensor output, etc.). Statistical methods or machine learning algorithms can be used to identify abnormal situations.

[0106] Save reference data: Record and store the state at system initialization as reference data, including but not limited to key parameters such as mirror position and laser emission angle.

[0107] Input verification: Ensure that all input parameters (such as temperature, salinity, pressure, wavelength, and pulse frequency) are within a reasonable range and comply with physical constraints.

[0108] Simulation test: Before actual operation, preliminary tests can be carried out through a simulator or historical data to verify the effectiveness of the set parameters and the expected performance of the system.

[0109] System self-check: Execute a comprehensive system self-check program to check whether the hardware connection is normal and whether the software algorithm is ready.

[0110] User confirmation: Provide a user interface for the user to confirm that all settings are correct, and then officially start the measurement process.

[0111] S200, emit laser pulses: The lidar emits laser pulses at a predetermined time interval while recording the emission time. .

[0112] In one embodiment of the present invention, it specifically includes the following steps:

[0113] Activate the laser source: Ensure that the laser source of the streak tube lidar is in a working state and configure it according to the working parameters (such as wavelength and pulse frequency ) set in S100.

[0114] Synchronization signal generation: Generate an accurate timestamp for each upcoming laser pulse, which will serve as the basis for subsequent distance calculations. A high-precision clock or GPS synchronization module can be used to ensure the accuracy of the timestamp.

[0115] Trigger laser emission: Trigger the emission of laser pulses at a predetermined time interval If the pulse frequency is set to 10,000 times per second (i.e., 10 kHz), a pulse will be emitted every 0.1 milliseconds.

[0116] The calculation formula for the time interval is as follows:

[0117] ;

[0118] Record the emission time : For each emission of a laser pulse, immediately record its exact emission time . This is usually achieved through a hardware timer to ensure microsecond-level accuracy.

[0119] Pulse propagation: The emitted laser pulse propagates forward at the speed of light c until it encounters the surface of the target object and is reflected. In an underwater environment, the light beam will be affected by refraction, scattering, and absorption, but these factors will be handled in subsequent steps.

[0120] Open the reflected echo reception window: Once the laser pulse is emitted, the system enters a short waiting period called the "reception window", during which it is ready to receive the reflected signal that may return from the target object.

[0121] Wavelength : The central wavelength of the laser, in nanometers (nm) or micrometers (μm), determines the behavior characteristics of light propagation in water.

[0122] Pulse frequency : The emission rate of the laser pulse, in hertz (Hz), that is, the number of pulses emitted per second.

[0123] Emission time : The exact moment of each laser pulse emission, in seconds (s) or smaller time units (such as microseconds μs).

[0124] Receiving window time: The time period from emission to the expected reception of the reflected signal, depending on the expected maximum measurement range and the speed of light c.

[0125] Maximum measurement distance (Based on the receiving window time)

[0126] ;

[0127] Where is the time length of the receiving window, considering the round-trip path, so it is divided by 2, represents the speed of light in a vacuum, = 299,792,458 (m / s).

[0128] S300, receiving the reflected signal: Receive the reflected signal from the surface of the target object and record the reception time .

[0129] In one embodiment of the present invention, the following are further included:

[0130] Start the receiving mode: Immediately enter the receiving ready state after the laser pulse is emitted. The system calculates the expected receiving window time based on the pre-set maximum measurement distance and the speed of light c .

[0131] ;

[0132] Where is the maximum measurement distance, is the speed of light in a vacuum;

[0133] Set the sensitivity: Adjust the sensitivity of the receiver to adapt to the low light intensity conditions unique to the underwater environment. This may involve settings such as gain control and filter selection to ensure that weak reflected signals can be captured while suppressing noise.

[0134] Detect the echo: The receiver starts to monitor the reflected signal from the surface of the target object and records the intensity of each received reflected signal and its corresponding timestamp .

[0135] Multi-channel reception: For complex scenarios or high-precision requirements, multi-channel reception technology can be used to simultaneously capture reflected signals in different directions, thereby improving the integrity and accuracy of the data.

[0136] Initial screening: The received signals are initially screened to remove data points that clearly do not conform to physical laws (such as signals outside the time window), reducing the subsequent processing burden.

[0137] Synchronization correction: If there are multiple sensors or receiving channels, time synchronization correction is required to ensure that the time reference of all data points is consistent.

[0138] Timestamp recording: Record the exact reception time for each valid reflected signal . This is one of the key parameters for subsequent distance calculation.

[0139] Signal feature extraction: In addition to time information, other signal features that are helpful for subsequent analysis should also be recorded, such as amplitude, frequency components, etc. These features can help distinguish real reflections from false echoes, especially in an environment with severe multipath reflections.

[0140] Reception time : The exact moment when each valid reflected signal arrives, in seconds (s) or a smaller time unit (such as microseconds μs)

[0141] Reflected signal intensity : The intensity of the reflected signal that changes with time, usually expressed in voltage or power, and is used to evaluate the signal quality.

[0142] Reception window time : The time period from transmission to the expected reception of the reflected signal, depending on the expected maximum measurement range and the speed of light c.

[0143] Maximum measurement distance : The maximum measurable distance calculated based on the reception window time.

[0144] S400, Calculate the raw distance:

[0145] ;

[0146] where is the speed of light in a vacuum, is the estimated value of the raw distance calculated based on time and the speed of light.

[0147] S500, Refraction correction: Apply Snell's law to correct the distance deviation caused by refraction. For the interface from air to water, determine the incident angle and the refraction angle , calculate the actual ranging value ;

[0148] ;

[0149] where and are the refractive indices in air and water respectively, and are the incident angle and the refraction angle respectively.

[0150] Calculate the actual ranging value the distance considering the refraction effect.

[0151] In one embodiment of the present invention, it specifically includes the following:

[0152] Determine the incident angle and the refraction angle :

[0153] Geometric relationship analysis: According to the installation angle of the lidar system and the relative position of the target object, determine the incident angle of the laser beam at the air-water interface . This angle can be calculated through the known sensor attitude information (such as pitch angle, roll angle) and the target position.

[0154] ;

[0155] where and are the distance differences in the vertical and horizontal directions respectively.

[0156] Apply Snell's law: Use Snell's law to calculate the refraction angle of light when it enters water from air ;

[0157] ;

[0158] Solve this equation to obtain :

[0159] ;

[0160] where is the refractive index of light in air (approximately 1), is the refractive index of light in water, which is determined by the S100 initialization settings.

[0161] Calculate the actual ranging value :

[0162] Consider the refraction path length: Since the propagation speed of light is different in different media, the original distance estimate value must be corrected to reflect the true ranging value . Specifically, the difference in the path lengths of light in the two media needs to be considered.

[0163] The distance traveled in air:

[0164] ;

[0165] Distance traveled in water:

[0166] ;

[0167] Actual ranging value Total distance considering refraction effects:

[0168] ;

[0169] Update model parameters: Update the distance model within the system according to the latest refraction angle and ranging value to ensure the consistency and accuracy of subsequent measurements.

[0170] Data verification: Conduct preliminary verification on the point cloud data after refraction correction, check for outliers or unreasonable results, and compare with other known reference data.

[0171] where represents the incident angle, represents the refraction angle, represents the refractive index of light in air, usually with a value of 1, represents the refractive index of light in water, which is determined by environmental conditions and has been determined in S100, represents the actual ranging value considering refraction effects.

[0172] S600, Multipath reflection suppression: Use digital filters or other signal processing methods to remove false echoes. This may involve filtering in the time domain or frequency domain, or machine learning-based methods to distinguish real and false signals.

[0173] For each point cloud data point, evaluate whether it may be generated by multipath reflection and decide whether to retain or exclude the point.

[0174] In one embodiment of the present invention, it further includes the following:

[0175] Preliminary screening: Filter out data points that clearly do not conform to physical laws based on a time window, such as signals beyond the maximum measurement distance.

[0176] Feature extraction: Extract key features from the received reflection signals, such as amplitude, frequency components, arrival time, etc. These features will be used for subsequent classification and decision-making.

[0177] Time domain analysis:

[0178] Peak detection: Identify multiple peaks in the reflection signal. Usually, the first peak corresponds to the direct reflection, and subsequent peaks may be caused by multipath reflection.

[0179] Delay difference: Calculate the time delay between different peaks to determine whether they come from different paths of the same target or different targets.

[0180] ;

[0181] where , represent the arrival times of the first and second peaks in seconds (s) or smaller time units (such as microseconds μs).

[0182] Intensity ratio: Compare the relative intensities of different peaks , if one peak is much weaker than the other, it may be a false echo.

[0183] ;

[0184] where , represent the reflection signal intensities of the first and second peaks, usually expressed in voltage or power.

[0185] Frequency domain analysis:

[0186] Fourier transform: Perform a fast Fourier transform (FFT) on the reflection signal and convert it to the frequency domain for analysis. Multipath reflections may introduce additional frequency components in the spectrum.

[0187] Spectrum characteristics: By analyzing spectrum characteristics such as bandwidth and main frequency position, real reflections and false echoes can be distinguished.

[0188] Design filters: Design digital filters based on known target characteristics (such as shape and material) and environmental conditions (such as water body type). Common filters include low-pass filters, high-pass filters, band-pass filters, etc.

[0189] Apply filters: Apply the designed filters to the reflection signal to remove unnecessary noise and false echoes and retain the most likely real reflection signal.

[0190] Train models: Use historical data sets to train machine learning models such as support vector machines (SVM), random forests (RF), neural networks (NN), etc. to automatically distinguish real reflections and false echoes.

[0191] Real-time classification: In actual operation, use the trained model to classify each received reflection signal to determine whether it is a real ranging signal.

[0192] Model selection: Select an appropriate machine learning algorithm according to the characteristics of the problem.

[0193] Training process: Use the training dataset to train the selected model and optimize the model parameters to minimize classification errors.

[0194] The following is a brief description of some commonly used models and their parameters:

[0195] Adopt Support Vector Machine (SVM):

[0196] Kernel function: Commonly used kernel functions include linear kernel, polynomial kernel, RBF (Radial Basis Function) kernel, etc.

[0197] Regularization parameter C: Controls the model complexity and prevents overfitting.

[0198] Kernel parameter γ: For the RBF kernel, γ controls the width of the Gaussian function.

[0199] ;

[0200] where is the kernel function, and are the parameters obtained through training.

[0201] Input the features of the newly received reflected signal into the trained model to obtain its classification result (true reflection or false echo);

[0202] Multi-feature fusion: Combine the analysis results in the time domain and frequency domain and the output of the machine learning model to comprehensively evaluate the authenticity of each reflected signal.

[0203] Threshold setting: Define a series of thresholds. When the reflected signal meets certain conditions (such as the delay difference is less than a certain value, the intensity ratio is lower than a certain ratio), it is determined as a false echo and excluded.

[0204] Data association: For the results of multiple scans, perform data association analysis to confirm whether there are consistent reflected signals, and further improve the discrimination accuracy.

[0205] S700, sensor drift correction:

[0206] Implement self-check: The automatic calibration module regularly checks the sensor status and compares the current readings with the reference benchmark (which can be the pre-stored standard value or the latest calibration information sent by the ground station).

[0207] Dynamic adjustment: If any deviation exceeds the preset threshold, the correction program is immediately started. The drift can be compensated by fine-tuning the position of the mirror or adjusting the laser emission angle.

[0208] Update model parameters: Update the refractive index model and other related parameters according to the latest calibration results to ensure the accuracy of subsequent measurements.

[0209] In one embodiment of the present invention, the following contents are further included:

[0210] Regular inspection: The automatic calibration module starts the self - inspection program according to the set calibration period (e.g., once per hour) or trigger conditions (e.g., when a significant change is detected).

[0211] Read the current state: Obtain the status data of each key component of the lidar system, including the position of the mirror, the laser emission angle, the readings of the temperature sensor, etc.

[0212] Load the reference benchmark: Load the reference benchmark data from the pre - stored standard values or the latest calibration information sent by the ground station. These benchmark data are usually obtained in a known stable environment and have been verified to be accurate and reliable.

[0213] Compare the differences: Compare the current state data with the reference benchmark to identify any changes or offsets that exceed the allowable range. This can be quantified by calculating the relative error as follows:

[0214] ;

[0215] where is the current reading, is the reference benchmark.

[0216] Judge whether correction is needed: If any offset exceeds the preset threshold, start the correction program immediately. The threshold can be set according to the requirements of specific applications. For example, for high - precision measurement tasks, a more stringent threshold can be set.

[0217] Fine - tune the position of the mirror: Use a driving device such as a precision motor or a piezoelectric ceramic to make fine adjustments to the position of the mirror to compensate for mechanical drift caused by temperature changes or other factors. Assume the angle to be adjusted is :

[0218] ;

[0219] where is the detected angle deviation, and k is the proportionality coefficient used to control the adjustment amplitude.

[0220] Adjust the laser emission angle: Compensate for the offset on the optical path by changing the direction or tilt angle of the laser emitter. Assume the angle to be adjusted is :

[0221] ;

[0222] where is the detected angle deviation, is the proportionality coefficient.

[0223] Update the refractive index model: Recalculate the refractive index in water based on the latest calibration results and update the refractive index model. If environmental conditions change (such as temperature, salinity), the refractive index needs to be adjusted accordingly: , and update the refractive index model. If environmental conditions change (such as temperature, salinity), the refractive index needs to be adjusted accordingly:

[0224] ;

[0225] where is the initial refractive index, , are the empirical coefficients of the effects of temperature and salinity respectively, , are the temperature and salinity under reference conditions respectively.

[0226] Update other relevant parameters: Ensure that all parameters dependent on the sensor state (such as timestamp synchronization in the distance calculation formula) are updated in a timely manner to ensure the consistency and accuracy of subsequent measurements.

[0227] Verify the calibration effect: After completing the dynamic adjustment, perform a self-check again and compare with the reference benchmark to confirm that all offsets are within the acceptable range.

[0228] Record the calibration log: Save the time, adjusted parameters, and results of each calibration for subsequent analysis and maintenance.

[0229] S800, Integration and optimization: Combine the calibrated and purified data points to form the final 3D point cloud image.

[0230] Further post-processing steps may be required, such as smoothing, feature extraction, etc., to enhance the image quality.

[0231] Output the result: Generate a high-quality underwater 3D map or model for subsequent research or decision support.

[0232] In one embodiment of the present invention, it specifically includes the following content:

[0233] Point cloud data aggregation: Combine all valid point cloud data points after refractive correction (S500) and multipath reflection suppression (S600) to form a complete 3D point cloud data set;

[0234] Ensure that each point contains its coordinates (x, y, z) and intensity information , where (x, y, z) represents the spatial coordinates of the point cloud data point, represents the reflection signal intensity;

[0235] Time synchronization processing: If data from multiple sensors or different time periods are used, time synchronization is required to ensure that the time bases of all data points are consistent.

[0236] Noise identification: Identify and remove noise points such as isolated points and outliers through statistical methods or machine learning-based algorithms.

[0237] Statistical filtering: For example, use the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm to identify noise points in low-density regions.

[0238] Machine learning-based method: Train a model to identify the normal point cloud distribution pattern and mark points deviating from these patterns as noise.

[0239] Smoothing processing: Apply various smoothing algorithms to reduce the irregularities in the point cloud and improve surface continuity and visual effects. Commonly used smoothing algorithms include:

[0240] Moving Least Squares (MLS):

[0241] ;

[0242] ;

[0243] where represents the Euclidean distance between the target point and the neighboring points , and represents the standard deviation, which controls the width and decay rate of the weight function.

[0244] where represents the position of the smoothed point, where is the position of the original point, is the weight function, usually based on distance decay, represents the number of neighboring points participating in the smoothing calculation.

[0245] Bilateral filter: Combine spatial distance and intensity difference for weighted averaging, smoothing the surface while preserving edge features.

[0246] Geometric feature extraction: Extract geometric features in the point cloud, such as planes, curvatures, boundaries, etc., for subsequent analysis or modeling.

[0247] Normal vector estimation: Calculate the normal vector of each point , which is used to describe the local surface direction.

[0248] ;

[0249] Among them is the current point, and is its neighboring point.

[0250] Curvature calculation: Evaluating the curvature change around each point in the point cloud helps identify different geometric structures.

[0251] Point cloud segmentation: Segmenting the point cloud according to geometric features or semantic information to separate independent objects or regions.

[0252] Common methods include:

[0253] Region growing method: Starting from a seed point, gradually expanding to adjacent similar points until the stopping condition is met.

[0254] Random forest classification: Using a trained classifier to distinguish different types of target objects.

[0255] Triangulated mesh generation: Converting point cloud data into a triangular mesh for visualization and further processing.

[0256] Common algorithms include Delaunay triangulation and Poisson reconstruction:

[0257] Delaunay triangulation: Ensuring that no two triangles share the same inscribed center to generate an optimal mesh.

[0258] Poisson reconstruction: Constructing a closed three-dimensional surface by solving the Poisson equation.

[0259] Texture mapping: If there is a corresponding optical image or other high-resolution data source, these data can be mapped onto the three-dimensional model to enhance the detail expression.

[0260] Filling holes: For holes or missing parts in the point cloud, interpolation methods are used to fill them to maintain the integrity of the model.

[0261] Optimizing display: Adjusting attributes such as the color and transparency of the model to make it more suitable for the requirements of specific application scenarios.

[0262] As Figure 1 shown, the present invention also proposes an underwater three-dimensional scanning imaging system based on a streak tube lidar, including:

[0263] Environmental perception and initialization module 101: According to the known environmental conditions, calculate and determine the initial refractive index, set the working parameters of the lidar, including wavelength and pulse frequency, and configure the automatic calibration function, setting the calibration period and trigger conditions;

[0264] Laser emission control module 102: Emitting laser pulses at a predetermined time interval and recording the exact time of each emission ;

[0265] Reflection signal receiving module 103: Receives the reflection signal from the surface of the target object and records its arrival time ;

[0266] Original distance calculation module 104: Calculates the original distance estimate based on time and the speed of light, and its calculation formula is as follows: ;

[0267] Where is the speed of light in vacuum, = 299,792,458 (m / s), is the original distance estimate calculated based on time and the speed of light;

[0268] Refraction correction module 105: Applies Snell's law to correct the distance deviation caused by refraction. For the interface from air to water, determines the incident angle and the refraction angle , and calculates the actual ranging value ;

[0269] Multipath reflection suppression module 106: Removes the false echoes caused by multipath reflection, retains the true reflection signals, and for each point cloud data point, evaluates whether it may be generated by multipath reflection and decides to retain or exclude the point;

[0270] Sensor drift correction module 107: The automatic calibration module regularly checks the sensor status, compares the current readings with the reference benchmark, and if any deviation exceeds the preset threshold, immediately starts the correction program;

[0271] Data integration and optimization module 108: Combines the corrected and purified data points to form the final three-dimensional point cloud image;

[0272] User interface and output module 109: Provides an intuitive operation interface, displays the finally generated three-dimensional map or model, and supports the export and sharing functions.

[0273] At least one embodiment disclosed by the present invention provides a storage medium storing non-temporary computer-readable instructions for executing the steps corresponding to one or more modules in the foregoing underwater three-dimensional scanning imaging system based on a streak tube lidar.

[0274] The computer program can be stored / distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with other hardware or as part of other hardware, but can also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems. Any reference signs in the claims shall not be construed as limiting the scope.

[0275] The above describes the embodiments of this example, but this example is not limited to the above specific implementation manners. The above specific implementation manners are merely illustrative rather than restrictive. Under the inspiration of this example, those of ordinary skill in the art can also make many forms, all of which fall within the protection scope of this example.

Claims

1. An underwater three-dimensional scanning imaging system based on streak tube laser radar, characterized in that: include: Environmental perception and initialization module: determines the initial refractive index according to environmental conditions, sets the wavelength and pulse frequency of the lidar, and configures the calibration cycle and trigger conditions of the automatic calibration function; Laser emission control module: emits laser pulses according to a predetermined pulse frequency and records the emission time; Reflection signal receiving module: receives the reflection signal from the surface of the target object and records its arrival time; Raw distance calculation module: calculates the raw distance estimate based on the launch time, launch time and speed of light; Refraction correction module: corrects the distance deviation caused by refraction. For the interface from air to water, it determines the angle of incidence and angle of refraction and calculates the actual distance measurement value. Multipath reflection suppression module: evaluates whether each point cloud data point is generated by multipath reflection, removes false echoes caused by multipath reflection, and retains the real reflection signal; The processing steps of the multipath reflection suppression module are as follows: Data preprocessing and feature extraction: Preliminary screening: Filter out data points that are obviously inconsistent with physical laws based on the time window; Feature extraction: extract key features from the received reflection signal; Identify multipath reflection signals: Time Domain Analysis: Peak detection: Identify multiple peaks in the reflected signal, the first peak corresponds to the direct reflection, while subsequent peaks are caused by multipath reflections; Delay difference: calculate the time delay between different peaks ; ； in , represents the arrival time of the first and second peaks; Intensity ratio: compare the relative intensities of different peaks , if one peak is much weaker than the other, it is a false echo; ； in , Indicates the reflected signal strength of the first and second peaks; Frequency Domain Analysis: Fourier transform: Perform fast Fourier transform on the reflected signal and convert it to the frequency domain for analysis; Spectral characteristics: Distinguish true reflections from false echoes by analyzing spectral characteristics; Use a digital filter to remove false echoes: Design filters: Design digital filters based on known target characteristics and environmental conditions. The filters include low-pass filters, high-pass filters, and band-pass filters. Apply filter: Apply the designed filter to the reflected signal to remove unnecessary noise and false echoes and retain the most likely true reflected signal; Sensor drift correction module: Regularly checks the sensor status, compares the current reading with the reference benchmark, and initiates the correction procedure if the drift exceeds the preset threshold; Data integration and optimization module: combines the corrected and purified data points to form the final 3D point cloud image; User interface and output module: provides an operation interface, displays the final generated 3D map or model, and supports export and sharing functions.

2. The underwater three-dimensional scanning imaging system based on streak tube laser radar according to claim 1, characterized in that: The initial refractive index in the environmental perception and initialization module is obtained by using the CTD probe to obtain the current water temperature. ,salinity and pressure ; Based on the obtained environmental conditions, the initial refractive index is determined using the empirical formula: ； in is the refractive index under reference conditions, represents the laser wavelength, where Represents the refractive index.

3. The underwater three-dimensional scanning imaging system based on streak tube laser radar according to claim 2, characterized in that: The incident angle in the refraction correction module is based on the installation angle of the lidar system and the relative position of the target object; The angle of incidence is calculated using the known sensor attitude information and target position; ； in and are the distance differences in the vertical and horizontal directions, respectively; Use Snell's law to calculate the angle of refraction of light when it passes from air into water ; ； Solve this equation to get : ； in is the refractive index of light in air, It is the refractive index of light in water, obtained from the environment perception and initialization module.

4. The underwater three-dimensional scanning imaging system based on streak tube laser radar according to claim 3, characterized in that: Calculate the actual distance value The specific steps are as follows: Distance traveled in air: ； Distance traveled in water: ； Actual distance measurement value Total distance after taking into account refraction: ； in Indicates the actual distance value after taking into account the effect of refraction. The distance traveled in the air. represents the distance traveled in water, is a raw distance estimate based on time and the speed of light.

5. The underwater three-dimensional scanning imaging system based on streak tube laser radar according to claim 4, characterized in that: The implementation steps of the sensor drift correction module are as follows: Implement self-test: The automatic calibration module starts the self-test procedure according to the set calibration cycle or trigger conditions to obtain the status data of each key component of the lidar system, including the reflector position, laser emission angle, and temperature sensor reading; Compare to reference data: Load reference data from pre-stored standard values ​​or the latest calibration information sent by the ground station, compare the current status data with the reference, and identify any changes or deviations outside the allowable range; By calculating the relative error To quantify: ； in is the current reading, It is a reference benchmark; Dynamic adjustment: If any deviation is found to exceed the preset threshold, the correction procedure is immediately initiated, using precision motors and piezoelectric ceramic drives to make subtle adjustments to the position of the mirror to compensate for mechanical drift caused by temperature factors; Assume that the angle to be adjusted is : ； in is the detected angular deviation, is the proportional coefficient, which is used to control the adjustment amplitude; Compensate for the deviation in the optical path by changing the direction and tilt angle of the laser transmitter; Assume that the angle to be adjusted is : ； in is the detected angular deviation, is the proportionality coefficient; Update model parameters: recalculate the refractive index in water based on the latest calibration results , and update the refractive index model; If environmental conditions change, the refractive index needs to be adjusted: ； in is the initial refractive index, , are the empirical coefficients of the effects of temperature and salinity, , are the temperature and salinity under reference conditions, Represents the temperature of the water body, Indicates the salinity of a body of water; All parameters that depend on the sensor status are updated in a timely manner; Verification and Recording: After completing the dynamic adjustment, perform self-check again and compare with the reference benchmark to confirm that all deviations are within the acceptable range, and save the time, adjustment parameters and results of each calibration.

6. The underwater three-dimensional scanning imaging system based on streak tube laser radar according to claim 5, characterized in that: The execution steps of the data integration and optimization module are as follows: Merge all valid point cloud data points after refraction correction and multipath reflection suppression to form a complete 3D point cloud data set; Make sure each point contains its coordinates (x,y,z) and intensity information , where (x, y, z) represents the spatial coordinates of the point cloud data points, Indicates the reflected signal strength; If multiple sensors or data from different time periods are used, time synchronization is required.

7. The underwater three-dimensional scanning imaging system based on streak tube laser radar according to claim 6, characterized in that: The Data Integration and Optimization module also performs the following steps: The algorithm identifies and removes isolated points and outlier noise points, trains the model to identify normal point cloud distribution patterns, and marks points that deviate from these patterns as noise. The smoothing algorithm reduces irregularities in the point cloud, combines spatial distance and intensity differences for weighted averaging, and smoothes the surface while retaining edge features. Extract geometric features from the point cloud and calculate the normal vector of each point : ； in is the current point, is its neighboring point; Evaluate the curvature change around each point in the point cloud, segment the point cloud based on geometric features or semantic information, and separate independent objects or regions; Delaunay triangulation and Poisson reconstruction are used to convert point cloud data into triangular meshes.

8. The underwater three-dimensional scanning imaging system based on streak tube laser radar according to claim 7, characterized in that: The calculation formula of the smoothing algorithm is as follows: ； ； in represents the point position after smoothing, where is the position of the original point, is a weight function based on distance decay, Indicates the number of neighboring points involved in the smoothing calculation.

9. A storage medium storing non-transitory computer-readable instructions, characterized in that: Used to execute steps corresponding to one or more modules in an underwater three-dimensional scanning imaging system based on a streak tube laser radar as described in any one of claims 1-8.

Images