Underwater imaging method and medium based on supercontinuum laser light source and dual-target positioning

Abstract

The present invention relates to the technical field of underwater imaging, and discloses an underwater imaging method and medium based on a supercontinuum laser light source and dual-target positioning, including: obtaining spectral range data of underwater imaging, three-dimensional coordinates of a target object, and image data, setting the output power of a supercontinuum laser according to the spectral range data, emitting a supercontinuum laser to the target object according to the output power and the three-dimensional coordinates, obtaining reflection data of the supercontinuum laser reflected on the surface of the target object, performing point cloud analysis on the reflection data to obtain three-dimensional point cloud data, performing data registration processing on the three-dimensional point cloud data and the image data to obtain comprehensive three-dimensional data, establishing a 3D image model based on the comprehensive three-dimensional data, performing model accuracy evaluation on the 3D image model, and confirming the model based on the evaluation result. The present invention also proposes an underwater imaging medium based on a supercontinuum laser light source and dual-target positioning. The present invention can improve the accuracy of underwater imaging.

Description

Technical Field

[0001] The present invention relates to the field of underwater imaging technology, and in particular to an underwater imaging method and medium based on a supercontinuum laser light source and dual-target positioning. Background Art

[0002] With the continuous development of underwater detection and imaging technology, the demand for high-precision and high-resolution imaging of underwater environments is becoming increasingly urgent. Traditional underwater imaging methods are often affected by factors such as water turbidity and light attenuation, and the imaging effect is not ideal. Therefore, studying new underwater imaging technologies and improving imaging quality have important practical application value in the fields of underwater detection, resource development, and environmental monitoring.

[0003] In recent years, supercontinuum laser technology has been widely used in spectroscopy, optical measurement and other fields due to its unique spectrum broadening output characteristics. The output pulse time of supercontinuum laser is longer than that of traditional lasers, and it is output in a continuous manner, with the characteristics of high energy, high repetition rate and high stability. When supercontinuum laser passes through solid, liquid or gas nonlinear media, supercontinuum spectrum is generated, which provides a new research direction for underwater imaging.

[0004] On the other hand, binocular vision positioning technology, as an effective three-dimensional measurement method, shows great potential in object positioning and scene reconstruction. Compared with monocular vision, binocular vision can provide more spatial information, thereby improving the accuracy and stability of positioning. In underwater environments, due to changes in light propagation characteristics, traditional visual positioning methods are often difficult to apply. Therefore, combining supercontinuum laser technology with binocular vision positioning technology is expected to solve a series of technical problems in underwater imaging. Summary of the invention

[0005] The present invention provides an underwater imaging method and medium based on a supercontinuum laser light source and binocular positioning, the main purpose of which is to improve the accuracy of underwater imaging.

[0006] To achieve the above object, the present invention provides an underwater imaging method based on a supercontinuum laser light source and dual-target positioning, comprising:

[0007] Acquire spectral range data of underwater imaging using a spectral imager pre-set underwater;

[0008] Using a pre-set binocular camera to obtain the three-dimensional coordinates of the target object through binocular positioning, and using the binocular camera to obtain image data of the target object;

[0009] Setting the output power of the supercontinuum laser according to the spectral range data, and emitting the supercontinuum laser to the target object according to the output power and the three-dimensional coordinates;

[0010] Acquire reflection data of the supercontinuum laser reflected on the surface of the target object by using the spectral imager, perform point cloud analysis on the reflection data, and obtain three-dimensional point cloud data;

[0011] Performing data registration processing on the three-dimensional point cloud data and the image data to obtain comprehensive three-dimensional data, and establishing a 3D image model based on the comprehensive three-dimensional data;

[0012] Performing model accuracy evaluation on the 3D image model;

[0013] If the 3D image model fails to pass the model accuracy assessment, adjusting the exposure time of the spectral imager, and re-performing the step of using the spectral imager to obtain the reflection data of the supercontinuum laser light source reflected on the surface of the target object;

[0014] If the 3D image model passes the model accuracy assessment, the 3D image model is output.

[0015] Optionally, the method of acquiring the spectral range data of underwater imaging by using a spectral imager pre-set underwater includes:

[0016] Fire a supercontinuum laser at a target object pre-placed underwater;

[0017] Acquiring the reflected light of the supercontinuum laser by using the spectral imager, recording the capture frequencies of reflected light of different wavelengths, and obtaining spectral distribution data;

[0018] The spectrum range is extracted from the spectrum distribution data according to a preset frequency threshold to obtain spectrum range data.

[0019] Optionally, the obtaining the three-dimensional coordinates of the target object by binocular positioning using a preset binocular camera includes:

[0020] Performing binocular positioning on the binocular camera to obtain internal and external parameters;

[0021] Performing epipolar correction on the binocular camera using the intrinsic reference;

[0022] The position of the target object is captured by the binocular camera after epipolar correction to obtain the matching point;

[0023] Disparity calculation and depth calculation are performed according to the matching points and the external parameters to obtain three-dimensional coordinates.

[0024] Optionally, acquiring image data of the target object by using the binocular camera includes:

[0025] Using the binocular camera to capture an image of a target object to obtain a target image;

[0026] Performing contrast enhancement processing on the target image to obtain a preprocessed image;

[0027] The pre-processed image is subjected to denoising processing to obtain image data.

[0028] Optionally, performing point cloud analysis on the reflection data to obtain three-dimensional point cloud data includes:

[0029] Performing outlier filtering on the reflection data to obtain pre-processed reflection data;

[0030] Extracting geometric features from the preprocessed reflection data to obtain geometric features of the reflection data;

[0031] The geometric features of the reflection data are subjected to three-dimensional analysis to obtain three-dimensional point cloud data.

[0032] Optionally, filtering outliers on the reflection data to obtain pre-processed reflection data includes:

[0033] Performing standardization processing on the reflection data to obtain standardized reflection data;

[0034] Performing outlier detection on the standardized reflection data to obtain an outlier detection result;

[0035] Setting a filtering threshold according to the outlier detection result;

[0036] The standardized reflection data is subjected to outlier filtering according to the filtering threshold to obtain preprocessed reflection data.

[0037] Optionally, extracting geometric features from the pre-processed reflection data to obtain geometric features of the reflection data includes:

[0038] Performing feature point recognition on the pre-processed reflection data to obtain feature points;

[0039] Calculating local features around the feature points;

[0040] Shape analysis and structural analysis are performed on the local features to obtain geometric features of reflection data.

[0041] Optionally, the performing data calibration processing on the three-dimensional point cloud data and the image data to obtain comprehensive three-dimensional data includes:

[0042] Extracting geometric features of the three-dimensional point cloud data to obtain three-dimensional geometric features, and extracting geometric features of the image data to obtain image geometric features;

[0043] Performing feature matching on the three-dimensional geometric features and the image geometric features using a feature matching algorithm to obtain a matching result;

[0044] Performing preliminary registration on the three-dimensional point cloud data and the image data according to the matching result to obtain preliminary registration data;

[0045] The preliminary registration data is globally optimized by minimizing the registration error to obtain comprehensive three-dimensional data.

[0046] Optionally, the performing model accuracy assessment on the 3D image model includes:

[0047] Extracting plane geometric data of the 3D image model;

[0048] Calculating the standard deviation of the plane geometric data;

[0049] If the standard deviation is greater than or equal to a preset standard deviation threshold, the model accuracy assessment is determined to have failed;

[0050] If the standard deviation is less than a preset standard deviation threshold, the model accuracy assessment is determined to have passed.

[0051] In order to solve the above problems, the present invention also provides a computer-readable storage medium, in which at least one computer program is stored. The at least one computer program is executed by a processor in an electronic device to implement the above-mentioned underwater imaging method based on supercontinuum laser light source and binocular positioning.

[0052] The embodiment of the present invention uses a spectral imager pre-set underwater to obtain spectral range data of underwater imaging, uses a pre-set binocular camera to obtain the three-dimensional coordinates of the target object through binocular positioning, and uses the binocular camera to obtain image data of the target object, sets the output power of the supercontinuum laser according to the spectral range data, and improves the quality of the reflection data obtained subsequently; emits a supercontinuum laser to the target object according to the output power and the three-dimensional coordinates, uses the spectral imager to obtain reflection data of the supercontinuum laser reflected on the surface of the target object, performs point cloud analysis on the reflection data to obtain three-dimensional point cloud data, and improves the efficiency of subsequent model building; performs data registration processing on the three-dimensional point cloud data and the image data to obtain comprehensive three-dimensional data; establishes a 3D image model according to the comprehensive three-dimensional data, and performs model accuracy evaluation on the 3D image model; if the 3D image model fails the model accuracy evaluation, adjusts the exposure time of the spectral imager, and re-executes the above step of using the spectral imager to obtain the reflection data of the supercontinuum laser light source reflected on the surface of the target object; if the 3D image model passes the model accuracy evaluation, outputs the 3D image model, and ensures the accuracy of the 3D model. Therefore, the underwater imaging method and medium based on supercontinuum laser light source and binocular positioning proposed in the present invention can improve the accuracy of underwater imaging. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A schematic flow chart of an underwater imaging method based on a supercontinuum laser light source and dual-target positioning provided by one embodiment of the present invention;

[0054] Figure 2 A schematic diagram of a process for performing point cloud analysis on reflection data provided by an embodiment of the present invention;

[0055] Figure 3 A schematic diagram of a data registration process provided by an embodiment of the present invention.

[0056] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0057] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0058] The embodiment of the present application provides an underwater imaging method based on a supercontinuum laser light source and dual-target positioning. The execution subject of the underwater imaging method based on a supercontinuum laser light source and dual-target positioning includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided in the embodiment of the present application. In other words, the underwater imaging method based on a supercontinuum laser light source and dual-target positioning can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms.

[0059] Reference Figure 1 FIG. 1 is a flow chart of an underwater imaging method based on a supercontinuum laser light source and dual-target positioning provided by an embodiment of the present invention. In this embodiment, the underwater imaging method based on a supercontinuum laser light source and dual-target positioning includes:

[0060] S1. Acquire spectral range data of underwater imaging using a spectral imager pre-set underwater.

[0061] In the embodiment of the present invention, the spectral imager is an advanced optical instrument that can capture the reflection or emission spectrum information of an object or scene at different wavelengths. Unlike traditional color cameras, spectral imagers can provide richer information because they can measure and record spectral data in a wide range of wavelengths from ultraviolet to infrared.

[0062] In the embodiment of the present invention, the spectral imager includes a push-scanning spectral imager, an interference spectral imager, and a dispersion spectral imager;

[0063] In detail, the push-broom spectroscopic imager is a hyperspectral imaging device, and its working principle is to complete two-dimensional spatial scanning through a planar array detector perpendicular to the direction of movement during the forward movement of the flying platform. In this configuration, the spatial scanning direction is consistent with the direction of movement of the remote sensing platform (scanning along the track). The push-broom imaging spectrometer uses gratings and prisms to split light, parallel to the direction of platform movement, to complete the scanning of the spectral dimension.

[0064] In detail, the interference spectrum imager is an instrument that uses the interference principle to obtain a series of interference patterns that vary with the optical path difference, and obtains the two-dimensional spatial image and one-dimensional spectral information of the target object through inversion technology. Its working principle is to divide the target light into two beams, and by controlling the optical path difference of the two beams of light, make them meet and interfere at the photosensitive element, thereby obtaining interference patterns under different optical path differences.

[0065] In detail, the dispersion type spectral imager is an instrument for spectral imaging based on the principle of dispersion. Its core spectroscopic element usually adopts a prism or a grating. When the light emitted by the target object passes through the front imaging system, it will be imaged at the slit. Then, these lights pass through the spectroscopic effect of the dispersion element, and lights of different wavelengths will be dispersed in different directions. Subsequently, after passing through the post-imaging system, these lights will be imaged on the detector, thereby obtaining information on each spectral band of the target object.

[0066] In the embodiment of the present invention, the spectral range data refers to laser wavelength data that can achieve an expected reflection effect in the current underwater environment.

[0067] In detail, the method of obtaining the spectral range data of underwater imaging by using a spectral imager pre-set underwater includes:

[0068] Fire a supercontinuum laser at a target object pre-placed underwater;

[0069] Acquiring the reflected light of the supercontinuum laser by using the spectral imager, recording the capture frequencies of reflected light of different wavelengths, and obtaining spectral distribution data;

[0070] The spectrum range is extracted from the spectrum distribution data according to a preset frequency threshold to obtain spectrum range data.

[0071] In detail, the supercontinuum laser is a special light source that can produce an extremely wide spectral range, usually covering hundreds of nanometers or even thousands of nanometers. This light source is generated by transmitting ultrashort pulse lasers with high peak power in nonlinear media (such as special optical fibers). During the transmission process, due to the combined effects of various nonlinear effects such as self-phase modulation, soliton self-frequency shift, Raman four-wave mixing, modulation instability, and high-order harmonics, the spectrum of the laser is greatly broadened, thereby forming a continuous spectrum output.

[0072] In detail, the preset frequency threshold includes a minimum frequency threshold and a maximum frequency threshold for capturing frequencies of different wavelengths of reflected light, and the spectral range extraction of the spectral distribution data according to the preset frequency threshold refers to extracting data in the spectral distribution data that is between the minimum frequency threshold and the maximum frequency threshold to obtain spectral range data.

[0073] In the embodiment of the present invention, the spectral range data of underwater imaging is acquired by using a spectral imager pre-arranged underwater, thereby providing necessary data for subsequent analysis.

[0074] S2. Using a preset binocular camera to obtain the three-dimensional coordinates of the target object through binocular positioning, and using the binocular camera to obtain image data of the target object.

[0075] In an embodiment of the present invention, the binocular camera is a camera system based on the binocular vision principle, which consists of two cameras and an image processing unit. The two cameras are usually placed in parallel. By capturing and analyzing images from multiple angles at the same time, the depth information of objects in the scene can be obtained to achieve three-dimensional reconstruction and depth perception. The working principle of the binocular camera is mainly completed by calculating the parallax between the two cameras. The farther the object is from the camera, the more obvious the difference between the images taken by the left and right cameras is. By measuring this difference, the binocular camera can calculate information such as the distance and shape of the object.

[0076] In detail, the method of obtaining the three-dimensional coordinates of the target object by binocular positioning using a pre-set binocular camera includes:

[0077] Performing binocular positioning on the binocular camera to obtain internal and external parameters;

[0078] Performing epipolar correction on the binocular camera using the intrinsic reference;

[0079] The position of the target object is captured by the binocular camera after epipolar correction to obtain the matching point;

[0080] Disparity calculation and depth calculation are performed according to the matching points and the external parameters to obtain three-dimensional coordinates.

[0081] In detail, the internal parameters represent the focal length, focal coordinates, and lens distortion parameters of the binocular camera, and the external parameters represent the relative rotation matrix and translation vector of the binocular camera.

[0082] In detail, the disparity calculation is an important concept in stereo vision, which is the process of determining the depth information of objects in an image by comparing images of the same scene captured from different perspectives by two cameras in a binocular camera system.

[0083] In detail, the depth calculation is a key process in the stereo vision system, which refers to calculating the distance of each pixel or feature point in the scene relative to the camera based on the disparity information between two images with different perspectives captured by the binocular camera. Depth calculation enables us to recover three-dimensional spatial information from two-dimensional images, thereby performing three-dimensional reconstruction of the scene or other spatial analysis.

[0084] In an embodiment of the present invention, the step of acquiring image data of a target object by using the binocular camera includes:

[0085] Using the binocular camera to capture an image of a target object to obtain a target image;

[0086] Performing contrast enhancement processing on the target image to obtain a preprocessed image;

[0087] The pre-processed image is subjected to denoising processing to obtain image data.

[0088] In detail, the contrast enhancement processing is a technique in image processing that aims to improve the visual quality of an image by enhancing the contrast differences between different regions in the image, making the details of the image clearer and more prominent. This processing can be applied to the global or local regions of an image to improve the visibility of the image and the accuracy of the analysis. Common contrast enhancement processing methods include histogram equalization, contrast stretching, local contrast enhancement, and gamma correction.

[0089] In detail, the denoising process is a basic technique in image processing and signal processing, which aims to remove or reduce noise from an image or signal interfered with by noise to restore the original quality of the image or signal. Noise may be caused by a variety of factors, including sensor defects, environmental interference, transmission errors, etc. Denoising is essential for applications such as improving image quality, enhancing image features, and improving image analysis and recognition. Common denoising methods include local value filtering, Gaussian filtering, bilateral filtering, and non-local mean denoising.

[0090] In the embodiment of the present invention, the image quality is improved by performing contrast enhancement processing and noise removal processing on the image.

[0091] In the embodiment of the present invention, the three-dimensional coordinates of the target object are acquired through binocular positioning using a pre-set binocular camera, and the image data of the target object is acquired using the binocular camera, thereby providing necessary data for subsequent processing.

[0092] S3. Setting the output power of the supercontinuum laser according to the spectral range data, and emitting the supercontinuum laser to the target object according to the output power and the three-dimensional coordinates.

[0093] In the embodiment of the present invention, the output power is set by analyzing the linear relationship between the spectrum wavelength and the output power of the supercontinuum laser.

[0094] In the embodiment of the present invention, the device emitting supercontinuum laser is a supercontinuum laser. This laser can generate laser output that is continuously distributed in a wide spectral range, and has the characteristics of high power, wide spectral range, and good stability. A supercontinuum laser is usually composed of a pump source, a nonlinear medium, and a spectrum broadening element. The pump source provides high-power laser light, and after passing through the nonlinear medium, the laser light undergoes spectrum broadening, and finally generates supercontinuum laser output.

[0095] In the embodiment of the present invention, the output power of the supercontinuum laser needs to be set taking into account the required spectral range data to ensure that the laser can generate sufficient power within the required wavelength range. The spectral range data defines the wavelength range that the spectrometer can detect, and the output power of the supercontinuum laser determines the intensity and energy distribution of the laser.

[0096] In detail, a spectrometer can be used to monitor the spectral distribution and power output of the laser in real time. By comparing the actual output spectral range data, the parameters of the laser can be continuously adjusted until a satisfactory output power is achieved.

[0097] In the embodiment of the present invention, the supercontinuum laser is emitted to the target object according to the output power and the three-dimensional coordinates, and the output power needs to be set by adjusting the control parameters of the laser. During the adjustment process, the power meter can be used for detection. The supercontinuum laser beam is guided to the three-dimensional coordinate position of the target object by using the beam guiding system.

[0098] In detail, the beam guiding system is used to guide and focus the emitted laser beam onto the target object and adjust it according to the shape of the object. It usually includes components such as reflectors, lenses, optical fibers, etc., which work together to ensure that the laser beam can be accurately transmitted to the required working area along a predetermined path. By accurately adjusting the components of the beam guiding system, the laser beam can be accurately focused and positioned to meet various application requirements.

[0099] In an embodiment of the present invention, after a supercontinuum laser is emitted to a target object according to the output power and the three-dimensional coordinates, the supercontinuum laser beam is adjusted according to a preset path so that the supercontinuum laser can scan the entire target object, thereby ensuring that a subsequent spectral imager can obtain reflection data of the target object of the website.

[0100] In an embodiment of the present invention, by setting the output power of the supercontinuum laser according to the spectral range data, and emitting the supercontinuum laser to the target object according to the output power and the three-dimensional coordinates, the reflection effect of the supercontinuum laser is improved, and the effect of subsequent imaging is improved.

[0101] S4. Using the spectral imager to obtain reflection data of the supercontinuum laser reflected on the surface of the target object, performing point cloud analysis on the reflection data to obtain three-dimensional point cloud data.

[0102] In the embodiment of the present invention, before using the spectral imager to obtain reflection data of the supercontinuum laser reflected on the surface of the target object, it is necessary to calibrate the spectral imager according to the three-dimensional coordinate data to ensure that the spectral imager is aimed at the target object.

[0103] In the embodiment of the present invention, the point cloud analysis is a data processing method based on computer vision and computer graphics, and the object processed is a three-dimensional data set composed of a large number of points, namely a point cloud. Each point contains position information in space and possible other attributes, such as color, normal vector, texture coordinates and intensity value, which are used to describe the shape and properties of the point.

[0104] In the embodiment of the present invention, refer to Figure 2 As shown, a schematic diagram of a process of performing point cloud analysis on reflection data is provided in an embodiment of the present invention, wherein performing point cloud analysis on the reflection data to obtain three-dimensional point cloud data includes:

[0105] S21, filtering outliers on the reflection data to obtain pre-processed reflection data;

[0106] S22, extracting geometric features from the pre-processed reflection data to obtain geometric features of the reflection data;

[0107] S23, performing three-dimensional analysis on the geometric features of the reflection data to obtain three-dimensional point cloud data.

[0108] In detail, the outlier filtering process is a data processing technique used to identify and remove outliers in a data set. Outliers refer to data points that have extreme values ​​or significantly deviate from normal trends compared to other data points. The purpose of filtering outliers is to make the data more accurate and reliable, and to reduce the impact of outliers on data analysis results.

[0109] In detail, the three-dimensional analysis refers to the operation of converting the geometric features of the reflection data into point cloud data in three-dimensional space, by performing key point detection on the geometric features of the reflection data, constructing point cloud data based on the detected key points, and assigning attributes to each generated point.

[0110] In detail, the key point detection is to identify meaningful feature points in geometric features, and these feature points usually correspond to vertices in the geometric image.

[0111] In detail, the extracting geometric features of the pre-processed reflection data to obtain geometric features of the reflection data includes:

[0112] Performing feature point recognition on the pre-processed reflection data to obtain feature points;

[0113] Calculating local features around the feature points;

[0114] Shape analysis and structural analysis are performed on the local features to obtain geometric features of reflection data.

[0115] In detail, feature points are identified through curvature analysis, normal vector variation, or grid-based methods.

[0116] In detail, the shape analysis refers to capturing shape information using volume descriptors, mesh features or point cloud features.

[0117] In detail, the structural analysis refers to analyzing the topological structure of the data, such as connectivity, number of holes, boundary length, etc.

[0118] In detail, the filtering outliers on the reflection data to obtain pre-processed reflection data includes:

[0119] Performing standardization processing on the reflection data to obtain standardized reflection data;

[0120] Performing outlier detection on the standardized reflection data to obtain an outlier detection result;

[0121] Setting a filtering threshold according to the outlier detection result;

[0122] The standardized reflection data is subjected to outlier filtering according to the filtering threshold to obtain preprocessed reflection data.

[0123] In the embodiment of the present invention, the filtering threshold needs to be set according to specific task requirements.

[0124] In the embodiment of the present invention, the reflection data of the supercontinuum laser reflected on the surface of the target object is obtained by using the spectral imager, and point cloud analysis is performed on the reflection data to obtain three-dimensional point cloud data, thereby providing necessary data for subsequent model building.

[0125] S5. Perform data calibration processing on the three-dimensional point cloud data and the image data to obtain comprehensive three-dimensional data, and establish a 3D image model based on the comprehensive three-dimensional data.

[0126] In the embodiment of the present invention, refer to Figure 3 As shown, a schematic diagram of a data registration process is provided in an embodiment of the present invention, wherein the data registration process is performed on the three-dimensional point cloud data and the image data to obtain comprehensive three-dimensional data, including:

[0127] S31, extracting geometric features of the three-dimensional point cloud data to obtain three-dimensional geometric features, and extracting geometric features of the image data to obtain image geometric features;

[0128] S32, performing feature matching on the three-dimensional geometric features and the image geometric features using a feature matching algorithm to obtain a matching result;

[0129] S33, performing preliminary registration on the three-dimensional point cloud data and the image data according to the matching result to obtain preliminary registration data;

[0130] S34, globally optimizing the preliminary registration data by minimizing the registration error to obtain comprehensive three-dimensional data.

[0131] In detail, the steps of extracting the geometric features of the three-dimensional point cloud data have been described above and will not be repeated here.

[0132] In detail, the feature matching algorithms include SIFT, SURF and ORB.

[0133] In detail, the preliminary registration refers to estimating a preliminary transformation model based on the three-dimensional point cloud data, which describes how one data set is transformed into another data set. Commonly used methods include iterative closest point (ICP) algorithm based on feature points, affine transformation, projection transformation, etc.

[0134] In detail, minimizing the registration error is a key goal in the data registration process, which refers to reducing the difference or error between one dataset (such as an image, point cloud, 3D model, etc.) when aligning the two datasets. This difference can be a mismatch in geometry, position, size, etc. The purpose of minimizing the registration error is to find an optimal transformation model so that the features in one dataset match the corresponding features in the other dataset as closely as possible.

[0135] S6. Evaluate the model accuracy of the 3D image model.

[0136] In the embodiment of the present invention, the step of evaluating the model accuracy of the 3D image model includes:

[0137] Extracting plane geometric data of the 3D image model;

[0138] Calculating the standard deviation of the plane geometric data;

[0139] If the standard deviation is greater than or equal to a preset standard deviation threshold, the model accuracy assessment is determined to have failed;

[0140] If the standard deviation is less than a preset standard deviation threshold, the model accuracy assessment is determined to have passed.

[0141] In detail, the standard deviation is an important concept in statistics, which is used to measure the degree of dispersion of a data set. It represents the square root of the average of the squares of the differences between each value and the average value. The larger the standard deviation, the greater the dispersion of the data set, that is, the more dispersed the data points are; the smaller the standard deviation, the smaller the dispersion of the data set, that is, the more concentrated the data points are.

[0142] In detail, the calculating the standard deviation of the plane geometric data includes:

[0143] The standard deviation is calculated using the following formula:

[0144]

[0145] Wherein, σ is the standard deviation, N is the number of samples of the plane geometry data, and x i is the i-th data in the plane geometric data, and μ is the mean value of the plane geometric data.

[0146] In detail, the preset standard deviation threshold needs to be set according to the requirement for model accuracy. The higher the requirement for model accuracy, the lower the standard deviation threshold.

[0147] S7: Determine whether the 3D image model passes the model accuracy assessment.

[0148] If the 3D image model fails the model accuracy assessment, executing S8, adjusting the exposure time of the spectral imager.

[0149] In the embodiment of the present invention, by adjusting the exposure time of the spectral imager, the accuracy of the reflection data acquired by the spectral imager is improved, thereby improving the accuracy of the established 3D image model.

[0150] In the embodiment of the present invention, after the exposure time of the spectral imager is adjusted, S4 is re-executed to obtain reflection data of the supercontinuum laser reflected on the surface of the target object using the spectral imager, and point cloud analysis is performed on the reflection data to obtain three-dimensional point cloud data.

[0151] If the 3D image model passes the model accuracy evaluation, executing S9, outputting the 3D image model.

[0152] In the embodiment of the present invention, when the 3D image model passes the model accuracy evaluation, the current 3D image model is confirmed to be the final model, and the 3D image model is output.

[0153] The present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor of an electronic device, the computer program can implement:

[0154] Acquire spectral range data of underwater imaging using a spectral imager pre-set underwater;

[0155] Using a pre-set binocular camera to obtain the three-dimensional coordinates of the target object through binocular positioning, and using the binocular camera to obtain image data of the target object;

[0156] Setting the output power of the supercontinuum laser according to the spectral range data, and emitting the supercontinuum laser to the target object according to the output power and the three-dimensional coordinates;

[0157] Acquire reflection data of the supercontinuum laser reflected on the surface of the target object by using the spectral imager, perform point cloud analysis on the reflection data, and obtain three-dimensional point cloud data;

[0158] Performing data registration processing on the three-dimensional point cloud data and the image data to obtain comprehensive three-dimensional data, and establishing a 3D image model based on the comprehensive three-dimensional data;

[0159] Performing model accuracy evaluation on the 3D image model;

[0160] If the 3D image model fails to pass the model accuracy assessment, adjusting the exposure time of the spectral imager, and re-performing the step of using the spectral imager to obtain the reflection data of the supercontinuum laser light source reflected on the surface of the target object;

[0161] If the 3D image model passes the model accuracy assessment, the 3D image model is output.

[0162] In the embodiments provided by the present invention, it should be understood that the disclosed medium and method can be implemented in other ways. For example, the medium embodiments described above are only illustrative.

[0163] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0164] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0165] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0166] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any attached figure mark in the claims should not be regarded as limiting the claims involved.

[0167] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0168] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

[0169] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. An underwater imaging method based on supercontinuum laser light source and binocular positioning, characterized in that: The method comprises: Acquire spectral range data of underwater imaging using a spectral imager pre-set underwater; Using a pre-set binocular camera to obtain the three-dimensional coordinates of the target object through binocular positioning, and using the binocular camera to obtain image data of the target object; Setting the output power of the supercontinuum laser according to the spectral range data, and emitting the supercontinuum laser to the target object according to the output power and the three-dimensional coordinates; Acquire reflection data of the supercontinuum laser reflected on the surface of the target object by using the spectral imager, perform point cloud analysis on the reflection data, and obtain three-dimensional point cloud data; Performing data registration processing on the three-dimensional point cloud data and the image data to obtain comprehensive three-dimensional data, and establishing a 3D image model based on the comprehensive three-dimensional data; Performing model accuracy evaluation on the 3D image model; If the 3D image model fails to pass the model accuracy assessment, adjusting the exposure time of the spectral imager, and re-performing the step of using the spectral imager to obtain the reflection data of the supercontinuum laser light source reflected on the surface of the target object; If the 3D image model passes the model accuracy assessment, the 3D image model is output.

2. The underwater imaging method based on supercontinuum laser light source and binocular positioning according to claim 1, characterized in that: The method of obtaining the spectral range data of underwater imaging by using a spectral imager pre-set underwater includes: Fire a supercontinuum laser at a target object pre-placed underwater; Acquiring the reflected light of the supercontinuum laser by using the spectral imager, recording the capture frequencies of reflected light of different wavelengths, and obtaining spectral distribution data; The spectrum range is extracted from the spectrum distribution data according to a preset frequency threshold to obtain spectrum range data.

3. The underwater imaging method based on supercontinuum laser light source and binocular positioning according to claim 1, characterized in that: The method of obtaining the three-dimensional coordinates of the target object by binocular positioning using a preset binocular camera includes: Performing binocular positioning on the binocular camera to obtain internal and external parameters; Performing epipolar correction on the binocular camera using the intrinsic reference; The position of the target object is captured by the binocular camera after epipolar correction to obtain the matching point; Disparity calculation and depth calculation are performed according to the matching points and the external parameters to obtain three-dimensional coordinates.

4. The underwater imaging method based on supercontinuum laser light source and binocular positioning as claimed in claim 1, characterized in that: The method of acquiring image data of a target object by using the binocular camera includes: Using the binocular camera to capture an image of a target object to obtain a target image; Performing contrast enhancement processing on the target image to obtain a preprocessed image; The pre-processed image is subjected to denoising processing to obtain image data.

5. The underwater imaging method based on supercontinuum laser light source and binocular positioning as claimed in claim 1, characterized in that: The performing point cloud analysis on the reflection data to obtain three-dimensional point cloud data includes: Performing outlier filtering on the reflection data to obtain pre-processed reflection data; Extracting geometric features from the preprocessed reflection data to obtain geometric features of the reflection data; The geometric features of the reflection data are subjected to three-dimensional analysis to obtain three-dimensional point cloud data.

6. The underwater imaging method based on supercontinuum laser light source and binocular positioning as claimed in claim 5, characterized in that: The filtering outlier processing of the reflection data to obtain pre-processed reflection data includes: Performing standardization processing on the reflection data to obtain standardized reflection data; Performing outlier detection on the standardized reflection data to obtain an outlier detection result; Setting a filtering threshold according to the outlier detection result; The standardized reflection data is subjected to outlier filtering according to the filtering threshold to obtain preprocessed reflection data.

7. The underwater imaging method based on supercontinuum laser light source and binocular positioning as claimed in claim 5, characterized in that: The extracting geometric features of the pre-processed reflection data to obtain geometric features of the reflection data includes: Performing feature point recognition on the pre-processed reflection data to obtain feature points; Calculating local features around the feature points; Shape analysis and structural analysis are performed on the local features to obtain geometric features of reflection data.

8. The underwater imaging method based on supercontinuum laser light source and binocular positioning as claimed in claim 1, characterized in that: The performing data registration processing on the three-dimensional point cloud data and the image data to obtain comprehensive three-dimensional data includes: Extracting geometric features of the three-dimensional point cloud data to obtain three-dimensional geometric features, and extracting geometric features of the image data to obtain image geometric features; Performing feature matching on the three-dimensional geometric features and the image geometric features using a feature matching algorithm to obtain a matching result; Performing preliminary registration on the three-dimensional point cloud data and the image data according to the matching result to obtain preliminary registration data; The preliminary registration data is globally optimized by minimizing the registration error to obtain comprehensive three-dimensional data.

9. The underwater imaging method based on supercontinuum laser light source and binocular positioning according to claim 1, characterized in that: The performing model accuracy evaluation on the 3D image model comprises: Extracting plane geometric data of the 3D image model; Calculating the standard deviation of the plane geometric data; If the standard deviation is greater than or equal to a preset standard deviation threshold, the model accuracy assessment is determined to have failed; If the standard deviation is less than a preset standard deviation threshold, the model accuracy assessment is determined to have passed.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the underwater imaging method based on a supercontinuum laser light source and binocular positioning as described in any one of claims 1 to 9 is implemented.

Images