Artificial fish reef empty volume calculation method based on convex hull volume of point cloud data

Abstract

The invention relates to the technical field of ocean engineering measurement and calculation, in particular to an artificial fish reef empty volume calculation method based on the convex hull volume of point cloud data. The method comprises the following steps: obtaining original point cloud data, processing the original point cloud data, calculating a convex hull of the point cloud data, calculating the volume of a convex polyhedron, and calculating the space volume of a fish reef cluster. A convex hull algorithm is adopted, the measurement process is relatively simplified, and only one-time measurement needs to be completed on a fish reef sea area; the calculation process is simple and efficient, and is suitable for empty volume calculation of large-scale fish reef clusters; especially for a hollow frame type fish reef without surrounding, the accuracy of empty volume calculation of the fish reef can be obviously improved; the method is high in flexibility, point cloud data acquisition and processing parameters can be adjusted according to actual requirements, and the requirements of different application scenes are met; the situation that the calculation error of the empty volume of the frame fish reef body is huge due to a conventional volume calculation method is avoided, and the objectivity and repeatability of a calculation result are improved.

Description

Technical Field

[0001] This invention relates to the field of marine engineering measurement technology, specifically to a method for calculating the volume of empty space of artificial reefs based on the convex hull volume of point cloud data. Background Technology

[0002] Artificial reefs are important facilities for improving the marine ecological environment and promoting the recovery of fishery resources. Accurately estimating the volume of air in artificial reefs is of great significance for assessing their effectiveness, optimizing their design, and making reasonable arrangements. Currently, most methods for estimating the volume of artificial reefs refer to the requirements of relevant specifications. For example, the method for estimating the volume of artificial reefs based on multibeam bathymetry data provided by Chinese Patent Publication No. CN113325424A uses underwater acoustic equipment to scan and identify artificial reefs. Based on the scan results, for artificial reefs with regular shapes, the volume enclosed by their outer surface is calculated according to their geometric dimensions, which is their volume. For artificial reefs with irregular shapes, the volume can only be calculated as their volume by referring to existing methods for calculating the volume of the reef. Although the above methods are effective, they have problems with insufficient accuracy or complex calculations when dealing with reefs with complex shapes. In particular, for those hollow frame-type artificial reefs without surface coverage, the calculated volume is closer to their drainage volume, but much smaller than the volume enclosed by their outer surface (i.e., the volume of emptiness), resulting in a huge difference between the calculated volume of emptiness of this type of artificial reef and its actual volume. Currently, in the absence of more practical and effective methods for identifying and calculating the volume of artificial reefs, there is an urgent need to develop a more accurate and efficient method for calculating the volume of artificial reefs based on multibeam bathymetry data or other underwater sonar data. Summary of the Invention

[0003] The technical problem to be solved by this invention is to overcome the shortcomings of the prior art and provide a method for calculating the volume of artificial reefs based on the convex hull volume of point cloud data.

[0004] The technical solution adopted in this invention is as follows:

[0005] A method for calculating the void volume of artificial reefs based on the convex hull volume of point cloud data includes the following steps:

[0006] S1. Acquisition of raw point cloud data: Using marine surveying equipment, raw point cloud data of the spatial location of the artificial reef and its surrounding sea area are collected from all directions and from multiple angles according to the predetermined measurement frequency and scanning angle.

[0007] S2. Processing of raw point cloud data: The acquired raw point cloud data is preprocessed in sequence through sound velocity correction, water level correction, position correction, data filtering, error data removal, target object extraction, and output of preprocessed data, so as to separate and extract point cloud data that only represents the spatial position of the target reef.

[0008] S3. Calculation of the convex hull of the point cloud data: The convex hull of the target reef point cloud dataset is calculated using QickHull's convex hull algorithm.

[0009] S4. Calculation of the volume of a convex polyhedron: The volume of the convex polyhedron is calculated using a GPU parallel computing architecture through tetrahedral volume accumulation addition.

[0010] S5. Calculation of the volume of voids in the artificial reef group: The calculated convex hull volume is used as the volume of voids of the target artificial reef. The voids of multiple artificial reefs are summed to obtain the volume of voids of the entire artificial reef group.

[0011] This technical solution acquires and processes raw point cloud data, employing convex hull algorithms and volume calculation methods to accurately calculate the volume of space in artificial reefs and their clusters, thereby more precisely assessing the spatial size and resource carrying capacity of the reefs. Specifically, this solution uses marine surveying equipment to acquire omnidirectional, multi-angle point cloud data of the artificial reef and its surrounding waters based on point cloud data. Preprocessing eliminates various errors and interferences, extracting only the point cloud data representing the target reef. The convex hull algorithm transforms the target reef point cloud dataset into a convex polyhedron, which encloses all the point cloud data of the reef, thus approximating its shape. Utilizing a GPU parallel computing architecture, the volume of the convex polyhedron is calculated through tetrahedral volume accumulation addition. This convex hull volume is used as the volume of space in the target reef, and then accumulated to obtain the volume of space in the entire reef cluster. The principle lies in using point cloud data to accurately reflect the spatial morphology of the reef and quantifying the volume of space in the reef through convex hull algorithms and volume calculation methods.

[0012] In addition, the artificial reef void volume calculation method based on the convex hull volume of point cloud data proposed above according to the present invention may also have the following additional technical features:

[0013] According to one embodiment of the present invention, in step S1, the marine surveying equipment includes a high-precision multibeam sonar system, which acquires high-quality raw point cloud data by measuring at a predetermined frequency and scanning angle within a certain measurement time period; the acquired raw point cloud data includes the three-dimensional coordinate information of each point. , and intensity information.

[0014] In this technical solution, the multibeam sonar utilizes the time difference and angle information between sound wave transmission and reception, combined with three-dimensional coordinate transformation, to determine the spatial location of each point and record intensity information. Through multi-angle, high-frequency scanning, it acquires comprehensive, high-precision raw point cloud data, covering detailed spatial information of the artificial reef and surrounding sea area. The point cloud representing only the artificial reef is separated from the data, and the minimum convex polyhedron enclosing the artificial reef point cloud is constructed using the convex hull algorithm to approximate the shape of the artificial reef. Finally, its spatial size is quantified through volume calculation methods to achieve accurate calculation of the space volume of the artificial reef.

[0015] According to an embodiment of the present invention, in step S2, the preprocessing of the raw point cloud data includes the following steps:

[0016] S21, sound velocity correction, including ray tracing and VSP correction;

[0017] S22, Water level correction, including tide level synchronization and datum surface correction;

[0018] S23. Position correction, including attitude correction and installation parameter input;

[0019] S24. Data filtering, including statistical data filtering and terrain filtering;

[0020] S25. Error data removal, including outliers in 3D coordinate information and intensity information;

[0021] S26. Target extraction: effectively distinguish the point cloud data of the artificial reef from the scattered points on the surrounding seabed.

[0022] S27. Output preprocessed data, including preprocessed 3D coordinate information and intensity information.

[0023] In this technical solution, sound velocity correction utilizes ray tracing and VSP correction to account for the differences in sound wave propagation speed in different media, correcting positioning errors caused by changes in sound velocity; water level correction employs tidal synchronization and datum plane correction to unify measurement data to the same datum, eliminating the influence of water level changes on coordinates; position correction uses attitude correction and installation parameter input to correct errors caused by equipment installation deviations and attitude changes; data filtering uses statistical and terrain filtering to remove data affected by noise and abnormal terrain; erroneous data removal targets outliers in three-dimensional coordinates and intensity information to ensure data reliability; target extraction utilizes the characteristic differences between artificial reefs and seabed scattered points for differentiation, and finally outputs preprocessed data for convex hull calculation and volume calculation.

[0024] According to an embodiment of the present invention, in step S3, the convex hull of the target reef point cloud dataset is calculated using QickHull's convex hull algorithm, which includes the following steps:

[0025] S31. Obtain the three-dimensional spatial point set P representing the artificial reef:

[0026] (1)

[0027] In the formula: Let n be the three-dimensional coordinate information of any point n, where n≥1;

[0028] S32. Find the point set P in three-dimensional space that has the maximum and minimum values. The extreme points of the coordinates include the leftmost, rightmost, frontmost, backmost, topmost, and bottommost points; select 4 non-coplanar points from the extreme points to form an initial tetrahedron. If a tetrahedron cannot be formed, the point set is coplanar and degenerates into a two-dimensional convex hull.

[0029] S33. Assign the remaining points to the outer side of each face of the initial tetrahedron. Determine the position by calculating the signed volume of the point and the face: if the volume > 0, the point is on the outer side of the face; if the volume = 0, the point is on the face; if the volume < 0, the point is on the inner side of the face.

[0030] S34. For each face F of the tetrahedron, check if its outer side contains any unassigned points; for each face F, calculate the perpendicular distance Distance from all points on its outer side to that face:

[0031] (2)

[0032] In the formula: N is the normal vector of surface F; A and C are the two edge vectors of the tetrahedron;

[0033] Connect point C to the edge of face F to form a new face and construct a new tetrahedral structure; redistribute the external points of the original face F to the outside of the new face, recursively process the new face until there are no remaining points on the outside of all faces; all the retained faces form the minimum convex polyhedron containing all data points, merge coplanar or overlapping triangular patches, and optimize the convex hull structure.

[0034] S35. Decompose the convex polyhedron into multiple tetrahedra; for each tetrahedron Ti, calculate its volume V using the scalar triplet product. i :

[0035] (3)

[0036] In the formula: B is an edge vector of the tetrahedron;

[0037] The volume V of all tetrahedra is obtained by summing them up:

[0038] (4)

[0039] In this technical solution, the QickHull convex hull algorithm constructs an initial structure based on the extreme point characteristics of a point set in geometric space, and then obtains the minimum convex polyhedron through continuous expansion and optimization. Specifically, firstly, the extreme points of the three-dimensional point set are determined, and an initial tetrahedron is constructed using these extreme points to provide the basic framework. Then, the positions of the remaining points are determined based on the signed volumes of the points and faces, and the points are assigned to the outer sides of the corresponding faces. By calculating the perpendicular distance between the points and faces, the point with the farthest distance is found, and a new face is constructed using this point. This process is repeated recursively to gradually construct the minimum convex polyhedron containing all data points. Finally, the convex polyhedron is decomposed into multiple tetrahedrons, and the volume of each tetrahedron is calculated using the scalar triple product and then summed to obtain the total volume of the convex polyhedron. Through geometric relationships and recursive processing, the minimum convex polyhedron that can enclose the point set is gradually approximated and determined, thereby realizing the volume calculation.

[0040] According to an embodiment of the present invention, in step S1, the original point cloud data of the spatial location of the artificial reef and its surrounding sea area are acquired, the spatial location information of the underwater artificial reef is acquired using a high-precision multibeam sonar system, and through data processing, all point cloud data outside the isobaths are removed based on the isobath method of the outer edge of the artificial reef patch.

[0041] In this technical solution, the sonar emits multi-beam sound waves underwater. When the sound waves encounter different media and objects in the artificial reef and its surrounding sea area, they are reflected back. By measuring the time difference from the emission to the reception of the sound waves and the direction of the beam, combined with information such as the speed of sound, the spatial location of each point underwater can be determined, thereby obtaining the original point cloud data. The method of removing point cloud data outside the isobaths based on the outer edge of the artificial reef patch utilizes the fact that isobaths can reflect underwater topographic features. The area where the artificial reef is located has its specific isobath range. Point cloud data outside this range has little correlation with the reef. Removing it can reduce invalid data interference and more accurately focus on the data related to the artificial reef.

[0042] According to an embodiment of the present invention, in step S32, four non-coplanar points are selected from the extreme points to form an initial tetrahedron. To address the uneven distribution of point cloud data, a dynamic weighting mechanism is introduced to optimize the initial tetrahedron construction process, including the following steps:

[0043] S321, Calculation of points in the point set With extreme points The Euclidean distance is used to determine the weighting coefficients based on the distance distribution. Select the four non-coplanar points with the highest weights as the vertices of the initial tetrahedron, and assign weight coefficients to them. The formula is:

[0044]

[0045] When recursively generating new surfaces, unassigned points with weights higher than the threshold β are processed first, and β is dynamically determined by the point cloud density distribution.

[0046] This technical solution introduces a dynamic weighting mechanism based on the uneven distribution of point cloud data. Euclidean distance reflects the spatial proximity of each point to extreme points; different distances indicate varying importance or representativeness of points in spatial distribution. Weighting coefficients are determined by calculating Euclidean distance, and points with higher weights are prioritized when constructing the initial tetrahedron, making the initial tetrahedron more representative of the overall distribution characteristics of the point set. When recursively generating new faces, unassigned points with weights higher than a threshold are processed first. This threshold is dynamically determined by the point cloud density distribution to ensure priority processing of key points in dense point cloud regions. This makes the convex hull construction process more adaptable to unevenly distributed point cloud data, improving the accuracy of convex hull calculation.

[0047] According to an embodiment of the present invention, in step S35, the volume V of the tetrahedron is implemented using a GPU parallel computing architecture, including the following steps:

[0048] S351. Divide the point cloud data into multiple subsets and allocate them to different thread blocks of the GPU;

[0049] S352. Each thread block independently calculates the convex polyhedron corresponding to its subset and decomposes it into tetrahedrons;

[0050] S353. Accumulate the tetrahedral volume of all thread blocks to the host side through a global memory synchronization mechanism;

[0051] S354. Perform redundant data verification on the calculation results. If the error exceeds η%, then redistribute the calculation tasks.

[0052] This technical solution achieves parallelization of computational tasks by dividing point cloud data into multiple subsets and allocating them to different thread blocks of the GPU; it simplifies complex problems into local computations by having each thread block independently compute convex polyhedra and decompose them into tetrahedra; it ensures data consistency by accumulating results through a global memory synchronization mechanism; and it finally implements redundant data verification and dynamic task allocation.

[0053] According to an embodiment of the present invention, in step S4, the volume of the convex polyhedron is calculated using a GPU parallel computing architecture, including the following steps:

[0054] S41. Based on the spatial distribution characteristics of point cloud data, use a spatial octree or KD tree to divide the data into blocks;

[0055] S42. By monitoring the execution time of each thread block of the GPU, the data allocation strategy is dynamically adjusted.

[0056] S43. Under the premise of ensuring calculation accuracy, half-precision floating-point numbers are used to accelerate some intermediate calculation processes, and the final volume accumulation time is restored to double precision to reduce accumulated error.

[0057] S44. The three stages of convex hull decomposition, tetrahedral volume calculation and result accumulation are decoupled into independent computing units, and pipeline parallelization is achieved through asynchronous memory copying and computational overlap.

[0058] This technical solution achieves rational data organization by using spatial octrees or KD trees to divide point cloud data into blocks based on the spatial distribution characteristics of the data; it dynamically adjusts the data allocation strategy and optimizes the utilization of computing resources by monitoring the execution time of each thread block of the GPU; while ensuring computational accuracy, it accelerates some intermediate calculation processes with half-precision floating-point numbers, and finally restores double precision to the volume accumulation and addition to reduce accumulated errors, thus balancing computational speed and accuracy; it decouples the three stages of convex hull decomposition, tetrahedral volume calculation, and result accumulation into independent computing units, and achieves pipeline parallelization through asynchronous memory copying and computational overlap, thereby improving computational efficiency.

[0059] According to an embodiment of the present invention, the calculation of the empty space of the artificial reef group in step S5 includes the following steps:

[0060] S51. The preprocessed point cloud data is segmented using a clustering algorithm to identify and separate point sets of different artificial reefs;

[0061] S52. Perform Boolean operations on the convex hulls of adjacent artificial reefs to eliminate the problem of repeated volume calculations caused by overlapping point cloud data.

[0062] S53. Introduce the reef material density parameter ρ and the seawater density parameter ρ_water, and calculate the buoyancy correction ΔV using Archimedes' principle;

[0063] S54. Based on the noise level of point cloud data and the error model of convex hull algorithm, output the confidence interval for calculating the empty square.

[0064] This technical solution uses a clustering algorithm to segment the preprocessed point cloud data, identifying and separating point sets of different artificial reefs. By performing Boolean operations on the convex hulls of adjacent artificial reefs, it eliminates the problem of repeated volume calculations caused by overlapping point cloud data, ensuring calculation accuracy. It introduces the reef material density parameter ρ and the seawater density parameter ρ_water, and calculates the buoyancy correction ΔV using Archimedes' principle, performing physical-level calibration of the calculation results. Based on the point cloud data noise level and the convex hull algorithm error model, it outputs the confidence interval for the empty volume calculation, reflecting the reliability of the calculation results from a statistical perspective.

[0065] According to one embodiment of the present invention, the method further includes the following steps:

[0066] S6. Visualization and Verification:

[0067] S61. The calculated convex hull model and void volume results are rendered through a 3D visualization platform, supporting interactive rotation, scaling and sectioning analysis.

[0068] S62. Generates standardized reports containing empty volume, calculation error, device parameters, and processing logs, and supports exporting to PDF or JSON format.

[0069] This technical solution renders the calculated convex hull model and void volume results through a 3D visualization platform, supporting interactive rotation, scaling, and sectioning analysis. This provides users with an intuitive and interactive way to observe and understand the calculation results in depth. By generating standardized reports containing void volume, calculation errors, equipment parameters, and processing logs, it systematically integrates key information for easy management and traceability. Furthermore, by supporting export to PDF or JSON format, it adapts to the data usage needs of different scenarios.

[0070] Compared with the prior art, the present invention has the following advantages:

[0071] Employing the convex hull algorithm simplifies the measurement process, requiring only a single measurement of the artificial reef area. The calculation process is simple and efficient, suitable for calculating the volumetric volume of large-scale artificial reef systems. Particularly for hollow-framed artificial reefs lacking surrounding structures, it significantly improves the accuracy of volumetric volume calculations. The method is highly flexible, allowing adjustment of point cloud data acquisition and processing parameters to meet the needs of different application scenarios. It avoids the significant errors in calculating the volumetric volume of framed artificial reefs caused by conventional volumetric calculation methods, thus improving the objectivity and repeatability of the calculation results. Attached Figure Description

[0072] Figure 1 This is a flowchart illustrating the principle of the method of the present invention.

[0073] Figure 2 This is the original point cloud measured data diagram of the present invention. Detailed Implementation

[0074] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0075] Example 1

[0076] like Figure 1As shown in the figure, this embodiment provides a method for calculating the void volume of artificial reefs based on the convex hull volume of point cloud data, including the following steps:

[0077] S1. Acquisition of Raw Point Cloud Data: Using marine surveying equipment, according to a predetermined measurement frequency and scanning angle, comprehensive and multi-angle acquisition of raw point cloud data of the artificial reef and its surrounding sea area is conducted. Figure 2 As shown;

[0078] S2. Processing of raw point cloud data: The acquired raw point cloud data is preprocessed in sequence through sound velocity correction, water level correction, position correction, data filtering, error data removal, target object extraction, and output of preprocessed data, so as to separate and extract point cloud data that only represents the spatial position of the target reef.

[0079] S3. Calculation of the convex hull of the point cloud data: The convex hull of the target reef point cloud dataset is calculated using QickHull's convex hull algorithm.

[0080] S4. Calculation of the volume of a convex polyhedron: The volume of the convex polyhedron is calculated using a GPU parallel computing architecture through tetrahedral volume accumulation addition.

[0081] S5. Calculation of the volume of voids in the artificial reef group: The calculated convex hull volume is used as the volume of voids of the target artificial reef. The voids of multiple artificial reefs are summed to obtain the volume of voids of the entire artificial reef group.

[0082] This technical solution acquires and processes raw point cloud data, employing convex hull algorithms and volume calculation methods to accurately calculate the volume of space in artificial reefs and their clusters, thereby more precisely assessing the spatial size and resource carrying capacity of the reefs. Specifically, this solution uses marine surveying equipment to acquire omnidirectional, multi-angle point cloud data of the artificial reef and its surrounding waters based on point cloud data. Preprocessing eliminates various errors and interferences, extracting only the point cloud data representing the target reef. The convex hull algorithm transforms the target reef point cloud dataset into a convex polyhedron, which encloses all the point cloud data of the reef, thus approximating its shape. Utilizing a GPU parallel computing architecture, the volume of the convex polyhedron is calculated through tetrahedral volume accumulation addition. This convex hull volume is used as the volume of space in the target reef, and then accumulated to obtain the volume of space in the entire reef cluster. The principle lies in using point cloud data to accurately reflect the spatial morphology of the reef and quantifying the volume of space in the reef through convex hull algorithms and volume calculation methods.

[0083] In addition, the artificial reef void volume calculation method based on the convex hull volume of point cloud data proposed above according to the present invention may also have the following additional technical features:

[0084] According to one embodiment of the present invention, in step S1, the marine surveying equipment includes a high-precision multibeam sonar system, which acquires high-quality raw point cloud data by measuring at a predetermined frequency and scanning angle within a certain measurement time period; the acquired raw point cloud data includes the three-dimensional coordinate information of each point. , and intensity information.

[0085] In this technical solution, the multibeam sonar utilizes the time difference and angle information between sound wave transmission and reception, combined with three-dimensional coordinate transformation, to determine the spatial location of each point and record intensity information. Through multi-angle, high-frequency scanning, it acquires comprehensive, high-precision raw point cloud data, covering detailed spatial information of the artificial reef and surrounding sea area. The point cloud representing only the artificial reef is separated from the data, and the minimum convex polyhedron enclosing the artificial reef point cloud is constructed using the convex hull algorithm to approximate the shape of the artificial reef. Finally, its spatial size is quantified through volume calculation methods to achieve accurate calculation of the space volume of the artificial reef.

[0086] According to an embodiment of the present invention, in step S2, the preprocessing of the raw point cloud data includes the following steps:

[0087] S21, sound velocity correction, including ray tracing and VSP correction;

[0088] S22, Water level correction, including tide level synchronization and datum surface correction;

[0089] S23. Position correction, including attitude correction and installation parameter input;

[0090] S24. Data filtering, including statistical data filtering and terrain filtering;

[0091] S25. Error data removal, including outliers in 3D coordinate information and intensity information;

[0092] S26. Target extraction: effectively distinguish the point cloud data of the artificial reef from the scattered points on the surrounding seabed.

[0093] S27. Output preprocessed data, including preprocessed 3D coordinate information and intensity information.

[0094] In this technical solution, sound velocity correction utilizes ray tracing and VSP correction to account for the differences in sound wave propagation speed in different media, correcting positioning errors caused by changes in sound velocity; water level correction employs tidal synchronization and datum plane correction to unify measurement data to the same datum, eliminating the influence of water level changes on coordinates; position correction uses attitude correction and installation parameter input to correct errors caused by equipment installation deviations and attitude changes; data filtering uses statistical and terrain filtering to remove data affected by noise and abnormal terrain; erroneous data removal targets outliers in three-dimensional coordinates and intensity information to ensure data reliability; target extraction utilizes the characteristic differences between artificial reefs and seabed scattered points for differentiation, and finally outputs preprocessed data for convex hull calculation and volume calculation.

[0095] According to an embodiment of the present invention, in step S3, the convex hull of the target reef point cloud dataset is calculated using QickHull's convex hull algorithm, which includes the following steps:

[0096] S31. Obtain the three-dimensional spatial point set P representing the artificial reef:

[0097] (1)

[0098] In the formula: Let n be the three-dimensional coordinate information of any point n, where n≥1;

[0099] S32. Find the point set P in three-dimensional space that has the maximum and minimum values. The extreme points of the coordinates include the leftmost, rightmost, frontmost, backmost, topmost, and bottommost points; select 4 non-coplanar points from the extreme points to form an initial tetrahedron. If a tetrahedron cannot be formed, the point set is coplanar and degenerates into a two-dimensional convex hull.

[0100] S33. Assign the remaining points to the outer side of each face of the initial tetrahedron. Determine the position by calculating the signed volume of the point and the face: if the volume > 0, the point is on the outer side of the face; if the volume = 0, the point is on the face; if the volume < 0, the point is on the inner side of the face.

[0101] S34. For each face F of the tetrahedron, check if its outer side contains any unassigned points; for each face F, calculate the perpendicular distance Distance from all points on its outer side to that face:

[0102] (2)

[0103] In the formula: N is the normal vector of surface F; A and C are the two edge vectors of the tetrahedron;

[0104] Connect point C to the edge of face F to form a new face and construct a new tetrahedral structure; redistribute the external points of the original face F to the outside of the new face, recursively process the new face until there are no remaining points on the outside of all faces; all the retained faces form the minimum convex polyhedron containing all data points, merge coplanar or overlapping triangular patches, and optimize the convex hull structure.

[0105] S35. Decompose the convex polyhedron into multiple tetrahedra; for each tetrahedron Ti, calculate its volume V using the scalar triplet product. i :

[0106] (3)

[0107] In the formula: B is an edge vector of the tetrahedron;

[0108] The volume V of all tetrahedra is obtained by summing them up:

[0109] (4)

[0110] In this technical solution, the QickHull convex hull algorithm constructs an initial structure based on the extreme point characteristics of a point set in geometric space, and then obtains the minimum convex polyhedron through continuous expansion and optimization. Specifically, firstly, the extreme points of the three-dimensional point set are determined, and an initial tetrahedron is constructed using these extreme points to provide the basic framework. Then, the positions of the remaining points are determined based on the signed volumes of the points and faces, and the points are assigned to the outer sides of the corresponding faces. By calculating the perpendicular distance between the points and faces, the point with the farthest distance is found, and a new face is constructed using this point. This process is repeated recursively to gradually construct the minimum convex polyhedron containing all data points. Finally, the convex polyhedron is decomposed into multiple tetrahedrons, and the volume of each tetrahedron is calculated using the scalar triple product and then summed to obtain the total volume of the convex polyhedron. Through geometric relationships and recursive processing, the minimum convex polyhedron that can enclose the point set is gradually approximated and determined, thereby realizing the volume calculation.

[0111] According to an embodiment of the present invention, in step S1, the original point cloud data of the spatial location of the artificial reef and its surrounding sea area are acquired, the spatial location information of the underwater artificial reef is acquired using a high-precision multibeam sonar system, and through data processing, all point cloud data outside the isobaths are removed based on the isobath method of the outer edge of the artificial reef patch.

[0112] In this technical solution, the sonar emits multi-beam sound waves underwater. When the sound waves encounter different media and objects in the artificial reef and its surrounding sea area, they are reflected back. By measuring the time difference from the emission to the reception of the sound waves and the direction of the beam, combined with information such as the speed of sound, the spatial location of each point underwater can be determined, thereby obtaining the original point cloud data. The method of removing point cloud data outside the isobaths based on the outer edge of the artificial reef patch utilizes the fact that isobaths can reflect underwater topographic features. The area where the artificial reef is located has its specific isobath range. Point cloud data outside this range has little correlation with the reef. Removing it can reduce invalid data interference and more accurately focus on the data related to the artificial reef.

[0113] According to an embodiment of the present invention, in step S32, four non-coplanar points are selected from the extreme points to form an initial tetrahedron. To address the uneven distribution of point cloud data, a dynamic weighting mechanism is introduced to optimize the initial tetrahedron construction process, including the following steps:

[0114] S321, Calculation of points in the point set With extreme points The Euclidean distance is used to determine the weighting coefficients based on the distance distribution. Select the four non-coplanar points with the highest weights as the vertices of the initial tetrahedron, and assign weight coefficients to them. The formula is:

[0115]

[0116] When recursively generating new surfaces, unassigned points with weights higher than the threshold β are processed first, and β is dynamically determined by the point cloud density distribution.

[0117] This technical solution introduces a dynamic weighting mechanism based on the uneven distribution of point cloud data. Euclidean distance reflects the spatial proximity of each point to extreme points; different distances indicate varying importance or representativeness of points in spatial distribution. Weighting coefficients are determined by calculating Euclidean distance, and points with higher weights are prioritized when constructing the initial tetrahedron, making the initial tetrahedron more representative of the overall distribution characteristics of the point set. When recursively generating new faces, unassigned points with weights higher than a threshold are processed first. This threshold is dynamically determined by the point cloud density distribution to ensure priority processing of key points in dense point cloud regions. This makes the convex hull construction process more adaptable to unevenly distributed point cloud data, improving the accuracy of convex hull calculation.

[0118] According to an embodiment of the present invention, in step S35, the volume V of the tetrahedron is implemented using a GPU parallel computing architecture, including the following steps:

[0119] S351. Divide the point cloud data into multiple subsets and allocate them to different thread blocks of the GPU;

[0120] S352. Each thread block independently calculates the convex polyhedron corresponding to its subset and decomposes it into tetrahedrons;

[0121] S353. Accumulate the tetrahedral volume of all thread blocks to the host side through a global memory synchronization mechanism;

[0122] S354. Perform redundant data verification on the calculation results. If the error exceeds η%, then redistribute the calculation tasks.

[0123] This technical solution achieves parallelization of computational tasks by dividing point cloud data into multiple subsets and allocating them to different thread blocks of the GPU; it simplifies complex problems into local computations by having each thread block independently compute convex polyhedra and decompose them into tetrahedra; it ensures data consistency by accumulating results through a global memory synchronization mechanism; and it finally implements redundant data verification and dynamic task allocation.

[0124] According to an embodiment of the present invention, in step S4, the volume of the convex polyhedron is calculated using a GPU parallel computing architecture, including the following steps:

[0125] S41. Based on the spatial distribution characteristics of point cloud data, use a spatial octree or KD tree to divide the data into blocks;

[0126] S42. By monitoring the execution time of each thread block of the GPU, the data allocation strategy is dynamically adjusted.

[0127] S43. Under the premise of ensuring calculation accuracy, half-precision floating-point numbers are used to accelerate some intermediate calculation processes, and the final volume accumulation time is restored to double precision to reduce accumulated error.

[0128] S44. The three stages of convex hull decomposition, tetrahedral volume calculation and result accumulation are decoupled into independent computing units, and pipeline parallelization is achieved through asynchronous memory copying and computational overlap.

[0129] This technical solution achieves rational data organization by using spatial octrees or KD trees to divide point cloud data into blocks based on the spatial distribution characteristics of the data; it dynamically adjusts the data allocation strategy and optimizes the utilization of computing resources by monitoring the execution time of each thread block of the GPU; while ensuring computational accuracy, it accelerates some intermediate calculation processes with half-precision floating-point numbers, and finally restores double precision to the volume accumulation and addition to reduce accumulated errors, thus balancing computational speed and accuracy; it decouples the three stages of convex hull decomposition, tetrahedral volume calculation, and result accumulation into independent computing units, and achieves pipeline parallelization through asynchronous memory copying and computational overlap, thereby improving computational efficiency.

[0130] According to an embodiment of the present invention, the calculation of the empty space of the artificial reef group in step S5 includes the following steps:

[0131] S51. The preprocessed point cloud data is segmented using a clustering algorithm to identify and separate point sets of different artificial reefs;

[0132] S52. Perform Boolean operations on the convex hulls of adjacent artificial reefs to eliminate the problem of repeated volume calculations caused by overlapping point cloud data.

[0133] S53. Introduce the reef material density parameter ρ and the seawater density parameter ρ_water, and calculate the buoyancy correction ΔV using Archimedes' principle;

[0134] S54. Based on the noise level of point cloud data and the error model of convex hull algorithm, output the confidence interval for calculating the empty square.

[0135] This technical solution uses a clustering algorithm to segment the preprocessed point cloud data, identifying and separating point sets of different artificial reefs. By performing Boolean operations on the convex hulls of adjacent artificial reefs, it eliminates the problem of repeated volume calculations caused by overlapping point cloud data, ensuring calculation accuracy. It introduces the reef material density parameter ρ and the seawater density parameter ρ_water, and calculates the buoyancy correction ΔV using Archimedes' principle, performing physical-level calibration of the calculation results. Based on the point cloud data noise level and the convex hull algorithm error model, it outputs the confidence interval for the empty volume calculation, reflecting the reliability of the calculation results from a statistical perspective.

[0136] According to one embodiment of the present invention, the method further includes the following steps:

[0137] S6. Visualization and Verification:

[0138] S61. The calculated convex hull model and void volume results are rendered through a 3D visualization platform, supporting interactive rotation, scaling and sectioning analysis.

[0139] S62. Generates standardized reports containing empty volume, calculation error, device parameters, and processing logs, and supports exporting to PDF or JSON format.

[0140] This technical solution renders the calculated convex hull model and void volume results through a 3D visualization platform, supporting interactive rotation, scaling, and sectioning analysis. This provides users with an intuitive and interactive way to observe and understand the calculation results in depth. By generating standardized reports containing void volume, calculation errors, equipment parameters, and processing logs, it systematically integrates key information for easy management and traceability. Furthermore, by supporting export to PDF or JSON format, it adapts to the data usage needs of different scenarios.

[0141] Example 2

[0142] Based on Example 1, such as Figure 1 and Figure 2As shown, this embodiment provides a method for calculating the spatial volume of artificial reefs based on the convex hull volume of point cloud data, including the following steps: using a multibeam echo sounder system to acquire raw point cloud data of the spatial location of the artificial reef and its surrounding sea area; preprocessing the acquired raw point cloud data to separate and extract point cloud data that only represents the spatial location of the target reef; selecting a high-resolution multibeam echo sounder system with a beam angle of not less than 0.5°×1.0°, such as the Kongsberg EM 2040C high-resolution multibeam system, with no less than 512 beams per ping; equipping an ApplanixPOS MV OceanMaster attitude sensor to assist in measuring attitude data, with a measurement accuracy better than 0.02°, an azimuth measurement accuracy better than 0.015°, and a positioning accuracy better than 0.1m. The entire scanning system was installed and statically calibrated strictly according to measurement specifications and the installation and calibration procedures recommended by the equipment manufacturer. Static calibration used the nine-bar method to calibrate roll / pitch deviations. Dynamic calibration involved setting up 12 acoustic control points spaced 50m apart in the survey area, and the calibration accuracy was verified through a patch test. Sound velocity profile data was collected in the survey area and input into the multibeam echo sounder, followed by dynamic system calibration. A full-coverage scan of the artificial reef area was conducted using the multibeam echo sounder, with the beam fan angle controlled to be less than 140° and the overlap of adjacent survey lines greater than 20%, acquiring spatial point cloud data of the artificial reef and its surrounding sea area. Attitude and navigation positioning information was recorded simultaneously. The convex hull of the target reef point cloud dataset was calculated; the volume of the convex polyhedron was calculated using tetrahedral volume accumulation; the calculated convex hull volume was used as the volume of the target reef's void volume, and the void volumes of multiple reefs were accumulated to obtain the void volume of the entire reef group.

[0143] In this embodiment, during the verification process, a cubic concrete reef with a side length of 3.5m and a theoretical volume of 42.875m³ was tested. The measured data are as follows:

[0144] Point cloud data: 128,750 points;

[0145] The calculated volume of the convex hull is 43.12 m³.

[0146] Calculation time: 2.7 seconds (CPU mode) / 0.12 seconds (GPU mode);

[0147] In a measured unit artificial reef consisting of 84 individual reefs with a total empty volume of 18,144 m³, and a steel frame frame, the measured data are as follows:

[0148] Total point cloud count: 4,275,890 points;

[0149] The traditional method calculates the volume of empty space per unit artificial reef as 12191 m³·empty.

[0150] The present invention calculates the volume of the empty space of the artificial reef as: 18945 m³·empty.

[0151] Total error: 32.8% for the traditional method; 4.4% for this invention;

[0152] The technical advantages of the measurement results obtained by the present invention compared with those of conventional methods are detailed in Table 1.

[0153] Table 1: Comparison of measurement results between the present invention and conventional methods

[0154]

[0155] As can be seen from the above examples, the technical solution of the present invention solves the shortcomings of the prior art in estimating the volume of air volume of irregular and complex artificial reefs, and can improve the calculation accuracy and efficiency.

[0156] This invention acquires 3D point cloud data of artificial reefs and surrounding waters using a high-resolution multibeam echo sounder. After preprocessing including tidal level correction, sound velocity correction, and noise filtering, a high-precision point cloud set characterizing the reef morphology is separated. An improved QickHull algorithm is used to construct the minimum convex hull polyhedron of the point cloud. The convex hull is decomposed into tetrahedral elements, and their volumes are accumulated through triple integration. Finally, the spatial volume of the convex hull represents the void volume of the artificial reef, achieving lossless quantitative calculation of complex reef structures. Its core lies in overcoming the dependence of traditional measurement methods on regular geometric assumptions through point cloud spatial envelope reconstruction and voxelization calculation, providing accurate 3D spatial measurement basis for marine ecological engineering. Although the invention has been described in detail with reference to the accompanying drawings and preferred embodiments, the invention is not limited thereto. Various equivalent modifications or substitutions can be made to the embodiments of the invention by those skilled in the art without departing from the spirit and essence of the invention, and such modifications or substitutions should be within the scope of the invention. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the invention should be covered within the protection scope of the invention. Therefore, the scope of protection of this invention should be determined by the scope of the claims.

Claims

1. A method for calculating the volume of an artificial reef based on the convex hull volume of point cloud data, characterized in that, It comprises the following steps: S1, acquisition of original point cloud data: using marine measurement equipment, according to the predetermined measurement frequency and scanning angle, the spatial position of artificial reef and its surrounding sea area is acquired by all-around and multi-angle data collection; S2, processing of original point cloud data: the acquired original point cloud data is preprocessed, and the point cloud data representing only the spatial position of the target artificial reef is extracted by sequentially passing through sound velocity correction, water level correction, position correction, data filtering, error data elimination, target object extraction and output of preprocessed data; S3, calculation of point cloud data convex hull: the convex hull algorithm of QickHull is used to calculate the convex hull of the target artificial reef point cloud data set; S4, calculation of convex polyhedron volume: the volume of the convex polyhedron is calculated by the tetrahedron volume accumulation method based on GPU parallel computing architecture; S5, calculation of artificial reef group volume: the calculated convex hull volume is taken as the volume of the target artificial reef, and the volumes of multiple artificial reefs are accumulated to obtain the volume of the entire artificial reef group. 2.The method of claim 1, wherein, In the S1, the marine measuring device includes a high-precision multi-beam sonar system, through in a certain measuring time period, according to the predetermined measuring frequency and scanning angle, to obtain high-quality original point cloud data; the obtained original point cloud data includes three-dimensional coordinate information of each point , and intensity information. 3.The method of claim 2, wherein, In S2, the preprocessing of the original point cloud data comprises the following steps: S21, sound velocity correction, including ray tracing and VSP correction; S22, water level correction, including tide level synchronization and datum surface correction; S23, position correction, including attitude correction and installation parameter input; S24, data filtering, including statistical data filtering and terrain filtering; S25, error data elimination, including three-dimensional coordinate information and abnormal value of intensity information; S26, target object extraction, effectively distinguishing the point cloud data of artificial reef from the surrounding seabed scatter points; S27, output of preprocessed data, including three-dimensional coordinate information and intensity information after preprocessing. 4.The method of claim 2, wherein, In S3, the convex hull of the target artificial reef point cloud data set is calculated by using the convex hull algorithm of QickHull, which comprises the following steps: S31, obtaining a three-dimensional space point set P representing an artificial reef: （1） In the formula: is the three-dimensional coordinate information of any point n, n≥1; S32. Find the point set P in three-dimensional space that has the maximum and minimum values. The extreme points of the coordinates include the leftmost, rightmost, frontmost, backmost, topmost, and bottommost points; select 4 non-coplanar points from the extreme points to form an initial tetrahedron. If a tetrahedron cannot be formed, the point set is coplanar and degenerates into a two-dimensional convex hull. S33, distributing the remaining points to the outside of each face of the initial tetrahedron, and judging the position by calculating the signed volume of the point and the face: if the volume > 0, the point is outside the face; if the volume = 0, the point is on the face; if the volume < 0, the point is inside the face; S34, checking whether the outside of each face F of the tetrahedron contains unassigned points; for each face F, calculating the perpendicular distance Distance of all points on the outside of the face to the face: （2） Wherein: N is the normal vector of face F; A and C are two edge vectors of the tetrahedron; Connecting point C and the edge of face F to form a new face and construct a new tetrahedron structure; reassigning the outside points of the original face F to the outside of the new face, recursively processing the new face until there are no remaining points on the outside of all faces; all retained faces constitute a minimum convex polyhedron containing all data points, and the coplanar or overlapping triangular facets are merged to optimize the convex hull structure; S35, decompose the convex polyhedron into a plurality of tetrahedrons; for each tetrahedron Ti, calculate its volume V using scalar triple product i : （3） Wherein: B is an edge vector of the tetrahedron; The volume V of all tetrahedrons is accumulated: （4）。 5.The method of claim 4, wherein, In the S1, the spatial position original point cloud data of the artificial fish reef and the surrounding sea area is acquired, the spatial position information of the underwater artificial fish reef is acquired by using a high-precision multi-beam sonar system, and through data processing, based on the artificial fish reef patch outer edge contour method, all point cloud data outside the contour is removed. 6.The method of claim 4, wherein, In the S32, four non-coplanar points are selected from the extreme points to form an initial tetrahedron, and a dynamic weighting mechanism is introduced to optimize the initial tetrahedron construction process for the case that the point cloud data is unevenly distributed, including the following steps: S321, Calculation of points in the point set With extreme points The Euclidean distance is used to determine the weighting coefficients based on the distance distribution. Select the four non-coplanar points with the highest weights as the vertices of the initial tetrahedron, and assign weight coefficients to them. The formula is: In the recursive generation of new surfaces, the unassigned points with a weight higher than a threshold value β are preferentially processed, and β is dynamically determined according to the point cloud density distribution. 7.The method of claim 4, wherein, In the S35, the volume V of the tetrahedron is realized by using a GPU parallel computing architecture, including the following steps: S351, the point cloud data is divided into multiple subsets and distributed to different thread blocks of the GPU; S352, each thread block independently calculates the convex hull sub-polyhedron corresponding to its subset and decomposes it into a tetrahedron; S353, through a global memory synchronization mechanism, the tetrahedron volumes of all thread blocks are accumulated to the host; S354, the calculation result is checked for redundant data, and if the error exceeds η%, the calculation task is redistributed. 8.The method of claim 1, wherein, In the S4, the volume of the convex polyhedron is calculated by using a GPU parallel computing architecture, including the following steps: S41, according to the spatial distribution characteristics of the point cloud data, an octree or KD tree is used for data blocking; S42, the execution time of each thread block of the GPU is monitored to dynamically adjust the data distribution strategy; S43, under the premise of ensuring the calculation accuracy, half-precision floating-point numbers are used to speed up part of the intermediate calculation process, and finally the volume is accumulated to restore the double-precision to reduce the accumulation error; S44, the convex hull decomposition, tetrahedron volume calculation and result accumulation are decoupled into independent calculation units, and pipeline parallelization is realized through asynchronous memory copying and calculation overlap. 9.The method of claim 1, wherein, In the S5, the fish reef group volume calculation includes the following steps: S51, the point cloud data after preprocessing is segmented by using a clustering algorithm to identify and separate the point sets of different fish reefs; S52, the convex hulls of adjacent fish reefs are subjected to Boolean operation to eliminate the problem of repeated volume calculation caused by point cloud data overlap; S53, the fish reef material density parameter ρ and the seawater density parameter ρ_water are introduced, and the buoyancy correction amount ΔV is calculated by using the Archimedes principle; S54, based on the noise level of the point cloud data and the error model of the convex hull algorithm, the confidence interval of the volume calculation is output. 10.The method of claim 1, wherein, The method further includes the following steps: S6, visualization and verification: S61, the convex hull model and the volume calculation result are rendered by a three-dimensional visualization platform, supporting interactive rotation, scaling and section analysis; S62, a standardized report containing the volume, calculation error, device parameters and processing log is generated, which can be exported in PDF or JSON format.

Images