A method for improving the first-time acceptance rate of shipborne sandstone

Abstract

The present application relates to a kind of methods for improving the acceptance rate of shipborne sandstone, comprising: collecting sandstone pile body before excavation and after excavation Two period land point cloud, transport ship empty state and two period cabin point cloud after loading;Two period land point cloud and two period cabin point cloud are carried out point cloud noise reduction, point cloud registration and three-dimensional modeling;The three-dimensional model corresponding to two period land point cloud is carried out rectangular grid integration and obtains land storage side;The three-dimensional model corresponding to two period cabin point cloud is carried out ship body interior inner protruding structure identification and rejection, and the volume difference after rejection is obtained according to the model volume difference;According to land storage side and loading side, calculate average side, deviation absolute value and relative deviation, output side quantity measurement consistency result or recalculation prompt.A kind of methods for improving the acceptance rate of shipborne sandstone designed by the present application, can solve the technical problems that land sandstone side quantity and cabin loading side quantity measurement benchmark are not consistent, deviation is difficult to control and affect the acceptance rate of first acceptance.

Description

Technical Field

[0001] This invention relates to the field of engineering surveying and bulk sand and gravel shipborne metering technology, specifically to a method for improving the first-time acceptance rate of shipborne sand and gravel. Background Technology

[0002] During the construction of ports, breakwaters, and related hydraulic engineering projects, sand and gravel materials typically undergo several stages, including storage in stockpiles, land excavation, loading onto ships, and ship transport. The accuracy of the sand and gravel volume directly impacts construction organization, material acceptance, and project settlement. Current on-site methods for estimating sand and gravel volume typically include weighing, density measurement, and changes in draft. While these methods are convenient, the results are easily affected by factors such as sand and gravel moisture content, sampling location within the stockpile, ship load conditions, and manual readings.

[0003] Current volumetric volume calculation methods typically treat the land-based and ship-cargo-based volumes as independent measurement objects. On the land-based side, volume is calculated based on local sampling density and transported weight; on the ship-cargo-based side, volume is calculated based on draft changes and post-loading density. However, due to uneven moisture distribution within the sand and gravel stockpile, external sampling density cannot consistently represent the overall density. Furthermore, sand and gravel mix during transport and loading, causing the post-loading density to differ from the land-based sampling density. Consequently, systematic discrepancies can easily arise between the land-based and ship-cargo-cargo volumes, leading to insufficient shipboard volume and consequently affecting the first-time acceptance rate.

[0004] Therefore, existing technologies have shortcomings and need to be improved and developed. Summary of the Invention

[0005] The present invention provides a method for improving the first-time acceptance rate of shipborne sand and gravel, which solves the technical problem in the prior art that the calculation benchmarks for land-based sand and gravel volume and ship-loaded volume are inconsistent and the deviation is difficult to control, thus affecting the first-time acceptance rate.

[0006] This invention provides a method for improving the first-time acceptance rate of shipborne sand and gravel, comprising:

[0007] Two phases of land area point clouds were collected before and after the excavation of the sand and gravel pile, and two phases of ship cabin point clouds were collected in the empty state of the transport ship and after loading.

[0008] Point cloud denoising, point cloud registration, and 3D modeling were performed on the two phases of land area point clouds and the two phases of ship cabin point clouds, respectively.

[0009] Based on the three-dimensional model corresponding to the two phases of land area point cloud, rectangular mesh integration is performed to obtain the land area storage area;

[0010] The internal protruding structures of the ship's hull were identified and removed from the three-dimensional models corresponding to the point clouds of the two phases of the ship's cabin, and the loading party was obtained based on the volume difference of the removed models.

[0011] The average volume, absolute value of deviation, and relative deviation are calculated based on the land storage party and the ship loading party. When the relative deviation and the absolute value of deviation are both within the preset error allowable range, the volume measurement consistency result is output. When at least one indicator exceeds the corresponding error allowable value, a recalculation prompt is output, including at least one of the following verification objects: scanning blind zone, registration error, modeling defect, and removal result of internal protruding structure of the hull.

[0012] Furthermore, two phases of land point cloud data were collected before and after the excavation of the sand and gravel pile, including:

[0013] Using the boundary of the sand and gravel pile as a reference, a closed-loop scanning route is laid out along the outside of the sand and gravel pile so that the scanning start point and the scanning end point coincide.

[0014] The laser scanner is moved along the closed loop scanning route to continuously scan the slope, top surface and bottom edge of the sand and gravel pile to obtain surface information of the excavation orientation of the sand and gravel pile.

[0015] The excavated gap area of ​​the sand and gravel pile is repeatedly scanned at close range and from multiple angles to increase the point cloud density of the gap area; the excavated gap area includes the depression area, the occlusion area and the elevation change area formed by the excavation operation; the repeated scanning is used to supplement the missing point cloud of the excavated gap area and make its point cloud density higher than that of the non-gap area.

[0016] The original point cloud before excavation and the original point cloud after excavation are output respectively. The original point cloud before excavation and the original point cloud after excavation include the three-dimensional coordinates of the points, the reflection intensity, and the scanning time.

[0017] Furthermore, point cloud denoising and point cloud registration are performed on the two phases of land area point clouds, including:

[0018] A statistical outlier filtering algorithm is used to denoise the original point cloud before excavation and the original point cloud after excavation, wherein the number of neighborhood statistical points is... It is 15, a multiple of the standard deviation. It is 2.0. This represents the number of nearest neighbor points selected with the point to be processed as the center. This represents the standard deviation multiple used to identify outliers;

[0019] Calculate the average distance between each point to be processed and its neighboring points. Points whose average distance exceeds twice the standard deviation are identified as noise points and removed to obtain two phases of land point cloud after noise reduction.

[0020] The iterative nearest point algorithm is used to register the two phases of land point clouds after noise reduction. The maximum number of iterations is 1000, the convergence threshold is 0.001m, and the corresponding point search radius is 0.1m.

[0021] During registration, firstly, fixed reference objects around the sand and gravel pile are selected to complete coarse registration. Then, the rotation matrix and translation vector are calculated by the iterative nearest point algorithm to unify the two phases of land point clouds into the same three-dimensional coordinate system. The fixed reference objects are ground control points, edges of surrounding structures, or fixed boundary features of the stockpile that can be identified in both phases of land point clouds and whose spatial positions do not change with the excavation of sand and gravel.

[0022] Furthermore, the step of performing rectangular mesh integration based on the 3D model corresponding to the two phases of land area point clouds to obtain the land area storage method includes:

[0023] Based on the registered two phases of land point clouds, normal vector estimation, surface reconstruction and meshing are performed to generate a 3D model before excavation and a 3D model after excavation.

[0024] The three-dimensional model before excavation and the three-dimensional model after excavation are projected onto the same two-dimensional plane and divided into multiple rectangular grid units according to the same bottom closed boundary.

[0025] Interpolate the point cloud elevations within each rectangular grid cell to obtain the representative elevation values ​​before and after excavation for that rectangular grid cell, and calculate the land storage volume using the following formula: In the formula, For land-based storage, This represents the total number of rectangular grid cells. The rectangular grid cell number. For the first The horizontal projected area of ​​each rectangular grid cell For the first The rectangular grid cells represent elevation values ​​before excavation. For the first Each rectangular grid cell represents an elevation value after excavation.

[0026] Furthermore, point clouds of the ship's hold were collected in two phases: when the transport ship was empty and after loading.

[0027] In the empty ship state, first scan the external outline of the transport ship around the deck, then enter the interior of the ship's cabin and perform partitioned scanning of the bottom, bulkhead and corner areas to obtain the original point cloud of the empty ship.

[0028] After the sand and gravel are loaded onto the ship, the same scanning route, scanning height and scanning angle as when scanning the ship in its empty state are maintained to scan the hull of the transport ship and the shape of the sand and gravel pile inside the ship's hold in order to obtain the original point cloud after loading.

[0029] The original point cloud of the empty ship and the original point cloud after loading are output respectively, serving as the data basis for volume comparison of the loading party.

[0030] Furthermore, point cloud denoising, point cloud registration, and 3D modeling are performed on the point clouds of the two phases of ship cabins, including:

[0031] The original point cloud of the empty ship and the original point cloud after loading are denoised by using a statistical outlier filtering algorithm or a radius filtering algorithm to remove noise points, outliers and isolated points generated during the scanning process.

[0032] The iterative nearest point algorithm is used to register the denoised empty ship point cloud and the loaded ship point cloud, so that the empty ship point cloud and the loaded ship point cloud are unified to the same three-dimensional coordinate system.

[0033] The Poisson surface reconstruction algorithm is used to generate an empty ship 3D model and a loaded ship 3D model. The empty ship 3D model is used to represent the ship's cabin space when it is not loaded with sand and gravel, and the loaded ship 3D model is used to represent the ship's cabin space and sand and gravel pile shape after it is loaded with sand and gravel.

[0034] Furthermore, the process of identifying and removing protruding internal structures from the 3D models corresponding to the point clouds of the two ship compartments, and obtaining the loading party based on the volume difference of the removed models, includes:

[0035] Using contour segmentation and mesh removal algorithms, the internal protruding structural volumes corresponding to beams, columns, bulkheads, stiffeners and pipes inside the ship's cabin are identified and removed to obtain the effective empty ship model and the effective model after loading.

[0036] The volume of the empty ship effective model and the volume of the loaded ship effective model are calculated using the rectangular mesh integration method, respectively, and the loading volume is calculated according to the following formula: In the formula, For the shipper, This is the volume of the effective model after loading onto the ship. This represents the volume of the effective empty ship model.

[0037] Furthermore, the average volume, absolute value of deviation, and relative deviation are calculated based on the land-based storage volume and the ship-loading volume, including:

[0038] The average volume is calculated using the following formula: In the formula, For average volume, For land-based storage, For the shipper;

[0039] Calculate the absolute value of the deviation using the following formula: In the formula, The absolute value of the deviation. For land-based storage, For the shipper;

[0040] The relative deviation is calculated using the following formula: In the formula, This is a relative deviation. The absolute value of the deviation. This is the average volume.

[0041] Furthermore, based on the preset error tolerance range, the system outputs the consistency result of the volume measurement or prompts for recalculation, including:

[0042] When the relative deviation is not greater than 1.5% and the absolute value of the deviation is not greater than 10m 3 At that time, the output results are consistent between the calculations of the land-based storage party and the ship-loading party;

[0043] When the preset application scenario is a high-precision control scenario for trade settlement, the relative deviation allowable value is set to no more than 1.0%;

[0044] When the relative deviation is greater than the corresponding allowable relative deviation value, or the absolute value of the deviation is greater than 10m 3 If this happens, a prompt to recalculate will be displayed.

[0045] Furthermore, after outputting the recalculation prompt, at least one of the following is corrected according to the recalculation prompt: scan point filling result, registration control point, filtering parameters, surface reconstruction parameters, integral grid size, and internal protrusion structure removal result of the hull. Based on the corrected data, the land storage volume, loading volume, average volume, absolute value of deviation, and relative deviation are recalculated.

[0046] Beneficial effects:

[0047] As can be seen from the above technical solutions, the present invention provides a method for improving the first-time acceptance rate of shipborne sand and gravel, which has the following beneficial effects:

[0048] 1. Improve the consistency of measurement between land-based storage parties and ship-based loading parties.

[0049] This application does not rely solely on sand and gravel density sampling, conveyor belt weight, or changes in ship draft for volume conversion. Instead, it establishes two-phase point cloud comparison relationships for the land-based stockpile and the transport ship's hold, respectively. On the land side, the actual excavation volume is calculated based on the elevation difference between the point clouds before and after excavation. On the ship's hold side, the actual loading volume is calculated based on the volume difference between the point clouds of an empty ship and those after loading. Although the two calculation results come from different scenarios, both are based on three-dimensional geometric volume changes, which reduces the impact of uneven moisture content, insufficient representativeness of local sampling, and density conversion errors on the measurement results. Because both ends employ point cloud acquisition, noise reduction, registration, modeling, and integral calculation techniques, the data processing logic is consistent, facilitating unified determination of average volume, absolute deviation, and relative deviation, thereby improving the comparability between the land-based stockpile and the ship's loading volume.

[0050] 2. Reduce the impact of scanning blind spots and local missing point clouds on volume calculation.

[0051] This application employs a closed-loop scanning route in the land-based aggregate acquisition phase, performing continuous and repeated scanning of the aggregate's slope, top surface, bottom edge, and excavation gap areas. This acquisition method ensures that the point cloud data covers the main geometric contours of the sand and gravel aggregate, particularly supplementing local point clouds in excavation gaps, depressions, and areas of abrupt elevation changes. For the ship compartment acquisition phase, this application acquires point clouds of the hull, bulkheads, and corner areas while the ship is empty. After loading, the scanning route, scanning height, and scanning angle remain consistent, resulting in a good spatial correspondence between the two phases of ship compartment point clouds. Through this acquisition strategy, the data foundation for subsequent registration and modeling is more complete, reducing the risk of voids in surface reconstruction, inaccurate volume integration, and misjudgments of the ship's structure due to missing local point clouds.

[0052] 3. Improve the accuracy of 3D model reconstruction by point cloud noise reduction and registration.

[0053] This application incorporates point cloud denoising and registration steps after field data acquisition, ensuring that the original point clouds are first noise-removed and then unified into the same 3D coordinate system. A statistical outlier filtering algorithm identifies isolated and outlier points through neighborhood distance statistics, reducing invalid points generated during scanning due to reflection anomalies, occlusion edges, and environmental interference. An iterative nearest-point algorithm further calculates the rigid transformation relationship between the two phases of point clouds using rotation matrices and translation vectors, ensuring spatial correspondence between point clouds before and after excavation, and between empty and loaded vessels. Since volume calculation inherently depends on the geometric differences between the two models, errors will directly accumulate in elevation or volume differences if the coordinate systems are not unified. This application, through progressive processing of denoising and registration, provides a more stable data foundation for subsequent 3D modeling and rectangular mesh integration.

[0054] 4. Improve the reliability of the loading party's calculations by eliminating protruding structures inside the hull.

[0055] The interior of a transport ship's hold typically contains structures such as beams, columns, bulkheads, stiffeners, and pipes. These structures do not represent the actual volume of sand and gravel loaded, but they are included as internal geometric objects in the point cloud model during 3D scanning and model reconstruction. If these protruding structures are not identified and removed, the calculation of the loading volume will be affected by the non-loading volume, potentially leading to inaccurate assessments of the effective loading space or the volume occupied by sand and gravel. This application identifies and removes protruding structures within the hull using contour segmentation and mesh elimination algorithms, allowing the volume difference calculation to be performed on the empty ship model and the loaded ship model under the same hold boundary, coordinate system, and integral mesh. This process brings the loading volume closer to the actual geometric volume occupied by sand and gravel, reducing interference from differences in hull structure in volume measurement.

[0056] 5. A quality closed loop is formed through deviation judgment and recalculation prompts.

[0057] After obtaining the land-based storage volume and the ship-loading volume, this application does not directly use a single calculation result as the acceptance basis. Instead, it further calculates the average volume, absolute value of deviation, and relative deviation, and compares them with the preset error allowable range. This process establishes a mutual verification relationship between the two independent calculation paths. When both the relative deviation and the absolute value of deviation are within the allowable range, it indicates that the two sets of volumes are consistent in an engineering measurement sense. When at least one indicator exceeds the limit, a recalculation prompt is output, and the results of scanning blind spots, registration errors, modeling defects, noise reduction intensity, and removal of protruding structures are reviewed. This closed-loop processing enables this application not only to complete the volume calculation but also to determine and correct the reliability of the calculation results, thereby improving the risk identification capability before sand and gravel loading and acceptance.

[0058] It should be understood that all combinations of the foregoing concepts and the additional concepts described in more detail below can be considered part of the inventive subject matter of this disclosure, provided that such concepts do not contradict each other.

[0059] The foregoing and other aspects, embodiments, and features of the teachings of the present invention will be more fully understood from the following description in conjunction with the accompanying drawings. Other additional aspects of the invention, such as features and / or beneficial effects of exemplary embodiments, will become apparent from the following description or may be learned through practice of specific embodiments according to the teachings of the present invention. Attached Figure Description

[0060] The accompanying drawings are not drawn to scale. In the drawings, each identical or nearly identical component shown in the various figures may be denoted by the same reference numeral. For clarity, not every component is labeled in each figure. Embodiments of various aspects of the invention will now be described by way of example and with reference to the accompanying drawings, wherein:

[0061] Figure 1 This is a flowchart illustrating a method for improving the first-time acceptance rate of shipborne sand and gravel in an embodiment of this application.

[0062] Figure 2 This is a flowchart of step S106 of a method for improving the first-time acceptance rate of shipborne sand and gravel in an embodiment of this application.

[0063] Figure 3 This is a flowchart of step S108 of a method for improving the first-time acceptance rate of shipborne sand and gravel in an embodiment of this application.

[0064] Figure 4 This is a schematic diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0065] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the scope of protection of the present invention. Unless otherwise defined, the technical or scientific terms used herein should have the ordinary meaning understood by those skilled in the art to which this invention pertains.

[0066] The terms "first," "second," and similar words used in the specification and claims of this patent application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, unless the context clearly indicates otherwise, the singular forms of "an," "a," or "the," etc., do not indicate a quantity limitation, but rather indicate the presence of at least one. Terms such as "comprising" or "including" mean that the element or object preceding "comprising" encompasses the features, integrals, steps, operations, elements, and / or components listed following "comprising" or "including," and do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; these relative positional relationships may change accordingly when the absolute position of the described object changes.

[0067] Current volumetric measurement methods typically treat the land-based and ship-cargo-based sides as independent measurement objects. On the land-based side, volumetric measurement relies on local sampling density and transported weight, while on the ship-cargo-based side, it relies on draft changes and post-loading density. However, due to uneven moisture distribution within the sand and gravel stockpile, external sampling density cannot consistently represent the overall density. Furthermore, sand and gravel mix during transport and loading, causing the post-loading density to differ from the land-based sampling density. Therefore, systematic discrepancies easily arise between land-based and ship-cargo-cargo volumes, leading to insufficient shipboard volume and consequently affecting the first-time acceptance rate.

[0068] Therefore, embodiments of the present invention provide a method for improving the first-time acceptance rate of shipborne sand and gravel, referring to... Figure 1 ,include:

[0069] Step S102: Collect two phases of land area point clouds before and after the excavation of the sand and gravel pile, and collect two phases of ship cabin point clouds in the empty state of the transport ship and after loading.

[0070] Step S104: Perform point cloud denoising, point cloud registration, and 3D modeling on the two phases of land area point clouds and the two phases of ship cabin point clouds respectively. 3D modeling includes: resampling, normal vector estimation, Poisson surface reconstruction, boundary clipping, hole repair, and mesh smoothing on the denoised and registered point clouds to obtain a triangular mesh model or surface model for volume integration.

[0071] Step S106: Perform rectangular mesh integration based on the 3D model corresponding to the two phases of land area point cloud to obtain the land area storage area.

[0072] Step S108: Identify and remove the internal protruding structures of the ship's hull from the three-dimensional models corresponding to the point clouds of the two phases of the ship's cabin, and obtain the loading party based on the volume difference of the removed models.

[0073] Step S110: Calculate the average volume, absolute value of deviation, and relative deviation based on the land storage volume and the loading volume. When the relative deviation and absolute value of deviation are both within the preset error allowable range, output the volume measurement consistency result. When at least one indicator exceeds the corresponding error allowable value, output a recalculation prompt including at least one of the following verification objects: scanning blind zone, registration error, modeling defect, and removal of internal protruding structures inside the hull.

[0074] The geometric changes of the sand and gravel stockpile before and after excavation, and the geometric changes of the transport ship after emptying and loading, are used as two independent but comparable data sources. Noise reduction, registration, and 3D modeling are then used to form a model basis for volume calculation. Subsequently, the excavation volume is obtained by rectangular mesh integration on the land side, and the loading volume is obtained on the ship side after removing internal protruding structures. The average volume, absolute value of deviation, and relative deviation are then used to determine whether the two calculation results are consistent. Thus, the two-stage point cloud differential calculations of the land stockpile and the ship's cargo hold are combined and used as a mutual verification volume calculation system.

[0075] Traditional methods often rely on land sampling density, transport weight, changes in ship draft, and post-loading density for conversion, which are easily affected by changes in sand and gravel moisture content and sampling location. This application does not simply replace the measuring tool, but establishes a geometric volume comparison relationship between the reduction in land area and the increase in ship hold for the same batch of sand and gravel. At the same time, it sets an allowable error range and recalculation prompts, so that the measurement results are no longer limited to a single output, but have a consistency judgment and anomaly verification mechanism.

[0076] By employing dual-source point cloud measurement and deviation assessment, the impact of density calculation errors on sand and gravel volume measurement can be reduced. The land-based stockpile volume reflects the actual volume of sand and gravel reduced after excavation from the stockpile, while the ship-loading volume reflects the actual volume occupied by sand and gravel after entering the ship's hold. Both originate from spatial geometric changes, thus providing a unified volume verification basis across different operational stages. When the deviations of both meet preset requirements, reliable data can be provided for the initial acceptance of ship-loaded sand and gravel. When the deviations exceed limits, it prompts a review of point cloud acquisition, model reconstruction, and ship structure deduction, thereby reducing the problem of insufficient ship-loaded volume due to measurement errors.

[0077] In some embodiments, collecting two phases of land area point clouds before and after the excavation of the sand and gravel pile includes:

[0078] Using the boundary of the sand and gravel pile as a reference, a closed-loop scanning route is laid out along the outside of the sand and gravel pile so that the scanning start point and the scanning end point coincide.

[0079] The laser scanner moves along a closed-loop scanning path to continuously scan the slope, top, and bottom edges of the sand and gravel pile to obtain surface information on the orientation of the excavation.

[0080] Close-range, multi-angle repeated scanning was performed on the excavated gap area of ​​the sand and gravel pile to increase the point cloud density of the gap area. The excavated gap area includes the depression area, occlusion area and elevation change area formed by the excavation operation. Repeated scanning is used to supplement the missing point cloud in the excavated gap area and make its point cloud density higher than that of the non-gap area.

[0081] The supplementary scanning involves repeatedly scanning the excavated gaps in the sand and gravel pile at close range and from multiple angles. During supplementary scanning, the center normal direction of the excavated gap area is used as the reference direction. The distance between the laser scanner and the surface to be scanned in the gap area is controlled between 0.5m and 3.0m; when the gap depth is greater than 1.5m or there is occlusion at the bottom of the gap, the distance between the laser scanner and the surface to be scanned at the bottom of the gap is preferably controlled between 0.5m and 1.5m. Supplementary scanning includes at least a forward scanning direction, a left oblique scanning direction, and a right oblique scanning direction. The angle between the forward scanning direction and the center normal direction is 0° to 15°, and the angles between the left oblique scanning direction and the right oblique scanning direction and the center normal direction are 30° to 60° respectively. In each scanning direction, the elevation angle of the laser scanner is controlled between -20° and 45°, and the same gap area is scanned at least twice to create locally refined point clouds in the recessed areas, occluded areas, and areas of abrupt elevation changes, which can be used for 3D modeling.

[0082] Output the original point cloud before excavation and the original point cloud after excavation respectively. The original point cloud before excavation and the original point cloud after excavation include the three-dimensional coordinates of the points, the reflection intensity, and the scanning time.

[0083] A closed-loop scanning route was established based on the boundary of the sand and gravel pile, ensuring that the scanning start and end points coincided, thus establishing a stable planar control boundary for point cloud acquisition. During the scanning process, the laser scanner moved along the closed route, continuously scanning the slope, top surface, and bottom edge of the pile to obtain relevant surface information of the excavated area. For excavated gaps, close-range, multi-angle repeated scanning was set to supplement point cloud data in recessed areas, occluded areas, and areas with abrupt elevation changes. The final output of two phases of raw point clouds includes 3D coordinates, reflection intensity, and scanning time, providing the original data foundation for subsequent noise reduction, registration, and modeling.

[0084] The land-based data collection targets were limited to the surface of the excavated pile that could be compared before and after excavation. The correspondence between the two point clouds was improved by using a closed-loop scanning route and repeated scanning of gap areas. The closed route was not simply a way of walking on site, but was used to reduce scanning position deviation and form a relatively stable data collection boundary. Repeated scanning of gap areas was not simply to increase the number of scans, but to locally enhance the point cloud density in the depressions and occlusion areas formed by sand and gravel excavation.

[0085] The irregular shape of the sand and gravel pile can easily lead to depressions, steep slopes, and obstructions in localized areas after excavation. If only a single-angle scan is performed, point cloud gaps may appear in the relevant areas, resulting in discontinuities on the surface of the 3D model. By using a closed-loop scanning route, the circumferential contour of the pile can be covered, and the drift of the acquisition position can be reduced. By performing close-range, multi-angle repeated scans of the excavation gap area, local details can be supplemented, improving the reliability of elevation interpolation.

[0086] In some embodiments, point cloud denoising and point cloud registration are performed on the two phases of land area point clouds, including:

[0087] A statistical outlier filtering algorithm is used to denoise the original point cloud before and after excavation. The algorithm includes a neighborhood statistical point count. It is 15, a multiple of the standard deviation. It is 2.0. This represents the number of nearest neighbor points selected with the point to be processed as the center. This represents the standard deviation multiple used to identify outliers.

[0088] Calculate the average distance between each point to be processed and its neighboring points. Points with an average distance exceeding twice the standard deviation are identified as noise points and removed, resulting in two phases of denoised land point clouds.

[0089] The iterative nearest point algorithm was used to register the two phases of land point clouds after denoising. The maximum number of iterations was 1000, the convergence threshold was 0.001m, and the search radius of the corresponding point was 0.1m.

[0090] During registration, firstly, fixed reference objects around the sand and gravel pile are selected to complete coarse registration. Then, the rotation matrix and translation vector are calculated by the iterative nearest point algorithm to unify the two phases of land point clouds into the same three-dimensional coordinate system. The fixed reference objects are ground control points, the edges of surrounding structures, or fixed boundary features of the stockpile that can be identified in both phases of land point clouds and whose spatial positions do not change with the excavation of sand and gravel.

[0091] A statistical outlier filtering algorithm was used to process the original point clouds before and after excavation. During processing, the nearest neighbor point was selected centered on the point to be processed, and the average distance between the point to be processed and its neighboring points was calculated. Points with an average distance exceeding a statistical threshold were removed as noise points. Subsequently, an iterative nearest-neighbor algorithm was used to register the two phases of land area point clouds, with limits on the maximum number of iterations, convergence threshold, and corresponding point search radius, so that the two phases of point clouds were unified to the same three-dimensional coordinate system.

[0092] The land point cloud processing is divided into two progressive stages: noise reduction and registration. The corresponding algorithm parameters are given for implementation: statistical outlier filtering is used to reduce the interference of isolated noise points on model reconstruction, and the iterative nearest point algorithm is used to eliminate the spatial position deviation between two acquisitions.

[0093] First, fixed reference objects around the sand and gravel pile are selected to complete coarse registration. Then, the rotation matrix and translation vector are calculated by the iterative nearest point algorithm to perform fine registration. This can avoid the iterative nearest point algorithm from falling into incorrect matching when the initial pose deviation is large, and can improve the spatial consistency between the two phases of land point clouds.

[0094] If outliers exist in the original point cloud, local spikes, holes, or erroneous surfaces can easily form during 3D modeling, thus affecting the mesh elevation values. If the two point clouds are not registered to the same coordinate system, the elevation difference before and after excavation will simultaneously include the actual excavation volume and coordinate offset errors. By first reducing noise and then registering, the elevation difference can primarily reflect the geometric changes caused by sand and gravel excavation, rather than scanning errors or coordinate deviations, thus improving the reliability of the landfill calculation results.

[0095] In some embodiments, rectangular mesh integration is performed based on the 3D model corresponding to the two phases of land area point clouds to obtain the land area storage area, referring to... Figure 2 ,include:

[0096] Step S1061: Based on the registered two phases of land point clouds, perform normal vector estimation, surface reconstruction and meshing processing to generate a three-dimensional model before excavation and a three-dimensional model after excavation.

[0097] Step S1062: Project the pre-excavation 3D model and the post-excavation 3D model onto the same 2D plane, and divide them into multiple rectangular mesh units according to the same bottom closed boundary.

[0098] Step S1063: Interpolate the point cloud elevation within each rectangular grid cell to obtain the representative elevation values ​​before and after excavation for that rectangular grid cell, and calculate the land storage volume according to the following formula: In the formula, For land-based storage, This represents the total number of rectangular grid cells. The rectangular grid cell number. For the first The horizontal projected area of ​​each rectangular grid cell For the first The rectangular grid cells represent elevation values ​​before excavation. For the first Each rectangular grid cell represents an elevation value after excavation.

[0099] In step S1061, a pre-excavation 3D model and a post-excavation 3D model are generated based on the registered two phases of land point clouds, including:

[0100] Point cloud resampling was performed on the registered point cloud before excavation and the point cloud after excavation to obtain resampled point clouds before excavation and resampled point clouds after excavation with point spacing meeting the preset modeling accuracy requirements.

[0101] Calculate the normal vector of each point in the resampled point cloud before excavation and the resampled point cloud after excavation, and unify the direction of the normal vector of adjacent points.

[0102] The Poisson surface reconstruction algorithm is used to reconstruct the surfaces of the resampled point cloud before excavation and the resampled point cloud after excavation after normal vector processing, respectively, to generate the initial surface model before excavation and the initial surface model after excavation.

[0103] Boundary trimming, hole repair, and mesh smoothing are performed on the initial surface model before excavation and the initial surface model after excavation to obtain the three-dimensional model before excavation and the three-dimensional model after excavation.

[0104] Among them, the reconstruction depth of the Poisson surface reconstruction algorithm is 10, the sampling density is 1.5, and the smoothing iteration number is 5; the boundary clipping is based on the closed boundary at the bottom of the sand and gravel pile; the hole repair is used to fill in the local missing mesh caused by scanning occlusion; and the mesh smoothing is used to weaken the impact of isolated elevation changes on subsequent volume integration.

[0105] Based on the registered point clouds of the two landmasses, two 3D models were generated: one before excavation and one after. These two models were then projected onto the same 2D plane and divided into multiple rectangular mesh elements according to the same bottom closed boundary. For each rectangular mesh element, the difference between the pre-excavation and post-excavation elevation values ​​was calculated, and this difference was multiplied by the horizontal projected area of ​​the mesh element. Finally, the summation over all mesh elements yielded the landmass volume, thus transforming the irregular volume of the landmass into a discrete integral problem under a regular mesh.

[0106] Instead of calculating the volumes of the two point cloud models separately and then simply subtracting them, volume integration is performed by placing the two point cloud models under the same bottom closed boundary and the same mesh generation rules. Rectangular mesh integration can directly map the local elevation differences between the models before and after excavation to the same plane position, which is suitable for engineering scenarios with irregular sand and gravel pile surfaces, obvious local excavation, and projectable overall boundaries. This method can preserve the elevation change information of different areas, so that the land stockpile not only reflects the overall volume difference but also corresponds to the spatial distribution of the excavated area.

[0107] By projecting the 3D model onto a rectangular grid in the same 2D plane, computational engineers can clearly define the contribution of each grid cell to the overall volume, facilitating the identification of anomalous regions, such as areas of abrupt elevation differences, missing point cloud areas, or abnormal boundary delineation. Compared to relying solely on the overall model volume, rectangular grid integration allows for easier adjustment of grid size to balance computational accuracy and efficiency, and also facilitates the location of verification objects when errors exceed limits.

[0108] In some embodiments, point clouds of the ship's hold are collected in two phases: when the transport ship is empty and after loading.

[0109] In an empty ship state, first scan the external outline of the transport ship around the deck, then enter the ship's interior and perform partitioned scanning of the bottom, bulkheads, and corner areas to obtain the original point cloud of the empty ship.

[0110] After the sand and gravel are loaded onto the ship, the same scanning route, scanning height, and scanning angle as when scanning the ship in its empty state are maintained to scan the hull of the transport ship and the shape of the sand and gravel pile inside the ship's hold in order to obtain the original point cloud after loading.

[0111] Output the original point cloud of the empty ship and the original point cloud after loading, respectively, as the data basis for volume comparison of the loading party.

[0112] In the empty ship state, the external outline of the transport ship is first scanned around it, and then the interior of the ship's hold is scanned in sections to obtain the original point cloud of the empty ship. After the sand and gravel are loaded onto the ship, the same scanning route, scanning height, and scanning angle as during the empty ship state scan, or after correction, are used to scan the hull of the transport ship and the shape of the sand and gravel piles inside the hold to obtain the original point cloud after loading. The point clouds of the ship's hold in both phases serve as the data basis for volume comparison of the loading party and are used for subsequent noise reduction, registration, modeling, and volume difference calculation.

[0113] The scanning conditions for the empty ship and the loaded ship are mapped to each other, making the point clouds of the two phases highly spatially comparable. The ship's compartments differ from land-based stockpiles; they contain bulkheads, corners, reinforcing structures, and obstructed areas. Significant differences in the scanning routes and viewing angles between the two phases will lead to inconsistencies in the coverage of the point clouds, affecting volume difference calculations. A baseline model of the ship's compartments is obtained through empty ship scanning, and a sand and gravel stockpile model is obtained through post-loading scanning, maintaining consistent scanning parameters. This establishes a stable dual-phase data foundation for calculations by the loading party. When on-site conditions cause deviations in the scanning route, scanning height, or scanning angle, fixed marker points on the hull or fixed structural features of the ship's compartments can be used to correct the original point clouds of the empty ship and the original point clouds of the loaded ship.

[0114] Empty ship point clouds are used to characterize the basic geometry of the ship's hold, while post-loading point clouds are used to characterize the geometry after the sand and gravel are loaded. Only when the acquisition range and coordinate conditions are relatively consistent can both reliably reflect the sand and gravel loading volume through registration and volume difference. By performing partitioned scanning of the hold's hull, bulkheads, and corner areas, model loss caused by blind spots in the hold can be reduced. Maintaining consistency in the scanning route, height, and angle before and after loading can reduce systematic errors introduced by differences in acquisition methods, thereby improving the reliability of the loading party's calculation results.

[0115] In some embodiments, point cloud denoising, point cloud registration, and 3D modeling are performed on the point clouds of the two ship cabins, including:

[0116] Statistical outlier filtering or radius filtering algorithms are used to reduce noise in the original point cloud of the empty ship and the original point cloud after loading, eliminating noise points, outliers and isolated points generated during the scanning process.

[0117] The iterative nearest point algorithm is used to register the denoised empty ship point cloud and the loaded ship point cloud, so that the empty ship point cloud and the loaded ship point cloud are unified into the same three-dimensional coordinate system.

[0118] The Poisson surface reconstruction algorithm was used to generate two 3D models: an empty ship and a loaded ship. The empty ship model was used to represent the ship's cabin space when no sand or gravel was loaded, while the loaded ship model was used to represent the ship's cabin space and the shape of the sand and gravel pile after loading.

[0119] The general process of generating 3D models of an empty ship and a ship after loading using the Poisson surface reconstruction algorithm is as follows: Using the denoised and registered point cloud as input data, the point cloud is first resampled to ensure its density meets the preset modeling accuracy requirements. Then, normal vectors are calculated based on the distribution of points in the neighborhood of each point, and the directions of the normal vectors of adjacent points are made consistent. Subsequently, the Poisson surface reconstruction algorithm is used to generate a continuous surface model based on the point cloud coordinates and corresponding normal vectors. Finally, the continuous surface model undergoes boundary trimming, hole repair, and mesh smoothing to obtain a 3D model suitable for volume integration. For land area point clouds, 3D models before and after excavation are generated; for ship cabin point clouds, 3D models of an empty ship and a ship after loading are generated. The specific process includes:

[0120] Point cloud resampling and normal vector estimation were performed on the denoised and registered empty ship point cloud and the point cloud after loading.

[0121] Based on the estimated normal vectors, the Poisson surface reconstruction algorithm is used to reconstruct continuous surfaces of the empty ship point cloud and the point cloud after loading, respectively, to obtain the initial surface model of the empty ship and the initial surface model after loading.

[0122] The initial surface model of the empty ship and the initial surface model after loading are respectively subjected to cabin boundary trimming, hole repair and mesh smoothing to generate the three-dimensional model of the empty ship and the three-dimensional model after loading.

[0123] The empty ship 3D model is used to represent the internal outline of the ship's hold when it is not loaded with sand and gravel, while the loaded ship 3D model is used to represent the internal outline of the ship's hold and the shape of the sand and gravel pile after it is loaded with sand and gravel.

[0124] Statistical outlier filtering or radius filtering algorithms are used to process the original point clouds of the empty ship and the original point clouds after loading, removing noise points, outliers, and isolated points. Statistical outlier filtering is used when noise points are isolated, while radius filtering is used when noise points are concentrated in locally sparse regions. Subsequently, the iterative nearest-point algorithm is used to register the denoised empty ship point clouds and the loaded ship point clouds, unifying the two sets of cabin point clouds into the same three-dimensional coordinate system. Finally, the Poisson surface reconstruction algorithm is used to generate three-dimensional models of the empty ship and the loaded ship, respectively. The empty ship 3D model represents the cabin space without sand and gravel, while the loaded ship 3D model represents the cabin space and sand and gravel pile shape after loading.

[0125] The interior of the ship's cabin presents complex reflections, numerous structural edges, and many occlusion areas, potentially resulting in isolated noise and locally sparse points in the original point cloud. Statistical outlier filtering and radius filtering can adapt to different noise distribution scenarios, respectively. The iterative nearest-point algorithm enables comparison of the empty ship's point cloud and the point cloud after loading in the same coordinate system, while Poisson surface reconstruction can form a continuous surface model from the discrete point cloud. This processing chain transforms the cabin point cloud from raw acquired data into a 3D model suitable for volumetric comparison.

[0126] Noise reduction minimizes the interference of invalid points on surface reconstruction, registration reduces spatial misalignment between point clouds in two phases, and Poisson surface reconstruction transforms discrete point clouds into continuous surfaces, providing a feasible model basis for subsequent identification of protruding structures and rectangular mesh integration. Due to the complex structure of the ship's cabin, directly calculating the volume of the original point cloud is easily affected by point cloud sparsity, noise, and occlusion. Through the above processing, the empty ship model and the loaded ship model can perform structural culling and volume difference calculations under a unified coordinate system, thereby improving the stability of the calculation results for the loaded ship.

[0127] In some embodiments, the internal protruding structures of the ship's hull are identified and removed from the 3D models corresponding to the point clouds of the two phases of the ship's cabins. The loading party is obtained based on the volume difference of the removed models, with reference to... Figure 3 ,include:

[0128] Step S1081: Using contour segmentation and mesh removal algorithms, identify and remove the internal protruding structural volumes corresponding to beams, columns, bulkheads, stiffeners and pipes inside the ship cabin to obtain the effective empty ship model and the effective model after loading.

[0129] Step S1082: Calculate the volume of the effective empty ship model and the volume of the effective model after loading using the rectangular mesh integration method, and calculate the loading volume according to the following formula: In the formula, For the shipper, This is the volume of the effective model after loading onto the ship. This represents the volume of the effective empty ship model.

[0130] Contour segmentation and mesh removal algorithms were employed to identify and remove the volumes of protruding structures such as beams, columns, bulkheads, stiffeners, and pipes within the ship's interior, resulting in effective empty-ship and loaded-ship models. Subsequently, the volumes of the empty-ship and loaded-ship models were calculated using the rectangular mesh integration method, and the volume difference between the two models under the same ship's boundary, coordinate system, and integration mesh was taken as the loading side. Protruding structures are mesh regions that bulge into the ship's interior relative to the ship's bulkhead reference surface. During removal, the protruding structures identified in the empty-ship model were used as the reference, and the corresponding mesh regions in the loaded 3D model were simultaneously subtracted.

[0131] The loading calculation does not directly compare the empty ship model and the loaded ship model. Instead, it first identifies and subtracts protruding structures within the hull before comparison. While the internal structures of the ship's compartments occupy space, they are not part of the sand and gravel loading volume. If this volume is not removed, the loading calculation may be affected by the hull structure. By using contour segmentation and mesh removal, protruding structures are uniformly excluded from both the empty ship model and the loaded ship model, allowing the volume difference to more effectively reflect the geometric changes caused by sand and gravel loading.

[0132] This paper proposes a method to address structural interference in ship hold measurement scenarios, reducing the impact of ship structure on loading calculations and improving the comparability between loading and land-based storage. Protruding structures such as beams, columns, bulkheads, stiffeners, and pipes create localized protrusions in the point cloud model. If mistakenly identified as loading space or sand / gravel volume, this can lead to deviations in volume calculations from the actual loaded volume. By identifying and removing these protruding structures, the effective empty ship model and the effective loaded model participate in volume comparison under the same rules, reducing the influence of non-sand / gravel structures on the loading process. This method is particularly suitable for transport ships with complex internal hold structures, helping to improve the accuracy of sand / gravel loading volume measurement.

[0133] In some embodiments, the average volume, absolute deviation, and relative deviation are calculated based on the land-based storage volume and the ship-loading volume, including:

[0134] The average volume is calculated using the following formula: In the formula, For average volume, For land-based storage, For the ship's loading party.

[0135] Calculate the absolute value of the deviation using the following formula: In the formula, The absolute value of the deviation. For land-based storage, For the ship's loading party.

[0136] The relative deviation is calculated using the following formula: In the formula, This is a relative deviation. The absolute value of the deviation. This is the average volume.

[0137] The average volume is used as the benchmark for deviation evaluation, rather than solely relying on either land-based storage or ship-loading volume. Land-based storage and ship-loading volumes originate from different measurement paths, and both may contain measurement errors. Using the average of these two values ​​as the benchmark reduces the dependence of deviation evaluation on a single measurement result. The absolute value of the deviation reflects the actual volume difference between the two measurement results, while the relative deviation reflects the proportion of this difference relative to the overall volume size.

[0138] Comparing only the difference between two volumes makes it difficult to determine whether the difference is acceptable under different total volume scales; calculating only the relative difference may ignore the impact of actual volume differences on project acceptance. By simultaneously calculating the average volume, the absolute value of the deviation, and the relative deviation, the measurement results can be evaluated from three perspectives: overall scale, absolute difference, and proportional difference. This calculation method can accommodate both large and small volume scenarios, making error judgment more engineering-adaptable and facilitating the transformation of point cloud measurement results into usable judgment indicators for project acceptance, thereby improving the objectivity of the acceptance process.

[0139] In some embodiments, outputting a consistency result of the volume measurement or a recalculation prompt based on a preset error allowable range includes:

[0140] When the relative deviation is no greater than 1.5% and the absolute value of the deviation is no greater than 10m 3 At that time, the output results are consistent between the calculations of the land-based storage party and the ship-loading party.

[0141] When the preset application scenario is a high-precision control scenario for trade settlement, the relative deviation allowable value is set to no more than 1.0%.

[0142] When the relative deviation is greater than the corresponding allowable relative deviation value, or the absolute value of the deviation is greater than 10m 3 If this happens, a prompt to recalculate will be displayed.

[0143] A dual-threshold judgment mechanism, employing both relative deviation and absolute deviation, is adopted. Relative deviation reflects the proportional error under different volume scales, while the absolute deviation controls the engineering impact of actual volume differences. Consistent results are only output when both thresholds are simultaneously satisfied, avoiding misjudgments caused by relying on a single indicator. For high-precision control scenarios in trade settlement, the allowable value for relative deviation is further tightened, enabling the method to adjust acceptance accuracy requirements according to the application scenario.

[0144] In engineering measurement, relying solely on relative deviation may lead to excessively large actual differences in small volume scenarios, while relying solely on absolute differences may fail to reflect proportional errors in large volume scenarios. By simultaneously setting allowable values ​​for both relative and absolute deviations, the measurement results can be subject to dual constraints. For scenarios with high accuracy requirements, such as trade settlement, further reducing the allowable value for relative deviation can improve the reliability of settlement data. Therefore, this application can not only calculate volume but also provide results that can be directly used for acceptance judgment.

[0145] In some embodiments, after outputting a recalculation prompt, at least one of the following is corrected based on the recalculation prompt: scan point filling result, registration control point, filtering parameters, surface reconstruction parameters, integral grid size, and internal protrusion structure removal result of the hull. Based on the corrected data, the land storage volume, loading volume, average volume, absolute value of deviation, and relative deviation are recalculated.

[0146] Sand and gravel storage areas and ship hull environments are characterized by irregularities, numerous obstructions, and complex structures. Initial scans or modeling results may contain localized gaps, registration discrepancies, or incorrectly excluded structures. By triggering a review process when deviations exceed limits, corrections can be made targeting specific error sources, rather than simply rejecting the calculation results. After re-executing point cloud denoising, registration, modeling, volume calculation, and deviation calculation, new volume results can be established based on the corrected data, thereby improving the reliability of the final output and increasing the first-time acceptance rate of shipborne sand and gravel.

[0147] Another embodiment of the present invention also provides an apparatus for improving the first-time acceptance rate of shipborne sand and gravel, comprising:

[0148] The data acquisition module is used to collect point clouds of the land area before and after the excavation of the sand and gravel pile, as well as point clouds of the ship's hold in the empty state and after loading.

[0149] The processing module is used to perform point cloud denoising, point cloud registration, and 3D modeling on the two phases of land area point clouds and the two phases of ship cabin point clouds, respectively.

[0150] The integration module is used to perform rectangular mesh integration based on the 3D model corresponding to the two phases of land area point cloud to obtain the land area storage area.

[0151] The identification module is used to identify and remove internal protruding structures inside the hull from the 3D models corresponding to the point clouds of the two phases of the ship's cabin, and to obtain the loading party based on the volume difference of the removed models.

[0152] The calculation module is used to calculate the average volume, absolute value of deviation, and relative deviation based on the land storage volume and the ship loading volume. When the relative deviation and absolute value of deviation are both within the preset error allowable range, the volume measurement consistency result is output. When at least one indicator exceeds the corresponding error allowable value, a recalculation prompt is output, including the results of removing at least one of the following: scanning blind zone, registration error, modeling defects, and internal protruding structures of the hull.

[0153] It should be noted that although several units or sub-units of the device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided and embodied by multiple units.

[0154] Based on the same inventive concept as the above method embodiments, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it enables the electronic device to implement the control method described in the above embodiments.

[0155] In one embodiment, the electronic device may be a server, and in this embodiment, the structure of the electronic device may be as follows: Figure 4 As shown, it includes a memory, a communication module, and one or more processors.

[0156] Memory is used to store computer programs executed by the processor. Memory can be mainly divided into a program storage area and a data storage area. The program storage area can store the operating system and programs required to run instant messaging functions, etc.; the data storage area can store various instant messaging information and operation instruction sets, etc.

[0157] Memory can be volatile memory, such as random access memory (RAM); memory can also be non-volatile memory, such as read-only memory, flash memory, hard disk drive (HDD), or solid-state drive (SSD); or memory can be any other medium capable of carrying or storing a desired computer program having the form of instructions or data structures and accessible by a computer, but is not limited thereto. Memory can be a combination of the above-mentioned types of memory.

[0158] A processor may include one or more central processing units (CPUs) or digital processing units, etc. A processor is used to implement the aforementioned data processing methods when it invokes computer programs stored in memory.

[0159] The communication module is used to communicate with terminal devices and other servers.

[0160] This application embodiment does not limit the specific connection medium between the above-described memory, communication module, and processor. This application embodiment... Figure 4 The memory and processor are connected via a bus, and the bus is in... Figure 4 The connections between other components are illustrated with arrows and are for illustrative purposes only, not as limiting information. Buses can be categorized as address buses, data buses, control buses, etc. For ease of description, Figure 4 The text uses only one arrow to describe it, but does not indicate that there is only one bus or one type of bus.

[0161] Based on the same inventive concept as the above-described method embodiments, embodiments of the present invention also provide a computer-readable storage medium for storing a computer program. When the computer program is run on a computer, it enables an electronic device to implement the control methods described in the above embodiments. The computer-readable storage medium can be a readable signal medium or a readable storage medium. A readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0162] Based on the same inventive concept as the above-described method embodiments, embodiments of the present invention also provide a computer program product. The computer program product includes a computer program that, when run on an electronic device, causes the electronic device to perform the steps of the control methods described above according to various exemplary embodiments of this application. The program product may take the form of any combination of one or more readable media. These computer program commands can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that the commands executed by the processor of the computer or other programmable data processing device generate a process for implementing... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0163] While the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Those skilled in the art can make various modifications and refinements without departing from the spirit and scope of the invention. Therefore, the scope of protection of the present invention shall be determined by the claims.

Claims

1. A method for improving the first-time acceptance rate of shipborne sand and gravel, characterized in that, include: Two phases of land area point clouds were collected before and after the excavation of the sand and gravel pile, and two phases of ship cabin point clouds were collected in the empty state of the transport ship and after loading. Point cloud denoising, point cloud registration, and 3D modeling were performed on the two phases of land area point clouds and the two phases of ship cabin point clouds, respectively. Based on the three-dimensional model corresponding to the two phases of land area point cloud, rectangular mesh integration is performed to obtain the land area storage area; The internal protruding structures of the ship's hull were identified and removed from the three-dimensional models corresponding to the point clouds of the two phases of the ship's cabin, and the loading party was obtained based on the volume difference of the removed models. The average volume, absolute value of deviation, and relative deviation are calculated based on the land storage party and the ship loading party. When the relative deviation and the absolute value of deviation are both within the preset error allowable range, the volume measurement consistency result is output. When at least one indicator exceeds the corresponding error allowable value, a recalculation prompt is output, including at least one of the following verification objects: scanning blind zone, registration error, modeling defect, and removal result of internal protruding structure of the hull.

2. The method for improving the first-time acceptance rate of shipborne sand and gravel according to claim 1, characterized in that, Two phases of land area point cloud data were collected, one before and one after the excavation of the sand and gravel pile, including: Using the boundary of the sand and gravel pile as a reference, a closed-loop scanning route is laid out along the outside of the sand and gravel pile so that the scanning start point and the scanning end point coincide. The laser scanner is moved along the closed loop scanning route to continuously scan the slope, top surface and bottom edge of the sand and gravel pile to obtain surface information of the excavation orientation of the sand and gravel pile. The excavated gap area of ​​the sand and gravel pile is repeatedly scanned at close range and from multiple angles to increase the point cloud density of the gap area; the excavated gap area includes the depression area, the occlusion area and the elevation change area formed by the excavation operation; the repeated scanning is used to supplement the missing point cloud of the excavated gap area and make its point cloud density higher than that of the non-gap area. The original point cloud before excavation and the original point cloud after excavation are output respectively. The original point cloud before excavation and the original point cloud after excavation include the three-dimensional coordinates of the points, the reflection intensity, and the scanning time.

3. The method for improving the first-time acceptance rate of shipborne sand and gravel according to claim 2, characterized in that, Point cloud denoising and point cloud registration are performed on the two phases of land area point clouds, including: A statistical outlier filtering algorithm is used to denoise the original point cloud before excavation and the original point cloud after excavation, wherein the number of neighborhood statistical points is... It is 15, a multiple of the standard deviation. It is version 2.

0. This represents the number of nearest neighbor points selected with the point to be processed as the center. This represents the standard deviation multiple used to identify outliers; Calculate the average distance between each point to be processed and its neighboring points. Points whose average distance exceeds twice the standard deviation are identified as noise points and removed to obtain two phases of land point cloud after noise reduction. The iterative nearest point algorithm is used to register the two phases of land point clouds after noise reduction. The maximum number of iterations is 1000, the convergence threshold is 0.001m, and the corresponding point search radius is 0.1m. During registration, firstly, fixed reference objects around the sand and gravel pile are selected to complete coarse registration. Then, the rotation matrix and translation vector are calculated by the iterative nearest point algorithm to unify the two phases of land point clouds into the same three-dimensional coordinate system. The fixed reference objects are ground control points, edges of surrounding structures, or fixed boundary features of the stockpile that can be identified in both phases of land point clouds and whose spatial positions do not change with the excavation of sand and gravel.

4. The method for improving the first-time acceptance rate of shipborne sand and gravel according to claim 3, characterized in that, The process of integrating rectangular meshes based on the 3D model corresponding to the two phases of land area point clouds to obtain the land area storage area includes: Based on the registered two phases of land point clouds, normal vector estimation, surface reconstruction and meshing are performed to generate a 3D model before excavation and a 3D model after excavation. The three-dimensional model before excavation and the three-dimensional model after excavation are projected onto the same two-dimensional plane and divided into multiple rectangular grid units according to the same bottom closed boundary. Interpolate the point cloud elevations within each rectangular grid cell to obtain the representative elevation values ​​before and after excavation for that rectangular grid cell, and calculate the land storage volume using the following formula: In the formula, For land-based storage, This represents the total number of rectangular grid cells. The rectangular grid cell number. For the first The horizontal projected area of ​​each rectangular grid cell For the first The rectangular grid cells represent elevation values ​​before excavation. For the first Each rectangular grid cell represents an elevation value after excavation.

5. The method for improving the first-time acceptance rate of shipborne sand and gravel according to claim 1, characterized in that, Collect point clouds of the ship's hold in both the empty and loaded states, including: In the empty ship state, first scan the external outline of the transport ship around the deck, then enter the interior of the ship's cabin and perform partitioned scanning of the bottom, bulkhead and corner areas to obtain the original point cloud of the empty ship. After the sand and gravel are loaded onto the ship, the same scanning route, scanning height and scanning angle as when scanning the ship in its empty state are maintained to scan the hull of the transport ship and the shape of the sand and gravel pile inside the ship's hold in order to obtain the original point cloud after loading. The original point cloud of the empty ship and the original point cloud after loading are output respectively, serving as the data basis for volume comparison of the loading party.

6. The method for improving the first-time acceptance rate of shipborne sand and gravel according to claim 5, characterized in that, Point cloud denoising, point cloud registration, and 3D modeling were performed on the point clouds of the two ship cabins, including: The original point cloud of the empty ship and the original point cloud after loading are denoised by using a statistical outlier filtering algorithm or a radius filtering algorithm to remove noise points, outliers and isolated points generated during the scanning process. The iterative nearest point algorithm is used to register the denoised empty ship point cloud and the loaded ship point cloud, so that the empty ship point cloud and the loaded ship point cloud are unified to the same three-dimensional coordinate system. The Poisson surface reconstruction algorithm is used to generate an empty ship 3D model and a loaded ship 3D model. The empty ship 3D model is used to represent the ship's cabin space when it is not loaded with sand and gravel, and the loaded ship 3D model is used to represent the ship's cabin space and sand and gravel pile shape after it is loaded with sand and gravel.

7. A method for improving the first-time acceptance rate of shipborne sand and gravel according to claim 6, characterized in that, The process of identifying and removing internal protruding structures within the hull of the 3D model corresponding to the point clouds of the two phases of the ship's cabin, and obtaining the loading party based on the volume difference of the removed models, includes: Using contour segmentation and mesh removal algorithms, the internal protruding structural volumes corresponding to beams, columns, bulkheads, stiffeners and pipes inside the ship's cabin are identified and removed to obtain the effective empty ship model and the effective model after loading. The volume of the empty ship effective model and the volume of the loaded ship effective model are calculated using the rectangular mesh integration method, respectively, and the loading volume is calculated according to the following formula: In the formula, For the shipper, This is the effective volume of the model after loading onto the ship. This represents the volume of the effective empty ship model.

8. The method for improving the first-time acceptance rate of shipborne sand and gravel according to claim 1, characterized in that, The average volume, absolute deviation, and relative deviation are calculated based on the land-based storage volume and the ship-loading volume, including: The average volume is calculated using the following formula: In the formula, For average volume, For land-based storage, For the shipper; The absolute value of the deviation is calculated using the following formula: In the formula, The absolute value of the deviation. For land-based storage, For the shipper; The relative deviation is calculated using the following formula: In the formula, This is a relative deviation. The absolute value of the deviation. This is the average volume.

9. A method for improving the first-time acceptance rate of shipborne sand and gravel according to claim 8, characterized in that, Output the consistency result of the volume measurement based on the preset error allowable range, or prompt for recalculation, including: When the relative deviation is not greater than 1.5% and the absolute value of the deviation is not greater than 10m 3 At that time, the output results are consistent between the calculations of the land-based storage party and the ship-loading party; When the preset application scenario is a high-precision control scenario for trade settlement, the relative deviation allowable value is set to no more than 1.0%; When the relative deviation is greater than the corresponding allowable relative deviation value, or the absolute value of the deviation is greater than 10m 3 If this happens, a prompt to recalculate will be displayed.

10. A method for improving the first-time acceptance rate of shipborne sand and gravel according to claim 9, characterized in that, After outputting a recalculation prompt, at least one of the following is corrected according to the recalculation prompt: scan point filling result, registration control point, filtering parameters, surface reconstruction parameters, integral mesh size, and internal protrusion structure removal result of the hull. Based on the corrected data, the land storage volume, loading volume, average volume, absolute value of deviation, and relative deviation are recalculated.

Images