CN112465976B - Storage yard three-dimensional map establishing method, inventory management method, equipment and medium - Google Patents
Storage yard three-dimensional map establishing method, inventory management method, equipment and medium Download PDFInfo
- Publication number
- CN112465976B CN112465976B CN202011467936.7A CN202011467936A CN112465976B CN 112465976 B CN112465976 B CN 112465976B CN 202011467936 A CN202011467936 A CN 202011467936A CN 112465976 B CN112465976 B CN 112465976B
- Authority
- CN
- China
- Prior art keywords
- tetrahedron
- model
- data
- image
- storage yard
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/05—Geographic models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T15/00—3D [Three Dimensional] image rendering
- G06T15/08—Volume rendering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
- G06T2207/10012—Stereo images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Computer Graphics (AREA)
- Remote Sensing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the invention provides a storage yard three-dimensional map establishing method, an inventory management method, equipment and a medium, wherein the storage yard three-dimensional map establishing method comprises the following steps: collecting sequence images of a pile position of a storage yard; acquiring point cloud data; generating digital elevation model data by using the point cloud data; forming an oblique image mapping model through the oblique image, and forming a building model, a container model and a facility model of a storage yard through mapping; and superposing the building model, the container model and the facility model of the storage yard to form a three-dimensional map of the storage yard. Due to the adoption of the scheme, the accurate positioning, the dynamic tracking, the process control and the visual management of port production elements and bulk cargos can be realized, the loading and unloading efficiency is improved, the operation cost is reduced, and the problems that the operation of a driver is not controlled, the field safety accident rate is high, the stacking is wrong, the loading is wrong, the operation instruction cannot be executed and the real-time monitoring is realized in the port production process are effectively solved, particularly the problem that the cargos are easily illegally stolen in the loading and unloading process.
Description
Technical Field
The invention relates to the technical field of digital image processing suitable for large goods yard management, in particular to a storage yard three-dimensional map establishing method, an inventory management method, equipment and a medium based on unmanned aerial vehicle surveying and mapping.
Background
The port area generally has the characteristics of large occupied area, various goods, more vehicles for coming in and going out and the like, and brings troubles to daily operators and port area managers for port management, such as uncontrollable operation of a truck operation driver in the port area, high site safety accident rate, wrong stack, wrong goods loading, incapability of executing operation instructions and real-time monitoring. Even if port areas put high labor costs, the following problems still exist in port management: personnel and vehicles entering and exiting the port area cannot quickly find the corresponding transaction places, so that the transaction efficiency is low and the customer satisfaction is low; meanwhile, the goods are complicated, the management of the goods is disturbed by accurately and quickly finding out the target goods and measuring and calculating the stock of various miscellaneous goods, and even more, the goods can be illegally stolen in the loading and unloading process.
Disclosure of Invention
The application aims to provide a yard three-dimensional map establishing method, a stock management method, equipment and a medium based on unmanned aerial vehicle surveying and mapping, which can realize accurate positioning, dynamic tracking, process control and visual management of port production elements and bulk cargos (coal, ore sand, grain and the like), improve loading and unloading efficiency, reduce operation cost, and effectively solve the problems of uncontrolled driver operation, high site safety accident rate, wrong stacking, wrong loading, incapability of executing operation instructions and real-time monitoring in the port production process, particularly the problem that cargos are easily illegally stolen in the loading and unloading process.
The embodiment of the invention provides a method for establishing a three-dimensional map of a storage yard, which comprises the following steps:
(1) collecting sequence images of a pile position of a storage yard;
the sequence images of the stacking position of the storage yard are acquired from a vertical angle and an inclination angle through a plurality of cameras carried on the unmanned aerial vehicle, all aerial photography aerial belts of the unmanned aerial vehicle are parallel to each other, the space between every two adjacent aerial photography aerial belts is equal, no less than 50% of overlapped detected areas are contained between any two adjacent sequence images in all aerial photography aerial belts, and no less than 50% of overlapped detected areas are contained between any two adjacent aerial photography aerial belts which are parallel to each other;
(2) acquiring point cloud data of a storage yard;
extracting homonymous feature points contained in an overlapping region in adjacent sequence images in the same aerial photography zone and homonymous feature points contained in an overlapping region in adjacent images in the adjacent aerial photography zone in the sequence images through an image matching algorithm, performing Euclidean reconstruction on the obtained homonymous feature points, performing triangularization processing in a three-dimensional space, and generating point cloud data for placing a stock dump from the sequence images of the stock dump;
(3) generating digital elevation model data by using point cloud data of a storage yard;
extracting characteristic points, lines and surfaces from the point cloud data, simultaneously obtaining volume information through a tetrahedral subdivision volume algorithm based on convex hull segmentation, and interpolating a network to generate digital elevation model data; obtaining an initial building model, an initial container model and an initial facility model of a storage yard;
(4) forming an oblique image mapping model through the oblique image, and extracting the oblique image of the building, the oblique image of the container and the oblique image of the storage yard facility from the oblique image;
(5) mapping the inclined image of the building, the inclined image of the container and the inclined image of the storage yard facility to an initial building model, an initial container model and an initial facility model through an inclined image mapping model to form a building model, a container model and a facility model of the storage yard;
(6) and superposing the building model, the container model and the facility model of the storage yard to form a three-dimensional map of the storage yard.
In an embodiment, the tetrahedron subdivision volume algorithm based on convex hull segmentation specifically includes:
(3.1) partitioning a tetrahedron based on convex hull partitioning;
(3.2) calculating normal vectors of four vertexes of the tetrahedron based on a fitted surface, and unifying the directions of the normal vectors;
(3.3) if the ray pointing along the normal vector of each vertex of one tetrahedron has an intersection point with the external sphere of the tetrahedron, determining that the tetrahedron is an external tetrahedron; otherwise, judging the tetrahedron to be an internal tetrahedron;
(3.4) the sum of the volumes of all the in vivo tetrahedra is the result volume.
In one embodiment, the step (3.1) of the tetrahedron subdivision based on convex hull segmentation specifically includes:
(3.1.1) constructing an initial tetrahedron based on the point cloud data obtained in the step (2) to form an initial tetrahedron grid;
(3.1.2), inserting scattered points in the point cloud data acquired in the step (2) as input points into a current tetrahedral mesh, and for the input points, using a random walking method to find a tetrahedron containing the input points;
a tetrahedron is first designated, and if the input point is located within the tetrahedron, walking is completed. If not, randomly assigning a triangle face, if the plane of the triangle face divides the tetrahedron from the input point, the next accessed tetrahedron is the adjacent tetrahedron sharing the triangle face; otherwise, traversing the other faces in a predetermined order until a face is found that separates the tetrahedron and the input point;
(3.1.3), finding the tetrahedron containing the input point, then dividing the tetrahedron into 4 small sub-tetrahedrons;
(3.1.4) selecting a visible face of the mesh if the input point is outside the current tetrahedral mesh; connecting the input point with three vertexes of the visible surface to form a new tetrahedron, and adding the new tetrahedron into the tetrahedral mesh;
(3.1.5) repeating the steps (3.1.2) to (3.1.4) until all scattered points in the point cloud data acquired in the step (2) are inserted into the tetrahedral mesh;
(3.1.6), verifying the effectiveness of the Delaunay triangulation;
the continuity of the Delaunay triangulation data structure, i.e. the adjacency of tetrahedrons, is first checked. The orientation of each tetrahedron and the correctness of the convex hull obtained by the Delaunay triangulation are then verified.
In one embodiment, in step (3.4), the sum of the volumes of all the in-vivo tetrahedra is the result volume, specifically, the result volume of the object can be obtained by calculating the sum of the volumes of all the tetrahedra according to the following formula:
In one embodiment, the step (4) of forming the oblique image mapping model by the oblique image includes performing orthorectification, image mosaicing, image processing, and digital orthoimage cropping on the oblique image in sequence to form digital orthoimage data.
In one embodiment, the method for building the three-dimensional map of the storage yard further comprises the following steps: collecting road network data, and forming road network vector data by using the digital orthographic image data; acquiring vector data, and forming an image data digital drawing by superposing the vector data and the digital orthographic image data; carrying out data layering and coordinate conversion on the vector data to form a vector map of a unified coordinate system; respectively carrying out coordinate conversion processing on the digital ortho-image data and the digital elevation model data to form data of a unified coordinate system; and carrying out data fusion on the building model, the container model, the facility model, the vector map, the digital orthographic image data of the unified coordinate system, the digital elevation model data of the unified coordinate system and the oblique image of the yard to form the three-dimensional map of the yard.
The embodiment of the invention also provides an inventory management method, which comprises the step of performing data interaction between the three-dimensional map of the storage yard established by using the storage yard three-dimensional map establishing method and the vehicle-mounted terminal stack running APP, the resource graphical scheduling operation monitoring management system and the basic information management system.
The embodiment of the invention also provides computer equipment which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor realizes the arbitrary method for establishing the three-dimensional map of the storage yard when executing the computer program, so that the technical problems that the operation vehicles in the prior art are frequently misplaced, the execution of scheduling instructions cannot be monitored in real time, the operation process of vehicles in ports cannot be monitored, and the change of goods cannot be updated in real time are solved.
The embodiment of the invention also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program for executing the arbitrary yard three-dimensional map building method, and solves the problems that the working vehicle is frequently misplaced, the execution of a scheduling instruction cannot be monitored in real time, the working process of the vehicle in a port cannot be monitored, and the change of goods cannot be updated in real time.
According to the yard three-dimensional map establishing method, the inventory management method, the equipment and the medium, the operation vehicle is quickly guided to quickly reach the yard stacking position through accurate positioning and navigation of the vehicle-mounted terminal and the yard stacking position, and real-time monitoring of the whole process of the operation vehicle in a port is realized; through the real-time seamless connection with the port production business system, the real-time transmission of operation instructions of dispatching personnel and mechanical drivers is realized, the wharf operation efficiency is effectively improved, and the intelligent management level of a wharf yard is greatly improved; meanwhile, real-time mapping of a port yard is achieved through unmanned aerial vehicle surveying and mapping, real-time updating and visual management of yard stacking position information are guaranteed, and the technical problems that in the prior art, operation vehicles often move in a staggered mode, scheduling instructions cannot be executed in real time, operation processes of vehicles in the port cannot be monitored, and cargo changes cannot be updated in real time are solved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principles of the invention. In the drawings:
fig. 1 is a schematic flow chart of a method for establishing a three-dimensional map of a yard according to an embodiment of the present invention;
fig. 2 is a schematic flow chart of a specific implementation of a method for building a three-dimensional map of a yard according to an embodiment of the present invention;
fig. 3 is a schematic diagram of image control point arrangement in an irregular area network according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a process for extracting data using point cloud data according to an embodiment of the present invention;
FIG. 5 is a schematic flow chart of obtaining three-dimensional model data according to an embodiment of the present invention;
FIG. 6 is a schematic view of a road network data collection process according to an embodiment of the present invention;
FIG. 7 is a schematic view of a vector data collection process provided by an embodiment of the present invention;
FIG. 8 is a schematic view of a vector map processing flow provided by an embodiment of the present invention;
FIG. 9 is a schematic view of a digital ortho image data processing flow according to an embodiment of the present invention;
FIG. 10 is a schematic diagram of a three-dimensional map processing flow provided by an embodiment of the invention;
fig. 11 is a flowchart illustrating an inventory management method based on a three-dimensional map of a yard according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the following embodiments and accompanying drawings. The exemplary embodiments and descriptions of the present invention are provided to explain the present invention, but not to limit the present invention.
In an embodiment of the present invention, a method for establishing a three-dimensional map of a yard is provided, as shown in fig. 1, the method includes:
step (1), collecting sequence images of a pile position of a storage yard;
step (2), point cloud data of a storage yard are obtained;
step (3), generating digital elevation model data by using the point cloud data of the storage yard;
step (4), forming an oblique image mapping model through the oblique image, and extracting the oblique image of the building, the oblique image of the container and the oblique image of the yard facility from the oblique image;
step (5), mapping the inclined image of the building, the inclined image of the container and the inclined image of the yard facility to an initial building model, an initial container model and an initial facility model through the inclined image mapping model to form a building model, a container model and a facility model of the yard;
and (6) superposing the building model, the container model and the facility model of the storage yard to form a three-dimensional map of the storage yard.
In one embodiment as shown in FIG. 2, sequential images of the yard berth are acquired for step (1). This project utilizes a plurality of cameras of settling on unmanned aerial vehicle load platform, from vertical angle and inclination collection piling position's sequence image simultaneously, in the collection process, each aerial photography of unmanned aerial vehicle is parallel to each other between the area, the interval between each adjacent aerial photography area is equal, include not less than 50% overlapping measured area between arbitrary two adjacent sequence images each other in each aerial photography area, also include not less than 50% overlapping measured area between arbitrary two adjacent aerial photography areas that are parallel to each other.
In view of the fact that ports are mostly arranged in urban areas and more buildings exist, the automatic matching difficulty of digital images caused by poor projection needs to be considered in aerial photography flight, and therefore course overlapping degree can be properly increased in the image acquisition process, and the quality and precision of aerial survey results can meet the requirements of post data processing. For example, for unmanned aerial vehicle shooting, the ground resolution of an aerial image of the unmanned aerial vehicle is not less than 0.08 m, and/or the average course overlapping degree is not less than 75%, the average lateral overlapping degree is not less than 50%, and/or the difference of the altitudes of adjacent pictures on the same route is not more than 30 m, the difference between the maximum altitude and the minimum altitude is not more than 50 m, and the difference between the actual altitude and the design altitude is not more than 50 m; when selecting the photo, the inclination angle of the photo is generally not more than 5 degrees, the maximum is not more than 12 degrees, the number of the photo which appears more than 8 degrees is not more than 10 percent of the total number, for the particularly difficult area, the inclination angle of the photo is generally not more than 8 degrees, the maximum is not more than 15 degrees, the number of the photo which appears more than 10 degrees is not more than 10 percent of the total number, the rotation angle of the photo is generally not more than 15 degrees, and the maximum is not more than 30 degrees on the premise that the course and the lateral overlapping degree of the photo meet the standard requirements; the number of the image slices reaching or approaching the maximum rotation deviation angle limit difference on one flight line is not required to continuously exceed three; the number of the image slices with the maximum deflection angle in one shooting area is not more than 10% of the total number of the image slices in the shooting area. The image collected by the project is clear, moderate in contrast, saturated in color, bright in color and consistent in tone, has rich levels, and can distinguish fine ground object images adaptive to ground resolution. According to the project field collection, the complementary shooting and the repeated shooting are avoided, the relative loopholes and the absolute loopholes in the aerial shooting are required to be compensated and shot in time, the digital camera of the previous aerial shooting flight is adopted for compensation shooting, and the two ends of the compensated shooting route carry out the complementary shooting according to two baselines beyond the loopholes according to the actual conditions.
This project adopts unmanned aerial vehicle aerial survey, need not carry out the photography partition. And designing a route according to 4 angular points in the range of the survey area. The method is characterized in that the ground station system is designed according to basic requirements of aerial photography, the four-corner coordinates of each photographic partition are respectively input into ground station software, the distance of a flight path and the direction of flight are set, and software provided by the project can automatically generate flight path files. After the photo meeting the requirements is obtained, the radiation correction and the geometric correction are carried out on the photo, and a combined image of the route and the photo is made.
For the stationing of the unmanned aerial vehicle aerial photography area, it should be noted that: 1) the area network should not include the route and the image pairs which do not meet the requirement of image overlapping, and should not have large cloud images, shadows and other image pairs which influence the connection of the route model; 2) laying a plane net or a plane high net, laying control points according to 8 base lines of the course and a base line at a lateral interval, and properly widening the number of the base lines in the vegetation covered hidden area; 3) the irregular area net is characterized in that flat high points are added at convex angles, and high points are added at concave angles; however, when the distance between the concave-convex corners exceeds 4 base lines, the concave-convex corners should be distributed with flat-high points, as shown in fig. 3, wherein the hollow circles represent the flat-high control points, and the solid circles represent the elevation control points. Area distribution points for some special cases, for example: 1) arranging points at the junction of the aerial photography area; control points at the joint of the aerial photography areas are laid at the overlapping joint of the air routes, and adjacent areas are as common as possible; if the public requirement can not be met, points are distributed respectively; 2) arranging points when the courses are not overlapped enough; when the course overlapping part is less than 53 percent of the overlapping degree, the digital topographic map is regarded as the relative aerial photograph leak, points are distributed respectively, and the digital topographic map is subjected to field compensation measurement at the absolute leak (when the aerial photograph can not be repaired); 3) arranging points when the side direction overlapping is not enough; when the side overlapping part is less than 15% of the overlapping degree, points are distributed respectively; if the overlapped part is larger than 1cm on the image slice, the image is clear, and no important ground object exists in the range, 2-3 elevation points can be additionally measured at the overlapped part, otherwise, the digital topographic map is actually measured at the part which is not overlapped enough.
For selecting thorns from image control points, firstly, selecting thorns from the image control points on intersection points of linear ground objects with good intersection angles and on the basis of targets which have small elevation changes and are convenient for joint measurement and clear images on adjacent images, wherein the targets are at the centers of point ground objects with images smaller than 0.2 mm; the spot position judgment accuracy is 0.1mm on the graph. The puncture hole at the photo site should be punctured through. When the prick point is deviated, the sheet should be replaced and pricked again or a description should be made; checking the spot position with a target after puncturing the hole, drawing a sketch and compiling a spot position description; secondly, when the point is selected to be higher than or lower than the ground by more than 0.2m, the point is supposed to be located above or below the threshold, and the height of the point to the larger reference ground is required to be higher and is noted to be 0.1 m; furthermore, the control sheet only finishes the thorn point sheet; the common points of the same route and the adjacent routes are only subjected to label conversion, and the common points of the adjacent areas are subjected to label conversion only for avoiding the error of the rotating prick; in addition, a flat high point anterior crown P and a high point anterior crown G in the image control points; and also comprises finishing of the image control points.
The precision of the image control point comprises the precision of an image control measurement plane and the precision of an image control measurement elevation. Wherein, the precision of the photo control measurement plane is that the error in the plane position of the plane control point and the flat height control point relative to the adjacent basic control point does not exceed 1/5 of the error in the plane position of the ground object point; the elevation precision of the photo control measurement is that the error in the elevation of the photo elevation control point and the flat height control point relative to the adjacent basic control point does not exceed 1/10 of the basic equal height distance.
The image control point joint measurement is carried out by adopting a GPS-RTK operation mode, and the specific requirements comprise: 1) RTK operation can adopt a CORS operation mode; 2) measuring by adopting a method of repeating observation for two times, and taking an average value of the three-dimensional space encryption; 3) under special conditions, such as the condition that a building cannot reach the top of the building for measurement, GPS signal difference, a wall corner which cannot reach the building and the like, point guiding method measurement can be adopted, namely, a pair of mutually through-looking map root points are arranged by using GPS-RTK, then a total station is matched with a small prism, a station is arranged on one map root point, and after the other point is checked to be correct, the coordinates of the image control points are determined.
In the embodiments illustrated in the present application, the measurement of the topographic map may be performed by prior art aerial triangulation. The digital photogrammetry workstation can be adopted, the photo data is used as the original data of image control encryption, relative orientation is carried out on the photo one by one, and the integral adjustment of the area network by the light beam method is carried out after the absolute orientation is carried out according to the area network, so that the encryption point result is obtained. When the encryption point selects the thorn, the standard point is selected when the connection point (standard point) required by the encryption point cannot be shared with the image control point; the connection point of the encryption itself should be selected near the position of the specified six standard points of 1, 3, 5, 2, 4 and 6, wherein the 1,2 points are selected within 1cm from the image main point, and if the individual difficulty is within 1.5cm, the 3, 4, 5 and 6 points should be consistent with the mapping orientation point, and the distance from the orientation line should be approximately equal and greater than 5 cm. When the lateral overlapping is too large, and the distance between the connecting point and the azimuth line value is not less than 5cm, the standard point is selected at a position 8-10cm away from the azimuth, or the distance passing through the main point and being perpendicular to the azimuth line is not more than 1 cm; when the side image overlap is too small, selecting points at the overlapping central line and the measurement accuracy is difficult to ensure, selecting points respectively, wherein the sum of the two points to the side overlapping central line is not more than 1.5 cm; the point is more than 1mm from each mark. In addition, the free graph edge should take care of the mapping area, and the point is selected outside the mapping range line. The rest of the details can be set according to the storage yard situation in actual use, and are not described herein again.
For the step (2), point cloud data of the storage yard is obtained; processing a sequence image of a stock yard stack position obtained by acquiring an aerial image of the unmanned aerial vehicle by using an image matching algorithm, extracting homonymous feature points contained in an overlapping region in adjacent sequence images in the same aerial image band and homonymous feature points contained in an overlapping region in adjacent images in the adjacent aerial image band in the sequence image, performing Euclidean reconstruction on the obtained homonymous feature points, performing triangularization processing in a three-dimensional space, and generating point cloud data for placing the stock yard by using the sequence image of the stock yard.
How to extract production DEM (digital elevation model) data and three-dimensional model data by using the point cloud data is explained below, and the overall flow can refer to fig. 4.
And (3) generating Digital Elevation Model (DEM) data by using the point cloud data of the storage yard.
Extracting characteristic points, lines and surfaces from the point cloud data of the storage yard acquired in the step (2), acquiring volume information through a tetrahedral subdivision volume algorithm based on convex hull segmentation, and generating DEM data through network construction interpolation; then, mapping the DEM grid points to an image three-dimensional model, and checking and observing to ensure that each DEM grid point is close to the ground; when the elevation error of the DEM grid point exceeds the limit and needs to be edited, a contour line and an elevation mark point with problems are searched for correction and measurement, or a characteristic point and a characteristic line are added, and the DEM data is generated by interpolation of a reconstruction network.
The tetrahedron subdivision volume algorithm based on convex hull segmentation specifically comprises the following steps:
step (3.1), dividing a tetrahedron based on convex hull segmentation;
in this application, convex hull refers to the set of points
The minimum convex polyhedron of the line segment formed by connecting any two points; the Delaunay rule is that no other scattered points are contained in a circumscribed sphere of a tetrahedron in the three-dimensional space; the degradation phenomenon is that in a floating-point number system, a topological error is caused because the numerical precision of a computer is limited.
For a convex hull, for example, in the initial building model of the yard, the convex hull is the smallest convex polyhedron containing a line segment connecting any two points in any one building model in the yard; in the initial container model, the convex hull is a minimum convex polyhedron comprising a line segment formed by connecting any two points in any container model; in the initial facility model, the convex hull is a minimum convex polyhedron including a line segment connecting any two points in any one facility model.
Then, the tetrahedron subdivision based on convex hull segmentation in step (3.1) specifically includes:
step (3.1.1), constructing an initial tetrahedron based on the point cloud data obtained in the step (2) to form an initial tetrahedron grid;
and (3.1.2) inserting scattered points in the point cloud data acquired in the step (2) as input points into the current tetrahedral mesh, and for the input points, searching for tetrahedrons containing the input points by using a random walking method. A tetrahedron is first designated, and if the input point is located within the tetrahedron, walking is completed. Randomly designating a triangle if not within the tetrahedron, if the plane of the triangle divides the tetrahedron and the input point (i.e. the tetrahedron and the input point are on both sides of the plane of the triangle), the next accessed tetrahedron being the neighboring tetrahedron sharing the triangle; otherwise, other faces are traversed in a predetermined order until a face is found that separates the tetrahedron and the input point.
Step (3.1.3), finding the tetrahedron containing the input point, then the tetrahedron is divided into 4 small sub-tetrahedrons.
Step (3.1.4), if the input point is located outside the current tetrahedral mesh, selecting a visible face of the mesh (i.e. the input point is at one side of the visible face), and connecting the input point and 3 vertices of the visible face to form a new tetrahedron to be added to the tetrahedral mesh; it is reminded here that when choosing the visible surface, it is endeavored to avoid making the newly generated tetrahedron long and narrow.
And (3.1.5) repeating the steps (3.1.2) to (3.1.4) until all scattered points in the point cloud data acquired in the step (2) are inserted into the tetrahedral mesh.
And (3.1.6) verifying the effectiveness of the Delaunay triangulation. The continuity of the Delaunay triangulation data structure, i.e. the adjacency of tetrahedrons, is first checked. The orientation of each tetrahedron and the correctness of the convex hull obtained by the Delaunay triangulation are then verified.
The improvement of the tetrahedron subdivision volume algorithm based on convex hull segmentation mainly lies in the walking method in the step (3.1.2). The traditional tetrahedron subdivision volume algorithm uses a linear walking method. For an input point P, a tetrahedron is specified, and if the input point P is located in the tetrahedron, walking is completed; if not, a ray is constructed with an origin that is a point in the current tetrahedron, labeled C, oriented C → P, and a tetrahedron intersecting the ray is located, the tetrahedron adjacent to and sharing the current tetrahedron being the next candidate tetrahedron containing P. This method, although fast, has the drawback that the next tetrahedron intersected by the ray when it passes through the vertex or edge of the tetrahedron is not the adjacent tetrahedron, resulting in a positioning error of P, which is effectively avoided by the random walk method.
Step (3.2), normal vectors of four vertexes of the tetrahedron are obtained, and the pointing directions of the normal vectors are uniformized;
since the tetrahedral subdivision is essentially the segmentation of the convex hull, the tetrahedral mesh obtained in step (3.1) includes extra-corporeal tetrahedrons as redundant data, which need to be removed. And because the scattered points in the point cloud data do not contain any topological information, and it is difficult to directly determine the external tetrahedron, the positions of the tetrahedrons are described by calculating the normal vector of the top points of the tetrahedrons of the scattered points, so that the external tetrahedrons are removed. The step (3.2) specifically comprises the following steps:
step (3.2.1), calculating normal vectors of four vertexes of the tetrahedron based on a fitted surface;
to calculate the normal vector of each of the vertices, a quadratic surface is used to fit the point to its K neighbors (meaning a data set containing n points under Euclidean distance)
Find K nearest points to a point). The quadratic surface parameters are calculated using a least squares method, and the quadratic surface equation can be expressed as:
z＝ax2+by2+cxy+dx+ey+f
for a given point of scatter (x)i,yi,zi) 1,2, …, N, minimizing the total error Q:
the solution of the problem can be summarized as the problem of solving the extremum of the hexabasic function Q (a, b, c, d, e, f), namely a, b, c, d, e, f satisfies:
therefore, parameters of the quadric surface equation can be obtained, the quadric surface equation is further obtained, and then the normal vector of the point can be calculated through partial differentiation.
Step (3.2.2), carrying out normal vector unification;
because the initial normal vector points differently from inside to outside, the problem that the initial normal vector needs to be adjusted in a consistent mode after being obtained can be modeled as optimization of a graph. The nodes of this graph are made up of scattered points, and if two nodes are close enough, an edge is constructed whose weight is given by ni*njIt is shown that the consistency of the normal vectors can be reduced to the problem of maximizing the total weight of the graph, and the main problem here is to select which pair of nodes in the connected graph is. Since the object surface can be considered as a single connected element, the graph should be connected. The minimum spanning tree is a simple connected graph connecting adjacent points, so that the minimum spanning tree is established based on the Euclidean distance of the scattered points. Since the density of the initial minimum spanning tree edges cannot meet the requirements of the text, edges need to be added, if one point in two nodes is a K neighbor of the other point, the two points are connected, and the graph generated by the method is called a Riemanniangraph (Riemanniangraph); then, a minimum spanning tree is generated based on the graph, a seed normal vector is selected, and the direction of the normal vector is propagated and adjusted in the minimum spanning tree according to a depth priority criterion. The adjustment rule is as follows: for two neighbors in the minimum spanning tree, if the dot product of the corresponding normal vectors is ni*njIf < 0, then niOr njShould be covered byAnd reversing.
The time complexity of this step is O (n)2) The selection of the number K of the neighbors is an important factor influencing the accuracy of the normal vector, and the K can be set to be an adjustable variable due to different complexity of the three-dimensional topology of the object. According to experience, K can take a small value in the case of most surfaces being relatively smooth; if the surface texture is complex, K may take on a larger value.
Step (3.3), if the ray pointing along the normal vector of each vertex of one tetrahedron has an intersection point with the external sphere of the tetrahedron, determining that the tetrahedron is an external tetrahedron; otherwise, judging the tetrahedron to be an internal tetrahedron;
after the normal vector of each vertex of the tetrahedron is obtained, the external tetrahedron and the internal tetrahedron can be judged through the direction of the normal vector. The normal vectors of the vertices of the tetrahedron in vivo all point away from the tetrahedron, while the normal vectors of the vertices of the tetrahedron in vitro point to the interior or vicinity of the tetrahedron. Experiments show that on the premise of obtaining a correct normal vector, an ideal result can be obtained by eliminating an external tetrahedron by using whether the normal vector of the vertex intersects with the external tetrahedron. The specific algorithm is as follows: traversing tetrahedrons, calculating an external sphere of each tetrahedron, then leading out a ray from each vertex of the tetrahedron along the direction of a normal vector, if the rays of 4 vertices have an intersection point (not including the origin of the ray) with the external sphere, considering the tetrahedron as an external tetrahedron, and not considering (if the rays of each vertex pointing along the normal vector and the external sphere of the tetrahedron have intersection points, namely the rays of 4 vertices pointing along the normal vector and the external sphere of the tetrahedron have 4 intersection points, judging the tetrahedron as an external tetrahedron); otherwise, the volume of the tetrahedron is calculated and the total volume is added.
The time complexity of the step is O (n), and a large number of programs verify that most external tetrahedrons can be removed by the tetrahedron subdivision volume algorithm based on convex hull segmentation under the premise that the basically correct normal vector can be obtained, so that the feasibility of the method is proved. And (3.4) the sum of the volumes of all the internal tetrahedrons is the result volume.
The resulting volume of the object is obtained by calculating the sum of the volumes of the tetrahedra according to the following formula:
since the traditional tetrahedron subdivision algorithm is mainly used in the process of constructing the surface network model, it is time-consuming to calculate the volume of the object by using the traditional tetrahedron subdivision algorithm. The method comprises the steps of selecting a Poisson reconstruction method to reconstruct a surface network model, expressing a surface reconstruction problem as a solution of a Poisson equation, reconstructing a seamless triangular approximation for the surface by estimating an indication function of the model and extracting an equivalent surface, so that a plurality of methods for fitting an implicit surface separate data into different regions to perform local fitting, and then further synthesizing the local fitting results by using a synthesis function.
Aiming at the process of firstly constructing a surface network model and then calculating the volume of an object in the traditional algorithm, the tetrahedral subdivision volume algorithm based on convex hull segmentation provided by the application is a method for directly calculating the volume of the object. The algorithm greatly improves the efficiency and enables real-time rapid volume calculation to be possible on the premise of ensuring the accuracy through the steps of tetrahedron subdivision based on convex hull segmentation, normal vector calculation, external tetrahedron elimination, volume calculation and the like.
And (4) forming an oblique image mapping model through the oblique images, and extracting the oblique images of the building, the container and the yard facility from the oblique images.
Wherein, forming the oblique image mapping model by the oblique image specifically comprises:
1) performing image orthorectification;
(1) adopting a piece-by-piece correction mode, preferentially selecting images with small projection difference of the center of the photo for splicing;
(2) the visual effect contradiction caused by the projection difference of buildings and tall trees on different air films is eliminated through reasonable and effective image mosaic;
(3) the integrity of the overpass, the integrity of the ground features and the correctness of the relative relationship are maintained.
2) Image embedding;
the generated ortho-images are corrected and mosaicing is required to ensure the integrity of linear and planar ground objects and realize seamless splicing of adjacent ortho-images. The embedded lines are drawn out along natural ground features as much as possible, such as image transformation positions of roads, rivers, tree seams and the like, so that the embedded lines are prevented from passing through houses, land parcels, ground features higher than the ground and the like, the images of the embedded ground features are ensured to be complete, continuous, seamless and consistent in vision, and no obvious photo splicing traces exist.
3) Processing an image;
the method specifically comprises the following steps:
3.1) introducing a null-triple encryption result to establish a test area file and restore the model.
And 3.2) defining the operation area of the single model, generating an epipolar line image, and matching the epipolar line image to form a matching point and an equal parallax curve. The working area should be determined as close as possible to the connecting line of the control points to prevent cracks from appearing between the image pairs.
3.3) checking the matching result, and carrying out interactive three-dimensional editing (area editing and point editing) processing as required. The key points are as follows: high-rise building areas, image fuzzy areas, shadow areas, large-area water areas, building dense areas, forest coverage areas, valley and ridge terrain transformation positions and the like. If local matching has a problem, the relative orientation should be returned, the relative orientation point is added at the part where the problem occurs, or the characteristic point and the characteristic line are added in the matching preprocessing.
3.4) digital differential correction
And correcting and resampling the image by adopting a differential correction method according to the internal and external orientation elements and the image resolution of the image to generate a single-model DOM.
3.5) adjustment of the hue or color
Before the image is embedded, the hue or color deviation between adjacent films is checked, and an image processing method is adopted to adjust the hue or color deviation according to needs so as to enable the hue or color deviation to be basically consistent.
3.6) mosaic splicing
Setting mosaic range according to the figure outline coordinate and appointing a file storage path. And executing the image mosaic command to automatically compose a DOM of the whole frame.
After the image embedding is finished, the generated DOM is carefully checked, the positions of the parts with image blurring and image omission in the boundary area are repaired, and the image embedding in the urban area is carried out by adopting a manual method, so that the image distortion and deformation of high-rise buildings caused by the inconsistency of the projection directions are prevented.
3.7) DOM Format conversion
And converting the DOM from the internal format to a GeoTIFF format.
3.8) edge joining inspection
The method comprises model edge connection and map edge connection inspection. Mainly checking whether the transition of the images at the joint of the model is natural or not and whether a blank gap exists or not; whether the brightness, contrast and color of the image at the joint of the picture frame are basically consistent or not, whether the distance of the same-name point is over-limit or not and the like.
4) Digital ortho-image (DOM) cropping
As shown in fig. 5, generating three-dimensional model data using the sequence images acquired in step (1) specifically includes: a) image enhancement, namely performing image enhancement such as denoising, restoration and the like on the sequence image; b) point cloud calculation and registration, namely analyzing the image and solving transformation parameters among frames; then, taking the public part of the scene as a reference, overlapping and matching the multi-frame images acquired at different time, angles and illumination intensities into a unified coordinate system to complete the registration of the images; then calculating corresponding translation vectors and rotation matrixes, and eliminating redundant information (point cloud calculation and registration can restrict the speed of three-dimensional reconstruction and also influence the fineness and the global effect of a final model); c) data fusion, namely the registered depth information is still point cloud data scattered and disordered in the space and only can show partial information of scenery, so that the point cloud data needs to be fused to obtain a more refined reconstruction model; d) generating a surface, namely generating a three-dimensional model surface by using a real scene picture acquired by an oblique image as a three-dimensional model map and texture mapping so as to construct a visual isosurface of an object; e) and outputting the three-dimensional result.
In addition, the embodiment of the method for establishing the three-dimensional map of the storage yard provided by the application further comprises the following steps:
1. road network data are collected, and road network vector data are formed by utilizing the digital orthographic image data. The method specifically comprises the steps of collecting and producing port and wharf road network data through a data tool by utilizing high-resolution digital ortho-image data to form a road network vector data result, and checking and correcting the road direction, the road name and whether the road is in a single-row or double-row state or not by combining a field verification mode. Finally, the road network data result is formed. A schematic flow diagram of a road network data acquisition process is shown in fig. 6.
2. And acquiring vector data, and forming an image data digital drawing by superposing the vector data and the digital orthographic image data. The digital orthographic image data with high resolution is utilized, the image data is digitally drawn through data tool superposition, the digital orthographic image data mainly comprises main geographic entity information such as roads, railways, buildings, vegetation, water systems, cargo accumulation ranges and the like, and basic facilities such as wharf ranges, breakwaters, introduction embankments, revetments, harbor ponds, anchor lands, storage yards, warehouses, railway and loading and unloading mechanical tracks, bridges and protective facilities can be preliminarily vectorized. And then, the field feature elements which can not be clearly identified through images, such as a dock range, a breakwater, a guide bank, a revetment, a harbor pool, an anchor site, a storage yard, a warehouse, a railway and loading and unloading mechanical track, a bridge, a protection facility and the like, and corresponding attribute information are collected and verified by the field. And identifying the feature points with the labeling function on the map, and obviously identifying to form the point of interest data. A schematic view of the vector data acquisition process is shown in fig. 7.
Meanwhile, data processing work such as reasonable layering, coordinate conversion, mapping and symbolization is carried out on the collected high-precision vector data (including roads, buildings, water systems, warehouses, interest points and the like), and a high-precision vector map with a unified coordinate system is formed. And the road data and the interest point data are formed into vector map data by symbolizing and superposing image data. The specific processing flow of the vector map can be seen in fig. 8.
3. And respectively carrying out coordinate conversion processing on the digital ortho-image data and the digital elevation model data to form data of a unified coordinate system. The digital orthogonal image data of the entire image data is subjected to coordinate conversion processing to form digital orthogonal image data of a unified coordinate system, as shown in fig. 9. The digital elevation model data of the whole image data is subjected to coordinate conversion processing to form digital elevation model data of a unified coordinate system, and the specific process can be referred to fig. 9.
4. And carrying out data fusion on the building model, the container model, the facility model, the vector map, the digital orthographic image data of the unified coordinate system, the digital elevation model data of the unified coordinate system and the oblique image of the yard to form the three-dimensional map of the yard. The method comprises the following steps of fully utilizing acquired and produced DOM data, DEM data, three-dimensional model data and oblique image data to construct a port live-action three-dimensional map, wherein the three-dimensional map comprises the following contents: three-dimensional models such as a port earth surface model, a port building model, a container and facility model, a yard model and the like. And forming a real three-dimensional map through data fusion of the DOM, the DEM, the oblique image and the three-dimensional model. The three-dimensional map processing flow is as shown in fig. 10.
The embodiment of the application also provides an inventory management method based on the three-dimensional map of the storage yard, which specifically comprises the step of carrying out data transmission between the three-dimensional map of the storage yard established by the method for establishing the three-dimensional map of the storage yard and a vehicle-mounted terminal stack running APP, a resource graphical scheduling operation monitoring and management system and a basic information management system, and referring to fig. 11.
The vehicle-mounted terminal stack running APP, the resource graphical scheduling operation monitoring and management system and the basic information management system form a visual storage yard system of a port.
1. Vehicle-mounted terminal runs buttress APP. Including vehicle-mounted terminal applications and hand-held jogging applications. The system mainly provides a function of accurately positioning, navigating and finding an operation area for a transport operation vehicle in a port and wharf, provides position service in the port for an operation driver in real time, is connected with a production service system, and receives an operation instruction issued by the system; meanwhile, updating and maintaining of the stack position information are achieved, and accurate management is carried out on the stack position. And (4) performing stack running operation according to the change condition of the stack position, and confirming the position on the stack position boundary, thereby automatically producing accurate data of the stack position boundary.
2. A resource graphical scheduling job monitoring management system. The method mainly comprises the steps of obtaining a three-dimensional model of a storage yard through unmanned aerial vehicle aerial imaging, displaying the three-dimensional model on a graphical interface, and calculating the volume of goods through a three-dimensional storage yard model so as to obtain the goods and the inventory of each storage yard; meanwhile, the monitoring and displaying of each resource in the port area and the area monitoring and alarming are also carried out. And the resource graphical scheduling operation monitoring and management system is provided with a software and hardware structure, so that interactive butt joint with a vehicle-mounted terminal stack running APP system operation instruction can be realized, and issuing transmission of a vehicle operation instruction and acquisition of position coordinates and a visual graph after stack position updating are realized.
3. A basic information management system. The method mainly manages all vehicle-mounted terminal equipment information, electronic fence information, issued instruction information, yard stacking position information, registered user information, department information, role authority information and the like.
The embodiment of the invention also provides computer equipment which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor realizes the arbitrary method for establishing the three-dimensional map of the storage yard when executing the computer program, so that the technical problems that the operation vehicles in the prior art are frequently misplaced, the execution of scheduling instructions cannot be monitored in real time, the operation process of vehicles in ports cannot be monitored, and the change of goods cannot be updated in real time are solved.
In particular, the computer device may be a computer terminal, a server or a similar computing device.
The embodiment of the invention also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program for executing the arbitrary yard three-dimensional map building method, and solves the problems that the working vehicle is frequently misplaced, the execution of a scheduling instruction cannot be monitored in real time, the working process of the vehicle in a port cannot be monitored, and the change of goods cannot be updated in real time.
In particular, computer-readable storage media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer-readable storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable storage medium does not include transitory computer readable media (transmyedia), such as modulated data signals and carrier waves.
It will be apparent to those skilled in the art that the modules or steps of the embodiments of the invention described above may be implemented by a general purpose computing device, they may be centralized on a single computing device or distributed across a network of multiple computing devices, and alternatively, they may be implemented by program code executable by a computing device, such that they may be stored in a storage device and executed by a computing device, and in some cases, the steps shown or described may be performed in an order different than that described herein, or they may be separately fabricated into individual integrated circuit modules, or multiple ones of them may be fabricated into a single integrated circuit module. Thus, embodiments of the invention are not limited to any specific combination of hardware and software.
The above description is only a preferred embodiment of the present invention, and is not intended to limit the present invention, and various modifications and changes may be made to the embodiment of the present invention by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (8)
1. A method for establishing a three-dimensional map of a storage yard is characterized by comprising the following steps:
(1) collecting sequence images of the pile position of the storage yard;
the sequence images of the stacking position of the storage yard are acquired from a vertical angle and an inclination angle through a plurality of cameras carried on the unmanned aerial vehicle, all aerial photography aerial belts of the unmanned aerial vehicle are parallel to each other, the space between every two adjacent aerial photography aerial belts is equal, no less than 50% of overlapped detected areas are contained between any two adjacent sequence images in all aerial photography aerial belts, and no less than 50% of overlapped detected areas are contained between any two adjacent aerial photography aerial belts which are parallel to each other;
(2) acquiring point cloud data of a storage yard;
extracting homonymous feature points contained in an overlapping region in adjacent sequence images in the same aerial photography zone and homonymous feature points contained in an overlapping region in adjacent images in the adjacent aerial photography zone in the sequence images through an image matching algorithm, performing Euclidean reconstruction on the obtained homonymous feature points, performing triangularization processing in a three-dimensional space, and generating point cloud data for placing a stock dump from the sequence images of the stock dump;
(3) generating digital elevation model data by using point cloud data of a storage yard;
extracting characteristic points, lines and surfaces from the point cloud data, simultaneously obtaining volume information through a tetrahedral subdivision volume algorithm based on convex hull segmentation, and interpolating a network to generate digital elevation model data; obtaining an initial building model, an initial container model and an initial facility model of a storage yard;
the tetrahedron subdivision volume algorithm based on convex hull segmentation specifically comprises the following steps:
(3.1) partitioning a tetrahedron based on convex hull partitioning;
(3.2) calculating normal vectors of four vertexes of the tetrahedron based on a fitted surface, and unifying the directions of the normal vectors;
(3.3) if the ray pointing along the normal vector of each vertex of one tetrahedron has an intersection point with the external sphere of the tetrahedron, determining that the tetrahedron is an external tetrahedron; otherwise, judging the tetrahedron to be an internal tetrahedron;
(3.4), the sum of the volumes of all the in vivo tetrahedra is the result volume;
step (3.1), the tetrahedron subdivision based on convex hull segmentation specifically comprises the following steps:
(3.1.1) constructing an initial tetrahedron based on the point cloud data obtained in the step (2) to form an initial tetrahedron grid;
(3.1.2), inserting scattered points in the point cloud data acquired in the step (2) as input points into a current tetrahedral mesh, and for the input points, using a random walking method to find a tetrahedron containing the input points;
designating a tetrahedron, if the input point is located in the tetrahedron, walking is completed, if the input point is not located in the tetrahedron, a triangular face is randomly designated, if the plane of the triangular face divides the tetrahedron and the input point, the next accessed tetrahedron is the adjacent tetrahedron sharing the triangular face; otherwise, traversing the other faces in a predetermined order until a face is found that separates the tetrahedron and the input point;
(3.1.3), finding the tetrahedron containing the input point, then dividing the tetrahedron into 4 small sub-tetrahedrons;
(3.1.4) selecting a visible face of the mesh if the input point is outside the current tetrahedral mesh; connecting the input point with three vertexes of the visible surface to form a new tetrahedron, and adding the new tetrahedron into the tetrahedral mesh;
(3.1.5) repeating the steps (3.1.2) to (3.1.4) until all scattered points in the point cloud data acquired in the step (2) are inserted into the tetrahedral mesh;
(3.1.6), verifying the effectiveness of the Delaunay triangulation;
firstly, checking the continuity of a Delaunay triangulation data structure, namely the adjacency relation of tetrahedrons, and then verifying the direction of each tetrahedron and the correctness of a convex hull obtained by the Delaunay triangulation;
(4) forming an oblique image mapping model through the oblique image, and extracting the oblique image of the building, the oblique image of the container and the oblique image of the storage yard facility from the oblique image;
(5) mapping the inclined image of the building, the inclined image of the container and the inclined image of the storage yard facility to an initial building model, an initial container model and an initial facility model through an inclined image mapping model to form a building model, a container model and a facility model of the storage yard;
(6) and superposing the building model, the container model and the facility model of the storage yard to form a three-dimensional map of the storage yard.
2. The method for building the three-dimensional map of the yard according to claim 1, wherein in step (3.4), the sum of the volumes of all the internal tetrahedrons is the result volume, specifically, the result volume of the object can be obtained by calculating the sum of the volumes of all the tetrahedrons according to the following formula:
3. the method for building the three-dimensional map of the storage yard according to claim 1, wherein the ground resolution of the aerial image of the unmanned aerial vehicle is not lower than 0.08 m; and/or, the average course overlapping degree is not less than 75%, and the average side overlapping degree is not less than 50%; and/or the difference of the heights of adjacent pictures on the same flight line is not more than 30 meters, the difference between the maximum height and the minimum height is not more than 50 meters, and the difference between the actual height and the designed height is not more than 50 meters.
4. The method for building the three-dimensional map of the storage yard according to any one of claims 1 to 3, wherein the forming of the oblique image mapping model by the oblique image specifically comprises:
and sequentially carrying out orthorectification, image mosaic, image processing and digital orthoimage cutting on the inclined image to form digital orthoimage data.
5. The yard three-dimensional map building method according to claim 3, further comprising:
collecting road network data, and forming road network vector data by using the digital orthographic image data;
acquiring vector data, and forming an image data digital drawing by superposing the vector data and the digital orthographic image data; carrying out data layering and coordinate conversion on the vector data to form a vector map of a unified coordinate system;
respectively carrying out coordinate conversion processing on the digital ortho-image data and the digital elevation model data to form data of a unified coordinate system;
and carrying out data fusion on the building model, the container model, the facility model, the vector map, the digital orthographic image data of the unified coordinate system, the digital elevation model data of the unified coordinate system and the oblique image of the yard to form the three-dimensional map of the yard.
6. An inventory management method, characterized by comprising the step of performing data interaction with a vehicle-mounted terminal stack running APP, a resource graphical scheduling operation monitoring management system and a basic information management system by using the three-dimensional map of the storage yard established by the storage yard three-dimensional map establishing method according to any one of claims 1 to 5.
7. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the yard three-dimensional map building method of any one of claims 1 to 5 when executing the computer program.
8. A computer-readable storage medium characterized in that the computer-readable storage medium stores a computer program for executing the three-dimensional map building method of a yard according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011467936.7A CN112465976B (en) | 2020-12-14 | 2020-12-14 | Storage yard three-dimensional map establishing method, inventory management method, equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011467936.7A CN112465976B (en) | 2020-12-14 | 2020-12-14 | Storage yard three-dimensional map establishing method, inventory management method, equipment and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112465976A CN112465976A (en) | 2021-03-09 |
| CN112465976B true CN112465976B (en) | 2021-07-20 |
Family
ID=74804026
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011467936.7A Active CN112465976B (en) | 2020-12-14 | 2020-12-14 | Storage yard three-dimensional map establishing method, inventory management method, equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112465976B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113034555B (en) * | 2021-03-19 | 2024-03-08 | 南京天巡遥感技术研究院有限公司 | Feature fine matching method based on minimum spanning tree and application |
| CN112802208B (en) * | 2021-03-30 | 2021-06-22 | 中国民用航空总局第二研究所 | Three-dimensional visualization method and device in terminal building |
| CN113792642B (en) * | 2021-09-08 | 2022-08-12 | 江苏省工程勘测研究院有限责任公司 | River and lake ecological management data processing method and system based on intelligent technology |
| CN115410390A (en) * | 2022-08-19 | 2022-11-29 | 河南泽阳实业有限公司 | Intelligent scheduling method for intelligent port |
| CN115358486B (en) * | 2022-09-20 | 2023-06-02 | 中咨数据有限公司 | Port freight volume prediction method, system and application based on stereoscopic satellite images |
| CN115937461B (en) * | 2022-11-16 | 2023-09-05 | 泰瑞数创科技(北京)股份有限公司 | Multi-source fusion model construction and texture generation method, device, medium and equipment |
| CN116342783B (en) * | 2023-05-25 | 2023-08-08 | 吉奥时空信息技术股份有限公司 | Live-action three-dimensional model data rendering optimization method and system |
| CN116342827B (en) * | 2023-05-26 | 2023-08-08 | 厦门精图信息技术有限公司 | Map element three-dimensional presentation system and method |
| CN117671167B (en) * | 2023-10-19 | 2024-05-28 | 兰州交通大学 | Heuristic DEM (digital elevation model) comprehensive method based on mountain shadow analysis |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110648398A (en) * | 2019-08-07 | 2020-01-03 | 武汉九州位讯科技有限公司 | Real-time ortho image generation method and system based on unmanned aerial vehicle aerial data |
| CN112529498A (en) * | 2020-12-08 | 2021-03-19 | 牟茹月 | Warehouse logistics management method and system |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104658039B (en) * | 2015-02-12 | 2018-01-30 | 南京市测绘勘察研究院股份有限公司 | A kind of city digital map three-dimensional modeling preparation method |
| CN106017320B (en) * | 2016-05-30 | 2018-06-12 | 燕山大学 | A kind of system of scattered groceries heap volume measuring method and realization the method based on image procossing |
| CN109035321A (en) * | 2017-06-09 | 2018-12-18 | 河北卓达建材研究院有限公司 | A kind of volume estimation method of building |
-
2020
- 2020-12-14 CN CN202011467936.7A patent/CN112465976B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110648398A (en) * | 2019-08-07 | 2020-01-03 | 武汉九州位讯科技有限公司 | Real-time ortho image generation method and system based on unmanned aerial vehicle aerial data |
| CN112529498A (en) * | 2020-12-08 | 2021-03-19 | 牟茹月 | Warehouse logistics management method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112465976A (en) | 2021-03-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112465976B (en) | Storage yard three-dimensional map establishing method, inventory management method, equipment and medium | |
| Carvajal-Ramírez et al. | Virtual reconstruction of damaged archaeological sites based on Unmanned Aerial Vehicle Photogrammetry and 3D modelling. Study case of a southeastern Iberia production area in the Bronze Age | |
| Agueera-Vega et al. | Reconstruction of extreme topography from UAV structure from motion photogrammetry | |
| CN111597666B (en) | Method for applying BIM to transformer substation construction process | |
| CN110285792A (en) | A kind of fine grid earthwork metering method of unmanned plane oblique photograph | |
| CN102506824A (en) | Method for generating digital orthophoto map (DOM) by urban low altitude unmanned aerial vehicle | |
| CN113570527B (en) | Fusion method of overwater and underwater three-dimensional point clouds | |
| KR20210077390A (en) | Analysis of corrosion and accumulation in coastal region using time series drone image | |
| US20230419501A1 (en) | Image analysis for aerial images | |
| Karan et al. | Digital modeling of construction site terrain using remotely sensed data and geographic information systems analyses | |
| CN114283070B (en) | Method for manufacturing terrain section by fusing unmanned aerial vehicle image and laser point cloud | |
| Kwak | Automatic 3D building model generation by integrating LiDAR and aerial images using a hybrid approach | |
| CN115423958A (en) | Mining area travelable area boundary updating method based on visual three-dimensional reconstruction | |
| Mao et al. | Precision evaluation and fusion of topographic data based on UAVs and TLS surveys of a loess landslide | |
| CN115797288A (en) | Method for calculating filling and excavating volume based on ground point cloud data | |
| Gu et al. | Surveying and mapping of large-scale 3D digital topographic map based on oblique photography technology | |
| Previtali et al. | Accurate 3D surface measurement of mountain slopes through a fully automated image-based technique | |
| Bagnolo et al. | UAV-based photogrammetry for archaeological heritage site survey and 3D modeling of the sardus pater temple (Italy) | |
| Batakanwa | Validating Uav-Sfm Photogrammetry Heights for Highway Topographic Surveying in Tanzania | |
| Švec et al. | Preparation of the Digital Elevation Model for Orthophoto CR Production | |
| CN116385686B (en) | Live-action three-dimensional model reconstruction method and system based on irregular oblique photography | |
| Charles | VALIDATING UNMANNED AERIAL VEHICLE PHOTOGRAMMETRY HEIGHTS, FOR HIGHWAY TOPOGRAPHIC SURVEYING IN TANZANIA. | |
| Agüera Vega et al. | Virtual reconstruction of damaged archaeological sites based on Unmanned Aerial Vehicle Photogrammetry and 3D modelling. Study case of a southeastern Iberia production area in the Bronze Age | |
| CN117968642A (en) | Soil measuring method based on oblique photography | |
| Danchenkov et al. | Comparative Analysis of the Unmanned Aerial Vehicles and Terrestrial Laser Scanning Application for Coastal Zone Monitoring |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |