BIM-based architectural viewpoint construction method, system, equipment and medium

By obtaining the top view and physical proportions of the building, determining the drone viewpoint and route, taking images and point cloud data, and adjusting the distance between the drone and the corridor, the problem of poor BIM model effect in multi-body buildings is solved, and the improvement is achieved High-precision building model construction.

CN117911643BActive Publication Date: 2024-11-08壹仟零壹艺数智(北京)科技有限公司
View PDF 9 Cites 0 Cited by

Patent Information

Application Number
CN202410098632.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-01-24
Publication Date
2024-11-08
Estimated Expiration
2044-01-24

AI Technical Summary

Technical Problem

The prior art is difficult to build an effective BIM building model in multi-body buildings, especially when two buildings are connected by corridors, resulting in poor model effects.

Method used

By obtaining the top view and the actual proportion of the building, the building viewpoint of the drone is determined, and the top view is used to construct the walking route of the drone, taking regional images and three-dimensional point cloud data, and adjusting the distance between the drone and the corridor to form a Accurate BIM building model.

Benefits of technology

The problem of poor BIM model effect at corridor junctions in multi-body buildings is solved, and the accuracy and accuracy of the model are improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117911643B_ABST
    Figure CN117911643B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of BIM building technology. A BIM-based building viewpoint construction method includes determining the building viewpoint of a drone according to the physical scale and the viewpoint distance; constructing a first walking route of the drone; obtaining a first area image and a second area image taken by the drone according to the first walking route and the building viewpoint, as well as first three-dimensional point cloud data corresponding to the first area image and second three-dimensional point cloud data corresponding to the second area image; determining the corridor length by using the second area image and the second three-dimensional point cloud data; adjusting the first walking route of the drone so that the distance between the drone and the corridor is equal to the viewpoint distance, obtaining a third area image taken by the drone, and third three-dimensional point cloud data corresponding to the third area image; obtaining a BIM building model according to the data of the first area image, the first three-dimensional point cloud data, the third area image and the third three-dimensional point cloud data. Improve the effect of the BIM building model.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of BIM building technology, and in particular to a method, system, equipment and medium for constructing a building viewpoint based on BIM. Background Art

[0002] At present, high-precision models of urban single-building projects are important reference objects for making decisions in urban management, project monitoring, safety assessment, etc. As a result, some 3D modeling methods for urban single-building projects have been derived. However, most of these methods are only applicable to single-building projects, and are not applicable to some multi-building projects, especially multi-building projects with different depths, such as two single-building projects connected by corridors.

[0003] In view of this, a method for building viewpoints based on BIM is provided. Summary of the invention

[0004] The BIM-based building viewpoint construction method of the present invention can solve the problem that an effective BIM building cannot be formed when there is a corridor between two buildings in a multi-building.

[0005] The first aspect of the present application provides a method for building a building viewpoint based on BIM, the method comprising:

[0006] Obtaining a top view of a building and the physical scale of said building;

[0007] Determining the building viewpoint of the drone according to the physical object proportion and the viewpoint distance, wherein the viewpoint distance is an optimal distance for the drone to photograph the building;

[0008] Using the top view, construct a first walking route of the drone;

[0009] According to the first walking route and the building viewpoint, a first area image and a second area image taken by the drone, as well as first three-dimensional point cloud data corresponding to the first area image and second three-dimensional point cloud data corresponding to the second area image are obtained, wherein the second area image indicates that the distance between the image taken in the second area image and the drone is greater than the optimal distance;

[0010] Determining the length of the corridor by using the second area image and the second three-dimensional point cloud data;

[0011] The first walking route of the UAV is adjusted so that the distance between the UAV and the corridor is equal to the viewpoint distance, and a second walking route of the UAV is obtained. According to the second walking route, a third area image taken by the UAV and third three-dimensional point cloud data corresponding to the third area image are obtained.

[0012] The first area image, the first three-dimensional point cloud data, the third area image and the third three-dimensional point cloud data are input into the BIM to obtain the BIM building model.

[0013] In some practicable manners, the step of obtaining a top view of a building and a physical scale of the building includes:

[0014] Obtaining a bird's-eye view and scale of the building according to a satellite map or an electronic map;

[0015] Determine the shape of the building using the top view;

[0016] Using a scale ruler, determine the physical proportions of the building.

[0017] In some practicable manners, the step of determining the building viewpoint of the drone according to the physical object scale and the viewpoint distance includes:

[0018] According to the camera parameters of the drone's camera, the optimal distance for the drone to shoot the building is obtained;

[0019] According to the physical proportion, the size of the building is obtained;

[0020] By using the size of the building and the optimal distance, the building viewpoint of the drone is obtained, where the building is located at the 11 o'clock direction of the clock.

[0021] In some practicable manners, the step of constructing the first walking route of the drone using the top view includes:

[0022] Using the top view, obtaining a quasi-regular shape of the top view;

[0023] According to the regular shape, a first walking route of the UAV is constructed.

[0024] In some practicable manners, the step of obtaining the first area image and the second area image taken by the drone according to the first walking route and the building viewpoint, as well as the first three-dimensional point cloud data corresponding to the first area image and the second three-dimensional point cloud data corresponding to the second area image includes:

[0025] Randomly select a point on the first walking route as the starting point and end point of the drone shooting;

[0026] Based on the starting point, a preset speed and a shooting time interval, the building is photographed step by step along the first walking route to obtain the first area image and the second area image;

[0027] Based on the laser radar of the UAV, the building is scanned along the first walking route to obtain the first three-dimensional point cloud data corresponding to the first area image and the second three-dimensional point cloud data corresponding to the second area image.

[0028] In some practicable manners, the step of photographing the building step by step along the first walking route based on the starting point, a preset speed and a shooting time interval to obtain the first area image and the second area image includes:

[0029] Obtain the texture and clarity of the current image, and the average of the texture and clarity of M images before the current image, and pre-set an upper error limit and a lower error limit for the average, to obtain the texture and clarity of the current image, and the range values ​​of the texture and clarity of M images, where M is an integer greater than 1;

[0030] Compare the texture and clarity of the current image with the M texture and clarity range values ​​to obtain a first image comparison result;

[0031] If the comparison result of the first image is: the texture and clarity of the current image are within the range of the texture and clarity of the M images, then marking the current image as the first region image;

[0032] If the comparison result of the first image is: the texture and clarity of the current image are not within the range of the texture and clarity of the M images, then marking the current image as the second region image;

[0033] Acquire the texture and clarity of N images after the current image marked as the second area image within a preset time period, obtain the average of the N textures and clarity, and pre-set an upper error limit and a lower error limit for the average to obtain N texture and clarity range values, wherein N is an integer greater than 1, and the N images are the number of images taken by the drone within a preset time period according to the flight speed of the drone;

[0034] Acquire the texture and clarity of the N+1th image after the N images, to obtain the N+1th texture and clarity;

[0035] Compare the N+1th texture and clarity with the N textures and clarity range values ​​to obtain a second image comparison result;

[0036] If the comparison result of the second image is: the texture and clarity of the N+1th image are within the range of the texture and clarity of the N images, then the N+1th image is marked as the second region image;

[0037] If the comparison result of the second image is that the texture and clarity of the N+1th image are not within the range of the textures and clarity of the N images, the N+1th image is marked as the first area image.

[0038] In some practicable manners, the step of scanning the building along the first walking route using the laser radar of the drone to obtain the first three-dimensional point cloud data corresponding to the first area image and the second three-dimensional point cloud data corresponding to the second area image further includes:

[0039] Acquire the three-dimensional point cloud data obtained by the laser radar of the drone when shooting the current image, to obtain current three-dimensional point cloud data;

[0040] Acquire three-dimensional point cloud data of M images before the current image is taken, and use a principal component analysis method to obtain a first plane formed by three principal component directions and corresponding eigenvalues ​​of the three-dimensional point cloud data of the M images;

[0041] Projecting the three principal component directions and corresponding eigenvalues ​​of the current three-dimensional point cloud data onto the first plane, determining whether the eigenvalues ​​are less than or equal to a set threshold, and obtaining a first feature comparison result;

[0042] If the first feature comparison result is less than or equal to a set threshold, marking the current image as the first area image;

[0043] If the comparison result of the first image is greater than a set threshold, marking the current image as the second area image;

[0044] Acquire three-dimensional point cloud data of N images after the current image marked as the second area image within a preset time period, and obtain a second plane formed by three principal component directions and corresponding eigenvalues ​​of the three-dimensional point cloud data of the N images;

[0045] After acquiring the N images, the three-dimensional point cloud data of the N+1th image is obtained to obtain the N+1th three-dimensional point cloud data;

[0046] Projecting the three principal component directions and corresponding eigenvalues ​​of the N+1th three-dimensional point cloud data onto the second plane, determining whether the eigenvalues ​​are less than or equal to a set threshold, and obtaining a second feature comparison result;

[0047] If the second feature comparison result is less than or equal to the set threshold, marking the N+1th image as the second area image;

[0048] If the second feature comparison result is greater than or equal to the set threshold, marking the N+1th image as the second area image;

[0049] A weighted calculation is performed on the first feature comparison result and the second feature comparison result so as to divide the image taken by the drone and determine whether the image taken by the drone belongs to the first area image or the second area image.

[0050] In some practicable manners, the step of determining the corridor length by using the second area image and the second three-dimensional point cloud data includes:

[0051] Acquire the three-dimensional point cloud data at the junction of the first area image and the second area image, and the two images with partially overlapping contents, and respectively obtain the corresponding positioning three-dimensional point cloud data and positioning image data;

[0052] Comprehensively analyzing the positioning three-dimensional point cloud data and the positioning image data, to obtain the initial positioning and the final positioning of the UAV, wherein, if the number of the corridor is one, the corridor connects two single buildings, and the two connection positions correspond to form the initial positioning and the final positioning;

[0053] The corridor length is determined using the initial position and the terminal position.

[0054] In some practicable methods, the step of adjusting the first walking route of the drone so that the distance between the drone and the corridor is equal to the viewpoint distance, obtaining the second walking route of the drone, and obtaining the third area image taken by the drone and the third three-dimensional point cloud data corresponding to the third area image according to the second walking route includes:

[0055] On the first walking route of the drone, the initial position and the end position are marked, and the image taken by the drone between the initial position and the end position is marked to obtain an image to be processed;

[0056] According to the initial position and the final position, the walking direction of the UAV is obtained;

[0057] Obtaining the second walking route according to the walking direction of the drone and the viewpoint distance;

[0058] According to the second walking route, the corridor image taken by the drone is acquired to obtain the third area image and third three-dimensional point cloud data corresponding to the third area image;

[0059] The image to be processed is deleted, and the third region image and the third three-dimensional point cloud data corresponding to the third region image are added into the time axis empty window period caused by deleting the image to be processed.

[0060] In some practicable manners, the step of inputting the data of the first area image, the first three-dimensional point cloud data, the third area image and the third three-dimensional point cloud data into the BIM to obtain the BIM building model comprises:

[0061] Acquire the first area image and the third area image formed by the drone, as well as the parameters of the first three-dimensional point cloud data and the third three-dimensional point cloud data, to obtain a calibrated target area image and target three-dimensional point cloud data, wherein the target area image and the target three-dimensional point cloud data form a corresponding relationship;

[0062] Performing feature extraction on the target area image and the target three-dimensional point cloud data to obtain target area image features and target three-dimensional point cloud data features respectively;

[0063] Matching the target area image features with the target three-dimensional point cloud data to obtain a corresponding relationship between the target area image features and the target three-dimensional point cloud data features;

[0064] According to the correspondence between the target area image features and the target three-dimensional point cloud data features, and the target area image features and the target three-dimensional point cloud data features, they are input into BIM to obtain the BIM building model.

[0065] A second aspect of the present application provides a BIM-based building viewpoint construction system, which is applied to the aforementioned BIM-based building viewpoint construction method, and the system includes:

[0066] An acquisition unit, used for acquiring a top view of a building and a physical scale of the building;

[0067] A determination unit, configured to determine a building viewpoint of the drone according to the physical object ratio and the viewpoint distance, wherein the viewpoint distance is an optimal distance for the drone to photograph the building;

[0068] A route unit, configured to construct a first walking route of the UAV by using the top view;

[0069] a data unit, for obtaining, according to the first walking route and the building viewpoint, a first area image and a second area image taken by the drone, as well as first three-dimensional point cloud data corresponding to the first area image and second three-dimensional point cloud data corresponding to the second area image, wherein the second area image indicates that the distance between the image taken in the second area image and the drone is greater than the optimal distance;

[0070] A length unit, used to determine the length of the corridor based on the second area image and the second three-dimensional point cloud data;

[0071] an adjusting unit, configured to adjust the first walking route of the drone so that the distance between the drone and the corridor is equal to the viewpoint distance, obtain a second walking route of the drone, and obtain a third area image taken by the drone and third three-dimensional point cloud data corresponding to the third area image according to the second walking route;

[0072] A result unit is used to input the data of the first area image, the first three-dimensional point cloud data, the third area image and the third three-dimensional point cloud data into the BIM to obtain the BIM building model.

[0073] A third aspect of the present application provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the aforementioned BIM-based building viewpoint construction method when executing the computer program.

[0074] A fourth aspect of the present application provides a computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the aforementioned BIM-based building viewpoint construction method.

[0075] Beneficial effects of the present invention:

[0076] The present application provides a method for constructing a building viewpoint based on BIM, which obtains a bird's-eye view of a building and the actual proportion of the building; determines the building viewpoint of a drone according to the actual proportion and the viewpoint distance, wherein the viewpoint distance is the optimal distance for the drone to photograph the building; constructs a first walking route of the drone using the bird's-eye view; obtains a first area image and a second area image photographed by the drone according to the first walking route and the building viewpoint, and obtains first three-dimensional point cloud data corresponding to the first area image and second three-dimensional point cloud data corresponding to the second area image, wherein the second area image indicates that the distance between the photographed image in the second area image and the drone is greater than the optimal distance; The length of the corridor is determined by using the second area image and the second three-dimensional point cloud data; the first walking route of the UAV is adjusted so that the distance between the UAV and the corridor is equal to the viewpoint distance, and the second walking route of the UAV is obtained, and according to the second walking route, the third area image taken by the UAV and the third three-dimensional point cloud data corresponding to the third area image are obtained; according to the data of the first area image, the first three-dimensional point cloud data, the third area image and the third three-dimensional point cloud data, the method for obtaining the BIM building model is input into BIM, which can solve the problem that the connection structure between two single buildings, such as the corridor, does not form a building viewpoint, resulting in a poor effect of the final BIM building model. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0078] Figure 1 A flowchart of a method for constructing a building viewpoint based on BIM according to the present invention;

[0079] Figure 2 It is a schematic diagram of a method for constructing a building viewpoint based on BIM in the present invention. DETAILED DESCRIPTION

[0080] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0081] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0082] In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined. In addition, the terms "installed", "connected" and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal connection of two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0083] Some terms involved in this application are explained below to facilitate understanding of this application:

[0084] BIM, Building Information Modeling, refers to a digital building design and management method that uses 3D modeling technology and database systems to integrate a building's geometric model, spatial relationships, material information, construction processes and other information.

[0085] Architectural viewpoint refers to the viewing angle or position chosen in architectural design or architectural photography. It determines the angle, distance and perspective effect of our observation and perception of buildings.

[0086] Principal Component Analysis, PCA, is a commonly used dimensionality reduction technology and data analysis method. It can transform high-dimensional data into low-dimensional data while retaining the information in the original data as much as possible.

[0087] Multi-body buildings are relative to single-body buildings. Multi-body buildings are mostly composed of at least two single-body buildings connected to each other, such as corridors. This corridor can be suspended in the air to form a connection between two single-body buildings, or it can be arranged on the ground to form a connection between two single-body buildings. Corridors are mostly not set at the edge of the outer surface of two single-body buildings for connection, and often have a certain depth. For example, the centers of the opposite surfaces of two single-body buildings are connected by corridors so that pedestrians can pass through this corridor and shuttle between the two single-body buildings. Corridors exist not only between two commercial buildings, but also between two factory buildings. Since corridors are mostly not at the edge of the opposite surfaces of single-body buildings, that is, the corridors are a certain distance away from the outer surface of the single-body building, when using drones for automatic shooting and three-dimensional modeling, the distance from the surface of the single building is far, resulting in the drone being unable to shoot the corridor at the optimal distance, which leads to poor modeling effect at the corridor.

[0088] It should be noted that when drones automatically photograph single buildings to build three-dimensional models, they mostly obtain the outline of the single building and then determine the intersection of the outer contour of the single building. For example, for a rectangular single building, the four corners are used as intersections, and the intersections are used to automatically generate a walking route around the single building. This method is difficult to capture the corridor between two single buildings. In addition, when drones face multi-body buildings, they mostly obtain the overall outline of the multi-body building and then automatically generate intersections, and use the lines connecting the intersections to form a walking route. Regardless of the method, the photography effect of the corridor is poor, resulting in a poor BIM building model.

[0089] Based on the above reasons, the present application provides a method for building viewpoints based on BIM.

[0090] like Figure 1 and Figure 2 As shown, the first aspect of the present application provides a method for building a building viewpoint based on BIM, the method comprising:

[0091] S100, obtaining a bird's-eye view of the building and the actual proportion of the building.

[0092] Specifically, obtaining a top view of a building and the physical proportions of the building includes steps S100 to S103.

[0093] S101, obtaining a bird's-eye view and scale of a building based on a satellite map or an electronic map.

[0094] S102, using the bird's-eye view, determine the shape of the building.

[0095] S103, using a scale to determine the physical proportions of the building.

[0096] First, use the current satellite map or electronic map to find the building to be modeled, and then get the top view of the building through the satellite map or electronic map. In addition, the satellite map or electronic map will be marked with a scale, so that the scale can be used to determine the ratio between the actual building and the map. Next, a simple conversion can be performed to get the size of the actual building's top view. The actual size is used as the data of the drone's flight route.

[0097] S200, determining the building viewpoint of the drone according to the actual object scale and the viewpoint distance.

[0098] Among them, the viewpoint distance is the optimal distance for the drone to shoot buildings.

[0099] S201, obtaining the optimal distance for the drone to shoot the building according to the camera parameters of the drone.

[0100] S202, obtaining the size of the building according to the actual proportion.

[0101] S203, using the size of the building and the optimal distance, obtain the building viewpoint of the drone where the building is located at the 11 o'clock direction of the clock.

[0102] Specifically, the drone is equipped with a camera and a laser radar, wherein the laser radar is used in subsequent steps to set the best distance for photographing buildings according to the camera's shooting parameters, that is, at this distance, the drone can obtain the best effect when photographing buildings. It should be noted that the best distance is not a specific value, but a range value, that is, within this range of values, a better effect can be obtained.

[0103] Furthermore, when the size of the building is obtained and the direction is indicated by the dial of a clock, the top and side surfaces of the building can be photographed at around 11 o'clock, which facilitates the modeling of the building. In other words, the drone is set at the 11 o'clock direction of the building on the clock, and the position of the drone is determined in combination with the optimal shooting distance. In this way, the drone maintains the 11 o'clock direction and the optimal distance, moves in a circumferential direction, and when photographing the building, a better shooting effect of the top and side surfaces can be obtained.

[0104] It should also be noted that if the height of the floor is high and the drone cannot capture the full view of the side of the building at the optimal shooting distance, the drone can be adjusted in shooting height after it travels around the building for one circle. The shooting angle of the drone is not limited to the 11 o'clock direction. The building can be located at the 9 o'clock direction of the clock to capture the side of the building.

[0105] Finally, it should be noted that if you want the drone to be at the 11 o'clock position, you can use the lidar to shoot lasers at the top and sides of the building respectively, and then determine the position of the drone based on the feedback signal. In this way, if the drone is at the 11 o'clock position and the building is at the 9 o'clock position of the clock, the angle at which the lidar receives the feedback signal can be used to determine the position.

[0106] S300,uses the top view to construct the first walking route of the drone.

[0107] Specifically, constructing the first walking route of the drone includes steps S301 and S302.

[0108] S301, using the top view, obtaining a quasi-regular shape of the top view.

[0109] S302: construct a first walking route of the drone according to the class regular shape.

[0110] In the above steps, a top view is obtained, and the edge of the top view is divided into quasi-regular shapes, such as quasi-rectangles, quasi-circles, etc., so that it is convenient to generate the walking route of the drone, that is, the edge line of the quasi-regular shape is used as a reference line. When the optimal distance is obtained, the two are added to form the walking route of the drone, that is, the first walking route. For another example, if the building is high, the height of the drone and the shooting angle can be adjusted (from the 11 o'clock direction to the 9 o'clock direction), and the drone walks along the first walking route again in a circumferential direction. This is repeated several times to obtain the contour data of the periphery of the building.

[0111] S400, obtaining a first area image and a second area image taken by a drone according to the first walking route and the building viewpoint, as well as first three-dimensional point cloud data corresponding to the first area image and second three-dimensional point cloud data corresponding to the second area image.

[0112] The second area image indicates that the distance between the captured image in the second area image and the drone is greater than the optimal distance.

[0113] Specifically, obtaining the regional image and the three-dimensional point cloud data includes steps S401 to S403.

[0114] S401, randomly selecting a point on the first walking route as the starting point and end point of the drone shooting.

[0115] S402, based on the starting point, a preset speed and a shooting time interval, gradually shoot the buildings along the first walking route to obtain a first area image and a second area image.

[0116] S403, based on the laser radar of the unmanned aerial vehicle, scan the building along the first walking route to obtain first three-dimensional point cloud data corresponding to the first area image and second three-dimensional point cloud data corresponding to the second area image.

[0117] It should be noted that after the drone is controlled to reach the first walking route, this point is used as the starting point and end point of the drone shooting. That is to say, the drone takes this point as the starting point, walks around in a circle, and then returns to this point. At this time, this point is the end point.

[0118] During the flight of the drone, the drone moves at a preset speed and then takes pictures at a preset shooting time interval, for example, the preset shooting time interval is 2 seconds. In this way, after the drone walks for one circle, it can obtain the first area image, the second area image, the first area image, the second area image and the first area image. That is to say, when the drone shoots the first building, the first area image is formed. Next, when the drone shoots the corridor between the second building and the first building, the area image is formed. Then the drone continues to walk and forms the first area image when shooting the second building. Next, the drone continues to walk and shoots the corridor between the second building and the first building to form the second area image. Finally, the drone shoots the first building and arrives at the end point to form the first area image, thereby completing the outline shooting of the first building, a corridor and the second building.

[0119] It should also be noted that the three-dimensional point cloud data formed by the laser radar scanning of the drone is consistent with the process of photographing the outline of the first building, a corridor and the second building, that is, the first three-dimensional point cloud data, the second three-dimensional point cloud data, the first three-dimensional point cloud data, the second three-dimensional point cloud data and the first three-dimensional point cloud data are obtained. In this way, the three-dimensional point cloud data and the regional image form a corresponding relationship for use in subsequent steps.

[0120] Finally, it should be noted that the two photographed corridors mentioned above refer to photographing the opposite sides of the two corridors, that is, the image depth formed by the second area image is higher than the image depth of the first area image, that is, when the drone photographs the corridor, it still walks according to the first walking route, but the corridor is located in the center between the two buildings. Therefore, when the drone photographs the corridor, the distance between the drone and the corridor is greater than the distance between the drone photographing the building facade, resulting in a difference between the second area image and the first area image. In other words, the distance between the image captured in the second area image and the drone is greater than the optimal distance.

[0121] In some embodiments, S403, based on the laser radar of the drone, the building is scanned along the first walking route to obtain first three-dimensional point cloud data corresponding to the first area image and second three-dimensional point cloud data corresponding to the second area image, including:

[0122] S4031, obtain the texture and clarity of the current image, and the mean of the texture and clarity of M images before the current image, and pre-set the upper and lower error limits for the mean, to obtain the texture and clarity of the current image, and the range values ​​of the texture and clarity of M images, where M is an integer greater than 1.

[0123] S4032, compare the texture and clarity of the current image with the M texture and clarity range values ​​to obtain a first image comparison result.

[0124] If the comparison result of the first image is: the texture and clarity of the current image are within the range of the texture and clarity of the M images, the current image is marked as the first region image.

[0125] If the comparison result of the first image is that the texture and clarity of the current image are not within the range of the texture and clarity of the M images, the current image is marked as a second region image.

[0126] Specifically, the drone cannot determine where the corridor is by itself during the process of moving and shooting along the first walking route. Therefore, the drone continuously shoots images while walking on the facade of the first building to form a first regional image. At the same time, it is also necessary to judge the texture and clarity of the current image taken. Prior to this, the texture and clarity of the M images taken previously have been averaged, that is, the texture and clarity in the N images are averaged, or weighted averaged, so as to obtain the mean of the M images. It should be noted that after obtaining the mean, since the facade of the wall will have errors even if it is a plane, it is necessary to increase the error on the basis of this mean, that is, to pre-set the upper and lower limits of the error, that is, after obtaining the mean, both the positive and reverse limits are increased, thereby forming a range value. In this case, the range value can form a certain inclusiveness to avoid the current image being not equal to the mean due to errors, and the texture and clarity of the current image can fall within the range of the texture and clarity of the M images. That is to say, the first image comparison result is processed by a judgment method, and the current image is divided into a first region image or a second region image by using the judgment of the first image comparison result. Exemplarily, after the current image is divided into the first region image, the current image will participate in the calculation of the mean value of the texture and clarity of the M images, that is, after the current image, the mean value of the texture and clarity of the M images becomes the mean value of the texture and clarity of the M+1 images. That is to say, when the next image of the current image is obtained, the texture and clarity of the next image are compared with the range value of the texture and clarity of the M+1 images, so as to judge whether the next image belongs to the first region image or the second region image.

[0127] On the contrary, if the current image is compared with the range values ​​of texture and clarity of the M images and does not fall within the range values, the current image will be marked as the second region image. Next, the next image of the current image is compared with the range values ​​of texture and clarity of the M images. If it still does not fall within the range values, it is determined that the next image belongs to the second region image. In other words, in this way, it is also possible to verify whether the mark of the current image is wrong, thereby increasing the accuracy of image marking.

[0128] It should be noted that if an image belongs to the second area image, but the images before and after this image belong to the first area image, this may be due to a problem with the facade of the building, which leads to a calculation error. At this time, the label of this image is adjusted from belonging to the second area image to the first area image.

[0129] In this embodiment, the judgment method of the first image comparison result is used to distinguish the first area image from the second area image, and then judge where the first walking route is the corridor. Of course, this judgment result cannot completely determine the corridor position, and further judgment is required in subsequent steps.

[0130] S4033, obtaining the texture and clarity of N images after the current image marked as the second area image within a preset time period, obtaining the mean of the N textures and clarity, and pre-setting the upper and lower error limits for the mean to obtain N texture and clarity range values, wherein N is an integer greater than 1, and the N images are the number of images taken by the drone within a preset time period according to the flight speed of the drone.

[0131] S4034, acquiring the texture and clarity of the N+1th image after the N images, and obtaining the N+1th texture and clarity.

[0132] S4035, compare the N+1th texture and clarity with the Nth texture and clarity range values ​​to obtain a second image comparison result.

[0133] If the comparison result of the second image is: the texture and clarity of the N+1th image are within the range of the texture and clarity of the M images, the N+1th image is marked as the second region image.

[0134] If the comparison result of the second image is: the texture and clarity of the N+1th image are not within the range of the texture and clarity of the M images, the N+1th image is marked as the first area image.

[0135] Specifically, S4033 to S4035 are used as transition judgment from the second region image to the first region image, and the judgment method is the same as the transition from the first region image to the second region image, which will not be repeated here.

[0136] It should be noted that the N+2 images need to be judged with the N texture and clarity range values, so as to verify whether the marking of the N+1 image is correct. This method has been described before and will not be repeated here.

[0137] It should also be noted that verifying the mark of the current image by the next image, or verifying the mark of the N+1 image by the N+2 image, is only for exemplary purposes. The current image and the next image can also be verified by the next image, or the marks of the N+2 image and the N+1 image can be verified by the N+3 image. The specific details can be adjusted according to actual needs, and the present application does not limit this.

[0138] Finally, it should be noted that the number of N images can be set according to needs, according to time, or according to the flight speed of the drone, and this application does not limit this.

[0139] In some embodiments, S403, based on the laser radar of the drone, scanning the building along the first walking route to obtain first three-dimensional point cloud data corresponding to the first area image and second three-dimensional point cloud data corresponding to the second area image further includes:

[0140] S4036, obtaining the three-dimensional point cloud data obtained by the laser radar of the drone when shooting the current image, and obtaining the current three-dimensional point cloud data.

[0141] S4037, obtaining the three-dimensional point cloud data of M images before the current image is taken, and using the principal component analysis method to obtain the first plane formed by the three principal component directions and corresponding eigenvalues ​​of the three-dimensional point cloud data of the M images.

[0142] S4038, projecting the three principal component directions and the corresponding eigenvalues ​​of the current three-dimensional point cloud data onto the first plane, determining whether the eigenvalues ​​are less than or equal to a set threshold, and obtaining a first feature comparison result.

[0143] If the first feature comparison result is less than or equal to the set threshold, the current image is marked as a first region image.

[0144] If the comparison result of the first image is greater than the set threshold, the current image is marked as the second region image.

[0145] Specifically, the lidar is used to measure distance and obtain three-dimensional point cloud data of the scene. It should be noted that the lidar in the present application collects three-dimensional point cloud data in real time during the flight of the drone. When taking the current image, the three-dimensional point cloud data obtained by the lidar of the drone can be used as a time marker, so that in the subsequent steps, the time point of the current image can be aligned with the three-dimensional point cloud data obtained at this time point, that is, the current three-dimensional point cloud data is obtained.

[0146] Next, as mentioned in the previous steps, the current three-dimensional point cloud data is marked in the first area image or the second area image using the same method of obtaining the range value of texture and clarity of M pictures. However, due to the different processing methods of three-dimensional point cloud data and images, this application uses the principal component analysis method to perform principal component analysis on the current three-dimensional point cloud data and the three-dimensional point cloud data of M images to obtain three principal component directions and corresponding eigenvalues, that is, the three-dimensional point cloud data is reduced in dimension. When the three-dimensional data is reduced to two or one dimension, it can be determined whether it is on the same plane by judging the variance of the data. The principal component analysis method used in this application is a conventional analysis method, and this application has not improved this principal component analysis method. After obtaining the first plane formed by the three principal component directions and corresponding eigenvalues ​​of the three-dimensional point cloud data of M images, the three principal component directions and corresponding eigenvalues ​​of the current three-dimensional point cloud data are projected on the first plane, and then the eigenvalues ​​are judged. If the eigenvalue of the smallest eigenvalue is close to 0, it means that the point cloud data is on a plane. In addition, the point cloud data can be determined to be still on the same plane by calculating the principal component direction and eigenvalue of the new point cloud data, and then judging whether the minimum eigenvalue is close to 0, where 0 is the threshold, which is only the threshold under ideal conditions. The actual threshold can be set as needed to tolerate a certain error, that is, when the minimum eigenvalue is less than a certain threshold, the three-dimensional point cloud data is considered to be on the same plane. It is a conventional method to determine whether the three-dimensional point cloud data collected at different times are on the same plane. The present application applies this method to the three-dimensional point cloud data collected by drones on the facades of walls and corridors to determine whether they are on the same plane.

[0147] If the current 3D point cloud data and the 3D point cloud data of the M images are on the same plane, it can be determined that the current 3D point cloud data belongs to the first region image and the current 3D point cloud data is marked. Otherwise, it is determined that the current 3D point cloud data belongs to the second region image and the current 3D point cloud data is marked. In addition, for the next 3D point cloud data formed by the next image, the marking of the current 3D point cloud data can also be verified in the same way as the previous next image verifies the current image, which will not be repeated here.

[0148] S4039, obtaining three-dimensional point cloud data of N images after the current image marked as the second region image within a preset time period, and obtaining a second plane formed by three principal component directions and corresponding eigenvalues ​​of the three-dimensional point cloud data of the N images.

[0149] S40310, after acquiring N images, the three-dimensional point cloud data of the N+1th image is obtained, thereby obtaining the N+1th three-dimensional point cloud data.

[0150] S40311, project the three principal component directions and corresponding eigenvalues ​​of the N+1th three-dimensional point cloud data onto the second plane, determine whether the eigenvalues ​​are less than or equal to a set threshold, and obtain a second feature comparison result.

[0151] If the second feature comparison result is less than or equal to the set threshold, the N+1th image is marked as the second area image.

[0152] If the second feature comparison result is greater than or equal to the set threshold, the N+1th image is marked as the second area image.

[0153] S40312, performing weighted calculation on the first feature comparison result and the second feature comparison result, so as to divide the image taken by the drone and determine whether the image taken by the drone belongs to the first area image or the second area image.

[0154] Specifically, S4039 to S40312 are transition judgments on the three-dimensional point cloud data corresponding to the second area image to the three-dimensional point cloud data of the first area image. The judgment method is the same as the transition of the three-dimensional point cloud data of the first area image to the three-dimensional point cloud data of the second area image, and will not be repeated here.

[0155] It should be noted that the marking verification of the previous three-dimensional point cloud data is also logically the same as the aforementioned image verification method, which will not be repeated here.

[0156] S500: Determine the length of the corridor by using the second area image and the second three-dimensional point cloud data.

[0157] Specifically, determining the corridor length includes steps S501 to S503.

[0158] S501, obtaining three-dimensional point cloud data at the junction of the first area image and the second area image, and two images with partially overlapping contents, and obtaining corresponding positioning three-dimensional point cloud data and positioning image data respectively.

[0159] S502, comprehensively analyzing the positioning three-dimensional point cloud data and the positioning image data to obtain the initial positioning and final positioning of the UAV.

[0160] Among them, if the number of corridors is one, the corridor connects two single buildings, and the two connection positions correspond to the initial positioning and the terminal position.

[0161] S503, determining the length of the corridor using the initial position and the end position.

[0162] Among them, in the above steps, the transition between the first area image and the second area image, that is, the three-dimensional point cloud data at the junction of the first area image and the second area image, has been obtained. Next, the last image of the first area image and the first image of the second area image are obtained, and part of the content of these two images will overlap, so as to obtain the positioning image data.

[0163] Then, first, feature points of the two images are extracted by feature extraction method. Next, the feature points of the two images are matched to obtain matching point pairs, and the essential matrix or basic matrix between the two images is calculated. These matrices describe the geometric relationship between the two images. Using the essential matrix or basic matrix, the attitude of the camera of the drone can be estimated, including the rotation matrix and translation vector of the camera. Next, using the matching point pairs and the estimation of the camera attitude, a triangulation operation can be performed, that is, the feature points on the image are associated with the corresponding three-dimensional point cloud data. Finally, by minimizing the reprojection error, the camera attitude can be optimized to obtain a more accurate camera position estimate, and then combined with the GPS positioning of the drone, the positioning of the drone is realized. That is, the junction of the first area image and the second area image is the initial positioning, and the junction of the second area image and the first area image is the terminal positioning. The length of the corridor can be determined by calculating the distance between the initial position and the terminal position. In addition, the length of the corridor can be obtained by combining the number of images of the second area, the time required to shoot these images, and the speed of the drone.

[0164] S600, adjust the first walking route of the UAV so that the distance between the UAV and the corridor is equal to the viewpoint distance, obtain the second walking route of the UAV, and based on the second walking route, obtain the third area image taken by the UAV and the third three-dimensional point cloud data corresponding to the third area image.

[0165] The process of obtaining the third area image photographed by the drone and the third three-dimensional point cloud data corresponding to the third area image includes steps S601 to S605.

[0166] S601, on a first walking route of the UAV, marking an initial position and an end position, and marking an image taken by the UAV between the initial position and the end position, to obtain an image to be processed.

[0167] S602, obtaining the walking direction of the UAV according to the initial position and the final position.

[0168] S603, obtaining a second walking route according to the walking direction of the drone and the viewpoint distance.

[0169] S604, according to the second walking route, obtain the corridor image taken by the drone, obtain the third area image, and the third three-dimensional point cloud data corresponding to the third area image.

[0170] S605, deleting the image to be processed, and filling the third region image and the third three-dimensional point cloud data corresponding to the third region image into the time axis empty window period caused by deleting the image to be processed.

[0171] Specifically, after obtaining the initial position and the end position, mark it on the first walking route. Next, mark the image taken by the drone between the initial position and the end position as the image to be processed. Since the flight of the drone requires direction and position, the walking direction of the drone can be obtained using the initial position and the end position. Next, if you want to get a better result of taking pictures of the corridor, you also need to ensure that the position of the drone is at the optimal distance. Therefore, the drone moves in the direction of the corridor so that the drone and the corridor are at the optimal distance (the distance between the drone and the corridor can be determined by laser radar. In addition, the movement of the drone to the corridor can also be measured by laser radar ranging. The drone moves according to the distance obtained by the laser radar feedback). Similarly, the drone is placed at the 11 o'clock direction of the corridor. In this way, the optimal distance and flight direction of the drone, that is, the second walking route, are obtained. Next, the drone works on the corridor along the second walking route, and then obtains the third area image and the third three-dimensional point cloud data corresponding to the third area image. Finally, since a better third area image than the second area image has been obtained, it is no longer meaningful to retain the second area image. Therefore, the image to be processed and the corresponding three-dimensional point cloud data are deleted, and the obtained third area image and the third three-dimensional point cloud data corresponding to the third area image are used to fill the time axis window period caused by deleting the image to be processed, so that the first area image and the third area image are complete.

[0172] It should be noted that the third area image includes not only the image of the corridor, but also the image of the adjacent faces of the two buildings taken when the drone flies to the corridor. It should also be noted that since there may be only one corridor adjacent to the adjacent faces between the two buildings, if the drone executes the second walking route once, the depth of the drone flight is determined by the optimal distance, which will cause the relative surface data above and below the corridor to be lost (the second area image will be deleted, and the corridor is in a suspended state. If the corridor is on the ground, only the relative surface data above is lost). Therefore, after determining the depth of the drone flight according to the optimal distance, the horizontal distance of this flight is recorded. Next, when the drone completes the shooting of the corridor, the distance between the drone and the ground is adjusted, and the depth distance recorded by the flight in the horizontal direction from the initial position is recorded, and then after flying along the second walking route, it returns to the end position. It can be understood that the initial position and the end position referred to here represent the position after the vertical height is adjusted. In addition, since the building is high, when it is necessary to shoot the side of the building at 9 o'clock, the second walking route after adjusting the vertical height can be executed. The present application does not limit when the drone takes the second walking route after adjusting the vertical height. After photographing the corridor, it can also photograph the images of the opposite sides of the two buildings and collect three-dimensional point cloud data.

[0173] S700: Input the first area image, the first three-dimensional point cloud data, the third area image and the third three-dimensional point cloud data into BIM to obtain a BIM building model.

[0174] Wherein, obtaining the BIM building model includes steps S701 to S704.

[0175] S701, obtaining the parameters of the first area image and the third area image, as well as the first three-dimensional point cloud data and the third three-dimensional point cloud data formed by the drone, and obtaining the calibrated target area image and target three-dimensional point cloud data.

[0176] Among them, the target area image and the target three-dimensional point cloud data form a corresponding relationship.

[0177] S702, performing feature extraction on the target area image and the target three-dimensional point cloud data to obtain target area image features and target three-dimensional point cloud data features respectively.

[0178] S703 , matching the target area image features with the target three-dimensional point cloud data to obtain a corresponding relationship between the target area image features and the target three-dimensional point cloud data features.

[0179] S704, inputting the corresponding relationship between the target area image features and the target three-dimensional point cloud data features, as well as the target area image features and the target three-dimensional point cloud data features into BIM to obtain a BIM building model.

[0180] Specifically, since the first area image and the third area image are acquired by shooting, and the first three-dimensional point cloud data and the third three-dimensional point cloud data are acquired by laser radar, if you want to form a three-dimensional model with high accuracy, you need to calibrate the image data and the three-dimensional point cloud data. The calibration method can be through feature matching or time matching to form a corresponding relationship. Next, the correspondence between the target area image features and the target three-dimensional point cloud data features, as well as the target area image features and the target three-dimensional point cloud data features, are input into BIM to obtain a BIM building model. In addition, if you want to quickly build a BIM building model, you can also directly input the first area image and the third area image, as well as the parameters of the first three-dimensional point cloud data and the third three-dimensional point cloud data into BIM.

[0181] It should be noted that the feature extraction is a conventional feature extraction method, and this application does not limit this.

[0182] In summary, the present application provides a method for building viewpoints based on BIM, which utilizes aerial photography of drones to form data corresponding to multiple building viewpoints, and utilizes a method combining images and three-dimensional point cloud data to improve the accuracy of the BIM building model, and can solve the problem of poor effect of the corridor between two buildings in the BIM building model.

[0183] A second aspect of the present application provides a BIM-based building viewpoint construction system, which is applied to the aforementioned BIM-based building viewpoint construction method, and the system includes:

[0184] An acquisition unit, used for acquiring a bird's-eye view of a building and a physical scale of the building;

[0185] A determination unit, used to determine the building viewpoint of the drone according to the real object scale and the viewpoint distance, wherein the viewpoint distance is the optimal distance for the drone to photograph the building;

[0186] A route unit, used to construct a first walking route of the UAV using the bird's-eye view;

[0187] A data unit is used to obtain a first area image and a second area image taken by a drone according to a first walking route and a building viewpoint, as well as first three-dimensional point cloud data corresponding to the first area image and second three-dimensional point cloud data corresponding to the second area image, wherein the second area image indicates that the distance between the image taken in the second area image and the drone is greater than the optimal distance;

[0188] A length unit, used for determining the length of the corridor according to the second area image and the second three-dimensional point cloud data;

[0189] The adjustment unit is used to adjust the first walking route of the UAV so that the distance between the UAV and the corridor is equal to the viewpoint distance, obtain the second walking route of the UAV, and obtain the third area image taken by the UAV and the third three-dimensional point cloud data corresponding to the third area image according to the second walking route;

[0190] The result unit is used to input the data of the first area image, the first three-dimensional point cloud data, the third area image and the third three-dimensional point cloud data into BIM to obtain a BIM building model.

[0191] A third aspect of the present application provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the aforementioned BIM-based building viewpoint construction method is implemented.

[0192] A fourth aspect of the present application provides a computer storage medium having a computer program stored thereon, which implements the aforementioned BIM-based building viewpoint construction method when the computer program is executed by a processor.

[0193] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for constructing a building viewpoint based on BIM, characterized in that: Methods include: Obtaining a top view of a building and the physical scale of said building; Determine the building viewpoint of the drone according to the physical object proportion and the viewpoint distance, wherein the viewpoint distance is the optimal distance for the drone to photograph the building; Using the top view, construct a first walking route of the drone; According to the first walking route and the building viewpoint, a first area image and a second area image taken by the drone, as well as first three-dimensional point cloud data corresponding to the first area image and second three-dimensional point cloud data corresponding to the second area image are obtained, wherein the second area image indicates that the distance between the image taken in the second area image and the drone is greater than the optimal distance; Determining the length of the corridor by using the second area image and the second three-dimensional point cloud data; The first walking route of the UAV is adjusted so that the distance between the UAV and the corridor is equal to the viewpoint distance, and a second walking route of the UAV is obtained. According to the second walking route, a third area image taken by the UAV and third three-dimensional point cloud data corresponding to the third area image are obtained. According to the data of the first area image, the first three-dimensional point cloud data, the third area image and the third three-dimensional point cloud data, the data are input into the BIM to obtain the BIM building model; wherein, The step of determining the corridor length by using the second area image and the second three-dimensional point cloud data comprises: Acquire the three-dimensional point cloud data at the junction of the first area image and the second area image, and the two images with partially overlapping contents, and obtain corresponding positioning three-dimensional point cloud data and positioning image data respectively; Comprehensively analyzing the positioning three-dimensional point cloud data and the positioning image data to obtain the initial position and the terminal position of the UAV, wherein, if the number of the corridor is one, the corridor connects two single buildings, and the two connection positions correspond to form the initial position and the terminal position; Determine the length of the corridor using the initial position and the terminal position; The step of adjusting the first walking route of the drone so that the distance between the drone and the corridor is equal to the viewpoint distance, obtaining the second walking route of the drone, and obtaining the third area image taken by the drone and the third three-dimensional point cloud data corresponding to the third area image according to the second walking route includes: On the first walking route of the drone, the initial position and the end position are marked, and the image taken by the drone between the initial position and the end position is marked to obtain an image to be processed; According to the initial position and the final position, the walking direction of the UAV is obtained; Obtaining the second walking route according to the walking direction of the drone and the viewpoint distance; According to the second walking route, the corridor image taken by the drone is acquired to obtain the third area image and third three-dimensional point cloud data corresponding to the third area image; The image to be processed is deleted, and the third region image and the third three-dimensional point cloud data corresponding to the third region image are added into the time axis empty window period caused by deleting the image to be processed.

2. The method for constructing a building viewpoint based on BIM according to claim 1, characterized in that: The step of obtaining the first area image and the second area image taken by the drone according to the first walking route and the building viewpoint, as well as the first three-dimensional point cloud data corresponding to the first area image and the second three-dimensional point cloud data corresponding to the second area image comprises: Randomly select a point on the first walking route as the starting point and end point of the drone shooting; Based on the starting point, a preset speed and a shooting time interval, the building is photographed step by step along the first walking route to obtain the first area image and the second area image; Based on the laser radar of the UAV, the building is scanned along the first walking route to obtain the first three-dimensional point cloud data corresponding to the first area image and the second three-dimensional point cloud data corresponding to the second area image.

3. The method for constructing a building viewpoint based on BIM according to claim 2, characterized in that: The step of photographing the building step by step along the first walking route based on the starting point, a preset speed and a shooting time interval to obtain the first area image and the second area image comprises: Obtain the texture and clarity of the current image, and the average of the texture and clarity of M images before the current image, and pre-set an upper error limit and a lower error limit for the average, to obtain the texture and clarity of the current image, and the range values ​​of the texture and clarity of M images, where M is an integer greater than 1; Compare the texture and clarity of the current image with the M texture and clarity range values ​​to obtain a first image comparison result; If the comparison result of the first image is: the texture and clarity of the current image are within the range of the texture and clarity of the M images, then marking the current image as the first region image; If the comparison result of the first image is: the texture and clarity of the current image are not within the range of the texture and clarity of the M images, then marking the current image as the second region image; Acquire the texture and clarity of N images after the current image marked as the second area image within a preset time period, obtain the average of the N textures and clarity, and pre-set an upper error limit and a lower error limit for the average to obtain N texture and clarity range values, wherein N is an integer greater than 1, and the N images are the number of images taken by the drone within a preset time period according to the flight speed of the drone; Acquire the texture and clarity of the N+1th image after the N images, to obtain the N+1th texture and clarity; Compare the N+1th texture and clarity with the N textures and clarity range values ​​to obtain a second image comparison result; If the comparison result of the second image is: the texture and clarity of the N+1th image are within the range of the texture and clarity of the N images, then the N+1th image is marked as the second region image; If the comparison result of the second image is that the texture and clarity of the N+1th image are not within the range of the textures and clarity of the N images, the N+1th image is marked as the first area image.

4. The method for constructing a building viewpoint based on BIM according to claim 3, characterized in that: The step of scanning the building along the first walking route based on the laser radar of the drone to obtain the first three-dimensional point cloud data corresponding to the first area image and the second three-dimensional point cloud data corresponding to the second area image also includes: Acquire the three-dimensional point cloud data obtained by the laser radar of the drone when shooting the current image, to obtain current three-dimensional point cloud data; Acquire three-dimensional point cloud data of M images before the current image is taken, and use a principal component analysis method to obtain a first plane formed by three principal component directions and corresponding eigenvalues ​​of the three-dimensional point cloud data of the M images; Projecting the three principal component directions and corresponding eigenvalues ​​of the current three-dimensional point cloud data onto the first plane, determining whether the eigenvalues ​​are less than or equal to a set threshold, and obtaining a first feature comparison result; If the first feature comparison result is less than or equal to a set threshold, marking the current image as the first area image; If the comparison result of the first image is greater than a set threshold, marking the current image as the second area image; Acquire three-dimensional point cloud data of N images after the current image marked as the second area image within a preset time period, and obtain a second plane formed by three principal component directions and corresponding eigenvalues ​​of the three-dimensional point cloud data of the N images; After acquiring the N images, the three-dimensional point cloud data of the N+1th image is obtained to obtain the N+1th three-dimensional point cloud data; Projecting the three principal component directions and corresponding eigenvalues ​​of the N+1th three-dimensional point cloud data onto the second plane, determining whether the eigenvalues ​​are less than or equal to a set threshold, and obtaining a second feature comparison result; If the second feature comparison result is less than or equal to the set threshold, marking the N+1th image as the second area image; If the second feature comparison result is greater than or equal to the set threshold, marking the N+1th image as the second area image; A weighted calculation is performed on the first feature comparison result and the second feature comparison result so as to divide the image taken by the drone and determine whether the image taken by the drone belongs to the first area image or the second area image.

5. The method for constructing a building viewpoint based on BIM according to claim 1, characterized in that: The step of inputting the data of the first area image, the first three-dimensional point cloud data, the third area image and the third three-dimensional point cloud data into the BIM to obtain the BIM building model comprises: Acquire the first area image and the third area image formed by the drone, as well as the parameters of the first three-dimensional point cloud data and the third three-dimensional point cloud data, to obtain a calibrated target area image and target three-dimensional point cloud data, wherein the target area image and the target three-dimensional point cloud data form a corresponding relationship; Performing feature extraction on the target area image and the target three-dimensional point cloud data to obtain target area image features and target three-dimensional point cloud data features respectively; Matching the target area image features with the target three-dimensional point cloud data to obtain a corresponding relationship between the target area image features and the target three-dimensional point cloud data features; According to the correspondence between the target area image features and the target three-dimensional point cloud data features, and the target area image features and the target three-dimensional point cloud data features, they are input into BIM to obtain the BIM building model.

6. A BIM-based building viewpoint construction system, characterized in that: The BIM-based building viewpoint construction method applied to any one of claims 1 to 5, the system comprising: An acquisition unit, used for acquiring a top view of a building and a physical scale of the building; A determination unit, configured to determine a building viewpoint of the drone according to the physical object ratio and the viewpoint distance, wherein the viewpoint distance is an optimal distance for the drone to photograph the building; A route unit, configured to construct a first walking route of the UAV by using the top view; a data unit, for obtaining, according to the first walking route and the building viewpoint, a first area image and a second area image taken by the drone, as well as first three-dimensional point cloud data corresponding to the first area image and second three-dimensional point cloud data corresponding to the second area image, wherein the second area image indicates that the distance between the image taken in the second area image and the drone is greater than the optimal distance; A length unit, used to determine the length of the corridor based on the second area image and the second three-dimensional point cloud data; an adjusting unit, configured to adjust the first walking route of the drone so that the distance between the drone and the corridor is equal to the viewpoint distance, obtain a second walking route of the drone, and obtain a third area image taken by the drone and third three-dimensional point cloud data corresponding to the third area image according to the second walking route; A result unit is used to input the first area image, the first three-dimensional point cloud data, the third area image and the third three-dimensional point cloud data into the BIM to obtain the BIM building model; wherein, The step of determining the corridor length by using the second area image and the second three-dimensional point cloud data comprises: Acquire the three-dimensional point cloud data at the junction of the first area image and the second area image, and the two images with partially overlapping contents, and obtain corresponding positioning three-dimensional point cloud data and positioning image data respectively; Comprehensively analyzing the positioning three-dimensional point cloud data and the positioning image data to obtain the initial position and the terminal position of the UAV, wherein, if the number of the corridor is one, the corridor connects two single buildings, and the two connection positions correspond to form the initial position and the terminal position; Determine the length of the corridor using the initial position and the terminal position; The step of adjusting the first walking route of the drone so that the distance between the drone and the corridor is equal to the viewpoint distance, obtaining the second walking route of the drone, and obtaining the third area image taken by the drone and the third three-dimensional point cloud data corresponding to the third area image according to the second walking route includes: On the first walking route of the drone, the initial position and the end position are marked, and the image taken by the drone between the initial position and the end position is marked to obtain an image to be processed; According to the initial position and the final position, the walking direction of the UAV is obtained; Obtaining the second walking route according to the walking direction of the drone and the viewpoint distance; According to the second walking route, the corridor image taken by the drone is acquired to obtain the third area image and third three-dimensional point cloud data corresponding to the third area image; The image to be processed is deleted, and the third region image and the third three-dimensional point cloud data corresponding to the third region image are added into the time axis empty window period caused by deleting the image to be processed.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the BIM-based building viewpoint construction method according to any one of claims 1 to 5 is implemented.

8. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the BIM-based building viewpoint construction method according to any one of claims 1 to 5 is implemented.

Citation Information

Patent Citations

  • Real-time three-dimensional reconstruction method based on unmanned aerial vehicle image

    CN109961497A

  • Intelligent lofting construction method based on unmanned aerial vehicle inclined image technology

    CN111951398A

  • Building template instantiation modeling method based on point cloud data

    CN115033967A

  • Building surveying and mapping method based on unmanned aerial vehicle remote sensing and unmanned aerial vehicle

    CN116295279A

  • Bridge route planning method and device

    CN116753962A