CN114255318A

CN114255318A - Method and device for building house type model, electronic equipment and storage medium

Info

Publication number: CN114255318A
Application number: CN202011017361.9A
Authority: CN
Inventors: 刘勉励; 曾翔
Original assignee: Guangdong Bozhilin Robot Co Ltd
Current assignee: Guangdong Bozhilin Robot Co Ltd
Priority date: 2020-09-24
Filing date: 2020-09-24
Publication date: 2022-03-29

Abstract

The embodiment of the invention discloses a method and a device for building a house type model, electronic equipment and a storage medium. The method comprises the following steps: carrying out plane segmentation on three-dimensional point cloud data of a target building, extracting each wall in the target building, and identifying the type of each wall; performing horizontal projection on the three-dimensional point cloud data of each wall, and determining the position of each wall in the target building based on the projection of each wall; determining the connection relation of each wall body based on the position of each wall body and the type of each wall body, and drawing a two-dimensional house type graph of the target building based on the position of each wall body, the connection relation of each wall body and the thickness of each wall body; and determining a three-dimensional house type model of the target building based on the two-dimensional house type graph. The method and the device can accurately and efficiently reconstruct the house type model, so that the effect of accurate positioning can be realized when the operation is carried out according to the reconstructed house type model.

Description

Method and device for building house type model, electronic equipment and storage medium

Technical Field

The embodiment of the invention relates to an image processing technology, in particular to a method and a device for building a house type model, electronic equipment and a storage medium.

Background

With the development of the technology, more and more attention is paid to the three-dimensional reconstruction work of the building, and people can check various information of the building based on the reconstructed building model so as to conveniently and visually know the building.

At present, the main mode of house type graph modeling is to use some software according to the two-dimensional design drawing of the house type graph to model the house type graph to obtain a three-dimensional BIM model. The model obtained in this way can represent the whole of the house type diagram, but due to various accidents possibly occurring in the construction process and the situation that the precision cannot reach the standard, the situation that the BIM model does not accord with the size of the actual house type diagram often occurs.

This problem causes a problem that when the construction robot performs work with reference to the BIM model, the work surface positioning is likely to fail due to an error.

Disclosure of Invention

The embodiment of the invention provides a method and a device for building a house type model, electronic equipment and a storage medium, which are used for accurately and efficiently rebuilding the house type model so as to realize the effect of accurate positioning when working according to the rebuilt house type model.

In a first aspect, an embodiment of the present invention provides a method for constructing a house type model, where the method includes:

carrying out plane segmentation on three-dimensional point cloud data of a target building, extracting each wall in the target building, and identifying the type of each wall;

horizontally projecting each wall plane, and determining the position of each wall in the target building based on each wall projection;

determining the connection relation of each wall body based on the position of each wall body and the type of each wall body, and drawing a two-dimensional house type graph of the target building based on the position of each wall body, the connection relation of each wall body and the thickness of each wall body;

and determining a three-dimensional house type model of the target building based on the two-dimensional house type graph.

In a second aspect, an embodiment of the present invention further provides a device for building a house type model, where the device includes:

the wall body extraction module is used for carrying out plane segmentation on the three-dimensional point cloud data of the target building, extracting each wall body in the target building and identifying the type of each wall body;

the wall body position determining module is used for horizontally projecting each wall body plane and determining the position of each wall body in the target building based on each wall body projection;

the two-dimensional graph acquisition module is used for determining the connection relation of each wall body based on the position of each wall body and each wall body type, and drawing a two-dimensional house type graph of the target building based on the position of each wall body, the connection relation of each wall body and the thickness of each wall body;

and the three-dimensional model determining module is used for determining a three-dimensional house type model of the target building based on the two-dimensional graph.

In a third aspect, an embodiment of the present invention further provides an electronic device, where the electronic device includes:

one or more processors;

storage means for storing one or more programs;

when the one or more programs are executed by the one or more processors, the one or more processors implement the method for constructing the house model according to any one of the embodiments of the present invention.

In a fourth aspect, the present invention further provides a storage medium containing computer-executable instructions, which when executed by a computer processor, are used to perform the method for building a house type model according to any one of the embodiments of the present invention.

According to the technical scheme of the embodiment of the invention, the three-dimensional point cloud data of the target building is subjected to plane segmentation, all wall bodies in the target building are extracted, the types of all wall bodies are identified, the plane of each wall body is subjected to horizontal projection, the position of each wall body in the target building is determined based on the projection of each wall body, the connection relation of each wall body is determined based on the position of each wall body and the type of each wall body, the two-dimensional house type diagram of the target building is drawn based on the position of each wall body, the connection relation of each wall body and the thickness of each wall body, and the three-dimensional house type model of the target building is determined based on the two-dimensional house type diagram, so that the accurate and efficient reconstruction of the house type model is realized, and the accurate positioning effect is realized when the operation is carried out according to the reconstructed house type model.

Drawings

FIG. 1 is a flow chart of a method for constructing a house type model according to a first embodiment of the present invention;

FIG. 2 is a flow chart of a method for constructing a house type model according to a second embodiment of the present invention;

fig. 3 is a schematic diagram of an indoor area image in the second embodiment of the present invention;

fig. 4 is a diagram illustrating a result of performing a close operation on an indoor area image according to a second embodiment of the present invention;

FIG. 5 is a diagram illustrating the result of performing an open operation on an image after a close operation according to a second embodiment of the present invention;

FIG. 6 is a schematic diagram illustrating a second embodiment of the present invention;

FIG. 7 is a graph showing the result after the flood fill operation in the second embodiment of the present invention;

FIG. 8 is an image after binarization and on operations in the second embodiment of the present invention;

fig. 9 is a schematic diagram of an image of a target indoor area in the second embodiment of the present invention;

FIG. 10 is a schematic diagram of a wall area image according to a second embodiment of the present invention;

FIG. 11 is a diagram illustrating a result of performing an open operation on a wall area image according to a second embodiment of the present invention;

FIG. 12 is a diagram illustrating the wall centerline extraction result in accordance with a second embodiment of the present invention;

FIG. 13 is a flowchart of a method for constructing a house type model according to a third embodiment of the present invention;

FIG. 14 is a schematic diagram of the centerline extraction result of the target wall in the third embodiment of the present invention;

FIG. 15 is a flowchart of a house type model constructing method according to a fourth embodiment of the present invention;

fig. 16 is a connection diagram of the walls of the target building according to the fourth embodiment of the present invention;

fig. 17 is a schematic view of the case where the walls intersect in the fourth embodiment of the present invention;

FIG. 18 is a schematic illustration of wall vertex coordinate determination in a fourth embodiment of the present invention;

fig. 19 is a two-dimensional house view of a target building in a fourth embodiment of the present invention;

FIG. 20 is a schematic diagram of a three-dimensional house type model of a target building according to a fourth embodiment of the present invention;

fig. 21 is a schematic structural diagram of a house type model building apparatus according to a fifth embodiment of the present invention;

fig. 22 is a schematic structural diagram of an electronic device in a sixth embodiment of the present invention.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.

Example one

Fig. 1 is a flowchart of a method for building a house type model according to an embodiment of the present invention, where the present embodiment is applicable to a case where a house type model is accurately built, the method may be executed by a building apparatus for a house type model, the building apparatus for a house type model may be implemented by software and/or hardware, and the building apparatus for a house type model may be configured on a computing electronic device, and specifically includes the following steps:

s110, carrying out plane segmentation on the three-dimensional point cloud data of the target building, extracting each wall in the target building, and identifying the type of each wall.

The target building may be any building that needs to be modeled, for example, a residential house, an office building for office, etc.

The three-dimensional point cloud data of the target building can be obtained by scanning the inside of the target building through an instrument such as a laser scanner. For example, taking a house where a target building is a living house as an example, the three-dimensional point cloud data of the target building may be acquired by using a laser scanner or the like in the house room.

The type of the wall body can be the category of the wall body, and the type of the wall body can be divided into an inner wall and an outer wall. Taking the target building as a residential house as an example, the inner wall can be the wall inside the house, i.e. all the walls in the room; the outer wall here may be a wall outside the house, i.e. the part of the wall that is enclosed outside the house.

And performing plane segmentation on the three-dimensional point cloud data of the target building to obtain wall bodies of all planes of the target building, and identifying the types of all the wall bodies according to preset processing rules based on all the determined wall bodies. The specific extraction of walls and the identification of the type of wall are described in more detail in the following examples.

And S120, performing horizontal projection on each wall plane, and determining the position of each wall in the target building based on the projection of each wall.

Illustratively, after the wall bodies are extracted and the types of the wall bodies are determined, horizontal projection is performed on planes of the wall bodies, the positions of the wall bodies in a target building can be determined according to the projection of the wall bodies, specifically, random sampling consistency linear fitting can be performed on three-dimensional point cloud data corresponding to the projection wall bodies to obtain a fitting linear equation of the three-dimensional point cloud data, the positions of the center points of the wall bodies in the target building can be obtained according to the fitting linear equation and the point cloud abscissa or ordinate of the center points of the wall bodies in the target building, and the positions of the wall bodies in the target building can be determined according to the positions of the center points of the wall bodies in the target building.

Thus, the accurate position coordinates of each wall in the target building can be obtained.

And S130, determining the connection relation of the walls based on the positions of the walls and the types of the walls, and drawing a two-dimensional floor plan of the target building based on the positions of the walls, the connection relation of the walls and the thicknesses of the walls.

For example, the connection relationship of the walls can be determined based on a preset calculation rule according to the position of each wall and the type of each wall, and the connection relationship of the walls can be determined according to the position of each wall and the type of each wall, which is described in detail in the following embodiments.

After the connection relation of each wall body is determined, the endpoint coordinate of each wall body can be determined according to the central point coordinate of each wall body, the connection relation graph of each wall body can be drawn according to the endpoint coordinate of each wall body and the connection relation of each wall body, and the two-dimensional house type graph of the target building can be drawn according to the connection relation graph of each wall body and the thickness of each wall body.

This results in an accurate two-dimensional floor plan of the target building.

And S140, determining a three-dimensional house type model of the target building based on the two-dimensional house type graph.

Illustratively, after a two-dimensional house type graph of the target building is obtained, a three-dimensional house type model of the target building can be drawn according to the endpoint coordinates of each wall and the height of each wall.

Through the mode, the effect of accurately and efficiently reconstructing the house type model can be realized, so that the effect of accurately positioning can be realized when the operation is carried out according to the reconstructed house type model.

Example two

Fig. 2 is a flowchart of a method for constructing a house type model according to a second embodiment of the present invention, and the second embodiment of the present invention may be combined with various alternatives in the above embodiments. In the embodiment of the present invention, optionally, the performing plane segmentation on the three-dimensional point cloud data of the target building, and extracting each wall in the target building includes: determining a normal vector of each three-dimensional point cloud data based on the three-dimensional point cloud data of the target building; performing plane segmentation on the three-dimensional point cloud data of the target building based on the normal vector of each three-dimensional point cloud data and the index of each segmentation unit to determine an indoor area image of the target building; overlapping and aligning the indoor area images, and determining a target indoor area image of a target building; and determining each wall of the target building based on the target indoor area image. The identifying the type of each wall comprises: performing closed operation on the target indoor area image to obtain each connected area; if the image coordinates of the center of the wall body are in the communicated area, determining that the wall body is an inner wall; and if the image coordinate of the center of the wall body is not in the communication area, determining that the wall body is an outer wall.

As shown in fig. 2, the method of the embodiment of the present invention specifically includes the following steps:

s210, determining normal vectors of the three-dimensional point cloud data based on the three-dimensional point cloud data of the target building; performing plane segmentation on the three-dimensional point cloud data of the target building based on the normal vector of each three-dimensional point cloud data and the index of each segmentation unit to determine an indoor area image of the target building; overlapping and aligning the indoor area images, and determining a target indoor area image of a target building; and determining each wall of the target building based on the target indoor area image.

Illustratively, the partition units may be components within the target building, such as, for example, a residential building, and may be, but are not limited to, wall surfaces, ceilings, floors, bay window tops, window bottoms, sill bottoms, door openings, window opening sides, and the like.

The index of the segmentation unit may be a height range, a length range, and a width range of the segmentation unit that comply with the national standard. Namely, when the three-dimensional point cloud data of each target building is subjected to plane segmentation, each segmentation unit is segmented according to what standard. For example, in the case of a target building as a residential building, the partition units may be, but are not limited to, a ceiling, a floor, a bay window ceiling, a beam floor, a door opening, a window opening, and the like, and in the case of a ceiling, the index of the ceiling may be that when the three-dimensional point cloud data of the target building respectively has a height, a width, and a length, the three-dimensional point cloud data in the area can be determined as the ceiling.

The indoor area may be an area included in the target building, for example, a house living in the target building, and the divided area may be, but is not limited to, a bedroom, a living room, a bathroom, a balcony, and the like.

The indoor area image may be an image of an area including a bedroom, a living room, a bathroom, a balcony, and the like.

The target indoor area image may be an image obtained by superimposing and aligning the indoor area images.

After the three-dimensional point cloud data of the target building are obtained, determining a normal vector of each three-dimensional point cloud data, which specifically may be:

(1) the three-dimensional point cloud data of the target building is preprocessed, and the preprocessing comprises conventional denoising, downsampling and the like, so that the more accurate three-dimensional point cloud data with noise removed can be obtained.

(2) And calculating the height of the ground and the ceiling of the target building according to the histogram statistics in the Z direction. And (3) calculating a rotation matrix of the three-dimensional point cloud data, rotating the three-dimensional point cloud data until the normal vector of each plane is parallel to the space coordinate system, and translating the point cloud until the ground height is 0, wherein the height of the ceiling can be determined to be H1 (for example, 3 meters).

(3) Calculating the normal vector of each three-dimensional point cloud data, and dividing all the three-dimensional point cloud data into 4 types according to the normal vector of each three-dimensional point cloud data, wherein the four types are respectively as follows: the normal vector is parallel to the X axis, the normal vector is parallel to the Y axis, the normal vector is parallel to the Z axis and is not parallel to the three coordinate axes.

After the normal vector of the three-dimensional point cloud data is obtained, according to the normal vector and indexes of each segmentation unit, plane segmentation can be performed on the three-dimensional point cloud data of the target building in sequence, and then an indoor area image of the target building is obtained.

The order may be from simple to complex, where simple may be such obvious elements as floors and ceilings, and complex may be bay window tops, beam sides, door openings and window openings. By adopting the mode to carry out division, the division units cannot be lost, and the division is more accurate.

The specific segmentation method comprises the following steps: the z-direction planes are classified according to size: floor (plane with width greater than 0.5m and height close to 0m), ceiling (with width greater than 0.5m and length greater than 1.5m and height close to H1), bay window top (with width greater than 0.4m, length greater than 1.0m, height greater than 1.5m, less than H1-0.1) and window bottom (with width greater than 0.06m, length greater than 0.7m, height less than 1.5m and greater than 0.4m), potential beam bottom (height greater than 2.0 m). In addition to these classifications using geometry, it is also necessary to calculate the ratio of the area of the plane occupied by the coincident area of the horizontal projection of each plane and the ground horizontal projection. The ratio is lower, and is less than 0.1, and the window holes and the bay window are regarded as the top surface and the bottom surface.

The planes in the x and y directions are classified into: the wall surface (the length is more than 0.8m, the height is more than 1.5m, the lower edge is less than 0.2m, or the length is more than 1.2m, the height is more than 0.45m plane), the beam side surface (the height is more than 0.08m, the lower edge is more than 2m), the opening side surface (the width is more than 0.06m, the upper edge is more than 1.8 m).

And after the segmentation is finished, combining the segmentation units to obtain the indoor area image. Specifically, the following steps can be performed: combining all the tunnel opening side surfaces of all the bay window top surfaces, bay window bottom surfaces and potential beam bottom surfaces, and if the tunnel openings can be formed, determining that the tunnel openings are the side surfaces of the tunnel openings; if not, the bay window top surface is considered a ceiling and the potential soffit surface is considered a soffit.

Calculating the maximum height Hz in the z-direction surface of the opening, and taking the point cloud above Hz and adding all the ceiling and the bottom surface of the beam to perform horizontal projection to obtain the reference of the free space area in the room, i.e. the indoor area image, such as the schematic diagram of the indoor area image shown in fig. 3.

It should be noted that all of the ceilings and soffit surfaces are added because of the possibility of having opening top surfaces higher than some of the ceiling and soffit surfaces.

After the indoor area image of the target building is obtained, the indoor area image is superposed and aligned to obtain the target indoor area image of the target building, wherein the purpose of superposing and aligning the indoor area image is to remove the problem of fine connected areas caused by noise in each indoor area.

Optionally, the aligning and superimposing the indoor area image to determine the target indoor area image of the target building may specifically be: and preprocessing the indoor area image to obtain a preprocessed indoor area image, and aligning and superposing all connected areas in the preprocessed indoor area image to obtain a target indoor area image of the target building.

For example, the preprocessing of the indoor area image may be to perform a closing operation on the indoor area image to obtain a result diagram of performing the closing operation on the indoor area image, which is described in fig. 4. Then, the horizontal and vertical opening operations are performed on fig. 8, and the fine connected regions (as part a in fig. 4) caused by noise between the indoor regions are removed, so as to obtain the result diagram of performing the opening operation on the image after the closing operation, which is described in fig. 5. The image in fig. 5 here has the connected components resulting from noise removed.

In fig. 5, there are 5 connected domains, and each connected domain is taken to perform the alignment operation, where the purpose of the alignment operation is: since there are jagged edges of the contours in fig. 5 (not shown in fig. 5, and the contours in fig. 5 are flat and not jagged), the alignment operation is to level the jagged boundaries of the contours in fig. 5 as in fig. 5.

Specifically, the irregular contour lines may be expanded by a certain number of pixels to obtain a flat boundary as shown in fig. 5. The dilated pixels may be determined according to the thickness of the wall and the scale, and are generally a ratio of the thickness value slightly larger than the thickness value of the wall to the scale. For example, if the wall thickness is 0.2m and the scale bar is 5mm for one pixel, the pixels of dilation here can be: 0.3/0.005.

After the indoor areas in the lower left corner of fig. 5 are expanded, taking the indoor area in the lower left corner of fig. 5 as an example, a plane related to a window opening in all planes with normal vectors x and y and a plane of a beam surface are removed, whether the horizontal projection of the plane satisfying the requirement and the expanded indoor area in the lower left corner of fig. 5 are overlapped or not is calculated, if so, a division straight line is drawn by using the plane, and a schematic drawing is drawn by using the division line shown in fig. 6.

Each of the segmentation modules in fig. 6 (where the segmentation module may be any one of the closed regions in fig. 6, such as the closed region B in fig. 6) is filled with water, and the filling result is subtracted from fig. 6 to obtain a mask of the segmentation module. The lower left corner area of fig. 5 is masked by using the mask, the number of pixels of the result of the masking process is calculated to be not 0 (namely, the number of pixels 255, and the default is only two pixels, namely, 0 and 255), and the contour information of the result of the masking process is compared, so that the label of the segmentation module is determined. And when the result of the masking processing is not 0 pixel number and the contour information of the masking processing result is consistent, the similarity of the processing result is considered to be high, the area is an area in the room, and when the result of the masking processing is not 0 pixel number and/or the contour information of the masking processing result is inconsistent, the similarity of the processing result is considered to be low, and the area is an area outside the room. And filling the areas of the segmentation modules with different similarities by using different colors to obtain a result schematic diagram after the flooding filling operation described in fig. 7.

And (4) carrying out binarization and opening operation on the image 7 to obtain the image after binarization and opening operation in the image 8. The alignment, the division line drawing and the flood filling operation are performed on each indoor area in fig. 5, so as to obtain an image of each indoor area after the flood filling operation, and the images of each indoor area after the flood filling operation are superposed, so as to obtain the target indoor area image shown in fig. 9.

The indoor area image is preprocessed, and then the indoor areas are aligned and superposed, so that a more accurate target indoor area image of the target building can be obtained.

According to the obtained target indoor area image, each wall of the target building can be obtained, and the specific method comprises the following steps: performing closed operation and expansion on the target indoor area image to obtain a wall area image in the target building; and respectively carrying out transverse opening operation and longitudinal opening operation on the wall area images with different sizes and structures to obtain the walls with different thicknesses in the transverse direction and the longitudinal direction of the target building.

Illustratively, the obtained image of the target indoor area of fig. 9 is closed, then expanded by a certain number of pixels (the number of pixels expanded here may be corresponding to the thickness of the wall, for example, assuming that the thickness of the outer wall is 20cm and the scale is 0.005 m/pixel), and then the image of the wall area is subtracted from the image of fig. 9, so as to obtain the schematic diagram of the image of the wall area as shown in fig. 10.

Different segmentation threshold values are set according to the thickness of each wall body in the graph 10, and the horizontal opening operation and the longitudinal opening operation of structures with different sizes are respectively carried out on the graph 10 based on the different segmentation threshold values, so that the wall bodies with different thicknesses are segmented.

For example, if the wall has a Size of 10cm, 17cm and 22cm, respectively, the segmentation threshold set here may be 15cm, 20cm and 25cm, so that first, using a segmentation threshold with d1 ═ 15cm, the morphologically processed kernel is set to Size (d1/scale +1, 1) in the horizontal direction, and Size (1, d1/scale +1) in the corresponding vertical direction. By performing the division operation on each wall in each wall area image in fig. 10, the wall with the thickness less than d1 can be divided, that is, the wall with the thickness less than or equal to 15m can be divided. In the above formula, d1 is the segmentation threshold, and scale is the scale.

Then, the division operation is performed on each wall in fig. 10 by using a division threshold value of 20cm, so that the wall with the thickness of d1-d2 can be divided, wherein d1 is 15cm, and d2 is 20cm, that is, the wall with the thickness of 17cm can be divided.

Then, the division operation is performed on each wall in fig. 10 by using a division threshold value of 25cm, so that the wall with the thickness of d2-d3 can be divided, wherein d2 is 20cm, and d3 is 25cm, that is, the wall with the thickness of 22cm can be divided.

According to the set different segmentation thresholds, the situation that two wall bodies with different thicknesses are directly connected and are judged as the same wall blocking by mistake can be avoided, and thus, the non-crossed rectangles in the transverse direction and the longitudinal direction can be obtained. The result diagram of the opening operation on the wall area image as shown in fig. 11 can be obtained, where the diagram a in fig. 11 is the result of performing the horizontal opening operation on fig. 10, the diagram b in fig. 11 is the result of subtracting the diagram a in fig. 11 from fig. 10, and the diagram a in fig. 11 is subtracted from fig. 10, so that the diagram b in fig. 11 can be obtained. The walls in fig. 11 are not subdivided.

Thus, the purpose of extracting each wall in the target building is achieved.

S220, performing closed operation on the target indoor area image to obtain each connected area; if the image coordinates of the center of the wall body are in the communicated area, determining that the wall body is an inner wall; and if the image coordinate of the center of the wall body is not in the communication area, determining that the wall body is an outer wall.

Illustratively, the types of wall bodies are divided into interior walls and exterior walls. Taking the target building as a residential house as an example, the inner wall can be the wall inside the house, i.e. all the walls in the room; the outer wall here may be a wall outside the house, i.e. the part of the wall that is enclosed outside the house.

When the target indoor area image of fig. 9 is obtained, the closing operation is performed on the target indoor area image of fig. 9, connected areas, for example, an indoor area a and an indoor area B as shown in fig. 9, and when the closing operation is performed on fig. 9, the indoor area a and the indoor area B are connected to form a connected area, and if the image coordinates of the central point of the wall body 1 between the indoor area a and the indoor area B are in the connected area, the wall body 1 is determined to be an inner wall. And if the image coordinate of the central point of the wall body 2 of the indoor area A is not in the connected area, the wall body 2 is an outer wall. This achieves the effect of accurately determining the types of the respective walls in the target building.

Specifically, the orientation of the outer wall may be determined according to the relationship between the outer wall and the connected component, for example, if the outer wall 1 is on the left side of the connected component, the orientation of the outer wall is on the left side.

Specifically, the image coordinates of the center point of each wall may be determined in such a manner that, for each independent rectangle in fig. 11, as a wall that is inseparable, the center point of the rectangle is used as the image center point of the wall, and the center coordinates of the two ends in the direction of the rectangular wall are used as the image coordinates of the two end points of the wall. The width of the rectangular wall in the vertical direction is taken as the image width of the wall.

It should be noted that the centerline of each rectangular wall is extracted, so that the wall centerline extraction result diagram shown in fig. 12 can be obtained. The circle inside the circle in fig. 12 may indicate an error occurring in the calculation process, and the error may be caused by the fact that the first etching is not completely performed and the second etching is performed when the on operation is performed. During subsequent calculations, the error is negligible.

It should be noted that the center line of each rectangular wall herein may be the center line of each rectangular wall in fig. 11, a and b.

And S230, performing horizontal projection on each wall plane, and determining the position of each wall in the target building based on each wall projection.

S240, determining the connection relation of the walls based on the positions of the walls and the types of the walls, and drawing a two-dimensional floor plan of the target building based on the positions of the walls, the connection relation of the walls and the thicknesses of the walls.

And S250, determining a three-dimensional house type model of the target building based on the two-dimensional house type graph.

According to the technical scheme of the embodiment of the invention, the normal vector of each three-dimensional point cloud data is determined through the three-dimensional point cloud data based on the target building; performing plane segmentation on the three-dimensional point cloud data of the target building based on the normal vector of each three-dimensional point cloud data and the index of each segmentation unit to determine an indoor area image of the target building; overlapping and aligning the indoor area images, and determining a target indoor area image of a target building; and determining each wall of the target building based on the target indoor area image, so that the aim of extracting each wall in the target building is fulfilled. Performing closed operation on the target indoor area image to obtain each connected area; if the image coordinates of the center of the wall body are in the communicated area, determining that the wall body is an inner wall; and if the image coordinates of the center of the wall body are not in the communication area, determining that the wall body is an outer wall, so that the effect of accurately determining the type of each wall body in the target building is realized.

EXAMPLE III

Fig. 13 is a flowchart of a method for constructing a house type model according to a third embodiment of the present invention, and the third embodiment of the present invention and various alternatives in the foregoing embodiments may be combined. In the embodiment of the present invention, optionally, the horizontally projecting the three-dimensional point cloud data of each wall, and determining the position of each wall in the target building based on the projection of each wall, include: horizontally projecting each wall plane, and linearly fitting the three-dimensional point cloud data corresponding to each projected wall plane to obtain a linear equation of each wall plane; substituting the horizontal coordinates of the point cloud of each wall center in the horizontal direction into a linear equation to obtain the vertical coordinates of the point cloud of each wall center in the horizontal direction; substituting the vertical coordinates of the point cloud of the centers of the walls in the vertical direction into a linear equation to obtain the horizontal coordinates of the point cloud of the centers of the walls in the vertical direction; determining point cloud coordinates of each wall body center point in the horizontal direction and point cloud coordinates of each wall body center in the vertical direction based on the point cloud abscissa and the point cloud ordinate of each wall body center in the horizontal direction and the point cloud ordinate and the point cloud abscissa of each wall body center in the vertical direction; and determining the position of each wall body in the target building based on the point cloud coordinates of the central points of the wall bodies in the horizontal direction and the point cloud coordinates of the central points of the wall bodies in the vertical direction.

As shown in fig. 13, the method of the embodiment of the present invention specifically includes the following steps:

s310, determining normal vectors of the three-dimensional point cloud data based on the three-dimensional point cloud data of the target building; performing plane segmentation on the three-dimensional point cloud data of the target building based on the normal vector of each three-dimensional point cloud data and the index of each segmentation unit to determine an indoor area image of the target building; overlapping and aligning the indoor area images, and determining a target indoor area image of a target building; and determining each wall of the target building based on the target indoor area image.

S320, performing closed operation on the target indoor area image to obtain each connected area; if the image coordinates of the center of the wall body are in the communicated area, determining that the wall body is an inner wall; and if the image coordinate of the center of the wall body is not in the communication area, determining that the wall body is an outer wall.

S330, horizontally projecting each wall plane, and linearly fitting the three-dimensional point cloud data corresponding to each projected wall plane to obtain a linear equation of each wall plane; substituting the horizontal coordinates of the point cloud of each wall center in the horizontal direction into a linear equation to obtain the vertical coordinates of the point cloud of each wall center in the horizontal direction; substituting the vertical coordinates of the point cloud of the centers of the walls in the vertical direction into a linear equation to obtain the horizontal coordinates of the point cloud of the centers of the walls in the vertical direction; determining point cloud coordinates of each wall body center point in the horizontal direction and point cloud coordinates of each wall body center in the vertical direction based on the point cloud abscissa and the point cloud ordinate of each wall body center in the horizontal direction and the point cloud ordinate and the point cloud abscissa of each wall body center in the vertical direction; and determining the position of each wall body in the target building based on the point cloud coordinates of the central points of the wall bodies in the horizontal direction and the point cloud coordinates of the central points of the wall bodies in the vertical direction.

Exemplarily, after extracting each wall in the target building and determining the type of each wall, performing horizontal projection on the extracted wall plane, performing random sampling consistency line fitting on the three-dimensional point cloud data of each wall subjected to projection, and calculating a linear equation. For the wall body in the horizontal direction, substituting the point cloud abscissa corresponding to the central point of the wall body in the image into the linear equation obtained by fitting to obtain the ordinate Y_{Outer cover}(exterior wall) or Y_{Inner 1}And Y_{Inner 2}(interior wall). Substituting the vertical coordinate of the point cloud corresponding to the central point of the wall in the image of the wall in the vertical direction into the linear equation obtained by fitting to obtain the X coordinate of the horizontal coordinate_{Outer cover}(exterior wall) or X_{Inner 1}And X_{Inner 2}(interior wall). Taking a horizontal wall as an example, the vertical coordinate of the outer wall needs to be calculated according to the result Y_{Outer cover}(exterior wall) the thickness of the exterior wall is offset 1/2 in a direction away from the interior of the house. Ordinate of inner wall ═ Y_{Inner 1}+Y_{Inner 2})/2. The actual thickness of the inner wall body is Y_{Inner 1}-Y_{Inner 2}|。

And determining the point cloud coordinates of the wall body center points in the horizontal direction and the point cloud coordinates of the wall body center points in the vertical direction according to the obtained point cloud abscissa and point cloud ordinate of the wall body center in the horizontal direction and the obtained point cloud ordinate and point cloud abscissa of the wall body center points in the vertical direction, and obtaining the position of each wall body in the target building according to the point cloud coordinates of the wall body center points in the horizontal direction and the point cloud coordinates of the wall body center points in the vertical direction. The accurate positioning of each wall body is realized, and the coordinates of the central point of each wall body are obtained.

It should be noted that, before performing step S330, the method further includes: converting the center line of each wall body to a point cloud coordinate system, and performing external expansion on the center line of each wall body for a preset distance to serve as an interesting area; intercepting three-dimensional point cloud data of a corresponding region of interest in the three-dimensional point cloud data to form three-dimensional point cloud data of the region of interest; performing normal vector calculation and segmentation on the three-dimensional point cloud data area of the region of interest, and extracting a vertical plane parallel to the central line of the wall body; and determining the target wall body based on the extracted vertical plane, the length requirement of the vertical plane wall body, the height requirement of the vertical plane wall body and the type of the wall body.

For example, the preset distance may be a preset distance.

The target wall body can be a wall body which meets the requirements and is obtained after the extracted wall body is deleted.

The walls meeting the requirements may be that, if the walls are interior walls, the number of vertical planes is greater than or equal to 2 (which is caused by the fact that there may be other obstacle planes in the room), and if the walls are exterior walls, the number of vertical planes is greater than or equal to 1, and the normal vector of the planes is parallel to the normal vector of the wall extracted from the image, and the walls which do not meet the requirements in the extracted walls are deleted, so that the target wall can be obtained.

For example, as shown in fig. 12, in the lower left corner area of fig. 12, the wall body of vertical line 1 is an inner wall, and has two surfaces, one is viewed from vertical line 0 to vertical line 1, and the other is viewed from vertical line 2 to vertical line 1. For the wall body with the vertical line 0, which is an outer wall, the wall body has 1 surface, and the surface is seen from the vertical line 1 to the vertical line 0.

It should be noted that, here, when looking at the surface of the wall, the user stands in the indoor area to look at the surface.

It should be noted that, after the target wall is obtained, subsequent calculation of the wall is performed based on the target wall.

The length requirement of the vertical planar wall may be a requirement for the length of the extracted vertical planar wall, for example, may be a length greater than a preset percentage of the length of the vertical planar wall, such as a length greater than 0.9 of the vertical planar wall.

Converting the central line of each wall body into a point cloud coordinate system, extending a preset distance (for example, 0.3m) to two ends in the direction perpendicular to the wall line to serve as an interesting region, and intercepting three-dimensional point cloud data of the corresponding wall body region in the original three-dimensional point cloud data to form three-dimensional point cloud data of the interesting region. And performing normal vector calculation and RANSAC segmentation on the three-dimensional point cloud data of the region of interest, and extracting all vertical planes parallel to the central line of the wall. And if the width of the extracted vertical plane in the horizontal direction is greater than 0.9 times the length of the center line of the wall body, and the height of the extracted vertical plane is greater than 2.0m, the extracted vertical plane is considered as the wall surface.

According to the extracted vertical plane and the type of the wall body, the wall body which does not meet the requirement can be deleted according to the rule of the wall body which meets the requirement. Namely, after the redundant wall in fig. 12 is deleted, the target wall center line extraction result schematic diagram in fig. 14 can be obtained.

Therefore, an accurate and qualified target wall is obtained, and a two-dimensional house type graph is constructed on the basis of the accurate and qualified target wall.

S340, determining the connection relation of the walls based on the positions of the walls and the types of the walls, and drawing a two-dimensional floor plan of the target building based on the positions of the walls, the connection relation of the walls and the thicknesses of the walls.

And S350, determining a three-dimensional house type model of the target building based on the two-dimensional house type graph.

According to the technical scheme of the embodiment of the invention, horizontal projection is carried out on each wall plane, and linear fitting is carried out on three-dimensional point cloud data corresponding to each wall plane subjected to projection to obtain a linear equation of each wall plane; substituting the horizontal coordinates of the point cloud of each wall center in the horizontal direction into a linear equation to obtain the vertical coordinates of the point cloud of each wall center in the horizontal direction; substituting the vertical coordinates of the point cloud of the centers of the walls in the vertical direction into a linear equation to obtain the horizontal coordinates of the point cloud of the centers of the walls in the vertical direction; determining point cloud coordinates of each wall body center point in the horizontal direction and point cloud coordinates of each wall body center in the vertical direction based on the point cloud abscissa and the point cloud ordinate of each wall body center in the horizontal direction and the point cloud ordinate and the point cloud abscissa of each wall body center in the vertical direction; and determining the position of each wall body in the target building based on the point cloud coordinates of the central points of the wall bodies in the horizontal direction and the point cloud coordinates of the central points of the wall bodies in the vertical direction, so that the accurate positioning of each wall body is realized, and the central point coordinates of each wall body are obtained.

Example four

Fig. 15 is a flowchart of a method for constructing a house type model according to a fourth embodiment of the present invention, and the fourth embodiment of the present invention and various alternatives in the foregoing embodiments may be combined. In the embodiment of the present invention, optionally, the determining the connection relationship of the walls based on the positions of the walls and the types of the walls includes: determining the walls connected in different directions based on the positions of the walls, and establishing the connection relation of the walls connected in different directions; and determining the walls connected in the same direction based on the positions of the walls, and establishing the connection relation of the walls connected in the same direction according to the types of the walls and the walls connected in the same direction. The two-dimensional house type graph of the target building is drawn based on the positions of the wall bodies, the connection relation of the wall bodies and the thickness of the wall bodies, and the two-dimensional house type graph comprises the following steps: determining the end point coordinates of each wall body based on the coordinates of the central point of each wall body; determining a connection relation graph of each wall of the target building based on the endpoint coordinates of each wall and the connection relation of each wall; and drawing a two-dimensional house type graph of the target building based on the connection relation graph and the thickness of each wall body.

As shown in fig. 15, the method of the embodiment of the present invention specifically includes the following steps:

s410, determining normal vectors of the three-dimensional point cloud data based on the three-dimensional point cloud data of the target building; performing plane segmentation on the three-dimensional point cloud data of the target building based on the normal vector of each three-dimensional point cloud data and the index of each segmentation unit to determine an indoor area image of the target building; overlapping and aligning the indoor area images, and determining a target indoor area image of a target building; and determining each wall of the target building based on the target indoor area image.

S420, performing closed operation on the target indoor area image to obtain each connected area; if the image coordinates of the center of the wall body are in the communicated area, determining that the wall body is an inner wall; and if the image coordinate of the center of the wall body is not in the communication area, determining that the wall body is an outer wall.

S430, horizontally projecting each wall plane, and linearly fitting the three-dimensional point cloud data corresponding to each projected wall plane to obtain a linear equation of each wall plane; substituting the horizontal coordinates of the point cloud of each wall center in the horizontal direction into a linear equation to obtain the vertical coordinates of the point cloud of each wall center in the horizontal direction; substituting the vertical coordinates of the point cloud of the centers of the walls in the vertical direction into a linear equation to obtain the horizontal coordinates of the point cloud of the centers of the walls in the vertical direction; determining point cloud coordinates of each wall body center point in the horizontal direction and point cloud coordinates of each wall body center in the vertical direction based on the point cloud abscissa and the point cloud ordinate of each wall body center in the horizontal direction and the point cloud ordinate and the point cloud abscissa of each wall body center in the vertical direction; and determining the position of each wall body in the target building based on the point cloud coordinates of the central points of the wall bodies in the horizontal direction and the point cloud coordinates of the central points of the wall bodies in the vertical direction.

S440, determining the walls connected in different directions based on the positions of the walls, and establishing the connection relation of the walls connected in different directions; and determining the walls connected in the same direction based on the positions of the walls, and establishing the connection relation of the walls connected in the same direction according to the types of the walls and the walls connected in the same direction.

For example, the opposite walls may be walls that are not aligned with the current wall, for example, as shown in fig. 14, the vertical line 8 is vertical, and the opposite walls of the vertical line 8 are the horizontal line 12 and the horizontal line 11.

The wall bodies connected in the same direction can be the wall bodies in the same direction as the current wall body, for example, as in fig. 14, the transverse line 9 is in the transverse direction, and the wall bodies in the same direction of the transverse line 9 are the transverse line 6 and the transverse line 8.

The image positions of the central points of the wall bodies are calculated, but point cloud coordinates of two ends of the wall bodies are unknown, and no relation is established among all the wall bodies. The connection relationship among the walls is established as follows:

(1) and determining the walls connected in different directions based on the positions of the walls, and establishing the connection relation of the walls connected in different directions.

The method comprises the steps of searching for the end point of the wall body adjacent to each end point of each wall body in a different direction, firstly setting a threshold value (set to be 0.4m in the experiment), taking the different-direction wall body farther away from the wall body as a target wall body connected with the target wall body within the threshold value, and enabling the center line of any one target wall body in the two target wall bodies not to be intersected with the extension line of the center line of the other target wall body.

When the threshold is set, the threshold is set according to the thickness of the two walls in different directions, and specifically, the threshold may be set according to the distance between the two walls in different directions calculated according to the pythagorean theorem. For example, as shown in fig. 14, the vertical line 8 and the horizontal line 12, and the upper end point of the vertical line 8 and the right end point of the horizontal line 12 are distances calculated according to the pythagorean theorem. In general, the threshold value may be set to be slightly larger than the distance calculated according to the Pythagorean theorem, and may be set to be 0.4m in the embodiment of the present invention. The specific threshold value can be set according to the user requirement, and is not limited here.

In this case, it is determined that the wall adjacent to the current wall may be, for example, as shown in fig. 14, if the vertical line 8 is the current wall, and the upper end point of the vertical line 8 is the current end point, the right end point of the horizontal line 12 and the left end point of the horizontal line 11 are both end points of another wall adjacent to the current end point (upper end point) of the current wall, that is, the number of end points adjacent to the current end point of the current wall is 2. According to the pythagorean theorem, the distance between the upper end point of the vertical line 8 and the right end point of the horizontal line 12 and the distance between the upper end point of the end point vertical line 8 and the left end point of the horizontal line 11 are calculated, if the distance between the upper end point of the vertical line 8 and the right end point of the horizontal line 12 is greater than the distance between the upper end point of the vertical line 8 and the left end point of the horizontal line 11, the image ordinate of the upper end point of the vertical line 8 is taken as the point cloud ordinate of the right end point of the horizontal line 12, namely, the wall body (the horizontal line 12) corresponding to the right end point of the horizontal line 12 is taken as the wall body connected with the target wall body (the vertical line 8).

And after the walls which are connected in different directions of the walls are determined, the connection relation of the walls which are connected in different directions is established.

(2) And determining the walls connected in the same direction based on the positions of the walls, and establishing the connection relation of the walls connected in the same direction according to the types of the walls and the walls connected in the same direction.

For the same-direction connected wall bodies, two cases are distinguished:

(a) adjacent walls of similar orientation, class, and approximate thickness without labels.

If a wall is deleted in fig. 12, if the No. 4 vertical wall is deleted, the

transverse walls

8 and 9, which are originally connected with the No. 4 vertical wall in an opposite direction, will lose an end point. At this time, the end points adjacent to the left end point of the transverse wall 8 and the right end point of the transverse wall 9 are both one, that is, one end point of the transverse wall 8 and the transverse wall 9 is lost, and no corresponding label is provided. Thus, after the calculation of the above (1), the longitudinal coordinates of the

lateral walls

8 and 9 are not determined.

For the end points (such as the horizontal line walls 8 and 9) of the wall body with only one end point, namely the same-direction wall body (outer wall or inner wall) with the same thickness and without labels (the coordinate in a certain direction is not determined after the calculation in the step (1)), if the coordinate difference in the wall body direction of the two end points (the left end point of the horizontal wall body 8 and the right end point of the horizontal wall body 9) is smaller than the maximum wall body thickness, the coordinate difference in the vertical wall body direction is smaller than 0.005m, and the wall body thickness difference is within 0.005 m. That is, the difference between the abscissa of the center line of the transverse wall 8 and the abscissa of the center line of the transverse wall 9 is smaller than the maximum wall thickness (for example, 30 cm), the difference between the ordinate of the center line of the transverse wall 8 and the ordinate of the center line of the transverse wall 9 is smaller than 0.005m, and the difference between the wall thickness of the transverse wall 8 and the wall thickness of the transverse wall 9 is smaller than 0.005 m. The

lateral walls

8 and 9 are connected as one wall.

And the coordinate on the coordinate axis of the central point of the combined wall body, which is perpendicular to the wall body direction, is equal to the sum of the coordinates of the two central points of the 1/2 original wall body, namely, the vertical coordinate of the central point of the combined wall body is half of the sum of the vertical coordinate of the original transverse wall body 8 and the vertical coordinate of the original transverse wall body 9.

Labels at two ends of the wall body (namely the serial numbers of the wall bodies) are replaced by labels at two ends with the farthest distance in the original wall body, namely, the combined transverse wall body 9 is replaced by the transverse wall body 8, namely, the combined transverse line wall body 8 and the transverse wall body 9 are collectively called as the transverse wall body 8. The thickness of the wall body is replaced by the thickness of a longer wall body, namely the thickness of the combined wall body is the thickness of the original wall body with the larger thickness, for example, the thickness of the original transverse wall body 8 is 10.1 centimeters, the thickness of the original transverse wall body 9 is 10.2 centimeters, and the thickness of the combined wall body is 10.2 centimeters.

(b) The adjacent collinear walls with labels in the same direction, the same type and the similar thickness are provided.

The above case (a) is for the end points of the wall bodies with the same direction and the same type of thickness close to each other without the label, and the case (b) is for the end points of the wall bodies (outer walls or inner walls) with the same direction and the same type with the existing label, if the distance between the two end points in the wall body direction is less than the maximum wall body thickness, the distance in the wall body vertical direction is less than 0.002m, and the wall body thickness difference is within 0.005m, the two wall bodies are connected into one wall body. For example, for the vertical wall 4 and the vertical wall 3, both the vertical wall 4 and the vertical wall 3 have labels, and if the difference between the abscissa of the center line of the vertical wall 4 and the abscissa of the center line of the vertical wall 3 is smaller than the maximum wall thickness, the difference between the ordinate of the center line of the vertical wall 4 and the ordinate of the center line of the vertical wall 3 is smaller than 0.002m, and the difference between the wall thickness of the vertical wall 4 and the wall thickness of the vertical wall 3 is smaller than 0.005m, the vertical wall 4 and the vertical wall 3 are combined into one wall.

And the coordinate on the coordinate axis of the central point of the combined wall body, which is vertical to the direction of the wall body, is equal to the sum of the coordinates of the two central points of the 1/2 original wall body. Namely, the abscissa of the central point of the combined wall body is half of the sum of the abscissa of the original vertical wall body 4 and the abscissa of the original vertical wall body 3.

The labels at the two ends of the wall body (namely the serial numbers of the wall bodies) are replaced by the labels at the two ends with the farthest distance in the original wall body, namely, the combined vertical wall body 4 is replaced by the vertical wall body 3, namely, the combined vertical wall body 4 and the combined vertical wall body 3 are collectively called as the vertical wall body 3. The thickness of the wall body is replaced by the thickness of a longer wall, namely the thickness of the combined wall body is the thickness of the original wall body with the larger thickness, for example, the thickness of the original vertical wall body 4 is 10.1 centimeters, the thickness of the original vertical wall body 3 is 10.2 centimeters, and the thickness of the combined wall body is 10.2 centimeters.

And after the walls connected in the same direction are determined, the connection relation of the walls connected in the same direction is established.

S450, determining the endpoint coordinates of each wall body based on the coordinates of the central point of each wall body; determining a connection relation graph of each wall of the target building based on the endpoint coordinates of each wall and the connection relation of each wall; and drawing a two-dimensional house type graph of the target building based on the connection relation graph and the thickness of each wall body.

For example, the connection relation graph may be a connection relation graph constructed based on connection relations of walls.

After the connection relation of the walls is determined, the endpoint coordinates of the walls can be determined according to the coordinates of the central points of the walls, and the connection relation graph of the walls is determined according to the endpoint coordinates of the walls and the connection relation of the walls.

Specifically, the endpoint coordinates of each wall are determined according to the coordinates of the center point of each wall, and the following two situations can be divided into:

(1) endpoint determination for labeled walls

For the wall with the label, the abscissa of the two end points of the transverse wall is replaced by the abscissa of the central point of the longitudinal wall with the end points corresponding to the label, for example, through the calculation in (1), it is determined that the vertical wall 11 corresponds to the label of the right end point of the transverse wall 8, the vertical wall 4 corresponds to the label of the left end point of the transverse wall, the abscissa of the right end point of the transverse wall 8 is the abscissa of the central point of the vertical wall 11, and the abscissa of the left end point of the transverse wall is the abscissa of the central point of the vertical wall 4.

And replacing the vertical coordinates of the two end points of the longitudinal wall body with the vertical coordinates of the central point of the transverse wall body of which the end points correspond to the labels. For example, through the calculation of (1) above, it is determined that the label corresponding to the upper end point of the longitudinal wall 4 is the transverse wall 9 (through the calculation of (1), the transverse wall 9 is located farther away from the end point of the longitudinal wall 4 in the transverse wall 8 and the transverse wall 9), and the label corresponding to the lower end point of the longitudinal wall 4 is the vertical wall 5, then the ordinate of the upper end point of the longitudinal wall 4 is the ordinate of the center point of the transverse wall 9, and the ordinate of the lower end point of the longitudinal wall 4 is the ordinate of the center point of the transverse wall 5.

(2) Endpointing for untagged walls

For a wall without a label, for example, in fig. 3, if the vertical wall No. 4 is deleted, and there is no wall that is adjacent to the right end point of the horizontal wall 9 in a different direction, the end point is used as a reference, 0.3m is expanded to the periphery of the horizontal direction to be used as an ROI area, the original point cloud data is captured, a horizontal plane whose normal vector is parallel to the wall direction is extracted, the plane is considered to be a section that limits the wall, the point cloud of the section is projected to the horizontal plane, and a straight line is fitted. And taking the coordinates of the intersection point of the central line of the wall body and the fitted straight line as the coordinates of the end point.

It should be noted that the terms "upper", "lower", "left" and "right" used herein are all directions in the physical connection relation in the embodiments of the present invention.

By the calculation method, the endpoint coordinates of each wall body can be obtained.

According to the end point coordinates of the walls and the connection relationship of the walls, the walls can be connected to obtain the connection relationship diagram of the walls of the target building as shown in fig. 16.

Optionally, the two-dimensional house type graph of the target building is drawn based on the connection relation graph and the thickness of each wall body, and specifically, the two-dimensional house type graph may be: determining each vertex coordinate of each wall body based on the connection relation graph, the central line of each wall body and the thickness of each wall body; and drawing a two-dimensional house type graph of the target building based on the vertex coordinates of the walls and the connection relation graph.

For example, in the obtained connection relation diagram of fig. 16, the center lines of the walls in fig. 14 are connected, but the length of the actual wall is not the length of the center line, and when the length of the center line in fig. 14 is used as the length of the wall for connection, the walls may intersect as shown in fig. 17, that is, two walls may have an intersection. Therefore, after obtaining the connection relationship diagram of each wall, it is necessary to determine the coordinates of 4 vertices of each wall according to the obtained connection relationship diagram of each wall, the center line of each wall, and the thickness of each wall, and connect the center lines according to the obtained coordinates of 4 vertices of each wall, thereby obtaining the two-dimensional house type diagram of the target building.

Specifically, the coordinates of 4 vertices of each wall are determined according to the connection relationship diagram and the thickness of each wall, which may be: the edge of the wall body in one direction, which is opposite to the far side of the adjacent wall body, is selected as the edge of the wall body in the direction, and the edge of the wall body in the other direction, which is opposite to the near side of the adjacent wall body, is selected as the edge of the wall body in the direction. Assuming an extended longitudinal wall, as shown in fig. 18. The z coordinate values of the four vertex angles of the lower surface of each wall body are 0, the xy coordinates of the four vertex angles of the upper surface are the same as those of the lower surface, and the z coordinate is the height of the ceiling. The four corners of the bottom surface were ordered in a counterclockwise sequence from ABCD as shown in fig. 18.

Taking the vertical wall shown in fig. 18 as an example, the abscissa of A, D is determined by the center line of the vertical wall and the width of the wall, and the ordinate is determined by the lower edge of the horizontal wall formed by a ' and D ', because fig. 17 is changed from fig. 17 to fig. 18, a ', B, C ' and D ' are moved to the right, and the vertexes A, D are respectively moved downward on the basis of the center line of the vertical wall by half the wall thickness of the horizontal wall, i.e., the ordinate is moved downward until they are on the same horizontal line as a ' and D ', and therefore, the ordinate of A, D is determined by the lower edge of the horizontal wall formed by a ' and D '.

The same reason B, C may be defined by the upper edges of adjacent anisotropic walls at the upper end of the longitudinal walls. The ordinate of C 'and D' is determined by the center line and the width of the wall body, the abscissa is determined by A, B, and A 'and B' have the same reason. For the wall end point without the adjacent anisotropic wall, the vertex angle coordinates at the two sides of the end point are directly determined by the end point coordinates and the wall width.

Thus, the coordinates of each vertex of each wall can be determined.

After determining the vertex coordinates of each wall, the two-dimensional house type graph of the target building as shown in fig. 19 can be obtained according to the vertex coordinates of each wall and the connection relation graph.

Fig. 19 is similar to fig. 10 in view, but differs from the same. Specifically, the wall contour lines in the two-dimensional floor plan obtained in fig. 19 are smooth, while the wall contour lines in fig. 10 are uneven, specifically referring to the circled portion in fig. 10, the circled portion in fig. 10 is smooth at the corresponding position in fig. 19.

And S460, determining a three-dimensional house type model of the target building based on the two-dimensional house type graph.

For example, after obtaining the two-dimensional house type graph, the coordinates of 8 vertexes of each wall body can be determined according to the endpoint coordinates of each wall body and the height of each wall body, and the three-dimensional house type model of the target building can be drawn according to the vertex coordinates of each wall body and the height of each wall body.

Specifically, the coordinates of the central line end points of the walls of the target building, the height of each wall, the coordinates of 8 vertices of each wall, and the like may be input into the three-dimensional modeling software, so that the three-dimensional house type model of the target building as shown in fig. 20 may be directly obtained.

According to the technical scheme of the embodiment of the invention, the walls which are connected in different directions are determined based on the positions of the walls, and the connection relation of the walls which are connected in different directions is established; determining walls connected in the same direction based on the positions of the walls, establishing a connection relation of the walls connected in the same direction according to the types of the walls and the walls connected in the same direction, and determining end point coordinates of the walls based on the coordinates of the center points of the walls; determining a connection relation graph of each wall of the target building based on the endpoint coordinates of each wall and the connection relation of each wall; and drawing a two-dimensional house type graph of the target building based on the connection relation graph and the thickness of each wall body, so that the accurate two-dimensional house type graph of the target building can be obtained, a three-dimensional house type model can be determined based on the accurate two-dimensional house type graph in the following process, and the effect of accurately and efficiently reconstructing the house type model is realized.

EXAMPLE five

Fig. 21 is a schematic structural diagram of a building apparatus of a house type model according to a fifth embodiment of the present invention, as shown in fig. 21, the apparatus includes: the system comprises a wall extraction module 31, a wall position determination module 32, a two-dimensional map acquisition module 33 and a three-dimensional model determination module 34.

The wall extraction module 31 is configured to perform plane segmentation on three-dimensional point cloud data of a target building, extract each wall in the target building, and identify the type of each wall;

the wall body position determining module 32 is configured to perform horizontal projection on each wall body plane, and determine the position of each wall body in the target building based on each wall body projection;

the two-dimensional graph acquisition module 33 is configured to determine a connection relationship between the walls based on the positions of the walls and the types of the walls, and draw a two-dimensional house type graph of the target building based on the positions of the walls, the connection relationship between the walls, and the thicknesses of the walls;

and the three-dimensional model determining module 34 is used for determining a three-dimensional house type model of the target building based on the two-dimensional map.

On the basis of the technical solution of the above embodiment, the wall extraction module 31 includes:

the normal vector determining unit is used for determining the normal vector of each three-dimensional point cloud data based on the three-dimensional point cloud data of the target building;

the indoor area image determining unit is used for carrying out plane segmentation on the three-dimensional point cloud data of the target building based on the normal vector of each three-dimensional point cloud data and the index of each segmentation unit to determine an indoor area image of the target building;

the target indoor area image determining unit is used for performing superposition alignment on the indoor area images and determining a target indoor area image of the target building;

and the wall body extraction unit is used for determining each wall body of the target building based on the target indoor area image.

On the basis of the technical solution of the above embodiment, the target indoor area image determining unit is specifically configured to:

preprocessing the indoor area image to obtain a preprocessed indoor area image, wherein a connected area caused by noise is removed from the preprocessed indoor area image; aligning and superposing all connected areas in the preprocessed indoor area image to obtain a target indoor area image of the target building; wherein the pre-processing comprises: and performing closing operation and opening operation on the indoor area image.

On the basis of the technical scheme of the embodiment, the wall body extraction unit is specifically used for:

performing closed operation and expansion on the target indoor area image to obtain a wall area image in the target building; and respectively carrying out transverse opening operation and longitudinal opening operation on the wall area images in different sizes and structures to obtain the walls with different thicknesses in the transverse direction and the longitudinal direction of the target building.

Optionally, the types of the wall body include: interior walls and exterior walls.

On the basis of the technical solution of the above embodiment, the wall extraction module 31 further includes:

the connected region acquisition unit is used for performing closed operation on the target indoor region image to obtain each connected region;

the first wall type determining unit is used for determining that the wall is an inner wall if the image coordinates of the center of the wall are in the communication area;

and the second wall type determining unit is used for determining that the wall is an outer wall if the image coordinate of the center of the wall is not in the communication area.

On the basis of the technical scheme of the embodiment, the device further comprises:

the interesting area determining module is used for converting the central line of each wall body into a point cloud coordinate system and performing external expansion on the central line of each wall body for a preset distance to serve as an interesting area;

the three-dimensional point cloud data determining module of the region of interest is used for intercepting the three-dimensional point cloud data of the corresponding region of interest in the three-dimensional point cloud data to form the three-dimensional point cloud data of the region of interest;

the vertical plane extraction module is used for performing normal vector calculation and segmentation on the three-dimensional point cloud data area of the region of interest and extracting a vertical plane parallel to the center line of the wall body;

and the target wall body determining module is used for determining a target wall body based on the extracted vertical plane, the length requirement of the vertical plane wall body, the height requirement of the vertical plane wall body and the type of the wall body.

On the basis of the technical solution of the above embodiment, the wall position determining module 32 includes:

the system comprises a linear equation determining unit, a data processing unit and a data processing unit, wherein the linear equation determining unit is used for horizontally projecting the three-dimensional point cloud data of each wall body and linearly fitting the projected three-dimensional point cloud data to obtain a linear equation of each wall body plane;

the first point cloud coordinate determination unit is used for substituting the point cloud horizontal coordinates of the centers of the wall bodies in the horizontal direction into the linear equation to obtain the point cloud vertical coordinates of the centers of the wall bodies in the horizontal direction;

the second point cloud coordinate determination unit is used for substituting the vertical coordinates of the point cloud of the centers of the walls in the vertical direction into the linear equation to obtain the horizontal coordinates of the point cloud of the centers of the walls in the vertical direction;

a third point cloud coordinate determination unit, configured to determine point cloud coordinates of each wall center point in the horizontal direction and point cloud coordinates of each wall center point in the vertical direction based on the point cloud abscissa and the point cloud ordinate of each wall center in the horizontal direction and the point cloud ordinate and the point cloud abscissa of each wall center in the vertical direction;

and the wall body position determining unit is used for determining the position of each wall body in the target building based on the point cloud coordinates of each wall body central point in the horizontal direction and the point cloud coordinates of each wall body central point in the vertical direction.

On the basis of the technical solution of the above embodiment, the two-dimensional map obtaining module 33 includes:

the first connection relation determining unit is used for determining the walls which are connected in different directions based on the positions of the walls and establishing the connection relation of the walls which are connected in different directions;

and the second connection relation determining unit is used for determining the walls connected in the same direction based on the positions of the walls and establishing the connection relation of the walls connected in the same direction according to the types of the walls and the walls connected in the same direction.

On the basis of the technical solution of the above embodiment, the two-dimensional map obtaining module 33 further includes:

the end point coordinate determination unit is used for determining the end point coordinates of each wall body based on the coordinates of the center point of each wall body;

the connection relation graph determining unit is used for determining the connection relation graph of each wall of the target building based on the endpoint coordinates of each wall and the connection relation of each wall;

and the two-dimensional house type graph determining unit is used for drawing the two-dimensional house type graph of the target building based on the connection relation graph and the thickness of each wall body.

On the basis of the technical solution of the above embodiment, the two-dimensional house type graph determining unit is specifically configured to:

determining each vertex coordinate of each wall body based on the connection relation graph, the central line of each wall body and the thickness of each wall body; and drawing a two-dimensional house type graph of the target building based on the vertex coordinates of each wall and the connection relation graph.

The device for building the house type model provided by the embodiment of the invention can execute the method for building the house type model provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

EXAMPLE six

Fig. 22 is a schematic structural diagram of an electronic apparatus according to a sixth embodiment of the present invention, as shown in fig. 22, the electronic apparatus includes a processor 70, a memory 71, an input device 72, and an output device 73; the number of the processors 70 in the electronic device may be one or more, and one processor 70 is taken as an example in fig. 22; the processor 70, the memory 71, the input device 72, and the output device 73 in the electronic apparatus may be connected by a bus or other means, and the bus connection is exemplified in fig. 22.

The memory 71 is a computer-readable storage medium, and can be used for storing software programs, computer-executable programs, and modules, such as program instructions/modules corresponding to the method for constructing the house type model in the embodiment of the present invention (for example, the wall extracting module 31, the wall position determining module 32, the two-dimensional map obtaining module 33, and the three-dimensional model determining module 34). The processor 70 executes various functional applications and data processing of the electronic device by executing software programs, instructions and modules stored in the memory 71, that is, implements the above-described method for building the house type model.

The memory 71 may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created according to the use of the terminal, and the like. Further, the memory 71 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some examples, the memory 71 may further include memory located remotely from the processor 70, which may be connected to the electronic device through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The input device 72 may be used to receive input numeric or character information and generate key signal inputs related to user settings and function controls of the electronic apparatus. The output device 73 may include a display device such as a display screen.

EXAMPLE seven

An embodiment of the present invention further provides a storage medium containing computer-executable instructions, which when executed by a computer processor, perform a method for constructing a house type model.

Of course, the storage medium provided by the embodiment of the present invention contains computer-executable instructions, and the computer-executable instructions are not limited to the method operations described above, and may also perform related operations in the method for building a house type model provided by any embodiment of the present invention.

From the above description of the embodiments, it is obvious for those skilled in the art that the present invention can be implemented by software and necessary general hardware, and certainly, can also be implemented by hardware, but the former is a better embodiment in many cases. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as a floppy disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a FLASH Memory (FLASH), a hard disk or an optical disk of a computer, and includes instructions for enabling a computer electronic device (which may be a personal computer, a server, or a network electronic device) to execute the methods according to the embodiments of the present invention.

It should be noted that, in the embodiment of the building apparatus of the house type model, the included units and modules are only divided according to the functional logic, but are not limited to the above division as long as the corresponding functions can be realized; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present invention.

It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A method for constructing a house type model is characterized by comprising the following steps:

2. The method of claim 1, wherein the performing planar segmentation on the three-dimensional point cloud data of the target building to extract each wall in the target building comprises:

determining a normal vector of each three-dimensional point cloud data based on the three-dimensional point cloud data of the target building;

performing plane segmentation on the three-dimensional point cloud data of the target building based on the normal vector of each three-dimensional point cloud data and the index of each segmentation unit to determine an indoor area image of the target building;

overlapping and aligning the indoor area images, and determining a target indoor area image of the target building;

and determining each wall of the target building based on the target indoor area image.

3. The method of claim 2, wherein said registering and overlaying said indoor area image to determine a target indoor area image of said target structure comprises:

preprocessing the indoor area image to obtain a preprocessed indoor area image, wherein a connected area caused by noise is removed from the preprocessed indoor area image;

aligning and superposing all connected areas in the preprocessed indoor area image to obtain a target indoor area image of the target building;

wherein the pre-processing comprises: and performing closing operation and opening operation on the indoor area image.

4. The method of claim 3, wherein said determining walls of said target structure based on said target indoor area image comprises:

performing closed operation and expansion on the target indoor area image to obtain a wall area image in the target building;

and respectively carrying out transverse opening operation and longitudinal opening operation on the wall area images in different sizes and structures to obtain the walls with different thicknesses in the transverse direction and the longitudinal direction of the target building.

5. The method of claim 3, wherein the type of wall comprises: interior and exterior walls;

the identifying the type of each wall comprises:

performing closed operation on the target indoor area image to obtain each connected area;

if the image coordinates of the center of the wall body are in the communicated area, determining that the wall body is an inner wall;

and if the image coordinate of the center of the wall body is not in the communication area, determining that the wall body is an outer wall.

6. The method of claim 1 or 5, wherein prior to said determining the location of each wall in the target building, the method further comprises:

converting the center line of each wall body to a point cloud coordinate system, and performing external expansion on the center line of each wall body for a preset distance to serve as an interesting area;

intercepting the three-dimensional point cloud data of the corresponding region of interest in the three-dimensional point cloud data to form three-dimensional point cloud data of the region of interest;

performing normal vector calculation and segmentation on the three-dimensional point cloud data area of the region of interest, and extracting a vertical plane parallel to the central line of the wall body;

and determining the target wall body based on the extracted vertical plane, the length requirement of the vertical plane wall body, the height requirement of the vertical plane wall body and the type of the wall body.

7. The method of claim 1, wherein the horizontally projecting each wall plane, and determining the location of each wall in the target structure based on each wall projection, comprises:

horizontally projecting each wall plane, and linearly fitting the three-dimensional point cloud data corresponding to each projected wall plane to obtain a linear equation of each wall plane;

substituting the horizontal coordinates of the point cloud of each wall center in the horizontal direction into the linear equation to obtain the vertical coordinates of the point cloud of each wall center in the horizontal direction;

substituting the vertical coordinates of the point cloud of the centers of the walls in the vertical direction into the linear equation to obtain the horizontal coordinates of the point cloud of the centers of the walls in the vertical direction;

determining the point cloud coordinates of the wall body center points in the horizontal direction and the point cloud coordinates of the wall body center points in the vertical direction based on the point cloud abscissa and the point cloud ordinate of the wall body center in the horizontal direction and the point cloud ordinate and the point cloud abscissa of the wall body center in the vertical direction;

and determining the position of each wall body in the target building based on the point cloud coordinates of the central points of the wall bodies in the horizontal direction and the point cloud coordinates of the central points of the wall bodies in the vertical direction.

8. The method of claim 6, wherein: the determining the connection relation of the walls based on the positions of the walls and the types of the walls comprises the following steps:

determining the walls connected in different directions based on the positions of the walls, and establishing the connection relation of the walls connected in different directions;

and determining the walls connected in the same direction based on the positions of the walls, and establishing the connection relation of the walls connected in the same direction according to the types of the walls and the walls connected in the same direction.

9. The method of claim 8, wherein the step of drawing the two-dimensional floor plan of the target building based on the positions of the walls, the connection relationship of the walls and the thickness of the walls comprises:

determining the end point coordinates of each wall body based on the coordinates of the central point of each wall body;

determining a connection relation graph of each wall of the target building based on the endpoint coordinates of each wall and the connection relation of each wall;

and drawing a two-dimensional house type graph of the target building based on the connection relation graph and the thickness of each wall body.

10. The method of claim 9, wherein the drawing a two-dimensional floor plan of the target building based on the connection relation map and the thickness of each wall body comprises:

determining each vertex coordinate of each wall body based on the connection relation graph, the central line of each wall body and the thickness of each wall body;

and drawing a two-dimensional house type graph of the target building based on the vertex coordinates of each wall and the connection relation graph.

11. A house pattern reconstruction apparatus, comprising:

12. An electronic device, characterized in that the electronic device comprises:

one or more processors;

storage means for storing one or more programs;

when executed by the one or more processors, cause the one or more processors to implement a method of constructing a house model according to any of claims 1-10.

13. A storage medium containing computer-executable instructions for performing the method of constructing a house model according to any one of claims 1-10 when executed by a computer processor.