A real estate surveying and mapping method based on image analysis

Abstract

The present invention discloses a real estate surveying and mapping method based on image analysis, comprising the following steps: step one: collecting a number of real estate images through an image acquisition device, and marking and screening the number of real estate images to obtain a number of groups of building images; wherein the image acquisition device is a camera or a drone; it relates to the field of real estate surveying and mapping technology, and solves the technical problem that the image cannot be filled after obstacles appear in positions in both left and right images; by extending the building outline, it is avoided that obstacles block the building during the process of real estate image acquisition, resulting in incomplete outline segments and inability to generate a building model, and the accuracy of the building frame diagram obtained after the extension processing is further improved, thereby improving the accuracy of the building model, and then being able to accurately reflect the actual graphic area of ​​the real estate.

Description

Technical Field

[0001] The invention belongs to the technical field of real estate surveying and mapping, and specifically is a real estate surveying and mapping method based on image analysis. Background Art

[0002] Real estate surveying and mapping refers to the activities of measuring, mapping and registering the shape, area and other elements of real estate. It is an important basis and means for real estate registration. With the development of computer vision and image processing technology, real estate surveying and mapping methods based on image analysis have gradually been widely used. These methods analyze the images of buildings or other real estate and use image processing technology and computer vision technology to achieve three-dimensional reconstruction and measurement of buildings or real estate.

[0003] Prior art (patent application with publication number CN116051613A) is a real estate surveying and mapping method based on image analysis. The patent uses linear transformation to make the polar lines of the left image and the right image on the same horizontal line; if there is a pixel point whose matching pixel points obtained in two stereo matchings are different, then the pixel point is an occlusion point, and the occlusion point is filled. However, supplementing it by swapping the left and right images is only applicable to symmetrical buildings, and if obstacles appear in the positions of the left and right images, filling is still impossible, which results in the inability to complete three-dimensional modeling in the subsequent drawing process, causing problems in the real estate drawing process. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a real estate surveying and mapping method based on image analysis, which is used to solve the technical problem that the image cannot be filled after obstacles appear in the positions of the left and right images. The present invention solves the above-mentioned problem by randomly selecting several fitting points on the incomplete line segment, fitting the several fitting points to obtain a fitting equation, and extending the endpoints without intersection in the incomplete line segment according to the fitting equation.

[0005] To achieve the above object, a first aspect of the present invention provides a real estate surveying and mapping method based on image analysis, comprising the following steps:

[0006] Step 1: a plurality of real estate images are collected by an image acquisition device, and the plurality of real estate images are marked and screened to obtain a plurality of groups of building images; wherein the image acquisition device is a camera or a drone;

[0007] Step 2: extract the building outline in the building image, obtain the outline line segments, and obtain the building frame diagram after preprocessing the outline line segments; calculate the building size according to the building panorama and scale parameters through image measurement technology to obtain building size data; wherein the building frame diagram is a closed figure composed of several building lines, the scale parameters include known sizes and proportions, etc., and the building size data is the size of the corresponding building line in the building frame diagram;

[0008] Step 3: After splicing several building frame drawings, a building panorama is obtained; the building size is calculated according to the building panorama and scale parameters through image measurement technology to obtain building size data; the building panorama and building size data are integrated to generate a building model, and stored in the model library;

[0009] Step 4: Integrate the building model with the geographic information system (GIS) to obtain the corresponding real estate survey map.

[0010] Step 1 of this embodiment performs label screening on a plurality of real estate images, including:

[0011] The location where the real estate image is collected is obtained, the location information of the corresponding building is obtained, the shooting direction of the real estate image to the building is identified, the real estate image is classified based on the shooting direction, and several real estate images corresponding to the building are screened according to the location information and the shooting direction, wherein the shooting directions include top view, front view, rear view, left view and right view.

[0012] For example: through the GPS module carried by the drone, the location where each real estate image was collected is obtained, the corresponding longitude and latitude coordinates are obtained, the center points of the locations of multiple real estate images are marked as location information, and the shooting direction of the real estate image to the building is identified based on the location information, and several real estate images corresponding to one location information are marked as a group of building images.

[0013] The real estate images of the same building are grouped into a group of building image data to facilitate the subsequent splicing of the building frame drawings.

[0014] The building frame diagram is obtained after preprocessing the contour line segments in step 2 of this embodiment, including:

[0015] S1: extracting a number of contour line segments corresponding to the real estate image and marking them as contour line segments L1, L2, ..., Ln, respectively; wherein n is a positive integer;

[0016] S2: Determine whether the two endpoints of the contour segment Li are both intersection points; where i∈[1,n];

[0017] If yes, the contour line segment Li is marked as a building line; if no, the contour line segment Li is marked as an incomplete line segment, and the incomplete line segment is extended to obtain the building line;

[0018] S3: Extract the closed figures formed by several building lines and mark them as building frame diagrams.

[0019] Preferably, the step of extending the incomplete line segments to obtain the building lines comprises:

[0020] Select several fitting points on the incomplete line segment, fit the fitting points to obtain a fitting equation, extend the endpoints without intersections in the incomplete line segment according to the fitting equation to obtain an extended line segment, repeat S2 to obtain several extended line segments, extract the line segment between the two intersections closest to the incomplete line segment in the extended line segment, and mark it as a building line.

[0021] By extending the incomplete line segments in the form of fitting equations, it is possible to adapt to the surface of the building and avoid situations where curved buildings are extended into straight lines, which would cause serious discrepancies between the building frame and the real estate.

[0022] Preferably, the step of splicing a plurality of building frame drawings to obtain a building panoramic view includes:

[0023] A number of building frame drawings corresponding to the building image data are extracted, and the building frame drawings are marked according to the shooting direction. Two building frame drawings of adjacent views are connected to each other, and common building lines are extracted from the building frame drawings of adjacent views for integration to obtain a building panorama, wherein the connection state can be determined according to the appearance of the building.

[0024] Preferably, the step of generating a building model by integrating the building panorama and the building dimension data comprises:

[0025] The building panorama is extracted, the building size data is matched with the building lines in the building panorama, the building panorama is scaled according to a preset ratio, and a three-dimensional geometric shape is generated based on the closed figure formed by the building panorama to obtain a building model.

[0026] Preferably, the integration of the building model with a geographic information system (GIS) comprises:

[0027] According to the location information of the corresponding building and matching it with the location in the geographic information system, the corresponding marked location is obtained, and the building model is imported into the marked location of the map data in the GIS.

[0028] Preferably, the step 2 further comprises: determining whether the building block diagram matches the building block diagram corresponding to the building model in the model library, if not, jumping to step 3, if yes, adjusting the proportion of the building model according to the building size data and then jumping to step 4;

[0029] The determining whether the building block diagram matches the building block diagram corresponding to the building model in the model library includes:

[0030] Extract several corresponding building block diagrams in the group of building images and several corresponding building block diagrams in the building model, and mark them as GTi and GKi respectively according to the view, and judge whether GTi and GKi match according to the image matching technology. If yes, the building block diagram matches the building block diagram corresponding to the building model in the model library, otherwise, the building block diagram does not match the building block diagram corresponding to the building model in the model library; where i = 1, 2, ..., n.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1. In the present invention, in the process of marking and screening the real estate images, the real estate images are classified, and the position of each building and several real estate images are matched to obtain the building image corresponding to the building. This method is convenient for subsequent splicing, and by extending the building outline, it is avoided that obstacles block the building during the real estate image acquisition process, resulting in incomplete outline segments and unable to generate a building model. The accuracy of the building frame diagram obtained after the extension process is further improved, thereby improving the accuracy of the building model, and then being able to accurately reflect the actual graphic area of ​​the real estate.

[0033] 2. In the present invention, by storing the completed building model in the model library, when generating the corresponding building model for the subsequent building, it can be determined whether several building block diagrams of the building model in the model library match the building block diagram of the building. If they match, the corresponding building model can be adjusted according to the building size data, and the adjusted building model can be directly used, thereby avoiding a large number of subsequent repeated modeling processes and further improving the efficiency of drawing. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0035] Figure 1 This is a schematic diagram of the process of Example 1 of the present invention;

[0036] Figure 2 This is a schematic diagram of the process of Example 2 of the present invention. DETAILED DESCRIPTION

[0037] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] See also Figure 1-Figure 2 The first aspect of the present invention provides a real estate surveying and mapping method based on image analysis, comprising the following steps:

[0039] Step 1: Collect a number of real estate images by using a drone, and mark and screen the real estate images to obtain a number of groups of building images;

[0040] Step 2: Extract the building outline in the building image through computer vision technology to obtain the outline line segments, and obtain the building frame diagram after preprocessing the outline line segments; calculate the building size according to the building panorama and scale parameters through image measurement technology to obtain the building size data; wherein the building frame diagram is a closed figure composed of several building lines, the scale parameters include known dimensions and proportions, etc., and the building size data is the size of the corresponding building line in the building frame diagram, that is, the length, width, height and other dimensions;

[0041] Step 3: After splicing several building frame drawings, a building panorama is obtained; the building size is calculated according to the building panorama and scale parameters through image measurement technology to obtain building size data; the building panorama and building size data are integrated to generate a building model, and stored in the model library;

[0042] Step 4: Integrate the building model with the geographic information system (GIS) to obtain the corresponding real estate survey map.

[0043] Step 1 of this embodiment performs label screening on a plurality of real estate images, including:

[0044] The location where the real estate image is collected is obtained, the location information of the corresponding building is obtained, the shooting direction of the real estate image to the building is identified, the real estate image is classified based on the shooting direction, and several real estate images corresponding to the building are screened according to the location information and the shooting direction, wherein the shooting directions include top view, front view, rear view, left view and right view.

[0045] For example: through the GPS module carried by the drone, the location where each real estate image was collected is obtained, the corresponding longitude and latitude coordinates are obtained, the center points of the locations of multiple real estate images are marked as location information, and the shooting direction of the real estate image to the building is identified based on the location information, and several real estate images corresponding to one location information are marked as a group of building images.

[0046] The real estate images of the same building are grouped into a group of building image data to facilitate the subsequent splicing of the building frame drawings.

[0047] In step 2 of this embodiment, extracting the building outline in the building image to obtain the outline segment;

[0048] The Canny function provided by OpenCV can be used to perform edge detection on each building image to obtain edge pixels in the image. Then, the Hough transform can be performed on the edge pixels in each building image according to the HoughLinesP function provided by OpenCV to obtain contour segments in the image.

[0049] Furthermore, the building frame diagram is obtained after preprocessing the contour line segments, including:

[0050] S1: extracting a number of contour line segments corresponding to the real estate image and marking them as contour line segments L1, L2, ..., Ln, respectively; wherein n is a positive integer;

[0051] S2: Determine whether the two endpoints of the contour segment Li are both intersection points, that is, they intersect with other contour segments Lj; where i, j∈[1, n], and i≠j;

[0052] If yes, the contour line segment Li is marked as a building line; if no, the contour line segment Li is marked as an incomplete line segment, and the incomplete line segment is extended to obtain the building line;

[0053] S3: Extract the closed figures formed by several building lines and mark them as building frame diagrams.

[0054] Specifically, in S2, the incomplete line segments are extended to obtain the building lines, including:

[0055] Select any 6 fitting points on the incomplete line segment, fit the 6 fitting points to obtain the fitting equation, extend the endpoints without intersection in the incomplete line segment according to the fitting equation to obtain the extended line segment, repeat S2 to obtain several extended line segments, extract the line segment between the two intersection points closest to the incomplete line segment in the extended line segment, and mark it as the building line.

[0056] By extending the incomplete line segments in the form of fitting equations, it is possible to adapt to the surface of the building and avoid situations where curved buildings are extended into straight lines, which would cause serious discrepancies between the building frame and the real estate.

[0057] In step 3 of this implementation, a panoramic view of a building is obtained by splicing several building frame drawings, including:

[0058] A number of building frame drawings corresponding to the building image data are extracted, and the building frame drawings are marked according to the shooting direction. Two building frame drawings of adjacent views are placed in a perpendicular state to each other, and common building lines are extracted from the building frame drawings of adjacent views and integrated to obtain a building panorama.

[0059] Furthermore, the building panorama and building dimension data are integrated to generate a building model, including:

[0060] The building panorama is extracted, the building size data is matched with the building lines in the building panorama, the building panorama is scaled according to a preset ratio, and a three-dimensional geometric shape is generated based on the closed figure formed by the building panorama to obtain a building model.

[0061] In step 4 of this embodiment, the building model is integrated with the geographic information system GIS, including:

[0062] Through the function geocode provided by ArcGIS, the location information of the corresponding building is matched with the location in the geographic information system to obtain the corresponding marked location, and the building model is imported into the marked location of the map data in GIS.

[0063] The second aspect of the present invention is different from the above embodiment in that step 2 further includes: determining whether the building block diagram matches the building block diagram corresponding to the building model in the model library, if not, jumping to step 3, if yes, adjusting the proportion of the building model according to the building size data and then jumping to step 4;

[0064] Specifically, step 2: extract the building outline in the building image by computer vision technology to obtain the outline line segments, and obtain the building frame diagram after preprocessing the outline line segments; calculate the building size according to the building panorama and scale parameters by image measurement technology to obtain the building size data; wherein the building frame diagram is a closed figure composed of several building lines, the scale parameters include known dimensions and proportions, etc., and the building size data is the size of the corresponding building line in the building frame diagram, that is, the length, width, height and other dimensions;

[0065] Determine whether the building block diagram matches the building block diagram corresponding to the building model in the model library. If not, jump to step three. If yes, adjust the proportion of the building model according to the building size data and then jump to step four.

[0066] Further, judging whether the building block diagram matches the building block diagram corresponding to the building model in the model library includes:

[0067] Extract the 5 corresponding building blocks in the group of building images and the 5 corresponding building blocks in the building model, and mark them as GTi and GKi according to the views, and determine whether GTi and GKi match according to the image matching technology; where i = 1, 2, ..., 5, and the views corresponding to GT1-GT5 and GK1-GK5 are front view, rear view, left view, right view and top view respectively

[0068] If all GTi and GKi match, it is considered that the building block diagram matches the building block diagram corresponding to the building model in the model library, and jump to step 4. If any GTi and GKi do not match, it is considered that the building block diagram does not match the building block diagram corresponding to the building model in the model library, and jump to step 3.

[0069] In this way, architectural models of multiple buildings with the same house type in a community can be quickly obtained to reduce the amount of calculation.

[0070] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A real estate surveying and mapping method based on image analysis, characterized in that: The following steps are involved: Step 1: a plurality of real estate images are collected by an image acquisition device, and the plurality of real estate images are marked and screened to obtain a plurality of groups of building images; wherein the image acquisition device is a camera or a drone; Step 2: Extract the building outline in the building image through computer vision technology to obtain the outline line segments, and obtain the building frame diagram after preprocessing the outline line segments; calculate the building size according to the building panorama and scale parameters through image measurement technology to obtain the building size data; wherein the building frame diagram is a closed figure composed of several building lines, and the building size data is the size of the corresponding building line in the building frame diagram; Step 3: After splicing several building frame drawings, a building panorama is obtained; the building panorama and building dimension data are integrated to generate a building model, and the model is stored in the model library; Step 4: Integrate the building model with the Geographic Information System (GIS) to obtain the corresponding real estate survey map; The step of obtaining a building frame diagram after preprocessing the contour line segments includes: S1: extracting a number of contour line segments corresponding to the real estate image and marking them as contour line segments L1, L2, ..., Ln, respectively; wherein n is a positive integer; S2: Determine whether the two endpoints of the contour segment Li are both intersection points; where i∈[1,n]; If yes, the contour line segment Li is marked as a building line; if no, the contour line segment Li is marked as an incomplete line segment, and the incomplete line segment is extended to obtain the building line; S3: extract the closed figures formed by several building lines and mark them as building frame diagrams; The step of extending the incomplete line segments to obtain the building lines comprises: Select several fitting points on the incomplete line segment, fit the fitting points to obtain a fitting equation, extend the endpoints without intersections in the incomplete line segment according to the fitting equation to obtain an extended line segment, repeat S2 to obtain several extended line segments, extract the line segment between the two intersections closest to the incomplete line segment in the extended line segment, and mark it as a building line.

2. The real estate surveying and mapping method based on image analysis according to claim 1, characterized in that: The tagging and screening of the plurality of real estate images includes: The location where the real estate image is collected is obtained, the location information of the corresponding building is obtained, the shooting direction of the real estate image to the building is identified, the real estate image is classified based on the shooting direction, and several real estate images corresponding to the building are screened according to the location information and the shooting direction.

3. The real estate surveying and mapping method based on image analysis according to claim 1, characterized in that: The building panoramic view is obtained by splicing several building frame drawings, including: A number of building frame drawings corresponding to the building image data are extracted, and the building frame drawings are marked according to the shooting direction. Two building frame drawings of adjacent views are connected to each other, and common building lines are extracted from the building frame drawings of adjacent views and integrated to obtain a building panorama.

4. The real estate surveying and mapping method based on image analysis according to claim 1, characterized in that: The step of generating a building model by integrating the building panorama and the building dimension data includes: The building panorama is extracted, the building size data is matched with the building lines in the building panorama, the building panorama is scaled according to a preset ratio, and a three-dimensional geometric shape is generated based on the closed figure formed by the building panorama to obtain a building model.

5. The real estate surveying and mapping method based on image analysis according to claim 1, characterized in that: The integration of the building model with the Geographic Information System (GIS) includes: According to the location information of the corresponding building and matching it with the location in the geographic information system, the corresponding marked location is obtained, and the building model is imported into the marked location of the map data in the GIS.

6. The real estate surveying and mapping method based on image analysis according to claim 1, characterized in that: The step 2 also includes: determining whether the building block diagram matches the building block diagram corresponding to the building model in the model library, if not, jumping to step 3, if yes, adjusting the proportion of the building model according to the building size data and then jumping to step 4.

Images