Building window effective area calculation method and system based on BIM model

Abstract

The invention relates to a building window effective area calculation method and system based on a BIM model, and relates to the technical field of building windows, and the method comprises the steps: obtaining a plurality of images of a current room in a building; the room comprises a plurality of building windows; selecting one of a plurality of images of the same building window shot from different angles to judge whether the type of the building window can be identified or not; based on a plurality of images of the same building window shot at different angles and the recognized building window type, extracting the image of the building window to obtain a building window image; recognizing geometric information of the building window based on the building window image; importing the identified window type and geometric information into a BIM model to obtain a corresponding window component; based on the type of the building window and the parameters of the window components, calling a corresponding window effective opening area algorithm to calculate to obtain the effective area of the current building window; and calculating the effective area of each building window of the current room, and summarizing the sum of the effective areas of the building windows of the current room to obtain the total effective area of the building windows of the current room.

Description

Technical Field

[0001] The present application relates to the technical field of BIM models, and in particular to a method and system for calculating the effective area of ​​building windows based on the BIM model. Background Art

[0002] With the widespread application of Building Information Modeling (BIM) technology, the design, construction and management efficiency of the construction industry has been significantly improved. BIM can not only provide detailed three-dimensional geometric information, but also integrate the attribute data of building components, thereby realizing the management and optimization of the entire life cycle of the building.

[0003] Traditional methods for calculating the effective area of ​​building windows usually rely on manual measurement and recording, which is not only time-consuming and labor-intensive, but also prone to human errors. In addition, manual measurement is even more difficult for building windows with complex structures or difficult to directly access. In recent years, with the development of image processing technology and machine learning, methods based on image recognition have gradually become a feasible solution. These methods can calculate the effective opening area of ​​building windows by taking photos of buildings and then using computer vision technology to identify and analyze the type and size of building windows.

[0004] Although existing image recognition-based technologies have improved the automation of building window area calculation to a certain extent, they still have some limitations. For example, images from a single angle may not fully reflect the actual status of building windows, which in turn leads to errors in building window area calculation. Summary of the invention

[0005] In order to at least partially solve the above technical problems, the present application provides a method and system for calculating the effective area of ​​building windows based on a BIM model.

[0006] In a first aspect, the present application provides a method for calculating the effective area of ​​building windows based on a BIM model, which adopts the following technical solution.

[0007] A method for calculating the effective area of ​​building windows based on a BIM model, characterized by comprising:

[0008] S101, obtaining a plurality of images of a current room in a building; the room includes a plurality of building windows;

[0009] S102, taking multiple images of the same building window from different angles, selecting one of them to determine whether the type of the building window can be identified; if so, executing S103; if not, selecting another image of the same building window to identify the type of the building window until the type of the building window is identified; the types of building windows include: casement windows, top-hung windows, center-hung windows, bottom-hung windows, horizontally-sliding windows, and side-sliding windows;

[0010] S103, extracting images of the building windows based on multiple images of the same building window taken at different angles and the identified building window types to obtain building window images;

[0011] S104, recognizing geometric information of the building window based on the building window image; the geometric information includes: window frame size, window sash opening direction and window sash opening angle;

[0012] S105, importing the identified window type and geometric information into the BIM model to obtain corresponding window components;

[0013] S106, based on the type of the building window and the parameters of the window component, calling the corresponding window effective opening area algorithm to calculate and obtain the effective area of ​​the current building window;

[0014] S107: Calculate the effective area of ​​each building window in the current room, summarize the total effective area of ​​the building windows in the current room, and obtain the total effective area of ​​the building windows in the current room.

[0015] Optionally, based on the type of the building window and the parameters of the window component, a corresponding window effective opening area algorithm is called to calculate the effective area of ​​the current building window, including:

[0016] The calculating of the effective opening area of ​​the window by calling a corresponding algorithm based on the type and the parameters of the window component includes:

[0017] If the type is a casement window, a top-hung window, a center-hung window or a bottom-hung window, it is determined whether the opening angle is greater than 70 degrees. If so, the effective opening area of ​​the window is equal to the area of ​​the window sash IfcSash; otherwise, the calculation formula is: effective opening area of ​​the window = area of ​​the window sash IfcSash * sinα, where α represents the opening angle;

[0018] If the type is a sliding window, the installation position of the sliding window is obtained. If the installation position is the roof, the calculation formula is: effective window opening area = min ((perimeter of the window sash IfcSash*0.5)*sliding distance, area of ​​the window sash IfcSash);

[0019] If the installation location is an exterior wall, the calculation formula is: Window effective opening area = min ((perimeter of window sash IfcSash*0.25)*sliding distance, area of ​​window sash IfcSash);

[0020] If the type is a side-sliding window, determine whether the side-sliding window has an adjacent side-sliding window. If so, the effective opening area of ​​the window is equal to the area of ​​the window sash IfcSash*0.5; otherwise, the effective opening area of ​​the window is equal to the area of ​​the window sash IfcSash.

[0021] Optionally, images of the same building window taken from multiple different angles are selected to determine whether the type of the building window can be identified, including:

[0022] Randomly select an image from the collected building window images taken from different angles;

[0023] Use a pre-trained machine learning model to identify building window types in selected images;

[0024] If the result output by the machine learning model shows that the building window type is successfully identified and the confidence exceeds the predetermined threshold, the type of the building window is obtained; wherein the confidence calculation formula is: C = ω 1 ×P max +ω 2 ×(1-H(P))+ω 3 ×(1-σ 2 ), where C is the confidence level; ω 1 ,ω 2 and 3 is the weight coefficient; P max is the probability of the most likely category output by the machine learning model; H(P) is the entropy of the probability distribution; σ 2 is the standard deviation of multiple prediction results;

[0025] If the recognition fails or the confidence level is lower than a predetermined threshold, another image is used.

[0026] Optionally, extracting an image of the building window based on multiple images of the same building window taken at different angles and the identified building window type to obtain the building window image includes:

[0027] Calculate the clarity score and angle score of each image of the same building window taken at different angles; wherein the angle score is determined based on the identified building window type, specifically including: determining the observation angle range according to the obtained building window type; for each image, calculating the position deviation between the center position of the building window in the image and the center of the image and the inclination angle of the building window in the image; and obtaining the angle score based on the matching degree of the position deviation and the inclination angle with the observation angle range; the angle score calculation formula is: Among them, S a Score the angle; the matching observation range is: [A min , A max ]; A is the actual observed angle; D is the center position deviation, T is the tilt angle; D max is the maximum position deviation allowed; T max is the maximum tilt angle allowed;

[0028] Selecting the image with the highest total score as the main image based on the clarity score and the angle score;

[0029] Extract feature points from images of the same building window taken from different angles;

[0030] Use feature matching algorithm to find the correspondence between feature points of the main image and other images;

[0031] Calculate the basic matrix and essential matrix based on the correspondence between feature points;

[0032] Register the main image using the fundamental matrix and the intrinsic matrix;

[0033] The image of the building window is extracted based on the registered main image to obtain the building window image.

[0034] Optionally, a feature matching algorithm is used to find the correspondence between feature points of the main image and other images, including:

[0035] Extract feature points from the main image and other images;

[0036] Describe the local area around each feature point to obtain the descriptor of the feature point;

[0037] Calculate the Euclidean distance between the main image feature point descriptor and all other image feature point descriptors;

[0038] Select the K closest feature points as matching points.

[0039] Optionally, the basic matrix and the essential matrix are calculated based on the correspondence between the feature points, including:

[0040] Filter out the internal point set from the matching feature point pairs;

[0041] Calculate the basic matrix based on the matching point pairs in the internal point set;

[0042] Get the camera intrinsic parameter matrix of the main image and other images;

[0043] Calculate the intrinsic matrix based on the camera intrinsic parameter matrix.

[0044] Optionally, filter out the inliers from the matched feature point pairs, including:

[0045] S601: Initialize the maximum number of iterations, threshold distance and minimum inlier ratio;

[0046] S602, randomly selecting a minimum number of point pairs from the matched feature point pairs;

[0047] S603, calculating a preliminary basic matrix F based on the selected point pairs;

[0048] S604, for all matched feature point pairs, calculating the reprojection error of each point pair under the basic matrix F;

[0049] S605, marking the point pairs whose reprojection errors are less than a preset threshold as inliers;

[0050] S606, recording the number of inliers and the corresponding inlier set in the current iteration;

[0051] S607, determining whether the number of inliers in the current iteration exceeds the maximum number of inliers recorded previously, and if so, updating the maximum number of inliers and the optimal inlier set;

[0052] S608, determining whether the maximum number of iterations has been reached, if not, returning to step S602 to continue iterating;

[0053] S609: If the maximum number of iterations is reached, output the final inlier point set as the screened inlier point set.

[0054] In the second aspect, the present application provides a building window effective area calculation system based on the BIM model, which adopts the following technical solution.

[0055] A building window effective area calculation system based on a BIM model, comprising:

[0056] A first processing module acquires a plurality of images of a current room in a building; the room includes a plurality of building windows;

[0057] The second processing module is used to: take multiple images of the same building window from different angles and select one of them to determine whether the type of the building window can be identified; the types of building windows include: casement windows, top-hung windows, center-hung windows, bottom-hung windows, horizontally-sliding windows and side-sliding windows;

[0058] The third processing module extracts the image of the building window based on multiple images of the same building window taken at different angles and the identified building window type to obtain a building window image;

[0059] A fourth processing module is used to: identify geometric information of the building window based on the building window image; the geometric information includes: window frame size, window sash opening direction and window sash opening angle;

[0060] A fifth processing module is used to: import the identified window type and geometric information into the BIM model to obtain a corresponding window component;

[0061] A sixth processing module is used to: call a corresponding window effective opening area algorithm based on the type of the building window and the parameters of the window component to calculate and obtain the effective area of ​​the current building window;

[0062] The seventh processing module is used to calculate the effective area of ​​each building window in the current room, summarize the sum of the effective areas of the building windows in the current room, and obtain the total effective area of ​​the building windows in the current room.

[0063] In a third aspect, the present application discloses an electronic device, comprising a memory and a processor, wherein the memory stores a computer program that is loaded by the processor and executes any of the above methods.

[0064] In a fourth aspect, the present application discloses a computer-readable storage medium storing a computer program that can be loaded by a processor and execute any of the above methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the embodiments of this specification 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 recorded in the embodiments of this specification. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0066] Figure 1 This is a flow chart of a method for calculating the effective area of ​​a building window based on a BIM model according to an embodiment of the present application;

[0067] Figure 2 This is a system block diagram of a building window effective area calculation system based on a BIM model in an embodiment of the present application;

[0068] In the figure, 201 is a first processing module; 202 is a second processing module; 203 is a third processing module; 204 is a fourth processing module; 205 is a fifth processing module; 206 is a sixth processing module; and 207 is a seventh processing module. DETAILED DESCRIPTION

[0069] The following is combined with Figure 1-2 The present application is further described with specific embodiments:

[0070] The present application embodiment discloses a method for calculating the effective area of ​​a building window based on a BIM model, comprising the following steps:

[0071] S101, obtaining a plurality of images of a current room in a building; the room includes a plurality of building windows;

[0072] S102, taking multiple images of the same building window from different angles, selecting one of them to determine whether the type of the building window can be identified; if so, executing S103; if not, selecting another image of the same building window to identify the type of the building window until the type of the building window is identified; the types of building windows include: casement windows, top-hung windows, center-hung windows, bottom-hung windows, horizontally-sliding windows, and side-sliding windows;

[0073] S103, extracting images of the building windows based on multiple images of the same building window taken at different angles and the identified building window types to obtain building window images;

[0074] S104, identifying geometric information of the building window based on the building window image; the geometric information includes: window frame size, window sash opening direction and window sash opening angle.

[0075] Specifically, the edge detection algorithm can be used to extract the edge contours of the window frame, window sash and other parts; for the building window structure with obvious corner points, the corner point detection algorithm is used to determine whether the pixel point is a corner point by calculating the grayscale changes of the pixel points in the image in various directions and the autocorrelation matrix, so as to identify the corner points of the window frame and the corner points where the window sash is opened. Determine the size of the window frame: According to the extracted window frame edge information, combined with the corresponding proportional relationship between the pre-calibrated image pixel size and the actual size, the number of pixels along the horizontal and vertical directions of the window frame edge is counted, and then converted into actual length units according to the proportion, so as to obtain the length and width of the window frame. Determine the opening direction of the window sash: Analyze the relative position and edge changes between the window sash and the window frame. If the edge of the window sash has obvious horizontal translation relative to the edge of the window frame and moves to one side, it can be determined as the opening direction of the side-sliding window; if the edge of the window sash has a rotation change around a certain axis point relative to the edge of the window frame, combined with the detected corner point position information, the rotation axis position and the rotation trend of the window sash are analyzed to determine whether it is a top-hung window, a middle-hung window or a bottom-hung window. Calculation of window sash opening angle: For window sashes that can be rotated and opened, the opening angle value of the window sash relative to the window frame is calculated based on the detected change in the angle of the window sash edge before and after opening relative to the window frame reference line (such as the horizontal or vertical edge of the window frame as a reference), and based on the acquired edge point coordinates, corner point coordinates and other information. In some embodiments, the existing prediction technology based on machine learning can also be used to obtain geometric information in the building window image, and the present invention does not specifically limit this.

[0076] S105, importing the identified window type and geometric information into the BIM model to obtain corresponding window components;

[0077] S106, based on the type of the building window and the parameters of the window component, calling the corresponding window effective opening area algorithm to calculate and obtain the effective area of ​​the current building window;

[0078] S107: Calculate the effective area of ​​each building window in the current room, summarize the total effective area of ​​the building windows in the current room, and obtain the total effective area of ​​the building windows in the current room.

[0079] Specifically, BIM model (Building Information Modeling) is an architectural design tool based on three-dimensional digital technology. It allows architects, engineers and project team members to create and manage building data in a virtual environment, thereby improving design efficiency. Building windows refer to openings on buildings for lighting, ventilation or providing views, usually consisting of window frames and openable sashes. According to the different opening methods, they can be divided into casement windows, top-hung windows, center-hung windows, bottom-hung windows, horizontal push windows and side-sliding windows. The sash of the casement window is connected to one side of the window frame by a hinge and can be rotated outward or inward to open. The opening method of the top-hung window is that the sash flips up from the bottom and the top is fixed to the window frame. The sash of the center-hung window opens from the middle position, and the upper and lower parts can rotate in different directions respectively. It is suitable for occasions where ventilation and light entry need to be controlled at the same time. The bottom-hung window is opposite to the top-hung window. The opening method of the building window is to flip down from the top and the bottom is fixed. It is often used in scenes that need to be rainproof and keep indoor ventilation. The horizontal push window sash can slide left and right or up and down, which is suitable for places with limited space. Side-pull windows are also called sliding windows. The window sashes slide in a direction perpendicular to the wall. Usually there are two or more window sashes that can move relative to each other to open the building window. The effective opening area of ​​the window refers to the effective area of ​​the building window that can actually be used for ventilation. This area depends on the opening method and opening degree of the building window.

[0080] Specifically, several images of the current room in the building are obtained, and one or more images of the same building window taken from multiple different angles are selected for type recognition. After successfully identifying the type of the building window, the building window image is extracted and optimized based on the multi-view images and known types, and the optimized building window image is used to identify key geometric parameters. After the identified window type and geometric information are imported into the BIM model, each building window can be represented in a digital form. Finally, according to the specific type of the building window and its parametric description in BIM, the corresponding algorithm is called to calculate its effective opening area, and the total value is obtained by summarizing the effective areas of all building windows in the entire room. Compared with the prior art, which directly performs image recognition to obtain parameter information of the building window, the present application improves the accuracy of geometric parameters obtained by building window image recognition and the accuracy of effective area calculation of the building window by first identifying the type of the building window, selecting the main image according to the type of the building window, and performing image registration.

[0081] In other embodiments, the building image is taken by a drone, and the method further includes:

[0082] Obtaining geographic location data of all buildings whose window areas are to be calculated;

[0083] Select the building closest to the drone’s starting point as the first stop;

[0084] Starting from the current building, determine whether there are other adjacent buildings that have not been visited. If so, add the building to the current inspection path and mark it as visited. Then set this building as the new starting point and repeat the above steps. If not, complete the current inspection path and record the path information.

[0085] Increase the number of path numbers M, select the unvisited building closest to the initial position as the starting point of the new path, and return to the previous step to continue the execution; continue the cycle until all buildings are included in the inspection path;

[0086] Obtain the total inspection time required for inspection based on the actual area and path information of each building;

[0087] Calculate the number of drones required based on the estimated total inspection time and the flight time of a single drone;

[0088] Upload the determined inspection route and related building information to the drone control system;

[0089] The control system instructs the drone to take off based on the received data and start the inspection mission along the predetermined path; during the inspection process, the drone takes images of the building, and when it identifies the building windows, it takes multiple images from different angles.

[0090] Specifically, when urban renewal is carried out, it is necessary to calculate the effective area of ​​building windows for large-scale buildings. In this application, the geographic location data of all buildings whose building window areas are to be calculated are obtained, and a building closest to the starting point of the drone is selected as the first stop; starting from the current building, it is determined whether there are other adjacent buildings that have not been visited; if so, the building is added to the current inspection path and marked as visited, and then it is set as the new starting point to continue the same steps; if not, the configuration of the current inspection path is completed and the relevant information is recorded. Increase the path number M and select the unvisited building closest to the initial position as the starting point of the new path, and repeat the above process until all buildings are included in the inspection path. The required number of drones is calculated based on the actual area and path information of each building. After that, the planned inspection path and related building information are uploaded to the drone control system, and the system instructs the drone to take off according to the received data, and starts the inspection task according to the predetermined path. During the inspection process, when the drone recognizes the building window, it will take multiple images at different angles, which improves the accuracy of window type and geometric information recognition and improves work efficiency.

[0091] As a specific implementation method of a method for calculating the effective area of ​​a building window based on a BIM model, a corresponding window effective opening area algorithm is called based on the type of the building window and the parameters of the window component to calculate the effective area of ​​the current building window, including:

[0092] The calculating of the effective opening area of ​​the window by calling a corresponding algorithm based on the type and the parameters of the window component includes:

[0093] If the type is a casement window, a top-hung window, a center-hung window or a bottom-hung window, it is determined whether the opening angle is greater than 70 degrees. If so, the effective opening area of ​​the window is equal to the area of ​​the window sash IfcSash; otherwise, the calculation formula is: effective opening area of ​​the window = area of ​​the window sash IfcSash * sinα, where α represents the opening angle;

[0094] If the type is a sliding window, the installation position of the sliding window is obtained. If the installation position is the roof, the calculation formula is: Window effective opening area = min ((the circumference of the window sash IfcSash * 0.5 * Push distance, area of ​​sash IfcSash);

[0095] If the installation location is an exterior wall, the calculation formula is: Window effective opening area = min ((perimeter of window sash IfcSash*0.25)*sliding distance, area of ​​window sash IfcSash);

[0096] If the type is a side-sliding window, determine whether the side-sliding window has an adjacent side-sliding window. If so, the effective opening area of ​​the window is equal to the area of ​​the window sash IfcSash*0.5; otherwise, the effective opening area of ​​the window is equal to the area of ​​the window sash IfcSash.

[0097] As a specific implementation of a method for calculating the effective area of ​​a building window based on a BIM model, images of the same building window are taken from multiple different angles, and one of the images is selected to determine whether the type of the building window can be identified, including:

[0098] Randomly select an image from the collected building window images taken from different angles;

[0099] Use a pre-trained machine learning model to identify building window types in selected images;

[0100] If the result output by the machine learning model shows that the building window type is successfully identified and the confidence exceeds the predetermined threshold, the type of the building window is obtained; wherein the confidence calculation formula is: C = ω 1 ×P max +ω 2 ×(1-H(P))+ω 3 ×(1-σ 2 ), where C is the confidence level; ω 1 ,ω 2 and 3 is the weight coefficient; P max is the probability of the most likely category output by the machine learning model; H(P) is the entropy of the probability distribution; σ 2 is the standard deviation of multiple prediction results;

[0101] If the recognition fails or the confidence level is lower than a predetermined threshold, another image is used.

[0102] Specifically, a single image is randomly selected from multiple images of building windows taken at different angles, and its type is identified using a pre-trained machine learning model, combined with a confidence assessment mechanism to ensure the reliability of the recognition result; if the probability of the most likely category output by the model is high, the entropy of the probability distribution is low, and the standard deviation of multiple prediction results is small, the calculated confidence C will be high, indicating that the recognition is not only accurate but also stable; once the confidence exceeds the predetermined threshold, the building window type is considered to be successfully identified. If the requirement is not met, another image is replaced and tried again until an image with a sufficiently high confidence is found. The success rate of window type recognition is improved by comprehensively considering indicators such as the maximum probability, entropy, and standard deviation predicted by the model.

[0103] As a specific implementation of a method for calculating the effective area of ​​a building window based on a BIM model, the image of the building window is extracted based on multiple images of the same building window taken at different angles and the identified building window type to obtain the building window image, including:

[0104] Calculate the clarity score and angle score of each image of the same building window taken at different angles; wherein the angle score is determined based on the identified building window type, specifically including: determining the observation angle range according to the obtained building window type; for each image, calculating the position deviation between the center position of the building window in the image and the center of the image and the inclination angle of the building window in the image; and obtaining the angle score based on the matching degree of the position deviation and the inclination angle with the observation angle range; the angle score calculation formula is: Among them, S a Score the angle; the matching observation range is: [A min , A max ]; A is the actual observed angle; D is the center position deviation, T is the tilt angle; D max is the maximum position deviation allowed; T max is the maximum tilt angle allowed;

[0105] Selecting the image with the highest total score as the main image based on the clarity score and the angle score;

[0106] Extract feature points from images of the same building window taken from different angles;

[0107] Use feature matching algorithm to find the correspondence between feature points of the main image and other images;

[0108] Calculate the basic matrix and essential matrix based on the correspondence between feature points;

[0109] Register the main image using the fundamental matrix and the intrinsic matrix;

[0110] The image of the building window is extracted based on the registered main image to obtain the building window image.

[0111] Specifically, in this application, the angle score determines the observation angle range based on the identified building window type, and is calculated in combination with the position deviation between the center position of the window in the image and the center of the image and the tilt angle. Based on the clarity score and the angle score, the image with the highest total score is selected as the main image, feature points are extracted from images taken at different angles, and the feature matching algorithm is used to find the corresponding relationship between the feature points of the main image and other images, and then the basic matrix and the essential matrix are calculated to align the main image, so that the final extracted building window image is more accurate and complete.

[0112] As one implementation of a method for calculating the effective area of ​​a building window based on a BIM model, a feature matching algorithm is used to find the corresponding relationship between feature points of a main image and other images, including:

[0113] Extract feature points from the main image and other images;

[0114] Describe the local area around each feature point to obtain the descriptor of the feature point;

[0115] Calculate the Euclidean distance between the main image feature point descriptor and all other image feature point descriptors;

[0116] Select the K closest feature points as matching points.

[0117] As one implementation of a method for calculating the effective area of ​​a building window based on a BIM model, a basic matrix and an essential matrix are calculated based on the corresponding relationship of feature points, including:

[0118] Filter out the internal point set from the matching feature point pairs;

[0119] Calculate the basic matrix based on the matching point pairs in the internal point set;

[0120] Get the camera intrinsic parameter matrix of the main image and other images;

[0121] Calculate the intrinsic matrix based on the camera intrinsic parameter matrix.

[0122] As one implementation of a method for calculating the effective area of ​​a building window based on a BIM model, an inner point set is selected from the matched feature point pairs, including:

[0123] S601: Initialize the maximum number of iterations, threshold distance and minimum inlier ratio;

[0124] S602, randomly selecting a minimum number of point pairs from the matched feature point pairs;

[0125] S603, calculating a preliminary basic matrix F based on the selected point pairs;

[0126] S604, for all matched feature point pairs, calculating the reprojection error of each point pair under the basic matrix F;

[0127] S605, marking the point pairs whose reprojection errors are less than a preset threshold as inliers;

[0128] S606, recording the number of inliers and the corresponding inlier set in the current iteration;

[0129] S607, determining whether the number of inliers in the current iteration exceeds the maximum number of inliers recorded previously, and if so, updating the maximum number of inliers and the optimal inlier set;

[0130] S608, determining whether the maximum number of iterations has been reached, if not, returning to step S602 to continue iterating;

[0131] S609: If the maximum number of iterations is reached, output the final inlier point set as the screened inlier point set.

[0132] The present application also provides a building window effective area calculation system based on a BIM model, comprising:

[0133] The first processing module 201 obtains a plurality of images of a current room in a building; the room includes a plurality of building windows;

[0134] The second processing module 202 is used to: select one of the images of the same building window taken from multiple different angles to determine whether the type of the building window can be identified; the types of building windows include: casement windows, top-hung windows, center-hung windows, bottom-hung windows, horizontally-sliding windows and side-sliding windows;

[0135] The third processing module 203 extracts the image of the building window based on multiple images of the same building window taken at different angles and the identified building window type to obtain a building window image;

[0136] The fourth processing module 204 is used to: identify geometric information of the building window based on the building window image; the geometric information includes: window frame size, window sash opening direction and window sash opening angle;

[0137] A fifth processing module 205 is used to: import the identified window type and geometric information into the BIM model to obtain a corresponding window component;

[0138] The sixth processing module 206 is used to: call the corresponding window effective opening area algorithm based on the type of the building window and the parameters of the window component to calculate the effective area of ​​the current building window;

[0139] The seventh processing module 207 is used to calculate the effective area of ​​each building window in the current room, summarize the sum of the effective areas of the building windows in the current room, and obtain the total effective area of ​​the building windows in the current room.

[0140] The embodiment of the present application also discloses an electronic device.

[0141] Specifically, the device includes a memory and a processor, and the memory stores a computer program that can be loaded by the processor and execute any of the above-mentioned methods for calculating the effective area of ​​building windows based on the BIM model.

[0142] The embodiment of the present application also discloses a computer-readable storage medium. Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and execute any of the above-mentioned methods for calculating the effective area of ​​building windows based on the BIM model, and the computer-readable storage medium includes, for example, various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0143] It should be noted that the above embodiments are only used to illustrate the present application and are not intended to limit the technical solutions described in the present application. Although the present application has been described in detail in this specification with reference to the above embodiments, a person of ordinary skill in the art should understand that a person of ordinary skill in the art can still modify or make equivalent substitutions to the present application, and all technical solutions and improvements thereof that do not depart from the spirit and scope of the present application should be included in the scope of the claims of the present application.

Claims

1. A method for calculating the effective area of ​​building windows based on a BIM model, characterized in that: include: S101, obtaining a plurality of images of a current room in a building; the room includes a plurality of building windows; S102, taking multiple images of the same building window from different angles and selecting one of them to determine whether the type of the building window can be identified; If yes, execute S103; If not, another image of the same building window is selected to identify the building window type until the building window type is identified; The types of building windows include: casement windows, top-hung windows, center-hung windows, bottom-hung windows, horizontal-sliding windows and side-sliding windows; S103, extracting images of the building windows based on multiple images of the same building window taken at different angles and the identified building window types to obtain building window images; S104, recognizing geometric information of the building window based on the building window image; the geometric information includes: window frame size, window sash opening direction and window sash opening angle; S105, importing the identified window type and geometric information into the BIM model to obtain corresponding window components; S106, based on the type of the building window and the parameters of the window component, calling the corresponding window effective opening area algorithm to calculate and obtain the effective area of ​​the current building window; S107: Calculate the effective area of ​​each building window in the current room, summarize the total effective area of ​​the building windows in the current room, and obtain the total effective area of ​​the building windows in the current room.

2. The method for calculating the effective area of ​​building windows based on the BIM model according to claim 1 is characterized in that: Based on the type of building window and the parameters of window components, the corresponding window effective opening area algorithm is called to calculate the effective area of ​​the current building window, including: The calculating of the effective opening area of ​​the window by calling a corresponding algorithm based on the type and the parameters of the window component includes: If the type is a casement window, a top-hung window, a center-hung window or a bottom-hung window, it is determined whether the opening angle is greater than 70 degrees. If so, the effective opening area of ​​the window is equal to the area of ​​the window sash IfcSash; otherwise, the calculation formula is: effective opening area of ​​the window = area of ​​the window sash IfcSash * sinα, where α represents the opening angle; If the type is a sliding window, the installation position of the sliding window is obtained. If the installation position is the roof, the calculation formula is: effective window opening area = min ((perimeter of the window sash IfcSash*0.5)*sliding distance, area of ​​the window sash IfcSash); If the installation location is an exterior wall, the calculation formula is: Window effective opening area = min ((perimeter of window sash IfcSash*0.25)*sliding distance, area of ​​window sash IfcSash); If the type is a side-sliding window, determine whether the side-sliding window has an adjacent side-sliding window. If so, the effective opening area of ​​the window is equal to the area of ​​the window sash IfcSash*0.5; otherwise, the effective opening area of ​​the window is equal to the area of ​​the window sash IfcSash.

3. The method for calculating the effective area of ​​building windows based on the BIM model according to claim 2 is characterized in that: Select one of the images of the same building window taken from multiple different angles to determine whether the type of the building window can be identified, including: Randomly select an image from the collected building window images taken from different angles; Use a pre-trained machine learning model to identify building window types in selected images; If the result output by the machine learning model shows that the building window type is successfully identified and the confidence exceeds the predetermined threshold, the type of the building window is obtained; wherein the confidence calculation formula is: C = ω1 × P max +ω2×(1-H(P))+ω3×(1-σ 2 ), where C is the confidence level, ω1, ω2 and ω3 are weight coefficients, and P max is the probability of the most likely category output by the machine learning model; H(P) is the entropy of the probability distribution; σ 2 is the standard deviation of multiple prediction results; If the recognition fails or the confidence level is lower than a predetermined threshold, another image is used.

4. The method for calculating the effective area of ​​building windows based on the BIM model according to claim 3 is characterized in that: The image of the building window is extracted based on multiple images of the same building window taken at different angles and the identified building window type to obtain the building window image, including: Calculate the clarity score and angle score of each image of the same building window taken at different angles; wherein the angle score is determined based on the identified building window type, specifically including: determining the observation angle range according to the obtained building window type; for each image, calculating the position deviation between the center position of the building window in the image and the center of the image and the inclination angle of the building window in the image; and obtaining the angle score based on the matching degree of the position deviation and the inclination angle with the observation angle range; the angle score calculation formula is: Among them, S a Score the angle; the matching observation range is: [A min , A max ]; A is the actual observed angle; D is the center position deviation, T is the tilt angle; D max is the maximum position deviation allowed; T max is the maximum permissible tilt angle; Selecting the image with the highest total score as the main image based on the clarity score and the angle score; Extract feature points from images of the same building window taken from different angles; Use feature matching algorithm to find the correspondence between feature points of the main image and other images; Calculate the basic matrix and essential matrix based on the correspondence between feature points; Register the main image using the fundamental matrix and the intrinsic matrix; The image of the building window is extracted based on the registered main image to obtain the building window image.

5. The method for calculating the effective area of ​​building windows based on the BIM model according to claim 4 is characterized in that: Use feature matching algorithms to find the correspondence between feature points in the main image and other images, including: Extract feature points from the main image and other images; Describe the local area around each feature point to obtain the descriptor of the feature point; Calculate the Euclidean distance between the main image feature point descriptor and all other image feature point descriptors; Select the K closest feature points as matching points.

6. The method for calculating the effective area of ​​building windows based on the BIM model according to claim 5 is characterized in that: The basic matrix and essential matrix are calculated based on the correspondence between feature points, including: Filter out the internal point set from the matching feature point pairs; Calculate the basic matrix based on the matching point pairs in the internal point set; Get the camera intrinsic parameter matrix of the main image and other images; Calculate the intrinsic matrix based on the camera intrinsic parameter matrix.

7. The method for calculating the effective area of ​​building windows based on the BIM model according to claim 6 is characterized in that: Filter out the internal point set from the matching feature point pairs, including: S601: Initialize the maximum number of iterations, threshold distance and minimum inlier ratio; S602, randomly selecting a minimum number of point pairs from the matched feature point pairs; S603, calculating a preliminary basic matrix F based on the selected point pairs; S604, for all matched feature point pairs, calculating the reprojection error of each point pair under the basic matrix F; S605, marking the point pairs whose reprojection errors are less than a preset threshold as inliers; S606, recording the number of inliers and the corresponding inlier set in the current iteration; S607, determining whether the number of inliers in the current iteration exceeds the maximum number of inliers recorded previously, and if so, updating the maximum number of inliers and the optimal inlier set; S608, determining whether the maximum number of iterations has been reached, if not, returning to step S602 to continue iterating; S609: If the maximum number of iterations is reached, output the final inlier point set as the screened inlier point set.

8. A building window effective area calculation system based on BIM model, characterized in that: include: A first processing module acquires a plurality of images of a current room in a building; the room includes a plurality of building windows; The second processing module is used to: select one of the images of the same building window taken from multiple different angles to determine whether the type of the building window can be identified; The types of building windows include: casement windows, top-hung windows, center-hung windows, bottom-hung windows, horizontal-sliding windows and side-sliding windows; The third processing module extracts the image of the building window based on multiple images of the same building window taken at different angles and the identified building window type to obtain a building window image; A fourth processing module is used to: identify geometric information of the building window based on the building window image; the geometric information includes: window frame size, window sash opening direction and window sash opening angle; A fifth processing module is used to: import the identified window type and geometric information into the BIM model to obtain a corresponding window component; A sixth processing module is used to: call a corresponding window effective opening area algorithm based on the type of the building window and the parameters of the window component to calculate and obtain the effective area of ​​the current building window; The seventh processing module is used to calculate the effective area of ​​each building window in the current room, summarize the sum of the effective areas of the building windows in the current room, and obtain the total effective area of ​​the building windows in the current room.

9. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program according to any one of the methods of claims 1 to 7 which is loaded and executed by the processor.

10. A method for calculating the effective area of ​​building windows in a functional space, characterized in that: include: Get all window components in the functional space; The effective area of ​​each window component is calculated respectively by the method described in any one of claims 1 to 9.

Images