An optimization method for generating 3D house layout models based on house surveying

Abstract

This invention relates to the field of model design technology, specifically a method for optimizing a generated 3D house model based on house surveying. The method includes: combining edge point coordinates through a weighted transformation to obtain the combined contour edge line; analyzing the contour edge line to divide it into multiple house units; obtaining logical sub-regions corresponding to each house unit based on the overlapping and discontinuous areas of each unit, combined with the projected area of ​​the contour edge line; identifying the interval area of ​​each logical sub-region, using the interval area to identify the boundary conditions of the logical sub-regions, and determining the target boundary conditions of the logical sub-regions; determining the offset region of each house unit based on the target boundary conditions of the logical sub-regions; and performing differential accumulation on the offset regions of adjacent house units, constructing a reference model of the current 3D house model using the differential accumulation path. This method achieves improved optimization efficiency and regional positioning accuracy for the house model.

Description

Technical Field

[0001] This invention relates to the field of model design technology, specifically a method for optimizing a generated 3D house model based on house surveying. Background Technology

[0002] In the field of 3D house model generation, house surveying is used to match building parameters, structure, geometric topology and composition to form a house model. However, the current method is biased towards manual surveying and experience-based description of the location of individual units in the house structure. This can lead to confusion during house model recognition, resulting in some coordinate errors and reduced accuracy in house model creation.

[0003] In the field of interior design, applying relevant technologies to decorate target homes has become an emerging trend. By combining architectural parameter information, identification of key structural points, geometric topology analysis, and matching of decorative elements, more intelligent, efficient, and personalized interior design solutions can be achieved.

[0004] For example, Chinese Patent Publication No. CN116702298A discloses a model construction method and system for interior decoration design, which is used to realize online processing of interior decoration design and improve the efficiency of interior decoration design. It includes: acquiring building parameter information and classifying the data, obtaining parameter identification data and identifying key points to obtain a set of structural key points; performing topological analysis on the target house, generating the target house's geometric topology and generating the geometric structure to obtain the target geometric structure; constructing an initial model to obtain an initial house model and acquiring interior space information; matching decorative elements to the interior space information to generate a set of decorative elements; matching decorative schemes to generate a target decorative scheme; filtering elements from the set of decorative elements to generate a target element set; matching the target element set to spatial locations to generate a set of spatial locations; and adjusting the parameters of the initial house model to generate the target house model.

[0005] For example, Chinese Patent Publication No. CN112948933A discloses a method for constructing a house model, a display method, a management device, and a storage medium. The method for constructing the house model includes: acquiring spatial data of the house and acquiring attribute data of the house; wherein the spatial data corresponds to each floor and each room, and the spatial data and attribute data of each room in the house correspond one-to-one, and the attribute data includes at least the resident information of each room; and generating a data model using the spatial data and attribute data.

[0006] Existing technologies describe the relative positions of decorative elements in house layout recognition and the coordinate differences that pop up when the house model is displayed using perspective commands. These house models mainly use the coordinate differences between decorative elements and the disassembled house to illustrate the accuracy of the current house model. However, these methods still have logical errors in the edge line configuration when the house model is disassembled, which makes the division of each house unit unclear, ultimately leading to increased errors in each area and reduced model accuracy. Summary of the Invention

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a method for optimizing a three-dimensional house model generated based on house surveying, including: S1, taking the connection attributes of the three-dimensional house model elements as the basis, extracting the edge point coordinates and topological relationships on multiple cross sections, and using a weight transformation method to perform multi-level combination of the edge lines of the model at different positions to obtain the combined contour edge lines.

[0008] S2 analyzes the contour edge lines, divides them into multiple housing units, and calculates the three-dimensional overlap of the contour edge lines of adjacent housing units to obtain the overlapping area and the broken area corresponding to each housing unit. Combined with the projected area of ​​the contour edge lines, the logical sub-regions corresponding to each housing unit are obtained.

[0009] S3, based on the logical sub-regions of each housing unit, identifies the interval area of ​​each logical sub-region, and uses the interval area to identify the boundary conditions of the current logical sub-region, and determines the target boundary conditions of the current logical sub-region.

[0010] S4. Based on the target boundary conditions of the current logical sub-region, perform spatial calculation on the currently accessed 3D house model, and determine the offset region of each house unit by the mapping region of each target boundary condition after spatial calculation.

[0011] S5 performs differential accumulation on the offset areas of adjacent house units, and constructs a reference model of the current three-dimensional house model using the differential accumulation path.

[0012] The beneficial effects of this invention are as follows: First, this invention, based on connection attributes, re-extracts the edge points on multiple cross-sections by combining them, thereby re-obtaining the contour edge lines at the corresponding positions and identifying the problem of blurred model edges when setting up the house model; and based on the projected area and density of the contour edge lines, it splices together to generate house units, and divides multiple logical sub-units in sequence by overlapping areas and broken areas within the house units, in order to verify whether the configuration of each position in the house model is reasonable, and whether the edge lines of the generated areas are configured accurately.

[0013] Second, this invention uses the interval area of ​​logical sub-regions as a benchmark to trigger boundary condition checks, and determines the target boundary conditions through similarity calculation and splicing relationship judgment. It further determines the setting accuracy of structural features such as walls, doors and windows in the house model, as well as the reliability of the target boundary conditions after verification.

[0014] Third, this invention determines the bias region by mapping the target boundary conditions, comparing data points (and classifying problem types), and generating a differential accumulation path by accumulating the differential regions of adjacent units. Combining the single access duration and the number of accumulations, the traversal cycle is dynamically adjusted, improving the model's update performance in multi-scenario processing, reducing manual intervention in model configuration, shortening the model optimization cycle, and improving efficiency. Attached Figure Description

[0015] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0016] Figure 1 This is a flowchart illustrating a method for optimizing a 3D house model generated from house surveying.

[0017] Figure 2 This is a flowchart illustrating step S1 of an optimization method for generating a 3D house model based on house surveying.

[0018] Figure 3 This is a flowchart illustrating step S2 of a method for optimizing a 3D house model generated based on house surveying.

[0019] Figure 4 This is a flowchart illustrating step S3 of a method for optimizing a 3D house model generated based on house surveying.

[0020] Figure 5 This is a flowchart illustrating step S4 of a method for optimizing a 3D house model generated based on house surveying.

[0021] Figure 6 This is a flowchart illustrating step S5 of a method for optimizing a 3D house model generated based on house surveying. Detailed Implementation

[0022] The embodiments of the present invention are described in detail below. The embodiments described below are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. Where specific techniques or conditions are not specified in the embodiments, they shall be performed in accordance with the techniques or conditions described in the literature in the art or in accordance with the product manual.

[0023] See Figure 1An optimization method for generating a 3D house model based on house surveying includes: S1, using the connection attributes of the elements of the 3D house model as a basis, extracting the coordinates of edge points and topological relationships on multiple cross sections, and using a weight transformation method to perform multi-level combination of the edge lines of the model at different positions to obtain the combined contour edge lines.

[0024] S2 analyzes the contour edge lines, divides them into multiple housing units, and calculates the three-dimensional overlap of the contour edge lines of adjacent housing units to obtain the overlapping area and the broken area corresponding to each housing unit. Combined with the projected area of ​​the contour edge lines, the logical sub-regions corresponding to each housing unit are obtained.

[0025] S3, based on the logical sub-regions of each housing unit, identifies the interval area of ​​each logical sub-region, and uses the interval area to identify the boundary conditions of the current logical sub-region, and determines the target boundary conditions of the current logical sub-region.

[0026] S4. Based on the target boundary conditions of the current logical sub-region, perform spatial calculation on the currently accessed 3D house model, and determine the offset region of each house unit by the mapping region of each target boundary condition after spatial calculation.

[0027] S5 performs differential accumulation on the offset areas of adjacent house units, and constructs a reference model of the current three-dimensional house model using the differential accumulation path.

[0028] The core objective of the current solution is to optimize the area of ​​each marked region within the house model, detect and correct potential errors by reassembling and reassembling them, and synchronize these identified problems to multiple 3D models after the house survey to identify areas prone to problems during the current survey.

[0029] Preferably, the connection attributes of the 3D house model represent the wall endpoints, corner points, and other points that need attention when connecting the models. These points are marked, and the current model connection relationship is identified to check whether there is a problem of reduced common area after the model is built.

[0030] like Figure 2 As shown, the implementation of step S1 includes: S11, obtaining the curve of the three-dimensional house model at the connection position as the basic line segment, and determining whether the position of the basic line segment includes multiple endpoints and multiple corner points.

[0031] S12, if they exist, use multiple endpoints and multiple corner points as the coordinates of the identified edge points, and record the topological relationship of each edge point.

[0032] S13, if it does not exist, then project the base line segment onto the centerline of the building structure at the connection point. Use the intersection of the projected center projection line and the current base line segment as the coordinates of the edge point, and record the topological relationship between the projection plane and the plane where the current base line segment is located. When there are no obvious geometric features on the base line segment, construct the edge point through the projection method to ensure that effective contour points can be extracted in all cases.

[0033] Using the current base line segment as the center, construct a center projection line perpendicular to the plane where the line segment is located; calculate the intersection of this projection line with adjacent structural surfaces such as another wall or roof; record the intersection point as the edge point, and then record the relationship between the projection line and the plane where the original base line segment is located; the three-dimensional coordinates of the edge point and its corresponding structural surface; store this information together with the normally extracted edge point part to complete the consistency of the current plane setting.

[0034] Preferably, the above-mentioned weight transformation method connects the edge points, sets the weight based on the connection attribute of the connected edge points, and further sets a weight value for the distance between the connected edge points to illustrate the geometric contour relationship of the house model in three-dimensional space, and outputs the connected boundary points as contour edge lines.

[0035] The method of obtaining the combined contour edge line through weight transformation includes: connecting edge points to form several candidate edges, setting weights based on the structure and length of the candidate edges; using the weights of each candidate edge to correct the candidate edges with the shortest path, and using the corrected candidate edges as the output contour edge line.

[0036] At this point, weights are set based on the structure of the candidate edge and the length value after connection. For example, basic weight values ​​are set according to different structures, and the reciprocal of the length of the candidate edge is used as the length-related weight. The sum of these two weights is used as the weight of the candidate edge. The calculated value is standardized and its dimensions are eliminated before setting. As for the basic weight, the weight values ​​corresponding to each structure of the current house model are set in advance.

[0037] When refining candidate edges using the shortest path, the system calculates the path with the smallest sum of candidate edge weights, ensuring that the current combination of candidate edges conforms to the main structural orientation. This path is then considered as the current output contour edge line. Subsequently, the system iterates through multiple planes and positions of the house model to complete the division of house units and the output of edge contours.

[0038] In one embodiment of the present invention, the logical sub-region includes a private area, a shared area, and an interval area. This step requires obtaining a clear, non-overlapping house boundary outline, calculating the projected area of ​​each area, and completing the logical region division of the model space to form a house area distribution map with multiple areas such as private area, shared area, and interval area. This map will further verify whether there are any abnormal issues in some area types of the logical sub-region after the current house model is continuously generated.

[0039] like Figure 3 As shown, the implementation of step S2 includes: S21, based on the obtained contour edge line, generating the projected area of ​​the contour edge line at the location of the contour edge line.

[0040] S22, using the density of the center point and the density of the vertices after the outline edge line is projected, the plane containing the outline edge line is spliced ​​according to the projected area to obtain multiple house units.

[0041] S23, Based on the transition value of adjacent house units on the common contour edge line, distinguish between overlapping areas and discontinuous areas of house units, and calculate the three-dimensional overlap of house units.

[0042] S24. Based on the three-dimensional overlap of the housing unit, the housing unit is divided into logical sub-regions by a secondary combination of overlapping areas and discontinuous areas.

[0043] When generating the projected area, the 3D contour edge lines can be projected onto the XY plane, i.e., the horizontal plane on which the houses are located. Then, the projected area of ​​each closed region formed by the contour edge lines is calculated using coordinate points. The density of each projected center coordinate and vertex coordinate is then calculated. For example, clustering is performed using the coordinates of multiple points projected on the contour edge lines to determine the coordinates of the current center point. The density of the center point is then determined using the number of corresponding points in the neighborhood of the center point. The density of vertices is calculated similarly. The vertices identified at this time are more likely to be quadrilateral vertices or other polygon vertices projected onto the 2D plane. The density values ​​are determined by the number of points at the center of the plane and the number of vertices. After determining the density values ​​using density clustering, the contour edge lines that can form closed planes are sequentially divided into house units according to the selected density values. Each house unit is considered as a region within the current house, or the house unit can be considered as the region after the entire house model is projected. The neighborhood radius used in density clustering can be set based on the average distance between multiple planes of the current house model to obtain multiple house units after the current house model is projected.

[0044] Preferably, after obtaining the density of the center point and the density of the vertices after projection, it is also necessary to verify the type of the area where the original outline edge line is located, that is, whether the current house model is a type of house surveying such as independent rooms, corridors or public areas.

[0045] Preferably, when calculating the transition value of the common contour edge line, if it is an overlapping area, the coordinate difference between the two endpoints on the common contour edge line is used for processing. That is, the coordinate difference is calculated sequentially, the average value of the coordinate difference is obtained, and the angle between adjacent house units at the boundary is determined. If the average value of the coordinate difference and the angle both meet the requirements of the current house unit combination, that is, the adjacent house units are in the normal house setting value, the current house unit is marked, and the ratio of its overlapping volume to the minimum bounding box volume is calculated. This value is output as the 3D overlap. If the average value of the coordinate difference and the angle do not meet the requirements, the corresponding house unit is directly output, and the 3D overlap is marked as 0, indicating that there is a problem of excessive gap or abnormal angle setting when splicing the current houses. At this time, the determination that the average value of the coordinate difference and the angle meet the requirements is set by the distance and angle between the corresponding house units during house surveying. That is, the current value is identified by the confidence interval of the average value of the coordinate difference and the angle in historical data, and the confidence interval is set with a 95% confidence level.

[0046] It should be noted that overlapping volume represents the intersection, while the minimum bounding box refers to the minimum bounding box of a house unit. It represents the volume of a bounding contour, which is the relative 3D union, that is, the ratio of the 3D area projected onto the 2D area. It is used to segment the overlap of the 3D contour in the broken lines and overlapping parts, and to explain the relative area ratio after projection. In essence, it uses density clustering after projection and historical data statistics to quickly identify house units, so as to realize the identification of multi-position splicing and aggregation in house units. Using 3D projection of 2D is used to quickly judge and identify the broken lines and overlapping parts in the preview scene of the model.

[0047] When there is a broken line area, the endpoint is checked using the common contour edge line. If the distance between the broken points is greater than the maximum allowable distance, it is considered that there is a broken line, and the three-dimensional overlap at the corresponding position is marked as 0. At this time, the maximum allowable distance is the upper limit of the current average coordinate difference to explain whether the broken line is reasonable.

[0048] Preferably, step S24 is implemented by: determining the number of overlapping areas in the current housing unit; if the number of overlapping areas is greater than or equal to two, calculating the area ratio between adjacent broken areas and overlapping areas; if the area ratio between adjacent broken areas and overlapping areas is less than a preset area ratio, deleting the corresponding broken area, and using the remaining overlapping areas and broken areas as the output logical sub-regions.

[0049] If the area ratio of an adjacent broken area to an overlapping area is greater than a preset area ratio, the corresponding overlapping area is deleted, and the remaining overlapping area and broken area are used as the output logical sub-region.

[0050] If the number of overlapping areas is less than two, the current overlapping area is expanded, and the boundary between the expanded overlapping area and the broken area is used to combine the broken area and the overlapping area to obtain the logical sub-area of ​​the current housing unit.

[0051] Preferably, the preset area ratio ranges from 0.2 to 0.5. When the current scene requires rich details, a value of 0.2 will be used. If it is necessary to avoid deleting the effective area, a value of 0.5 will be used. In this case, if the focus is on recognizing residential building models, a value of 0.3 will be used as the current threshold to ensure that small broken lines are deleted and the main structure is preserved.

[0052] In one embodiment of the present invention, after obtaining the logical sub-regions of each housing unit, the size of the area between the logical sub-regions is identified, and the data format of the combination of type, boundary conditions, etc. of each logical sub-region is determined. After processing these data formats, the target boundary conditions of each logical sub-region are marked.

[0053] The boundary conditions output at this point are used to indicate whether the ownership type is private or shared, the function type is one of several types such as corridor or elevator shaft, and the corresponding coordinates and connection angles, to illustrate the current setup of the building model. The boundaries of each logical sub-region are labeled, and the boundary conditions are then displayed in numerical form. Using the difference values ​​between multiple boundary conditions, the target boundary condition that clearly exhibits anomalies is selected.

[0054] After obtaining the gap area, abnormal gaps are identified to eliminate invalid gap areas, such as... Figure 4 As shown, the implementation of step S3 also includes: S31, using the interval area of ​​each logical sub-region, for abnormal intervals between logical sub-regions, triggering each boundary condition to check for abnormal intervals, forming an interval check sequence.

[0055] S32, determine the splicing relationship between the interval investigation sequences. When the splicing relationship of any group of logical sub-regions in the interval investigation sequence is consistent, splice each logical sub-region in a continuous hierarchical form, and use the boundary conditions of the spliced ​​logical sub-region as the target boundary conditions of the current logical sub-region.

[0056] S33, when the splicing relationship is inconsistent, any set of logical sub-regions are spliced ​​together in a cyclic splicing manner, the intersection of the boundary conditions of the spliced ​​logical sub-regions is calculated, and the corresponding intersection is output as the target boundary condition of the current logical sub-region.

[0057] Preferably, the above-mentioned splicing relationship is generally represented by various relative relationships during splicing, such as nested sub-regions, obvious boundaries between sub-regions, intersection or overlap between sub-regions, sub-regions differentiated by grid, and sub-regions forming closed loops. If the extracted logical sub-regions satisfy these splicing relationships, and any set of extracted data can achieve consistent representation content, then the logical sub-regions are combined according to their relative relationships and multi-level splicing. The final output boundary conditions of each logical sub-region are relative to the overall boundary conditions under this splicing. As for inconsistent scenarios, it is necessary to iterate and splice multiple times to obtain the intersection to explain the target splicing conditions at the current splicing time.

[0058] Preferably, the aforementioned abnormal intervals are identified by comparing the interval area with the initial dimensions of the current 3D house model. If the currently extracted interval area is inconsistent with the initial dimensions, it is considered that there are abnormal intervals, and the currently generated 3D model may have problems such as incorrect dimension marking.

[0059] The implementation method of boundary condition investigation in step S31 includes: based on the boundary conditions of each logical sub-region, calculating the maximum difference of the logical sub-regions under the same boundary type according to the boundary type corresponding to the boundary conditions, and performing similarity calculation on each logical sub-region based on the maximum difference under the same boundary type.

[0060] After calculation, the minimum difference under the same boundary type is determined from multiple logical sub-regions. The logical sub-regions corresponding to the minimum difference are then combined to obtain the interval investigation sequence.

[0061] Preferably, the maximum difference mentioned above can be calculated based on the interval area to obtain a group of logical sub-regions that are far apart. These logical sub-regions are then calculated using cosine similarity based on the calculated maximum difference. Multiple logical sub-regions with a similarity value greater than 0.6 are taken as the calculated logical sub-regions. The corresponding logical sub-regions are then divided into multiple groups of data according to the minimum difference under the same boundary type. These combined logical sub-regions are regarded as an interval screening sequence.

[0062] At this point, the correlation between logical sub-regions on boundary conditions is quantified by similarity measurement, providing a basis for combination. When identifying the same boundary type using the maximum difference in interval area, the main focus is on identifying defects in the current model for different region markers. Then, for these defects, similarity clustering is used to indicate problems with the boundary conditions in the current scene. After processing according to similarity values, the region with the smallest difference is prioritized for investigation, reducing the cost of model identification and ultimately achieving the correction of the 3D house model.

[0063] In one embodiment of the present invention, such as Figure 5 As shown, step S4 is implemented as follows: S41, identify the mapping region of each target boundary condition, and traverse the mapping region by the ratio of the number of contour edge lines in the mapping region to the total number of contour edge lines. At this time, the target boundary conditions with high ratios can be used as priority identification data for subsequent house model identification. The identified data includes boundary condition redundancy, insufficient boundary condition identification, and boundary condition identification conflicts between different functional areas. These problems are centrally represented, and a bias region related to the corresponding problem is generated. When the ratio of the number of contour edge lines in the mapping region to the total number of contour edge lines is large, it indicates that the current mapping region is a relatively important region, and a small value indicates that the corresponding region is a local small region. At this time, it will be processed in order according to the size of the ratio, and the corresponding bias will be combined into multiple logical sub-regions in the current house model in the form of continuous mapping regions in the form of three or more consecutive regions to complete the processing.

[0064] S42, after traversing the mapping region, form data point pairs according to the ratio of the number of contour edge lines in the mapping region to the total number of contour edge lines. Compare the parameter offsets of adjacent mapping regions under the data point pairs to determine the output offset region. The data point pairs are sorted based on this ratio so that each mapping region corresponds to a ratio.

[0065] When identifying the bias region in step S42, the implementation method includes: in response to the input mapping region, saving the mapping region with the data point pairs corresponding to the mapping region, and determining the problem type of the data point pairs.

[0066] If the problem type is boundary condition redundancy, the output is the region where the maximum offset value is located after the difference between the current input mapping region and the adjacent mapping regions.

[0067] If the problem type is one with missing boundary conditions, calculate the average offset value of the mapped region based on the current location of the mapped region, and use the center region corresponding to the average offset value as the output offset region.

[0068] If the problem type involves boundary condition conflicts, the mapping region corresponding to the conflicting boundary conditions will be used as the output bias region.

[0069] At this point, the offset value is obtained by comparing multiple points corresponding to the contour edge line on the mapped area with the expected house model to determine the coordinate errors. Then, errors at multiple locations are identified. For example, for boundary condition redundancy, the difference between the current mapped area and the adjacent mapped areas is calculated to find different mapped positions. The local area corresponding to the maximum offset value is found and the corresponding area is output as the offset area. When the boundary condition is missing, it is necessary to identify the central area where the average offset value is located when the condition is missing, that is, the center position of the current input mapped area. Based on the access to this central position, the relevant mapped area at this central position is finally readjusted to adjust the house model on the offset area. As for boundary condition conflicts, it is only necessary to directly handle the conflict positions. These positions will have obvious coordinate point errors, that is, the house model of this part needs to be adjusted in real time according to the access situation.

[0070] In one embodiment of the present invention, during differential accumulation, differential accumulation is performed at the location of the offset region. During differential accumulation, the coordinate errors before and after spatial calculation on the offset region are used to form a differential accumulation path in the form of error points. The accessibility of the relevant regions is identified by the areas traversed by the differential accumulation path to determine whether there are any accuracy setting problems at each location after the house is spliced ​​and reassembled.

[0071] like Figure 6 As shown, the implementation of step S5 includes: S51, performing spatial merging and matching based on the offset areas of adjacent housing units, obtaining the combined cumulative difference set, and sorting the data in the cumulative difference set to obtain the differential accumulation path.

[0072] S52, connect each housing unit by the single visit duration and cumulative number of visits at each location of the differential cumulative path, and determine the traversal period of the differential path.

[0073] S53 updates the current house model according to the traversal cycle and outputs the updated house model as the reference model.

[0074] The implementation methods for determining the traversal period of the differential path include: if the cumulative number of differential paths is less than the preset number of times, then the traversal period is determined based on the current location of the house model and the duration of a single visit.

[0075] If the cumulative number of differential paths is greater than the preset number of times, and the duration of a single access is greater than the preset access duration, then the traversal period is determined by the difference between the current single access duration and the preset access duration.

[0076] The preset count value set above is used to describe the required number of times the current house model is accessed by normal users. This preset count value can be based on the number of user visits, setting a confidence interval, and using the lower limit of the confidence interval as the preset count value to indicate the minimum number of times a user needs to access and view the house model. The preset access duration is set in the same way as the preset count value, using the lower limit of the confidence interval. It can be based on user access data from the past three months, taking the lower limit of the 95% confidence interval as the preset count value and preset access duration. This represents the minimum time period required for a user to access and view a house model. Then, this count value and access duration are used to set the traversal cycle of the 3D model to check in real time whether there are any corresponding errors in the rendered house model and update it promptly.

[0077] It should be noted that the single access duration and cumulative number of visits are used to describe the relative situation when the house model is generated. The single access duration directly indicates the display time required for the current model under fast preview. If the time is too short, there may be a problem that the model is not fully generated and displayed. The quantified model deviation is used to describe the deviation of some parts that may not be loaded within a specific time period under fast model generation. As for the cumulative number of visits, it emphasizes the number of times the house model is accessed. Based on the cumulative number of visits, the deviation of the model not being loaded is accumulated. By accumulating and aggregating these deviations, we can know the path of the differential relative position that the current model can appear when it is displayed in the scenario of fast preview of the house. This is convenient for subsequent model optimization and adjustment of its generation speed and the relative content filling time of the model display.

[0078] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention, which are still covered within the protection scope of the present invention.

Claims

1. A method for optimizing a generated 3D floor plan model based on house surveying, characterized in that, include: S1, based on the connection attributes of the three-dimensional house model elements, extracts the edge point coordinates and topological relationships on multiple cross sections, and performs multi-level combination of the edge lines of the model at different positions through weight transformation to obtain the combined contour edge lines; S2, analyze the contour edge lines, divide into multiple house units, and calculate the three-dimensional overlap of the contour edge lines of adjacent house units to obtain the overlapping area and the broken area corresponding to each house unit. Combined with the projected area of ​​the contour edge lines, obtain the logical sub-region corresponding to each house unit. Logical sub-regions include private indoor areas, shared common areas, and partitioned areas; Step S2 can be implemented in the following ways: S21, Based on the obtained contour edge line, generate the projected area of ​​the contour edge line at the location of the contour edge line; S22, using the density of the center point and the density of the vertices after the outline edge line is projected, the plane containing the outline edge line is spliced ​​according to the projected area to obtain multiple house units; S23, Based on the transition value of adjacent house units on the common contour edge line, distinguish between the overlapping area and the discontinuous area of ​​the house units, and calculate the three-dimensional overlap of the house units. S24, based on the three-dimensional overlap of the housing unit, the housing unit is divided into logical sub-regions by a secondary combination of overlapping areas and discontinuous areas; S3, based on the logical sub-regions of each housing unit, identifies the interval area of ​​each logical sub-region, and uses the interval area to identify the boundary conditions of the current logical sub-region, and determines the target boundary conditions of the current logical sub-region. S4. Based on the target boundary conditions of the current logical sub-region, perform spatial calculation on the currently accessed 3D house model, and determine the offset region of each house unit by the mapping region of each target boundary condition after spatial calculation. S5 performs differential accumulation on the offset areas of adjacent house units, and constructs a reference model of the current three-dimensional house model using the differential accumulation path.

2. The method for optimizing a 3D floor plan model generated based on house surveying according to claim 1, characterized in that, The implementation methods for step S1 include: S11, obtain the curve of the 3D house model at the connection position as the basic line segment, and determine whether the position of the basic line segment includes multiple endpoints and multiple corner points; S12, if they exist, use multiple endpoints and multiple corner points as the coordinates of the identified edge points, and record the topological relationship of each edge point; S13, if it does not exist, then project the basic line segment onto the center line of the building structure at the connection position, and use the intersection of the projected center projection line and the current basic line segment as the coordinates of the edge point, and record the topological relationship between the projection plane and the plane where the current basic line segment is located as the edge point.

3. The method for optimizing a generated 3D floor plan model based on house surveying according to claim 2, characterized in that, The implementation methods of step S24 include: Determine the number of overlapping areas in the current housing unit. If the number of overlapping areas is greater than or equal to two, calculate the area ratio between the adjacent broken area and the overlapping area. If the area ratio between the adjacent broken area and the overlapping area is less than the preset area ratio, delete the corresponding broken area and use the remaining overlapping area and broken area as the output logical sub-region. If the area ratio of an adjacent broken area to an overlapping area is greater than a preset area ratio, the corresponding overlapping area is deleted, and the remaining overlapping area and broken area are used as the output logical sub-region. If the number of overlapping areas is less than two, the current overlapping area is expanded, and the boundary between the expanded overlapping area and the broken area is used to combine the broken area and the overlapping area to obtain the logical sub-area of ​​the current housing unit.

4. The method for optimizing a generated 3D floor plan model based on house surveying according to claim 1, characterized in that, The implementation of step S3 also includes: S31, using the area of ​​the interval between each logical sub-region, for abnormal intervals between logical sub-regions, trigger the investigation of each boundary condition with abnormal intervals, forming an interval investigation sequence; S32, determine the splicing relationship between the interval investigation sequences. When the splicing relationship of any group of logical sub-regions in the interval investigation sequence is consistent, splice each logical sub-region in a continuous hierarchical form, and use the boundary conditions of the spliced ​​logical sub-region as the target boundary conditions of the current logical sub-region. S33, when the splicing relationship is inconsistent, any set of logical sub-regions are spliced ​​together in a cyclic splicing manner, the intersection of the boundary conditions of the spliced ​​logical sub-regions is calculated, and the corresponding intersection is output as the target boundary condition of the current logical sub-region.

5. The method for optimizing a generated 3D floor plan model based on house surveying according to claim 4, characterized in that, The implementation methods for boundary condition checking in step S31 include: Based on the boundary conditions of each logical sub-region, the maximum difference between logical sub-regions under the same boundary type is calculated according to the boundary type corresponding to the boundary conditions. Based on the maximum difference under the same boundary type, the similarity of each logical sub-region is calculated. After calculation, the minimum difference under the same boundary type is determined from multiple logical sub-regions. The logical sub-regions corresponding to the minimum difference are then combined to obtain the interval investigation sequence.

6. The method for optimizing a generated 3D floor plan model based on house surveying according to claim 1, characterized in that, Step S4 can be implemented in the following ways: S41, Identify the mapping region of each target boundary condition, and perform traversal processing on the mapping region by the ratio of the number of contour edge lines in the mapping region to the total number of contour edge lines; S42, after traversing the mapped region, form data point pairs according to the ratio of the number of contour edge lines in the mapped region to the total number of contour edge lines. Then, compare the parameter offsets of the adjacent mapped regions under the data point pairs to determine the output offset region.

7. The method for optimizing a generated 3D floor plan model based on house surveying according to claim 6, characterized in that, The implementation methods of step S42 include: In response to the input mapping region, the mapping region is saved with the data point pairs corresponding to the mapping region, and the problem type of the data point pairs is determined. If the problem type is boundary condition redundancy, the output is the region where the maximum offset value is located after the difference between the current input mapping region and the adjacent mapping regions. If the problem type is one with missing boundary conditions, calculate the average offset value of the mapping region based on the current location of the mapping region, and use the center region corresponding to the average offset value as the output offset region. If the problem type involves boundary condition conflicts, the mapping region corresponding to the conflicting boundary conditions will be used as the output bias region.

8. The method for optimizing a generated 3D floor plan model based on house surveying according to claim 1, characterized in that, Step S5 can be implemented in the following ways: S51, perform spatial merging and matching based on the offset areas of adjacent housing units, obtain the combined cumulative difference set, and sort the data in the cumulative difference set to obtain the differential accumulation path; S52, connect each housing unit by the single visit duration and cumulative number of visits at each location of the differential cumulative path to determine the traversal period of the differential path; S53 updates the current house model according to the traversal cycle and outputs the updated house model as the reference model.

9. The method for optimizing a generated 3D floor plan model based on house surveying according to claim 8, characterized in that, The implementation methods for determining the traversal period of the differential path include: if the cumulative number of differential paths is less than the preset number of times, then the traversal period is determined based on the current location of the house model and the duration of a single visit; If the cumulative number of differential paths is greater than the preset number of times, and the duration of a single access is greater than the preset access duration, then the traversal period is determined by the difference between the current single access duration and the preset access duration.

Images