Point cloud and house type image automatic matching method based on through view point cloud features

Abstract

The invention discloses a point cloud and house type image automatic matching method based on through point cloud features, and relates to the technical field of electrical digital processing, and the method comprises the steps: obtaining target point cloud data through an interaction point cloud collection device; collecting a target house type image, and obtaining a house type image feature set; based on the target point cloud data, horizontal section extraction is carried out, a target section point set is obtained, and the target section point set comprises multiple groups of layer height point sets; selecting any layer height point set in the target section point set, and obtaining target contour features based on a straight line detection algorithm; performing initial matching according to the house type image feature set and the target contour features; on the basis of the target tangent plane point set, through-view point cloud feature recognition is carried out, and target through-view features are obtained; and according to the target intervisibility features and the house type image feature set, performing displacement optimization on the initial matching result, obtaining optimal matching, and generating a matching result. Therefore, the technical effects of automatic matching and improvement of matching calculation efficiency are achieved.

Description

Technical Field

[0001] The invention relates to the technical field of electronic digital processing, and in particular to a method for automatically matching a point cloud with a floor plan based on line-of-sight point cloud features. Background Art

[0002] Point cloud is a three-dimensional data representation composed of a large number of discrete points, which is used in the fields of architecture, interior design, map making, etc. By matching point clouds, the relationship between two-dimensional floor plans and point clouds is established, which is of great significance to the three-dimensional digitization of indoor spaces. The existing matching methods have technical problems such as low automation and low matching calculation efficiency. Summary of the invention

[0003] The purpose of this application is to provide a method for automatically matching point clouds and floor plans based on line-of-sight point cloud features, so as to solve the technical problems of low automation and low matching calculation efficiency in the prior art.

[0004] In view of the above technical problems, the present application provides a method for automatically matching point clouds and floor plans based on line-of-sight point cloud features.

[0005] In a first aspect, the present application provides a method for automatically matching point clouds and floor plans based on viewpoint cloud features, wherein the method comprises:

[0006] Interactive point cloud acquisition equipment obtains target point cloud data; collects target floor plan and obtains floor plan feature set; based on the target point cloud data, performs horizontal section extraction to obtain target section point set, the target section point set includes multiple groups of floor height point sets; selects any group of floor height point sets in the target section point set, and obtains target contour features based on a straight line detection algorithm; performs initial matching based on the floor plan feature set and the target contour features; performs line-of-sight point cloud feature recognition based on the target section point set to obtain target line-of-sight features; performs displacement optimization on the initial matching result based on the target line-of-sight features and the floor plan feature set to obtain the best match and generate a matching result.

[0007] In a second aspect, the present application also provides a point cloud and floor plan automatic matching system based on the viewpoint cloud feature, wherein the system comprises:

[0008] A data acquisition module, the data acquisition module is used to interact with the point cloud acquisition device to acquire target point cloud data; an image feature extraction module, the image feature extraction module is used to acquire the target floor plan and acquire the floor plan feature set; a horizontal section extraction module, the horizontal section extraction module is used to perform horizontal section extraction based on the target point cloud data to acquire the target section point set, and the target section point set includes multiple groups of floor height point sets; a contour detection module, the contour detection module is used to select any group of floor height point sets in the target section point set, and acquire the target contour feature based on the straight line detection algorithm; an initial matching module, the initial matching module is used to perform initial matching with the target contour feature according to the floor plan feature set; a line of sight feature recognition module, the line of sight feature recognition module is used to perform line of sight point cloud feature recognition based on the target section point set to acquire the target line of sight feature; an optimal matching module, the optimal matching module is used to perform displacement optimization on the initial matching result according to the target line of sight feature and the floor plan feature set, acquire the optimal match, and generate a matching result.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0010] The target point cloud data is acquired through the interactive point cloud acquisition device; the target floor plan is acquired to obtain the floor plan feature set; based on the target point cloud data, the horizontal section is extracted to obtain the target section point set, which includes multiple sets of floor height point sets; any set of floor height point sets in the target section point set is selected, and the target contour features are acquired based on the straight line detection algorithm; initial matching is performed based on the floor plan feature set and the target contour features; based on the target section point set, the line of sight point cloud feature recognition is performed to obtain the target line of sight features; based on the target line of sight features and the floor plan feature set, the initial matching results are optimized by displacement, the optimal match is acquired, and the matching results are generated. Thus, the technical effect of automatic matching and improving the matching calculation efficiency is achieved.

[0011] The above description is only an overview of the technical solution of the present application. In order to more clearly illustrate the technical means of the present application and to implement it in accordance with the contents of the specification, and to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The embodiments of the present invention and the following brief description are illustrated in conjunction with the accompanying drawings, which are described as follows:

[0013] Figure 1 This is a flow chart of the method for automatically matching point clouds and floor plans based on the feature of the point cloud of the present application;

[0014] Figure 2A schematic diagram of the process of obtaining a target section point set in the automatic matching method between point cloud and floor plan based on the feature of the through-view point cloud in this application;

[0015] Figure 3 This is a structural diagram of the point cloud and floor plan automatic matching system based on the line-of-sight point cloud features of this application.

[0016] Explanation of reference numerals: data acquisition module 11 , image feature extraction module 12 , horizontal section extraction module 13 , contour detection module 14 , initial matching module 15 , visual feature recognition module 16 , optimal matching module 17 . DETAILED DESCRIPTION

[0017] The present application solves the technical problems of low automation and low matching calculation efficiency faced by the prior art by providing a method for automatically matching point clouds and floor plans based on line-of-sight point cloud features.

[0018] The solution in the embodiment of the present technology, in order to solve the above problems, adopts the following overall ideas:

[0019] First, the interactive point cloud acquisition device is used to obtain the target point cloud data. Next, the target floor plan is collected to obtain the feature set of the floor plan. Based on the target point cloud data, horizontal section extraction is performed to obtain the target section point set. This section point set includes multiple sets of point sets with different floor heights. Then, one of these floor height point sets is selected, and the target contour features are obtained using the straight line detection algorithm. Subsequently, the feature set of the user floor plan is initially matched with the target contour features. Then, the line-of-sight point cloud feature recognition is performed based on the target section point set. Then, according to the target line-of-sight features and the floor plan feature set, the initial matching results are optimized by displacement, the optimal matching results are generated, and the point cloud data is accurately matched with the floor plan. Thus, the technical effect of automated matching and improved matching calculation efficiency is achieved.

[0020] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the drawings and specific implementation methods of the specification. It should be noted that the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. It should also be noted that for the sake of ease of description, only the parts related to the present invention are shown in the drawings, rather than all of them.

[0021] Embodiment 1

[0022] like Figure 1 As shown, the present application provides a method for automatically matching point clouds and floor plans based on viewpoint cloud features, the method comprising:

[0023] S100: interactive point cloud acquisition device, obtaining target point cloud data;

[0024] Point cloud acquisition equipment refers to equipment used to acquire point cloud data, including laser scanners, LiDAR equipment or 3D cameras, etc. The original data acquired by the point cloud acquisition equipment is processed with point cloud data processing software to obtain target point cloud data.

[0025] Optionally, first, prepare the point cloud acquisition device and ensure that the point cloud acquisition device is connected and in working condition. Then, place the point cloud acquisition device in the target area where point cloud data needs to be collected. Ensure that the device can obtain the required coverage. Next, start the acquisition device to start data acquisition and create point cloud data. According to the device type, device parameters and target area acquisition requirements, set different parameters such as sampling density, scanning angle, etc. Then, store the collected point cloud data in the internal memory of the device or on the connected computer. Finally, import the collected raw point cloud data into the point cloud data processing software. The point cloud data processing software is used to remove noise, align data at different scanning positions, create three-dimensional models, extract features, etc. The target point cloud data can be generated through the point cloud data processing software.

[0026] S200: Collect the target floor plan and obtain a floor plan feature set;

[0027] The target floor plan refers to the floor plan of the target floor plan that needs to be matched with the point cloud, and the target floor plan includes the outline, edges, corners, doors, windows and other feature information of the target floor plan. Specifically, a floor plan feature set is obtained, and the information in the floor plan is extracted as digital features through feature extraction, and the extracted features are saved as a digital data set to form a floor plan feature set. This includes the outline of the room, the size and position of the walls, windows and doors, and the marking of the purpose of the room.

[0028] S300: extracting a horizontal section based on the target point cloud data to obtain a target section point set, wherein the target section point set includes multiple groups of layer height point sets;

[0029] Optionally, horizontal slice extraction is performed on the target point cloud data. By extracting point cloud data points on multiple horizontal planes at different heights, compression of the point cloud data is achieved, reducing the number of point cloud data points that need to be processed subsequently, which has the technical effect of improving processing efficiency and reducing processing computing power consumption.

[0030] Optionally, the target section point set includes multiple sets of floor height point sets, and the multiple sets of floor height point sets contain all target point cloud data features. The target point cloud data features correspond one-to-one to multiple floor plan features in the floor plan feature set.

[0031] Further, such as Figure 2 As shown, based on the target point cloud data, horizontal section extraction is performed to obtain a target section point set, and step S300 further includes:

[0032] Obtain design parameter information of the target area;

[0033] Based on the design parameter information, determining the height of the horizontal section;

[0034] A section thickness threshold is selected, and a horizontal section is extracted according to the height of the horizontal section to obtain the target section point set.

[0035] The target area design parameter information refers to the design parameters of the house, building, project, etc. corresponding to the target floor plan, specifically including door height, window height, opening height, etc. The target area design parameter information is used to determine the horizontal section height. The target area design parameter information can be obtained by field measurement, design document extraction, etc.

[0036] Optionally, the horizontal section is a plane located at a specific height of the target area, which is used to analyze and display the cross-sectional information at a specific height. The height of the horizontal section is determined by an expert system or professional technicians based on the design parameter information of the target area. Among them, the selection of the horizontal section is based on the principle of including as many features as possible. The height of the horizontal section includes multiple height values, corresponding to the elevations of multiple horizontal sections, and the elevation value is the relative height in the point cloud coordinate system.

[0037] Optionally, the horizontal section has a certain thickness, and the section thickness threshold refers to the set thickness value of the horizontal section, which is used to ensure the number of data points in the target section point set extracted by the horizontal section and the multiple sets of layer height point sets, thereby forming a relatively clear and complete edge, which is convenient for subsequent contour recognition based on the target section point set.

[0038] S400: Select any layer height point set in the target section point set, and obtain the target contour feature based on the straight line detection algorithm;

[0039] Optionally, the target contour feature refers to the edge feature formed by any set of layer height points. Exemplary, it includes contour shape, contour forming area, etc. The target contour feature is used for initial matching between the point cloud and the target floor plan. The target contour represents the overall feature information of the larger scale and rougher in the point cloud, which helps to quickly understand the features related to the target point cloud contour and improve the speed and efficiency of initial matching.

[0040] Optionally, before obtaining the target contour features based on the straight line detection algorithm. First, use an edge detection algorithm (such as Canny edge detection) to detect the edges of the layer height point set. Then, use a straight line detection algorithm (such as Hough transform) to detect the straight lines in the layer height point set. Obtain multiple contour straight lines, and the closed figure surrounded by the multiple contour straight lines forms the target contour image. Then, extract its related features, such as length, angle, position, etc. Obtain multiple features to form the target contour features.

[0041] Specifically, line detection refers to processing edge point sets to determine possible lines in them. Common line detection algorithms include Hough Transform and RANSAC (Random Sample Consensus) algorithms. The parameters of a line or the start and end points of a line segment can be estimated based on the distribution characteristics of the edge point set. Line detection based on Hough transform is used as an example to detect lines in a layer of high-level point sets. The principle of Hough transform is as follows:

[0042] In the image plane, a point can be represented by the horizontal and vertical coordinates; and a straight line in the image plane can be represented by the distance ρ from the origin and the angle θ of the vertical line, and thus a straight line in the image plane can be represented by a point in the θ-ρ plane. Therefore, a point in the image plane can be represented by a curve composed of the points corresponding to the family of straight lines passing through it in the θ-ρ plane. According to the above characteristics, a straight line passing through two points in the image plane corresponds to the intersection of the two curves in the θ-ρ plane, which is a straight line that is collinear with the two points in the image plane.

[0043] Specifically, first, a two-dimensional plane coordinate system is established in the layer height point set, and for each edge point, multiple first coordinates and second coordinates are obtained; then, the line equation is transformed to obtain multiple first line parameters and second line parameters, where the first line parameters and the second line parameters are expressed in polar coordinate form, that is, the distance ρ and the angle θ are used to represent the line, and multiple parameters correspond to multiple lines passing through the edge point; then, according to the first line parameters and the second line parameters, intersection analysis is performed, which means that according to multiple parameters, multiple straight line families passing through multiple edge points are converted into multiple lines in the parameter space, where the straight line families correspond to the lines in the parameter space one by one; then, an accumulator is created to count the intersections of the lines in the parameter space. For each edge point, according to its parameter value, an addition operation is performed at the corresponding position in the accumulator; finally, the parameter values ​​with higher accumulation times in the accumulator are found, and these values ​​correspond to straight lines with higher probabilities in the original image, and these values ​​with high accumulation times in the parameter space are converted back to the two-dimensional plane coordinate space to obtain the detected straight line.

[0044] S500: Performing initial matching based on the floor plan feature set and the target contour feature;

[0045] Initial matching refers to the pre-matching based on the floor plan feature set and the target contour feature, which is used to determine whether the point cloud data and the target floor plan are based on the same target floor plan. The point cloud data that has passed the adaptability judgment is quickly and roughly aligned with the target floor plan. That is, the point cloud data and the target floor plan are matched in the same coordinate system.

[0046] Further, according to the initial matching of the floor plan feature set and the target contour feature, step S500 further includes:

[0047] Performing contour completion on the target contour feature to obtain the contour area;

[0048] According to the outline area, a preliminary degree of fit with the target floor plan is calculated, and a degree of fit is determined according to a preset degree of fit threshold;

[0049] If the preliminary fitness is satisfied, an initial match is performed based on the geometric center of the target outline and the target floor plan.

[0050] Optionally, the degree of fit is determined by calculating the area of ​​a closed figure formed by the target contour features, wherein the contour area is the area of ​​a closed figure formed by the target contour features, which involves scaling and proportional conversion of the coordinate system.

[0051] Optionally, the preliminary fitness of the target floor plan is determined by calculating the ratio of the outline area to the floor plan area. Exemplarily, the calculation formula is as follows:

[0052]

[0053] Among them, α is the initial fitness, a is the scale conversion coefficient of the floor plan coordinates, S is the floor plan area, and s is the contour area. The initial fitness indicates the similarity between the point cloud data and the target floor plan.

[0054] Optionally, the preliminary fitness of the target floor plan can also be obtained by calculating the aspect ratio of the outline area to the floor plan area.

[0055] The preliminary fitness is compared with the preset fitness threshold. If the preliminary fitness meets or exceeds the preset threshold, it means that the point cloud data fits the target floor plan. An initial match can be performed based on the geometric center of the target contour and the target floor plan. The point cloud data is aligned with the target floor plan.

[0056] Furthermore, initial matching is performed based on the geometric center of the target outline and the target floor plan, and the steps further include:

[0057] Calculate and obtain the geometric center of the target outline and the geometric center of the target floor plan;

[0058] Aligning the geometric center of the target outline with the geometric center of the target floor plan, and traversing and rotating the target outline or the target floor plan to calculate the rough fit degree;

[0059] The rotation angle corresponding to the best one among the plurality of rough matching degrees of fitness is selected as the initial matching result.

[0060] Among them, the geometric center refers to a special point in a geometric figure, which can be the centroid, circumcenter, incenter or orthocenter, etc. Different geometric figures have different calculation methods to determine the position of the geometric center. For example, for a convex polygon, first, determine the vertex coordinates of the polygon. Then, use the horizontal coordinate formula and the vertical coordinate formula to calculate the arithmetic mean of the horizontal and vertical coordinates to obtain the coordinates of the geometric center point. For a non-convex polygon, it can be decomposed into convex polygons and the geometric center point of each convex polygon is calculated. Finally, the mean of all geometric center points is taken as the geometric center point of the non-convex polygon.

[0061] Optionally, align the geometric center of the target contour with the geometric center of the target floor plan. This alignment process ensures that the center positions of the two are consistent through translation. Then, traverse and rotate with the aligned geometric center as the rotation center to calculate the rough fit between the point cloud data and the target floor plan. The rough fit measures the similarity of the point cloud data with the floor plan at different rotation angles. Exemplarily, the overlap ratio of the point cloud data and the target floor plan is calculated and used as the rough fit.

[0062] Optionally, the target contour or target floor plan is traversed and rotated to calculate the rough matching degree. The rotation angle is determined based on the rough matching accuracy and timeliness requirements. Multiple rotation angles generate multiple rough matching positions and multiple rough matching degrees.

[0063] Optionally, the rotation angle corresponding to the best one among multiple rough fit degrees is selected as the initial matching result. First, a plane rectangular coordinate system is constructed, the first coordinate is the rotation angle, and the second coordinate is the rough fit degree; then, multiple rough matching positions and multiple rough fit degrees are marked as multiple data points in the coordinate system, and curve fitting is performed to generate a rough fit degree curve; then, the data point with a slope of zero and a maximum rough fit degree on the rough fit degree curve is selected as the initial matching result. Among them, the initial matching result includes the rotation angle and the translation vector of the geometric center of the target contour.

[0064] S600: Based on the target section point set, perform line-of-sight point cloud feature recognition to obtain target line-of-sight features;

[0065] The feature recognition of the through-line point cloud refers to the feature recognition of multiple groups of layer height point sets in the target section point set to obtain all the features of the target section point set. Specifically, first, feature extraction is performed on each group of layer height point sets. Feature extraction includes detecting shape, edge, curvature, normal, height, density and other attributes related to the point cloud. These features help to describe different aspects of the point cloud data. Then, the features of the groups of layer height point sets are fused to obtain all the features of the target section point set. Fusion includes summing, averaging, weighted averaging or other suitable methods of feature vectors to integrate different feature information. Then, after the feature fusion, some repeated or redundant features that exist are screened and deduplicated to ensure that each feature is unique, or the most representative features are selected. The through-line point cloud feature recognition obtains all the matching-related features in the point cloud, so that subsequent feature-based matching can be carried out.

[0066] S700: performing displacement optimization on the initial matching result according to the target visibility feature and the floor plan feature set, obtaining the best match, and generating a matching result.

[0067] Further, according to the target visibility feature and the floor plan feature set, the initial matching result is optimized by displacement to obtain the best match and generate a matching result. Step S700 also includes:

[0068] Based on the initial matching result, generating a plurality of pending matching positions under different displacement vectors;

[0069] Evaluate each of the pending matching positions and calculate a matching quality index;

[0070] According to the plurality of matching quality indicators, a position with the best matching quality is selected as the optimal match.

[0071] Optionally, multiple pending matching positions are generated under different displacement vectors. The displacement vector reflects different transformation operations such as translation and rotation to search for possible matching positions. The multiple different displacement vectors are generated by random generation after a preset displacement step size.

[0072] Each pending matching position is evaluated and the matching quality index is calculated. The matching quality index comprehensively considers the calculation results of parameter values ​​including matching error, feature point alignment, shape similarity, etc. The matching quality index is used to measure the matching quality of each pending matching position. The pending matching position with the highest matching quality index is the best match.

[0073] Furthermore, each of the pending matching positions is evaluated and a matching quality index is calculated, and the steps further include:

[0074] Based on the floor plan feature set, feature weighting is performed to obtain a feature weight value set;

[0075] Calculating the deviation of each corresponding feature in each of the pending matching positions to obtain a deviation calculation result;

[0076] The quality assessment channel is activated, and matching quality assessment is performed according to the deviation calculation result and the feature weight set to generate the matching quality index.

[0077] Optionally, feature weighting refers to setting weights for corresponding quality evaluation parameters according to the importance of different features in matching, and the weights reflect the importance of different features to the matching quality. The feature weight set reflects the corresponding relationship between multiple quality evaluation parameters and weights.

[0078] Optionally, the quality assessment channel is a mathematical model built based on a feature weight set, which is used to perform a weighted summation of the deviation values ​​of each feature to calculate a matching quality index. The matching quality index reflects the matching quality of the point cloud data and the target floor plan given a pending matching position. Matching quality assessment can be used to more accurately evaluate the quality of each pending matching position, so that the best matching position can be selected to ensure accurate alignment of the point cloud data with the target floor plan.

[0079] Furthermore, the automatic matching of point cloud and floor plan based on the viewpoint cloud feature of the present application also includes:

[0080] According to the matching results, point cloud processing is performed to clean up data points with large deviations in the point cloud;

[0081] Based on 3D visualization technology, point cloud visualization is performed according to the point cloud processing results.

[0082] Optionally, data points with large deviations in the point cloud may be collection error points or noise points in the acquisition of target point cloud data. Specifically, by setting a data point deviation threshold and making judgments and cleaning based on the deviation threshold and the deviation value of the data point, the point cloud can be denoised and smoothed while retaining the original information of the point cloud to the maximum extent, thereby ensuring the visualization presentation effect of the point cloud.

[0083] Based on the point cloud processing results, the point cloud data is visualized using 3D visualization technology. This helps users better understand and analyze the point cloud data and the matching between it and the target floor plan. 3D visualization is achieved through 3D modeling software or point cloud visualization tools.

[0084] Optionally, the 3D visualization includes coloring the point cloud according to the deviation value between the data points in the point cloud and the target floor plan, providing an intuitive display of the matching results.

[0085] In summary, the method for automatically matching point clouds and floor plans based on viewpoint cloud features provided by the present invention has the following technical effects:

[0086] The target point cloud data is acquired through the interactive point cloud acquisition device; the target floor plan is acquired to obtain the floor plan feature set; based on the target point cloud data, the horizontal section is extracted to obtain the target section point set, which includes multiple sets of floor height point sets; any set of floor height point sets in the target section point set is selected, and the target contour features are acquired based on the straight line detection algorithm; initial matching is performed based on the floor plan feature set and the target contour features; based on the target section point set, the line of sight point cloud feature recognition is performed to obtain the target line of sight features; based on the target line of sight features and the floor plan feature set, the initial matching results are optimized by displacement, the optimal match is acquired, and the matching results are generated. Thus, the technical effect of automatic matching and improving the matching calculation efficiency is achieved.

[0087] Embodiment 2

[0088] Based on the same concept as the method for automatically matching point cloud and floor plan based on the feature of the point cloud in the embodiment, Figure 3 As shown, the present application also provides a point cloud and floor plan automatic matching system based on the viewpoint cloud feature, the system comprising:

[0089] The data acquisition module 11 is used to interact with the point cloud acquisition device to acquire target point cloud data;

[0090] An image feature extraction module 12 is used to collect a target floor plan and obtain a floor plan feature set;

[0091] A horizontal section extraction module 13 is used to extract horizontal sections based on the target point cloud data to obtain a target section point set, wherein the target section point set includes multiple groups of layer height point sets;

[0092] The contour detection module 14 is used to select any set of layer height points in the target section point set, and obtain the target contour features based on the straight line detection algorithm;

[0093] An initial matching module 15, configured to perform an initial matching between the floor plan feature set and the target contour feature;

[0094] A visual feature recognition module 16 is used to perform visual point cloud feature recognition based on the target section point set to obtain the target visual feature;

[0095] The optimal matching module 17 is used to perform displacement optimization on the initial matching result according to the target visibility feature and the floor plan feature set, obtain the best match, and generate a matching result.

[0096] Furthermore, the horizontal section extraction module 13 also includes:

[0097] A parameter acquisition unit, used to acquire design parameter information of a target area;

[0098] A height setting unit, used to determine the height of the horizontal section based on the design parameter information;

[0099] The thickness setting unit is used to select a section thickness threshold, extract the horizontal section according to the height of the horizontal section, and obtain the target section point set.

[0100] Furthermore, the initial matching module 15 further includes:

[0101] A completion unit, used to complete the contour of the target contour feature to obtain the contour area;

[0102] A matching determination unit, used to calculate the initial matching degree with the target floor plan according to the outline area, and to make a matching degree determination according to a preset matching degree threshold;

[0103] The center matching unit is used to perform initial matching based on the geometric center of the target outline and the target floor plan if the preliminary fitness is satisfied.

[0104] Furthermore, the central matching unit also includes:

[0105] A center acquisition unit, used for calculating and acquiring the geometric center of the target outline and the geometric center of the target floor plan;

[0106] A rotation adaptation unit, used for aligning the geometric center of the target outline with the geometric center of the target floor plan, and traversing and rotating the target outline or the target floor plan to calculate a rough matching degree;

[0107] The optimal selection unit is used to select the rotation angle corresponding to the optimal one among the plurality of rough matching fitnesses as the initial matching result.

[0108] Furthermore, the optimal matching module 17 also includes:

[0109] An offset position setting unit, used to generate a plurality of pending matching positions under different displacement vectors based on the initial matching result;

[0110] A quality evaluation unit, used to evaluate each of the pending matching positions and calculate a matching quality index;

[0111] The optimal matching unit is used to select a position with the best matching quality as the optimal match according to the multiple matching quality indicators.

[0112] Furthermore, the quality assessment unit also includes:

[0113] A unit, configured to perform feature weighting based on the floor plan feature set to obtain a feature weight value set;

[0114] A unit, used to calculate the deviation degree of each corresponding feature in each of the pending matching positions, and obtain a deviation calculation result;

[0115] The unit is used to activate the quality assessment channel, perform matching quality assessment according to the deviation calculation result and the feature weight set, and generate the matching quality index.

[0116] Furthermore, the system also includes:

[0117] A unit, used for performing point cloud processing according to the matching result, and cleaning data points with large deviations in the point cloud;

[0118] The unit is used to perform point cloud visualization based on the point cloud processing results based on the 3D visualization technology.

[0119] It should be understood that the embodiments mentioned in this specification focus on their differences from other embodiments. The specific embodiments in the aforementioned embodiment one are also applicable to the point cloud and floor plan automatic matching system based on line-of-sight point cloud features described in embodiment two. For the sake of brevity of the specification, they will not be further elaborated here.

[0120] It should be understood that the embodiments disclosed in this application and the above description can enable those skilled in the art to use this application to implement this application. At the same time, this application is not limited to the above-mentioned embodiments, and obvious modifications and variations of the embodiments mentioned in this application also fall within the scope of the principles of this application.

Claims

1. The point cloud and floor plan automatic matching method based on the viewpoint cloud feature is characterized by: The method comprises: Interactive point cloud acquisition equipment to obtain target point cloud data; Collect the target floor plan and obtain the floor plan feature set; Based on the target point cloud data, horizontal section extraction is performed to obtain a target section point set, wherein the target section point set includes multiple groups of layer height point sets; Select any set of layer height points in the target section point set, and obtain the target contour features based on the straight line detection algorithm; Performing an initial match between the floor plan feature set and the target contour feature; Based on the target section point set, performing line-of-sight point cloud feature recognition to obtain target line-of-sight features; According to the target visibility feature and the floor plan feature set, the initial matching result is optimized by displacement to obtain the best match and generate a matching result.

2. The method according to claim 1, characterized in that Based on the target point cloud data, horizontal section extraction is performed to obtain a target section point set, including: Obtain design parameter information of the target area; Based on the design parameter information, determining the height of the horizontal section; A section thickness threshold is selected, and a horizontal section is extracted according to the height of the horizontal section to obtain the target section point set.

3. The method according to claim 2, characterized in that Performing an initial match between the floor plan feature set and the target contour feature, including Performing contour completion on the target contour feature to obtain the contour area; According to the outline area, a preliminary degree of fit with the target floor plan is calculated, and a degree of fit is determined according to a preset degree of fit threshold; If the preliminary fitness is satisfied, an initial match is performed based on the geometric center of the target outline and the target floor plan.

4. The method according to claim 3, characterized in that Based on the geometric center of the target outline and the target floor plan, an initial match is performed, including: Calculate and obtain the geometric center of the target outline and the geometric center of the target floor plan; Aligning the geometric center of the target outline with the geometric center of the target floor plan, and traversing and rotating the target outline or the target floor plan to calculate the rough fit degree; The rotation angle corresponding to the best one among the plurality of rough matching degrees of fitness is selected as the initial matching result.

5. The method according to claim 4, characterized in that According to the target visibility feature and the floor plan feature set, the initial matching result is optimized by displacement to obtain the best match and generate a matching result, including: Based on the initial matching result, generating a plurality of pending matching positions under different displacement vectors; Evaluate each of the pending matching positions and calculate a matching quality index; According to the plurality of matching quality indicators, a position with the best matching quality is selected as the optimal match.

6. The method according to claim 5, characterized in that Each of the pending matching positions is evaluated and a matching quality index is calculated, including: Based on the floor plan feature set, feature weighting is performed to obtain a feature weight value set; Calculating the deviation of each corresponding feature in each of the pending matching positions to obtain a deviation calculation result; The quality assessment channel is activated, and matching quality assessment is performed according to the deviation calculation result and the feature weight set to generate the matching quality index.

7. The method according to claim 1, characterized in that The method further comprises: According to the matching results, point cloud processing is performed to clean up data points with large deviations in the point cloud; Based on 3D visualization technology, point cloud visualization is performed according to the point cloud processing results.

8. The point cloud and floor plan automatic matching system based on the point cloud feature of the viewpoint is characterized by: The system comprises: A data acquisition module, wherein the data acquisition module is used to interact with the point cloud acquisition device to acquire target point cloud data; An image feature extraction module, wherein the image feature extraction module is used to collect a target floor plan and obtain a floor plan feature set; A horizontal section extraction module, the horizontal section extraction module is used to extract a horizontal section based on the target point cloud data to obtain a target section point set, wherein the target section point set includes a plurality of layer height point sets; A contour detection module, the contour detection module is used to select any set of layer height points in the target section point set, and obtain the target contour features based on a straight line detection algorithm; An initial matching module, the initial matching module is used to perform an initial matching between the floor plan feature set and the target contour feature; A visual feature recognition module, wherein the visual feature recognition module is used to perform visual point cloud feature recognition based on the target section point set to obtain the target visual feature; The optimal matching module is used to perform displacement optimization on the initial matching result according to the target visibility feature and the floor plan feature set, obtain the best match, and generate a matching result.

Images