Method and System for Building Surveying and Mapping

Abstract

This application relates to the field of measurement technology, specifically to a method and system for building surveying and mapping. The method includes: collecting point cloud data of a target building at each preset measurement station location; obtaining the normal consistency coefficient, distance eigenvalue, and distance significance coefficient of each point cloud data to perform denoising processing on the collected point cloud data; presetting each overlapping area and obtaining the building dispersion coefficient of each overlapping area; obtaining each point cloud set of all point cloud data within each overlapping area, and by comprehensively analyzing the distribution of the curvature and distance significance coefficient of all point cloud data in each point cloud set of each overlapping area, and combining the building dispersion coefficient, obtaining the distribution coefficient of each overlapping area; and then combining the point cloud registration algorithm to perform registration on the point cloud data to obtain the complete point cloud data of the target building and obtain the building surveying and mapping model of the target building. The purpose of this application is to improve the accuracy of building surveying and mapping by removing noise points and improving the registration accuracy.

Description

Technical Field

[0001] This application relates to the field of measurement technology, and particularly to a method and system for building surveying and mapping. Background Art

[0002] Building surveying and mapping refers to the precise measurement and recording of information such as the shape, size, position, and structure of buildings using surveying instruments and technical means. Three-dimensional laser surveying and mapping technology plays an increasingly important role in building surveying and mapping due to its advantages of high precision, high efficiency, and non-contact measurement.

[0003] The building surveying and mapping process generally includes data acquisition, data preprocessing, and model construction. Among them, the data preprocessing link is crucial for the efficiency and accuracy of the surveying and mapping results. Due to the influence of the accuracy of the acquisition equipment, external environmental interference, and the reflection characteristics of the building surface, there is noise in the acquired point cloud data. In addition, the richness of details and the complex structure characteristics of the building surface will also have a significant impact on the processing of the point cloud data. However, when the existing technology processes the point cloud data, only conventional methods are used, and the above interference factors are not fully analyzed, resulting in difficulty in obtaining point cloud data with higher precision, which is not conducive to subsequent model construction. Summary of the Invention

[0004] In view of the above, it is necessary to provide a method and system for building surveying and mapping. Compared with the traditional building surveying and mapping method, by removing noise points and improving the registration accuracy, the accuracy of building surveying and mapping is improved:

[0005] In a first aspect, an embodiment of the present application provides a method for building surveying and mapping, and the method includes the following steps:

[0006] Collect point cloud data of a target building at each preset survey station position;

[0007] Obtain the nearest neighbor point cloud data of each point cloud data collected at each of the preset survey station positions. By analyzing the distribution density of each point cloud data and all its nearest neighbor point cloud data, perform preliminary denoising on each point cloud data; by analyzing the direction difference of the normal vectors between each point cloud data and its nearest neighbor point cloud data, obtain the normal consistency coefficient of each point cloud data; by analyzing the spatial distribution of each point cloud data and all its nearest neighbor point cloud data, obtain the distance eigenvalue of each point cloud data, and combine the distance eigenvalue to obtain the distance significance coefficient of each point cloud data; use the distance significance coefficient to perform denoising processing on the point cloud data collected at each of the preset survey station positions respectively;

[0008] Preset each overlapping area, and obtain the building discrete coefficient of each overlapping area by analyzing the curvature of each location on the surface of the target building corresponding to each overlapping area and the discrete degree of the distance significant coefficient of all point cloud data in each overlapping area; obtain each point cloud set of all point cloud data in each overlapping area by using a point cloud segmentation algorithm, and obtain the distribution coefficient of each overlapping area by comprehensively analyzing the distribution of the curvature and distance significant coefficient of all point cloud data in each point cloud set of each overlapping area, combined with the building discrete coefficient; through the distribution coefficient, in combination with a point cloud registration algorithm, register the point cloud data collected at all the preset measuring station positions to obtain the complete point cloud data of the target building;

[0009] An architectural surveying and mapping model of the target building is obtained through the complete point cloud data.

[0010] In one embodiment, the process of obtaining the normal consistency coefficient is:

[0011] Obtaining the fitting plane of all neighboring point cloud data of each point cloud data, and using the normal vector of the fitting plane as the normal vector of each point cloud data;

[0012] The mean of the cosine similarity of the normal vectors between each point cloud data and all its neighboring point cloud data is taken as the normal consistency coefficient of each point cloud data.

[0013] In one embodiment, the process of obtaining the distance feature value is:

[0014] The distance between each point cloud data and the fitting plane is recorded as a first distance, and the distance between each neighboring point cloud data of each point cloud data and the fitting plane is recorded as a second distance;

[0015] The ratio of the first distance to the second distance is calculated, and the sum of all the ratios corresponding to each point cloud data is used as the distance feature value of each point cloud data.

[0016] In one embodiment, the process of obtaining the distance significance coefficient is:

[0017] Mapping the normal consistency coefficient to a positive number;

[0018] The distance significance coefficient is the ratio of the distance characteristic value to the positive number.

[0019] In one embodiment, the method of performing denoising processing on the point cloud data collected at each of the preset measuring station positions is:

[0020] All point cloud data collected at any of the preset measuring station positions are arranged in descending order according to the distance significance coefficient; and point cloud data of a preset proportion are taken as noise points and removed from the point cloud data.

[0021] In one of the embodiments, the process of obtaining the building dispersion coefficient is as follows:

[0022] Denote the dispersion degree of the curvatures of all the point cloud data in each overlapping area as the first dispersion degree, denote the dispersion degree of the distance significant coefficients of all the point cloud data in each overlapping area as the second dispersion degree, and take the sum of the first dispersion degree and the second dispersion degree as the building dispersion coefficient of each overlapping area.

[0023] In one of the embodiments, the process of obtaining the distribution coefficient is as follows:

[0024] Combine the curvature and the distance significant coefficient of each point cloud data in each overlapping area to form a two-dimensional feature data point of each point cloud data;

[0025] Obtain the dispersion coefficient of each point cloud set by analyzing the dispersion degree of the two-dimensional feature data points of all the point cloud data in each point cloud set;

[0026] Take the fusion result of the mean value of the dispersion coefficients of all the point cloud sets in each overlapping area and the building dispersion coefficient as the distribution coefficient of each overlapping area.

[0027] In one of the embodiments, the process of obtaining the dispersion coefficient is as follows:

[0028] Denote the average value of the curvatures of all the point cloud data in each point cloud set as the first average value, and denote the average value of the distance significant coefficients of all the point cloud data in each point cloud set as the second average value;

[0029] Combine the first average value and the second average value to form a two-dimensional central data point, and take the dispersion degree of the distance between the two-dimensional feature data points of all the point cloud data in each point cloud set and the two-dimensional central data point as the dispersion coefficient of each point cloud set.

[0030] In one of the embodiments, during the process of registering the point cloud data collected at all the preset measuring station positions, denote the overlapping area between any two adjacent measuring station positions as the overlapping area to be analyzed, and take the product of the normalized value of the distribution coefficient of the overlapping area to be analyzed and a preset value as the number of sampling points when the point cloud matching algorithm registers the point cloud data collected at the two adjacent measuring station positions.

[0031] In a second aspect, the embodiments of the present application further provide a system for building surveying and mapping, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of the method for building surveying and mapping described in any one of the above are implemented.

[0032] The present application has at least the following beneficial effects:

[0033] By analyzing the characteristics of the noise points generated during the acquisition of point cloud data due to the accuracy limitations of the 3D laser scanner itself, the interference of the surrounding environment of the target building, and the reflection characteristics of the laser scanning on the surface material of the target building, and based on the differences between the normal vector directions of the point cloud data and the distribution of the point cloud data in space, the distance significance coefficient is obtained to reflect the possibility that the point cloud data is a noise point. Analyzing from multiple perspectives improves the denoising accuracy of the point cloud data, which is beneficial to improving the accuracy and reliability of feature point extraction in the subsequent point cloud registration process;

[0034] Furthermore, when performing point cloud registration, considering the influence of the number of sampling points in the point cloud registration algorithm on the registration efficiency and accuracy, by analyzing the curvature of the surface of the target building corresponding to the overlapping area and the distribution of the distance significance coefficients of the point cloud data in the overlapping area, the richness of details and the complex structure characteristics of the surface of the target building corresponding to the overlapping area are evaluated, and then an appropriate number of sampling points is selected according to the actual situation, which can improve the registration accuracy while ensuring the registration efficiency; constructing a building surveying and mapping model for the point cloud data after denoising and registration processing improves the accuracy of building surveying and mapping. Description of the Drawings

[0035] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required to be used in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0036] Figure 1 It is a flowchart of the steps of a method for building surveying and mapping provided by an embodiment of the present application;

[0037] Figure 2 It is a schematic diagram of the acquisition process of the distribution coefficient;

[0038] Figure 3 It is a schematic diagram of the acquisition process of the number of sampling points. Detailed Embodiments

[0039] In the description of the embodiments of the present application, words such as "exemplary", "or", and "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design solution described as "exemplary" or "for example" in the embodiments of the present application should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Exactly speaking, using words such as "exemplary", "or", and "for example" aims to present relevant concepts in a specific manner.

[0040] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this application belongs. The terms used in this application are only for the purpose of describing specific embodiments and are not intended to limit this application. It should be understood that unless otherwise specified in this application, " / " means "or".

[0041] In addition, it should be noted that the terms "first" and "second" in this application are used to distinguish similar objects and are not used to describe a specific order or sequence.

[0042] The following specifically describes the specific solutions of the method and system for building surveying and mapping provided by this application in conjunction with the accompanying drawings.

[0043] Please refer to Figure 1 , which shows a flowchart of the steps of the method for building surveying and mapping provided by an embodiment of this application. The method includes the following steps:

[0044] Step 1, collect point cloud data of the target building at each preset measuring station position.

[0045] In this embodiment, a teaching building is used as the target building for building surveying and mapping. Technical personnel use a total station and a digital camera to conduct on-site surveys to understand the relevant information of the target building and its surrounding areas, so as to master the shape of the target building and the possible shielding conditions around it. Further determine the number of measuring stations required during the three-dimensional scanning process and the specific positions of the measuring stations.

[0046] In this embodiment, the number of measuring stations is 15, and the number and specific positions of the measuring stations need to be determined according to the actual situation.

[0047] Install the three-dimensional laser scanner at the measuring station position. First, set the scanning mode to panoramic scanning, and the corresponding resolution is 12 mm. The purpose is to determine the scope and orientation of the target building, and then adjust the resolution to 2 mm for fine scanning to obtain the point cloud data of the target building.

[0048] In this embodiment, a Riegl VZ-1000 type three-dimensional laser scanner is used to collect point cloud data. The emission frequency of the Riegl VZ-1000 type three-dimensional laser scanner can reach 300,000 points per second, and the maximum scanning range can reach 1400 meters. It supports two working modes: vertical scanning and horizontal scanning, which helps to obtain more accurate point cloud data.

[0049] After a round of scanning by the 3D laser scanner, the collected point cloud data is inspected for quality. If missing or significantly abnormal point cloud data is found in a local area, supplementary measurement is performed on the corresponding area to improve the integrity and usability of the point cloud data. Scanning continues at the next measurement station position until data collection is completed at all measurement station positions. Point cloud data is collected for a partial area of the building at each measurement station position, and there needs to be an overlapping area between the building areas collected at adjacent measurement station positions for subsequent registration of the point cloud data.

[0050] In this embodiment, the size of the overlapping area between two adjacent measurement station positions is set to 30% of the minimum value of the scanning area areas of the two adjacent measurement station positions. Here, 30% is only one embodiment of this application, and the implementer can set its specific value by themselves.

[0051] Step 2: Obtain the nearest neighbor point cloud data of each point cloud data collected at each of the preset measurement station positions. By analyzing the distribution density of each point cloud data and all its nearest neighbor point cloud data, preliminary denoising is performed on each point cloud data. By analyzing the direction difference of the normal vectors between each point cloud data and its nearest neighbor point cloud data, the normal consistency coefficient of each point cloud data is obtained. By analyzing the spatial distribution of each point cloud data and all its nearest neighbor point cloud data, the distance eigenvalue of each point cloud data is obtained. Combining the distance eigenvalue, the distance significance coefficient of each point cloud data is obtained. Through the distance significance coefficient, denoising processing is respectively performed on the point cloud data collected at each of the preset measurement station positions.

[0052] After the collection of point cloud data is completed at all measurement station positions, a series of processing steps are required to improve the integrity and accuracy of the point cloud data and ensure the integrity and reliability of the subsequent constructed model.

[0053] During the collection process of the point cloud data, due to the accuracy limitation of the 3D laser scanner itself, the interference of the surrounding environment of the target building, and the influence of the reflection characteristics of the laser scanning on the surface material of the target building, noise inevitably exists in the collected point cloud data, thus affecting the accuracy of the surveying and mapping results. Therefore, it is necessary to perform denoising processing on the collected point cloud data. Taking the point cloud data collected at the f-th measurement station position as an example.

[0054] The point cloud noise generated due to insufficient accuracy of the 3D laser scanner itself or the interference of water vapor and floating objects in the air appears as sparse and scattered isolated points, which are significantly far from the main body of the target building in terms of spatial position, and the point cloud density around them is low. Therefore, in this embodiment, a denoising algorithm based on spatial distribution (Statistical OutlierRemoval, SOR) is used to perform preliminary denoising processing on the point cloud data. Among them, the SOR algorithm is a well-known technology and will not be elaborated in this application.

[0055] In addition, due to the influence of the reflection characteristics of the surface material of the target building on laser scanning, point cloud noise is usually generated, and relatively dense and randomly distributed noise point clusters are usually formed near the surface of the target building. Depending on the degree of diffuse reflection interference, the noise point clusters may overlap with the point cloud data of the target building within a certain range. However, the distribution of such noise points is relatively random and irregular compared to the point cloud data of the target building. Since the distribution of the point cloud data of the target building is relatively regular and the normal vector directions of the point cloud data of the target building are also relatively unified, while the normal vector directions of the noise points are random, there is a large difference between the normal vectors of the noise points and the normal vectors of the point cloud data of the target building, and there is a certain distance between the noise points and the surface of the target building.

[0056] Based on the above analysis, the K-nearest neighbor algorithm is used to obtain the nearest neighbor point cloud data of any point cloud data, the fitting plane of all the nearest neighbor point cloud data of the any point cloud data is obtained, the normal vector of the fitting plane is used as the normal vector of the any point cloud data, and according to the same acquisition method as the normal vector of the any point cloud data, the normal vectors of the nearest neighbor point cloud data of the any point cloud data are obtained. The cosine similarity between the normal vectors of the any point cloud data and its nearest neighbor point cloud data is calculated respectively, and the mean value of all the obtained cosine similarities is used as the normal vector consistency coefficient of the any point cloud data. The smaller the value of the normal vector consistency coefficient, the greater the direction difference between the normal vector of the any point cloud data and the normal vectors of the other nearest neighbor point cloud data. Among them, the calculation of the K-nearest neighbor algorithm and the normal vector of the fitting plane are both well-known technologies and will not be elaborated in this application.

[0057] In this embodiment, when the number of nearest neighbor point cloud data is too small, the estimation of the normal vector is easily affected by noise, resulting in inaccurate normal vector estimation. When the number of nearest neighbor point cloud data is too large, the nearest neighbor point cloud data of the any point cloud data may span different geometric feature regions, resulting in the fitting plane being affected by multiple different geometric features, thus causing inaccurate normal vector estimation. Therefore, the number of nearest neighbor point cloud data is 20. On the basis of satisfying that the value range of the number of nearest neighbor point cloud data is [20, 30], the implementer can set the value of the number of nearest neighbor point cloud data by himself.

[0058] In this embodiment, the least squares method is used to obtain the fitting plane.

[0059] Furthermore, since there is a certain distance between the noise points and the surface of the target building, and the distribution of the point cloud data of the target building is more dense and numerous, the distance between the point cloud of the target building and the fitting plane is smaller than the distance between the noise points and the fitting plane. The distance between any point cloud data and the fitting plane is recorded as the first distance, and the distance between each neighboring point cloud data of the any point cloud data and the fitting plane is recorded as the second distance. The ratio of the first distance to the second distance is calculated, and the sum of all the ratios is obtained as the distance characteristic value of any point cloud data. The larger the distance characteristic value, the farther the distance between any point cloud data and the surface of the target building is.

[0060] In this embodiment, the distance between the point cloud data and the fitting plane is a vertical distance.

[0061] Furthermore, the distance significance coefficient of any point cloud data is obtained by combining the normal consistency coefficient and the distance eigenvalue of any point cloud data, and the expression is:

[0062] ; In the formula, Represents the distance significance coefficient of the i-th point cloud data; Represents the distance feature value of the i-th point cloud data; Represents the normal consistency coefficient of the i-th point cloud data; exp( ) represents an exponential function with a natural constant as the base, which is used to avoid the denominator being 0.

[0063] It should be noted that: the farther the distance between the i-th point cloud data and the surface of the target building is, and the greater the direction difference between the normal vectors of the i-th point cloud data and its neighboring point cloud data is, the more likely the i-th point cloud data is to be a noise point.

[0064] All point cloud data collected at the f-th measuring station are arranged in descending order according to the distance significance coefficient, and the point cloud data of the preset proportion are taken as noise points and removed from all point cloud data collected at the f-th measuring station.

[0065] In this embodiment, the value of the preset ratio is 1%, and the value of the preset ratio is preset by humans and can be set by the implementer. This application does not impose any special restrictions. For example, when the number of point clouds collected at the fth measuring station position is 10,000, the first 100 point cloud data are used as noise points.

[0066] According to the method of eliminating noise points in all point cloud data collected at the f-th measuring station position, noise points in all point cloud data collected at the remaining measuring station positions are eliminated.

[0067] Step 3: Preset each overlapping area. By analyzing the bending degree of each point on the surface of the target building corresponding to each overlapping area and the dispersion degree of the distance significant coefficients of all the point cloud data within each overlapping area, obtain the building dispersion coefficient of each overlapping area. Use the point cloud segmentation algorithm to obtain each point cloud set of all the point cloud data within each overlapping area. By comprehensively analyzing the distribution of the curvature and distance significant coefficients of all the point cloud data in each point cloud set of each overlapping area and combining with the building dispersion coefficient, obtain the distribution coefficient of each overlapping area. Based on the distribution coefficient and combined with the point cloud registration algorithm, register the point cloud data collected at all the preset station positions to obtain the complete point cloud data of the target building.

[0068] The noise points in the point cloud data will interfere with the accuracy of point cloud registration. Especially when the point cloud density is uneven or the density of the noise points is similar to that of the point cloud of the target building, it is easy to cause incorrect feature matching, thus affecting the alignment effect between the point clouds. By removing the noise points, the accuracy and reliability of feature point extraction in the point cloud registration process can be improved, so that the features on the surface of the target building can be accurately identified and utilized, thereby improving the accuracy of point cloud registration.

[0069] Since the point cloud data collected at each station position only contains a partial area of the entire target building, it is necessary to obtain the complete point cloud data through registration and at the same time remove some redundant data. Process and analyze the point cloud data of the overlapping area between any two adjacent station positions. This application uses the Four-Point Congruent Sets (4PCS) algorithm for point cloud registration. Among them, the number of sampling points of the 4PCS algorithm has a significant impact on the efficiency and accuracy of the registration result. When the number of sampling points is smaller, the registration efficiency is higher, but there is a defect of insufficient accuracy. On the contrary, when the number of sampling points is larger, the registration accuracy is higher, but the registration efficiency will decrease significantly.

[0070] Taking the overlapping area between the f-th and the (f + 1)-th station positions as an example, denote the overlapping area between the f-th and the (f + 1)-th station positions as the overlapping area to be analyzed. By analyzing the richness of the detailed features on the surface of the target building corresponding to the overlapping area to be analyzed and the complex features of the target building structure, optimize and adjust the number of sampling points of the 4PCS algorithm. In building surveying and mapping, the bending degree of the building surface and the difference characteristics of the normal vector directions at each point on the surface can reflect the richness of the detailed features on the building surface.

[0071] Based on the above analysis, by analyzing the degree of change in the curvature and distance significance coefficient of all point cloud data within the overlapping region, the richness of the detailed features of the surface of the target building is obtained. Among them, the magnitude of the curvature is used to reflect the degree of bending of the surface of the target building. In the region with more detailed features, the normal vector directions of the point cloud data are more different, and the distance between the point cloud data and the fitting plane is more unstable, so the distance significance coefficient is more unstable. Denote the degree of dispersion of the curvature of all point cloud data within the overlapping region to be analyzed as the first degree of dispersion, and denote the degree of dispersion of the distance significance coefficient of all point cloud data within the overlapping region to be analyzed as the second degree of dispersion. Take the sum of the first degree of dispersion and the second degree of dispersion as the building dispersion coefficient of the overlapping region to be analyzed. The larger the building dispersion coefficient, the more detailed information there is on the surface of the target building corresponding to the overlapping region to be analyzed. Among them, the calculation of the curvature is a well-known technology and will not be elaborated in this application.

[0072] In this embodiment, the calculation method for the degree of dispersion of the curvature of all point cloud data within the overlapping region is as follows: Randomly arrange the curvatures of all point cloud data within the overlapping region to form a curvature sequence, and use the Higuchi algorithm to process the curvature sequence. Take the obtained fractal dimension as the degree of dispersion of the curvature of all point cloud data within the overlapping region. Use the same calculation method as the degree of dispersion of the curvature of all point cloud data within the overlapping region to calculate the degree of dispersion of the distance significance coefficient of all point cloud data within the overlapping region. Among them, the Higuchi algorithm is a well-known technology and will not be elaborated in this application. As another implementation manner, on the basis of being able to measure the uneven distribution degree of the curvature of all point cloud data within the overlapping region and the uneven distribution degree of the distance significance coefficient of all point cloud data within the overlapping region, the implementer can use other existing technologies for measurement, such as standard deviation, variance, coefficient of variation, etc. This application does not make special restrictions.

[0073] Furthermore, obtain the complex features of the structure of the target building. Compared with directly analyzing the spatial positions of the point cloud data and the distance relationships between the point cloud data, since the degree of bending of the surface of the target building, the change situation of the normal vector directions of the point cloud data, and the distance relationship between the point cloud data and the fitting plane are deeply considered, through the differences in curvature and the differences in distance significance coefficients, the complex features of the structure of the target building can be more accurately reflected.

[0074] Furthermore, use a point cloud segmentation algorithm to segment the point cloud data of the overlapping region to be analyzed to obtain each point cloud set. The structural forms of the target building present diverse characteristics, some structures are relatively simple, and some structures have high complexity. By analyzing the structural form characteristics of the surface of the target building corresponding to different point cloud sets, obtain the complexity of the overlapping region to be analyzed.

[0075] In this embodiment, a point cloud segmentation algorithm based on region growing is used to segment the point cloud data in the overlapping region. The point cloud segmentation algorithm based on region growing is a well-known technology, which will not be elaborated in this application. Implementers can select other feasible algorithms to segment the point cloud data in the overlapping region by themselves.

[0076] The curvature and distance significance coefficient of each point cloud data are combined to form the two-dimensional feature data points of each point cloud data. In the target building structure with a relatively simple shape, the distribution of the two-dimensional feature data points of the point cloud data is relatively concentrated, while in the target building structure with a relatively complex shape, the distribution of the two-dimensional feature data points of the point cloud data is generally more discrete.

[0077] Based on the above analysis, taking any point cloud set as an example, the average value of the curvatures of all point cloud data in the any point cloud set is denoted as the first average value, the average value of the distance significance coefficients of all point cloud data in the any point cloud set is denoted as the second average value, the first average value and the second average value are combined to form two-dimensional center data points, and the distances between the two-dimensional feature data points and the two-dimensional center data points of each point cloud data in the any point cloud set are calculated respectively. The dispersion degree of the distances between the two-dimensional feature data points and the two-dimensional center data points of all point cloud data in the any point cloud set is used as the dispersion coefficient of the point cloud set. The larger the dispersion coefficient, the more complex the shape of the target building structure corresponding to the any point cloud set.

[0078] In this embodiment, the distance between the two-dimensional feature data points and the two-dimensional center data points is the Euclidean distance.

[0079] In this embodiment, the dispersion degree is the variance. As other implementation manners, when it can measure the uneven degree of the distribution of the distances between the two-dimensional feature data points and the two-dimensional center data points of all point cloud data in the any point cloud set, implementers can use other existing technologies for measurement, such as standard deviation, coefficient of variation, etc. This application does not make special restrictions.

[0080] Furthermore, the average value of the dispersion coefficients of all point cloud sets in the overlapping region to be analyzed is denoted as the complexity average value, and the fusion result of the complexity average value and the building dispersion coefficient is used as the distribution coefficient of the overlapping region to be analyzed. The larger the distribution coefficient, the more complex the shape of the target building corresponding to the overlapping region to be analyzed. The schematic diagram of the acquisition process of the distribution coefficient is as Figure 2 shown.

[0081] It should be understood that: Fusion means combining multiple independent variables together in a way that enhances the overall effect, such as an addition relationship, a multiplication relationship, etc. Implementers can make limitations according to the actual situation.

[0082] In this embodiment, the cumulative value of the complex mean and the building dispersion coefficient is used as the distribution coefficient of the overlapping area to be analyzed.

[0083] In another embodiment, the product of the complex mean and the building dispersion coefficient is used as the distribution coefficient of the overlapping area to be analyzed.

[0084] If the shape of the target building is more complex, more sampling points need to be set during the processing of the 4PCS algorithm to improve the registration accuracy. Conversely, if the structure of the target building is simpler, fewer sampling points can be set to improve the registration efficiency.

[0085] Based on the above analysis, the product of the normalized value of the distribution coefficient of the overlapping area to be analyzed and a preset value is used as the number of sampling points when the 4PCS algorithm performs registration on the point cloud data collected at the f-th and (f + 1)-th station positions. Among them, to avoid too few sampling points resulting in the inability to capture the overall structure of the point cloud, a threshold needs to be set. When the product is less than the preset threshold, the preset threshold is used as the number of sampling points when the 4PCS algorithm performs registration on the point cloud data collected at the f-th and (f + 1)-th station positions. The schematic diagram of the acquisition process of the number of sampling points is as Figure 3 shown.

[0086] In this embodiment, to avoid too many sampling points resulting in too long a running time of the 4PCS algorithm, the preset value is taken as 1000, and the preset threshold is taken as 100. The values of the preset value and the preset threshold are both preset manually, and the implementer can set them by himself / herself. This application does not make special restrictions.

[0087] According to the method for obtaining the number of sampling points when the 4PCS algorithm performs registration on the point cloud data collected at the f-th and (f + 1)-th station positions, obtain the number of sampling points when the 4PCS algorithm performs registration on the point cloud data collected at any two adjacent station positions.

[0088] Based on the number of sampling points of the 4PCS algorithm, use the 4PCS algorithm to perform point cloud registration to obtain the complete point cloud data of the target building. The 4PCS algorithm is a well-known technology, and the specific process is not described in detail in this application.

[0089] Step 4, obtain the building surveying and mapping model of the target building through the complete point cloud data.

[0090] In building surveying and mapping, the larger the building area and the higher the scanning resolution, the more point cloud data is obtained. Although point cloud registration methods can effectively improve the integrity of point cloud data, they may also cause data redundancy problems, thereby affecting the speed and efficiency of subsequent modeling. To solve the data redundancy problem, it is necessary to streamline the point cloud data. In this embodiment, the "uniform sampling" function of Geomagic Studio software is used to adjust the point cloud data by uniform thinning, specifically including: randomly removing some points from the overly dense point cloud and readjusting the point cloud spacing to ensure the uniformity of point cloud distribution, so as to achieve the purpose of data streamlining.

[0091] Based on the point cloud data after denoising, registration and streamlining, a three-dimensional building model is constructed. In this embodiment, the "wrap" function of Geomagic Studio software is used to convert the point cloud data into a polygon mesh model, and the CAD solid model is restored by spatial triangle approximation. Due to the geometric topological relationship of the building itself or occlusion reasons, there may be a situation where the point cloud data of some surfaces of the target building cannot be collected. Therefore, there may be holes in the constructed polygon mesh model, and the holes are filled by the "fill holes" command of Geomagic Studio software to obtain a complete three-dimensional building model.

[0092] Furthermore, based on the three-dimensional building model, a building elevation view is generated. The appearance and external structure information of the target building are shown in the building three-dimensional view. Since the amount of point cloud data is extremely large and it takes a long time for the drawing software to process the data, in order to improve work efficiency, it is necessary to divide the three-dimensional building model. In this embodiment, the division tool provided by Geomagic Studio software is used for division, and AutoCAD software is used to draw the building elevation view of each local three-dimensional building model obtained by division respectively, and then all the building elevation views are spliced to obtain the building surveying and mapping model.

[0093] Based on the same inventive concept as the above method, an embodiment of the present application also provides a system for building surveying and mapping, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above methods for building surveying and mapping.

[0094] In summary, in the present application, by analyzing the characteristics of noise points generated during the acquisition process of point cloud data due to the accuracy limitations of the three-dimensional laser scanner itself, the interference of the surrounding environment of the target building, and the influence of the reflection characteristics of the laser scan on the surface material of the target building, and based on the differences between the normal vector directions of the point cloud data and the distribution of the point cloud data in space, a distance significance coefficient is obtained to reflect the possibility that the point cloud data is a noise point. Analyzing from multiple perspectives improves the denoising accuracy of the point cloud data, which is beneficial to improving the accuracy and reliability of feature point extraction in the subsequent point cloud registration process;

[0095] Furthermore, when registering the point cloud data, considering the influence of the number of sampling points in the point cloud registration algorithm on the registration efficiency and accuracy, by analyzing the curvature of the surface of the target building corresponding to the overlapping area and the distribution of the distance significance coefficients of the point cloud data within the overlapping area, the richness of details and the complex structural characteristics of the surface of the target building corresponding to the overlapping area are evaluated. Then, according to the actual situation, an appropriate number of sampling points is selected, which can improve the registration accuracy while ensuring the registration efficiency; constructing a model for the point cloud data that has been denoised and registered to obtain a building surveying and mapping model improves the accuracy of building surveying and mapping.

[0096] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing the specified logical function. In some alternative implementations, the functions marked in the block may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description. Sometimes, there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. Each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0097] For those skilled in the art, it is obvious that the present application is not limited to the details of the above-described exemplary embodiments, and the present application can be implemented in other specific forms without departing from the basic features of the present application. Therefore, from any point of view, the above-described embodiments of the present application should be regarded as exemplary and non-limiting.

Claims

1. A method for architectural surveying, characterized in that The method comprises the following steps: Collect point cloud data of the target building at each preset measuring station location; Obtain each neighboring point cloud data of each point cloud data collected at each of the preset measuring station positions, and perform preliminary denoising on each point cloud data by analyzing the distribution density of each point cloud data and all of its neighboring point cloud data; obtain the normal consistency coefficient of each point cloud data by analyzing the direction difference of the normal vector between each point cloud data and its neighboring point cloud data; obtain the distance characteristic value of each point cloud data by analyzing the spatial distribution of each point cloud data and all of its neighboring point cloud data, and obtain the distance significance coefficient of each point cloud data by combining the distance characteristic value; perform denoising on the point cloud data collected at each of the preset measuring station positions respectively by using the distance significance coefficient; Preset each overlapping area, and obtain the building discrete coefficient of each overlapping area by analyzing the curvature of each location on the surface of the target building corresponding to each overlapping area and the discrete degree of the distance significant coefficient of all point cloud data in each overlapping area; obtain each point cloud set of all point cloud data in each overlapping area by using a point cloud segmentation algorithm, and obtain the distribution coefficient of each overlapping area by comprehensively analyzing the distribution of the curvature and distance significant coefficient of all point cloud data in each point cloud set of each overlapping area, combined with the building discrete coefficient; through the distribution coefficient, in combination with a point cloud registration algorithm, register the point cloud data collected at all the preset measuring station positions to obtain the complete point cloud data of the target building; An architectural surveying and mapping model of the target building is obtained through the complete point cloud data.

2. The method for building surveying according to claim 1, characterized in that: The process of obtaining the normal consistency coefficient is as follows: Obtaining the fitting plane of all neighboring point cloud data of each point cloud data, and using the normal vector of the fitting plane as the normal vector of each point cloud data; The mean of the cosine similarity of the normal vectors between each point cloud data and all its neighboring point cloud data is taken as the normal consistency coefficient of each point cloud data.

3. The method for architectural surveying as claimed in claim 2, characterized in that: The process of obtaining the distance feature value is as follows: The distance between each point cloud data and the fitting plane is recorded as a first distance, and the distance between each neighboring point cloud data of each point cloud data and the fitting plane is recorded as a second distance; The ratio of the first distance to the second distance is calculated, and the sum of all the ratios corresponding to each point cloud data is used as the distance feature value of each point cloud data.

4. The method for building surveying according to claim 1, characterized in that: The process of obtaining the distance significant coefficient is as follows: Mapping the normal consistency coefficient to a positive number; The distance significance coefficient is the ratio of the distance characteristic value to the positive number.

5. The method for building surveying according to claim 1, characterized in that: The method for performing denoising processing on the point cloud data collected at each of the preset measuring station positions is: All point cloud data collected at any of the preset measuring station positions are arranged in descending order according to the distance significance coefficient; and point cloud data of a preset proportion are taken as noise points and removed from the point cloud data.

6. The method for building surveying according to claim 1, characterized in that: The process of obtaining the building dispersion coefficient is as follows: The degree of discreteness of the curvature of all point cloud data in each overlapping area is recorded as the first discreteness, the degree of discreteness of the distance significance coefficient of all point cloud data in each overlapping area is recorded as the second discreteness, and the sum of the first discreteness and the second discreteness is taken as the building discrete coefficient of each overlapping area.

7. The method for building surveying according to claim 1, characterized in that: The process of obtaining the distribution coefficient is as follows: The curvature and distance significance coefficient of each point cloud data in each overlapping area are used to form a two-dimensional feature data point of each point cloud data; By analyzing the dispersion degree of all two-dimensional feature data points of the point cloud data in each point cloud set, the dispersion coefficient of each point cloud set is obtained; The fusion result of the dispersion coefficient mean of all point cloud sets in each overlapping area and the building dispersion coefficient is used as the distribution coefficient of each overlapping area.

8. The method for architectural surveying according to claim 7, characterized in that: The process of obtaining the dispersion coefficient is as follows: The average value of the curvature of all the point cloud data in each point cloud set is recorded as the first average value, and the average value of the distance significance coefficient of all the point cloud data in each point cloud set is recorded as the second average value; The first average value and the second average value are used to form a two-dimensional central data point, and the dispersion of the distance between the two-dimensional feature data points of all point cloud data in each point cloud set and the two-dimensional central data point is used as the dispersion coefficient of each point cloud set.

9. The method for building surveying according to claim 1, characterized in that: In the process of registering the point cloud data collected at all the preset measuring station positions, the overlapping area under any two adjacent measuring station positions is recorded as the overlapping area to be analyzed, and the product of the normalized value of the distribution coefficient of the overlapping area to be analyzed and the preset value is used as the number of sampling points when the point cloud matching algorithm registers the point cloud data collected at any two adjacent measuring station positions.

10. A system for architectural surveying, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method for building surveying according to any one of claims 1 to 9 are implemented.

Images