CN115953604A - Real estate geographic information mapping data acquisition method - Google Patents
Real estate geographic information mapping data acquisition method Download PDFInfo
- Publication number
- CN115953604A CN115953604A CN202310231199.8A CN202310231199A CN115953604A CN 115953604 A CN115953604 A CN 115953604A CN 202310231199 A CN202310231199 A CN 202310231199A CN 115953604 A CN115953604 A CN 115953604A
- Authority
- CN
- China
- Prior art keywords
- point
- data
- feature
- acquisition
- real estate
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000013507 mapping Methods 0.000 title claims abstract description 37
- 239000013598 vector Substances 0.000 claims abstract description 66
- 238000012216 screening Methods 0.000 claims abstract description 12
- 238000010586 diagram Methods 0.000 claims description 62
- 239000006185 dispersion Substances 0.000 claims description 16
- 238000013480 data collection Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 abstract description 2
- 230000002159 abnormal effect Effects 0.000 description 15
- 230000008569 process Effects 0.000 description 11
- 230000008859 change Effects 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000013450 outlier detection Methods 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 238000013519 translation Methods 0.000 description 3
- 238000005259 measurement Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to the technical field of electric digital processing, in particular to a real estate geographic information mapping data acquisition method. The method comprises the following steps: acquiring a point cloud data set of the real estate building, and screening feature points according to Thiessen polygons corresponding to data points in the acquired images, normal vectors of the data points in the acquired images and normal vectors of key points in the acquired images; obtaining a feature descriptor based on the HOG operator corresponding to each feature point in each acquisition graph, the gray value of each feature point, the Euclidean distance between each feature point and the feature point in the preset neighborhood thereof, and the structural similarity between the structural graph corresponding to each feature point and the structural graph corresponding to the feature point in the preset neighborhood of each feature point; and matching the point cloud data set of the real estate building with the standard point cloud data set based on the feature descriptor, and further determining the position information of the point cloud data of the real estate building. The invention improves the accuracy and the credibility of the real estate building geographic information mapping data acquisition.
Description
Technical Field
The invention relates to the technical field of electric digital processing, in particular to a real estate geographic information mapping data acquisition method.
Background
The bearing form of the mapping geographic information basic data is diversified, and the mapping geographic information basic data can be various types of data, such as satellite images, aerial images, various scale maps and the like, and the images can form a grid map database, a digital elevation model database, a orthographic image database and the like.
For real estate buildings, surveying and mapping data of the buildings are often acquired through three-dimensional models, the traditional real estate buildings generate corresponding three-dimensional models through professional software to collect the surveying and mapping data, an orthoimage is needed, DEM data are needed to correct the orthoimage, three-dimensional modeling is conducted through CAD data, and the implementation process is complicated. At present, a three-dimensional scanner is generally used to collect mapping data by using a point cloud technology of real estate buildings. The point cloud is a massive point set of surface features of the real estate building, the denser the point cloud is, the more the image detail information is reflected, but the preliminarily acquired building facade mapping point cloud data may contain wrong or abnormal data points, the existence of the data points will influence the matching result of the point cloud data of the real estate building and the standard point cloud data, further the acquisition result of the whole mapping data of the point cloud data of the real estate building generates errors, the accuracy of the mapping data influences the credibility of the geographic information storage result, and therefore the point cloud data in the three-dimensional model needs to be processed, and the accuracy and stability of the acquisition of the real estate geographic information mapping data are improved.
Disclosure of Invention
In order to solve the problem of low accuracy in the existing method for collecting the geographic information mapping data of the real estate building, the invention aims to provide a method for collecting the geographic information mapping data of the real estate, and the adopted technical scheme is as follows:
the invention provides a real estate geographic information mapping data acquisition method, which comprises the following steps:
acquiring a point cloud data set of a real estate building;
acquiring each acquisition image based on the spatial coordinates of each data point in the point cloud data set; obtaining the characteristic stability of each data point in the point cloud data set according to the Thiessen polygon corresponding to each data point in each acquisition image, the normal vector of the key point in the acquisition image where each data point is located and the normal vector of the acquisition image where each data point is located; screening feature points based on the feature stability;
constructing a structure chart corresponding to each feature point based on each feature point and the feature points in a preset neighborhood, and constructing a point cloud descriptor of each feature point based on a HOG operator corresponding to each feature point in each acquisition graph, a normal vector of the feature point in the acquisition graph where each feature point is located and the gray value of each feature point; obtaining a feature descriptor of each feature point according to the Euclidean distance between each feature point and the feature point in the preset neighborhood of the feature point, the point cloud descriptor of each feature point, the point cloud descriptor of the feature point in the preset neighborhood of each feature point, and the structural similarity between the structural diagram corresponding to each feature point and the structural diagram corresponding to the feature point in the preset neighborhood of each feature point;
matching the characteristic points in the point cloud data set of the real estate building with the data points in the standard point cloud data set based on the characteristic descriptors of the characteristic points in the point cloud data set of the real estate building and the characteristic descriptors of the data points in the standard point cloud data set, and determining the position information of the point cloud data of the real estate building based on the matching result.
Preferably, the obtaining each acquisition map based on the spatial coordinates of each data point in the point cloud data set includes:
the acquisition diagrams comprise a transverse acquisition diagram and a longitudinal acquisition diagram;
the original point of a space coordinate system is an O point, three coordinate axes of the coordinate system are respectively an X axis, a Y axis and a Z axis, each plane parallel to the plane YOZ is marked as a first plane, each plane parallel to the plane XOZ is marked as a second plane, the first plane of the abscissa of any data point in the point cloud data set of the real estate building is used as a transverse acquisition graph, and the second plane of the ordinate of any data point in the point cloud data set of the real estate building is used as a longitudinal acquisition graph.
Preferably, the obtaining the feature stability of each data point in the point cloud data set according to the tesson polygon corresponding to each data point in each acquisition map, the normal vector of the key point in the acquisition map where each data point is located, and the normal vector of the acquisition map where each data point is located includes:
for the a-th data point in the point cloud dataset:
according to the Thiessen polygons corresponding to the a-th data point in each acquisition map and the Thiessen polygons corresponding to the data points in the preset neighborhood of the a-th data point in each acquisition map, obtaining the transverse angle difference and the longitudinal angle difference of the a-th data point corresponding to the data points in the preset neighborhood of the a-th data point;
calculating the area difference of the Thiessen polygon corresponding to the a-th data point in the transverse acquisition map and the Thiessen polygon corresponding to each data point in the preset neighborhood of the a-th data point in the transverse acquisition map, and recording the area difference as the transverse area difference; calculating the area difference of the Thiessen polygon corresponding to the a-th data point in the longitudinal acquisition diagram and the Thiessen polygon corresponding to each data point in the preset neighborhood of the a-th data point in the longitudinal acquisition diagram, and recording the area difference as the longitudinal area difference; calculating the spatial dispersion of the a-th data point based on the transverse area difference, the longitudinal area difference, the transverse angle difference and the longitudinal angle difference, wherein the transverse area difference, the longitudinal area difference, the transverse angle difference and the longitudinal angle difference are in positive correlation with the spatial dispersion;
respectively calculating cosine similarity of a normal vector of each key point in each acquisition graph where the a-th data point is located and a normal vector of each acquisition graph where the a-th data point is located, and obtaining structural stability of the a-th data point based on the cosine similarity, wherein the cosine similarity and the structural stability are in positive correlation;
and taking the ratio of the structural stability to the spatial dispersion as the characteristic stability of the a-th data point.
Preferably, the method for acquiring the transverse angle difference between the a-th data point and each data point in the preset neighborhood comprises the following steps:
respectively calculating the absolute value of the difference value between the maximum internal angle of the a-th data point in the corresponding Thiessen polygon in the transverse acquisition diagram and the maximum internal angle of the ith data point in the corresponding Thiessen polygon in the transverse acquisition diagram, and recording the absolute value as the difference of the maximum internal angles; respectively calculating the absolute value of the difference value between the minimum internal angle of the a-th data point in the corresponding Thiessen polygon in the transverse acquisition diagram and the minimum internal angle of the ith data point in the corresponding Thiessen polygon in the transverse acquisition diagram, and recording the absolute value as the difference of the minimum internal angles; the ith data point is a data point in a preset neighborhood of the a-th data point; and recording the sum of the difference of the maximum internal angle and the difference of the minimum internal angle as the transverse angle difference corresponding to the ith data point in the preset neighborhood of the ith data point.
Preferably, the screening of feature points based on the feature stability degree includes: and acquiring the minimum value of the characteristic stability of all data points in the standard point cloud data set, taking the minimum value as a screening threshold value, and taking the data points with the characteristic stability of the point cloud data set of the real estate building greater than or equal to the screening threshold value as the characteristic points.
Preferably, the constructing a point cloud descriptor of each feature point based on the HOG operator corresponding to each feature point in each acquisition map, the normal vector of the feature point in the acquisition map where each feature point is located, and the gray value of each feature point includes:
for the qth feature point:
calculating the sum value of the HOG operator in the transverse acquisition graph where the q-th characteristic point is located and the HOG operator in the longitudinal acquisition graph where the q-th characteristic point is located;
calculating the normal vector of the q-th feature point in the transverse acquisition graph where the q-th feature point is located and the variance of the normal vectors of all feature points in the preset neighborhood of the q-th feature point, and recording the variance as a first variance; calculating the normal vector of the qth characteristic point in the longitudinal acquisition image of the qth characteristic point and the variances of the normal vectors of all the characteristic points in the preset neighborhood of the qth characteristic point, recording the variances as second variances, and taking the sum of the first variances and the second variances as the normal vector distribution variance of the qth characteristic point;
and combining the sum, the normal vector distribution variance and the gray value of the qth characteristic point together to be used as a point cloud descriptor of the qth characteristic point.
Preferably, the feature descriptor of each feature point is calculated by the following formula:
wherein ,
for the feature descriptor of the qth feature point,
the distance weight of the kth characteristic point in the preset neighborhood of the qth characteristic point is calculated,
the Euclidean distance between the qth characteristic point and the kth characteristic point in the preset neighborhood is taken as the Euclidean distance,
the number of feature points in the preset neighborhood of the qth feature point,
in order to preset the radius of the neighborhood,
is the mean value of Euclidean distances between the qth characteristic point and all the characteristic points in the preset neighborhood,
is a point cloud descriptor of the qth characteristic point,
is a point cloud descriptor of the kth characteristic point in the preset neighborhood of the qth characteristic point,
is a natureThe constant number is a constant number,
in order to preset the adjustment parameters, the adjustment parameters are set,
a structure diagram corresponding to the qth characteristic point,
a structure diagram corresponding to the kth characteristic point in the preset neighborhood of the qth characteristic point,
is a structural drawing
And structure diagram
The structural similarity of (a) to (b),
the absolute value sign is taken.
Preferably, the constructing a structure diagram corresponding to each feature point based on each feature point and the feature points in the preset neighborhood thereof includes: and respectively connecting each feature point with each feature point in a preset neighborhood by taking each feature point as a center to obtain a structure chart corresponding to each feature point.
The invention has at least the following beneficial effects:
the method comprises the steps of firstly obtaining a point cloud data set of the real estate building, determining the characteristic stability of each data point in the point cloud data set of the real estate building, wherein the characteristic stability considers the distribution condition of the data points in each acquisition image, and avoiding the influence on point cloud data matching caused by the local density reduction of adjacent point cloud data points due to scanning angles when a scanner scans an area with severe surface change of the real estate building; according to the Euclidean distance between each feature point and the feature point in the preset neighborhood, the point cloud descriptor of each feature point, the point cloud descriptor of the feature point in the preset neighborhood of each feature point, the structural diagram corresponding to each feature point and the structural diagram corresponding to the feature point in the preset neighborhood of each feature point, the feature descriptor of each feature point is constructed, the influence of the distance is considered by the feature descriptor, data points with different distances are analyzed, the problem that data points with longer distances are ignored when the descriptor is constructed in the prior art is solved, the structural similarity of the feature point and the data points in the neighborhood is also considered by the feature descriptor, accurate feature information can be obtained in a local area with a complex structural line, the influence of isolated interference points is avoided, and the matching precision between subsequent data points is improved; according to the invention, the point cloud data is matched based on the feature descriptors of the feature points in the point cloud data set of the real estate building and the feature descriptors of the data points in the standard point cloud data set, so that the position information of the point cloud data of the real estate building is further determined, the interference of abnormal points on the matching result is eliminated, and the accuracy and the reliability of the acquisition of the geographic information mapping data of the real estate building are improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flow chart of a real estate geographic information mapping data collection method according to an embodiment of the present invention;
fig. 2 is a structural diagram corresponding to the qth feature point.
Detailed Description
To further illustrate the technical means and effects of the present invention for achieving the predetermined objects, a method for collecting geographic information mapping data of real estate according to the present invention is described in detail below with reference to the accompanying drawings and preferred embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following describes a specific scheme of the real estate geographic information mapping data acquisition method provided by the present invention in detail with reference to the accompanying drawings.
The embodiment of the method for collecting the geographic information mapping data of real estate comprises the following steps:
the embodiment provides a real estate geographic information mapping data acquisition method, as shown in fig. 1, the real estate geographic information mapping data acquisition method of the embodiment includes the following steps:
s1, acquiring a point cloud data set of the real estate building.
The specific scenario addressed by the present embodiment is as follows: the method comprises the steps of obtaining a three-dimensional model of the real estate building by using a three-dimensional scanner, obtaining a corresponding point cloud data set from the three-dimensional model, matching the obtained point cloud data set with a standard point cloud data set, eliminating the influence of abnormal data points or error data points on the collection of the real estate building surveying and mapping data, and improving the collection precision of the real estate building surveying and mapping data.
The acquisition mode of point cloud data generally is through three-dimensional laser scanner acquisition, three-dimensional laser scanner can be the three-dimensional coordinate information on the acquisition real estate building surface of large tracts of land high resolution, utilize Trimble three-dimensional scanner to scan real estate building in this embodiment, at Trimble three-dimensional laser scanning's in-process, often can receive the influence of factors such as object shielding, illumination is inhomogeneous around the real estate building, there is the scanning blind spot in the region that causes complicated shape object easily, form the hole, simultaneously because the scanning measuring range is limited, usually can't once only carry out complete measurement to real estate building, only can acquire real estate building's local data in a scanning position, must change the position and carry out scanning measurement many times. In this embodiment, a plurality of scanning positions are first determined according to the appearance characteristics of the real estate building, data collected by all the scanning positions can include all data of the surface of the real estate building, a Trimble three-dimensional scanner is sequentially placed at each scanning position or a Trimble three-dimensional scanner is placed at each scanning position, a plurality of initial data with different coordinate systems are obtained, therefore, coordinate transformation is required to be performed to unify all point cloud data to the same coordinate system, the coordinate transformation can be achieved by using a rotation matrix, the process is a known technology, and redundant description is not repeated here. In order to improve the accuracy of the point cloud data and eliminate the noise effect, the point cloud data needs to be preprocessed, in this embodiment, an outlier detection method is used to preprocess the acquired point cloud data, and outlier detection is a known technology, and is not described here again.
Therefore, a point cloud data set of the real estate building can be obtained according to outlier detection and used for being matched with a standard point cloud data set subsequently.
S2, acquiring each acquisition image based on the spatial coordinates of each data point in the point cloud data set; obtaining the characteristic stability of each data point in the point cloud data set according to the Thiessen polygon corresponding to each data point in each acquisition image, the normal vector of the key point in the acquisition image where each data point is located and the normal vector of the acquisition image where each data point is located; and screening the characteristic points based on the characteristic stability.
The purpose of this embodiment is to carry out accurate matching with the point cloud data set of real estate building and standard point cloud data set, and carry out the collection of real estate building geographic information survey and drawing data through point cloud data set matching result. Because the probability of the change of the data points corresponding to the internal positions of the building is extremely small, the data points are not necessarily matched completely, so the embodiment matches the data points with certain characteristics, respectively obtains the horizontal acquisition map and the vertical acquisition map corresponding to the point cloud data set of the real estate building, respectively analyzes the horizontal acquisition map and the vertical acquisition map, obtains the characteristic stability of each data point, screens normal data points based on the characteristic stability, matches the point cloud data set of the real estate building with the standard point cloud data set, if the data points are normal point cloud data, the matching result of the normal data points is adjacent to the spatial positions or the matching result of the data points with the same dimensionality is stable, and if the data points are abnormal or wrong point cloud data, the matching results of the data points at the adjacent spatial positions or the same dimensionality are different; for example, data points located on the real estate building structure line, there are data points belonging to the same dimension as the data points, and the matching results of the data points have certain similarity. And fusing different matching results to obtain final screening point cloud data, and obtaining initial values in the ICP algorithm in a self-adaptive manner for different areas to improve the accuracy and stability of point cloud matching.
The point cloud data set of the real estate building comprises normal data points and abnormal data points, the reason for generating the abnormal data points is that the scanning angle of a scanner does not reach an ideal angle, or the data points are located at the top angle and the complex structure of the real estate building, direction changes with large differences exist among the data points in the areas, and abnormity easily occurs in the scanning process, so that before the point cloud data set of the real estate building is matched with the standard point cloud data set, characteristic point extraction needs to be carried out on the point cloud data set, and on the premise that the matching precision is not reduced, the interference of the abnormal data to a matching result is reduced.
Each data point in the point cloud data set comprises coordinate information in a three-dimensional space, when an abscissa is fixed, a plurality of data points exist on the same plane, the data points are located in the same space plane in a real estate building, an origin of a space coordinate system is an O point, three coordinate axes of the coordinate system are respectively an X axis, a Y axis and a Z axis, each plane parallel to a plane YOZ is marked as a first plane, each plane parallel to a plane XOZ is marked as a second plane, the first plane of the abscissa of any data point in the point cloud data set passing through the real estate building is used as a transverse acquisition graph, and the second plane of the ordinate of any data point in the point cloud data set passing through the real estate building is used as a longitudinal acquisition graph; thus, a plurality of horizontal acquisition maps and a plurality of vertical acquisition maps are obtained, and one horizontal acquisition map or one vertical acquisition map may contain a plurality of data points.
The real estate building comprises a large number of structure lines, and the structure lines are represented as line segments with different lengths and slopes and distributed in each collection graph according to the design requirements of the real estate building. The vertex angle of the real estate building is an area where a plurality of structure lines intersect and is also an area where local characteristics in the real estate building change greatly, for example, the normal vector of a data point on the same plane in the vertex angle area changes violently, so that a large weight can be given to a matching point pair formed by the data points in the vertex angle area of the real estate building, because the matching point pair formed by the data points and standard point cloud data is obviously different from a matching point pair corresponding to a surrounding adjacent data point, the matching point pair cannot be replaced by other data points; and local feature changes of adjacent point cloud data points on the same plane are small in local areas with flat real estate buildings, so that directions of normal vectors of the data points in the areas are almost consistent, and after matching point pairs formed by the data points are replaced by matching point pairs formed by the adjacent data points, local matching results cannot be changed remarkably. In addition, in the horizontal acquisition map or the vertical acquisition map, if a data point is located in a region with a large local change at a top angle, and the adjacent data points are also large and possibly located in the region with the large local change, the local features of the data points in the three-dimensional model corresponding to the point cloud data set of the real estate building are more remarkable. Based on this, for the a-th data point in the point cloud dataset of the real estate building: any data point in the point cloud data set is simultaneously positioned in a transverse acquisition picture and a longitudinal acquisition picture, so that a transverse acquisition picture where the a-th data point is positioned and a longitudinal acquisition picture where the a-th data point is positioned are obtained; considering that the shape of the polygon in the Thiessen polygon reflects the distribution characteristics of the adjacent data points in each acquisition surface, the smaller the area of the polygon is, the denser the distribution of the data points in the polygon is; thus is paired withObtaining a Thiessen polygon corresponding to the a-th data point in the transverse acquisition map according to the a-th data point and the data points in the preset neighborhood thereof in the transverse acquisition map where the a-th data point is positioned
Simultaneously acquiring a Thiessen polygon corresponding to each data point in a preset neighborhood of the a-th data point in the transverse acquisition map, wherein the Thiessen polygon corresponding to the ith data point in the preset neighborhood of the a-th data point in the transverse acquisition map is
(ii) a Obtaining Thiessen polygons
Obtaining the Thiessen polygon from the maximum internal angle and the minimum internal angle
The maximum internal angle and the minimum internal angle in the process of calculating the Thiessen polygon
The maximum internal angle of the middle and the Tassen polygon are
Calculating the Thiessen polygon by taking the absolute value of the difference value of the maximum internal angles as the difference of the maximum internal angles
The minimum interior angle and the Thiessen polygon of
Recording the absolute value as the difference of the minimum internal angles, and recording the sum of the difference of the maximum internal angles and the difference of the minimum internal angles as the transverse angle difference corresponding to the ith data point in the preset neighborhood of the ith data point; for the longitudinal acquisition map where the a-th data point is located, acquiring a Thiessen polygon corresponding to the a-th data point in the longitudinal acquisition map according to the a-th data point and data points in a preset neighborhood of the a-th data point
And simultaneously acquiring a Thiessen polygon corresponding to each data point in the preset neighborhood of the a-th data point in the longitudinal acquisition map, wherein the Thiessen polygon corresponding to the j-th data point in the preset neighborhood of the a-th data point in the longitudinal acquisition map is
Similarly, by analogy with the method, the longitudinal angle difference corresponding to the jth data point in the a-th data point and the preset neighborhood is obtained; respectively extracting key points from a transverse acquisition image where the a-th data point is located and a longitudinal acquisition image where the a-th data point is located by adopting an SIFT algorithm to obtain a plurality of key points in the transverse acquisition image where the a-th data point is located and a plurality of key points in the longitudinal acquisition image where the a-th data point is located; the SIFT algorithm and the thiessen polygon obtaining process are well-known technologies, and are not described in detail herein. The preset neighborhood in the embodiment is a circular area with the radius r as the radius taken by taking a data point as a central point, the radius r is the radius of the preset neighborhood, the value of the r is set to be 9 in the embodiment, and an implementer can set the radius according to specific conditions in specific application. The greater the similarity between the normal vector of the acquisition map where the data point is located and the normal vector of the key point in the acquisition map where the data point is located, the more likely the data point is to be a normal data point, and the more stable the structure. Therefore, in this embodiment, the structural stability of each data point is determined based on the cosine similarity between the normal vector of the acquired map where each data point is located and the normal vector of the key point in the acquired map where each data point is located, and the spatial dispersion of each data point is determined based on the area difference between the tesson polygon corresponding to each data point in each acquired map and each tesson polygon corresponding to each data point in the preset neighborhood of each data point in each acquired map, the transverse angle difference between each data point and each data point in the preset neighborhood of each data point, and the longitudinal angle difference between each data point and each data point in the preset neighborhood of each data point; and then according to the structural stability and the characteristic stability, constructing the characteristic stability for representing the stability of the characteristic of the data point on the corresponding acquisition graph in the point cloud data set and the characteristic stability of the a-th data pointThe specific expression is as follows:
By adopting the method, the characteristic stability of each data point in the point cloud data set of the real estate building can be obtained; in a point cloud data set of a real estate building, the more uneven the distribution of data points adjacent to a spatial position is, the more complex the area where the data points are located is, the worse the stability of the data points is, and the more inaccurate the point cloud matching result is; the area where the key point in the real estate building is located is often an area with remarkable structural features, and abnormal data points are not easy to appear in the scanning process, so that the more similar the data points are to the key points, the stronger the feature stability of the data points is.
Further, the minimum value of the feature stability of all data points in the standard point cloud data set is obtained according to the steps, and the minimum value is used as a screening threshold value
The smaller the stability of the point cloud data set of the real estate building, the more likely the data points are abnormal, so that the stability of the point cloud data set of the real estate building is smaller than that of the point cloud data set of the real estate building
Determining the data points to be abnormal data points, deleting all the abnormal data points, recording the rest data points in the point cloud data set of the real estate building as feature points, namely, recording the feature stability of the point cloud data set of the real estate building to be more than or equal to
The data points are judged as normal data points, the normal data points are marked as characteristic points, and matching is carried out based on the characteristic points. The characteristic stability degree deletes abnormal data points in the point cloud data set of the real estate building by evaluating the characteristic stability degree of the data points, so that the condition that the scanner scans the area with severe surface change of the real estate building is avoidedIn time, the local density of adjacent data points is reduced due to the scanning angle, and the influence of the coordinate error of the data points on the matching of point cloud data is increased.
Therefore, data points in the point cloud data set of the real estate building are screened, and all feature points are obtained.
S3, constructing a structure diagram corresponding to each feature point based on each feature point and the feature points in the preset neighborhood of the feature point, and constructing a point cloud descriptor of each feature point based on the HOG operator corresponding to each feature point in each acquisition diagram, the normal vector of the feature point in the acquisition diagram where each feature point is located and the gray value of each feature point; and obtaining the feature descriptors of the feature points according to the Euclidean distance between the feature points and the feature points in the preset neighborhood, the point cloud descriptors of the feature points in the preset neighborhood of the feature points, and the structural drawing corresponding to the feature points in the preset neighborhood of the feature points.
In the embodiment, feature points are screened in step S2, and then the embodiment obtains a feature descriptor of each feature point for matching calculation with a data point in a standard point cloud data set. In the matching process of the traditional ICP algorithm, features are generally calculated by utilizing a small number of data points which are close to each other in the neighborhood of the feature point, namely when the distance between the qth feature point and the data point in the preset neighborhood is far, the feature descriptor of the qth feature point is not influenced by the corresponding neighborhood data point; when the distance between the qth characteristic point and a data point in a preset neighborhood is close, the corresponding neighborhood data point has a larger influence on the characteristic descriptor of the qth characteristic point; in this way, data points with a large distance are ignored, and in a large area on the surface of the real estate building with a complicated structure, the characteristics of the data points can be influenced by the data points with a large distance, for example, an abnormal data point appears at the vertex of the corner of the carved window, so that the vertex of the corner on the same structure line can be influenced although the distance is long. Considering that the similarity of the structure chart constructed by taking the feature point as the center and the structure chart constructed by taking the neighborhood feature point as the center can reflect the consistency degree of the structure characteristics of the local area where the feature point is located; therefore, in the embodiment, in combination with the similarity between the distance weight between the feature point and the feature point in the preset neighborhood and the structure diagram, a feature descriptor of the feature point is constructed, and the feature descriptor is used for representing feature information of the feature point in the point cloud data set of the real estate building.
For the qth feature point:
respectively connecting the feature point with each feature point in the preset neighborhood by taking the feature point as a center to obtain a structure diagram
Marking as a structure diagram corresponding to the qth feature point, as shown in fig. 2, q in the diagram represents the qth feature point, and k in the diagram represents the kth feature point in the preset neighborhood of the qth feature point; similarly, by analogy with the above method, a structure diagram corresponding to each feature point in a preset neighborhood of the q-th feature point is obtained; then, respectively calculating the structural similarity of the structural diagram corresponding to the qth characteristic point and the structural diagram corresponding to each characteristic point in a preset neighborhood of the qth characteristic point, wherein the structural similarity is obtained by adopting the conventional SSIM structural similarity algorithm; respectively obtaining the Euclidean distance between the qth characteristic point and each characteristic point in a preset neighborhood; the calculation method of the euclidean distance is a well-known technique, and will not be described in detail herein. Respectively acquiring the HOG operator in the transverse acquisition diagram where the q-th feature point is located and the HOG operator in the longitudinal acquisition diagram where the q-th feature point is located, and calculating the sum value of the HOG operator in the transverse acquisition diagram where the q-th feature point is located and the HOG operator in the longitudinal acquisition diagram where the q-th feature point is located
(ii) a Acquiring a normal vector of a q-th feature point in a transverse acquisition image where the q-th feature point is located and a normal vector of each feature point in a preset neighborhood of the q-th feature point, calculating variances of the normal vector of the q-th feature point in the transverse acquisition image where the q-th feature point is located and the normal vectors of all feature points in the preset neighborhood of the q-th feature point, and marking the variances as first variances; obtaining the normal vector of the q-th feature point in the longitudinal acquisition picture of the q-th feature point and the normal vector of each feature point in the preset neighborhood of the q-th feature point, and calculatingMarking the variance of the normal vector of the q-th feature point in the longitudinal acquisition image where the q-th feature point is located and the variances of the normal vectors of all feature points in the preset neighborhood of the q-th feature point as a second variance, and marking the sum of the first variance and the second variance as the second variance
As the variance of the normal vector distribution of the qth feature point. Will be provided with
The variance of the normal vector distribution of the qth feature point
The gray value of the qth characteristic point
Combined together as point cloud descriptor for the qth feature point, i.e. will
As a point cloud descriptor of the qth feature point, the acquisition process of the HOG operator is a known technology, which is not described herein more than once, and the value of m in this embodiment is 31, that is, the value is
The operator is a 31-dimensional vector; the distribution variance and the gray value of the normal vector are one-dimensional data, so that the point cloud descriptor is a 33-dimensional vector and a feature descriptor
Also a 33-dimensional feature vector.
The reason why the feature descriptors are constructed in this embodiment is that in the ICP matching process, the closest point corresponding to the q-th feature point is found in the standard point cloud set, and the closest point and the feature points certainly satisfy the similarity degree of some features, so that the features of the q-th feature point in the point cloud data set of the real estate building can be characterized by using the normal vector, the gray value and the structural similarity of the q-th feature point. The feature descriptor of the qth feature point is specifically:
wherein ,
for the feature descriptor of the qth feature point,
the distance weight of the kth characteristic point in the preset neighborhood of the qth characteristic point is calculated,
the Euclidean distance between the qth characteristic point and the kth characteristic point in the preset neighborhood is taken as the Euclidean distance,
the number of feature points in the preset neighborhood of the qth feature point,
in order to preset the radius of the neighborhood,
is the mean value of Euclidean distances between the qth characteristic point and all the characteristic points in the preset neighborhood,
is a point cloud descriptor of the qth characteristic point,
is a point cloud descriptor of the kth characteristic point in the preset neighborhood of the qth characteristic point,
is a natural constant and is a natural constant,
in order to preset the adjustment parameters, the adjustment parameters are set,
is the qth feature pointA corresponding structure diagram is shown in the figure,
a structure diagram corresponding to the kth characteristic point in the preset neighborhood of the qth characteristic point,
is a structural drawing
And structure diagram
The structural similarity of (a) to (b),
the absolute value sign is taken.
The preset tuning parameter is introduced to prevent the denominator from being 0, and in the present embodiment, the preset tuning parameter is set
The value of (b) is 0.01, which can be set by the practitioner as the case may be in a particular application.
The structural significance of the qth characteristic point is represented, the significance of the structure of the qth characteristic point and the adjacent data point in the local area of the three-dimensional model of the real estate building is represented in the point cloud data set of the real estate building,
the larger the illustration is of the structure
And structure diagram
The more similar, the higher the structural feature consistency of the local area where the qth feature point is located, the more prominent the features in the three-dimensional model are.
The closer the distance between a certain feature point and the feature point in the neighborhood, the greater the influence of the corresponding neighborhood feature point on the feature point, and the closer the distance isThe smaller the influence on the feature point is; the structural similarity between the feature points and the feature points in the neighborhood is larger, the more stable the region where the feature points are located is, the more remarkable the features are, and the more accurate the matching result is. The feature descriptors are feature vectors formed by feature points carrying image information in the point cloud data sets, and the influence of feature points which are closer to the q-th feature point feature descriptors in the neighborhood on the q-th feature point feature descriptors is larger, namely the feature descriptors are
The larger the size of the tube is,
the larger; the structural similarity of the corresponding structure chart of the q-th characteristic point and the neighborhood characteristic point is larger, the degree of significance of the characteristic of the local area where the q-th characteristic point is located is higher, namely
The larger the point cloud data set, the higher the structural feature consistency of the local area where the q-th feature point is located, and the more prominent the point cloud data set is in the real estate building. The characteristic descriptor has the advantages that a distance weight which cannot be ignored is obtained for characteristic points with different distances, and the problem that long-distance data points are ignored when the descriptor is constructed by the traditional ICP is solved; the feature descriptor considers the structural similarity of the structural graph corresponding to the feature points and the structural graph corresponding to the feature points in the neighborhood, accurate feature information can be obtained in a local area with a complex structural line, the influence of isolated interference points is avoided, and the accuracy of a subsequent matching result is improved.
Thus, by adopting the method, the feature descriptor of each feature point can be obtained.
And S4, matching the characteristic points in the point cloud data set of the real estate building with the data points in the standard point cloud data set based on the characteristic descriptors of the characteristic points in the point cloud data set of the real estate building and the characteristic descriptors of the data points in the standard point cloud data set, and determining the position information of the point cloud data of the real estate building based on the matching result.
In this embodiment, a feature descriptor of each feature point is obtained in step S3, the point cloud data set of the real estate building is used as an input of a matching algorithm, the point cloud data set of the real estate building is matched with the standard point cloud data set, and the position information of the point cloud data of the real estate building is determined based on a matching result.
The improved matching process in this embodiment is as follows: inputting a point cloud data set and a standard point cloud data set of a real estate building, respectively extracting a feature descriptor of a feature point in the point cloud data set of the real estate building and a feature descriptor of a data point in the standard point cloud data set, forming a matching point pair by the most similar data points of the feature descriptors, measuring the similarity of the feature descriptors of the matching point pair through a target function of the matching point pair, wherein the target function specifically comprises the following steps:
wherein ,
in order to be the objective function of the target,
the number of feature points in the point cloud dataset for a real estate building,
is the number of data points in the standard point cloud data set, y is the y-th characteristic point in the point cloud data set of the real estate building,
as the first in a standard point cloud data set
The number of data points is, for example,
a feature descriptor for the y-th feature point in the point cloud dataset for the real estate building,
first in a point cloud data set for real estate buildings
A descriptor of the characteristics of a data point,
feature descriptor for the y-th feature point and the
Similarity of feature descriptors for several data points. In this embodiment, the similarity between the two feature descriptors is represented by cosine similarity, and the process of calculating cosine similarity is a known technique and is not described in detail.
Rotation matrix calculation using SVD
And translation vector
Concentrating point cloud data of real estate building according to rotation matrix
And translation vector
Performing transformation, namely matching the transformed point cloud data set with a standard point cloud data set, and continuously iterating until an iteration stop condition is met, wherein the iteration stop condition is set to be that the iteration frequency reaches 50 times, and in specific application, an implementer can set the iteration stop condition according to specific conditions; after iteration stops, the rotation matrix is output
And translation vector
. The matching effect of the matching point pair is optimal when the target function takes the maximum value, the matching cost of the data point is minimum, the corresponding matching result when the target function takes the maximum value is obtained, the corresponding matching result when the target function takes the maximum value is recorded as the optimal matching result, and the data point matching method and the device are implemented by aiming at the data pointsAnd performing fusion processing on the multiple characteristic values to obtain the matching characteristics of the data points.
In the embodiment, the point cloud data set and the standard point cloud data set of the real estate building are used as the input of the improved ICP algorithm, the actual information of the data points in the collected point cloud data set is obtained according to the matching result, and the collection of the mapping data of the real estate building is completed.
Obtaining an ICP matching algorithm flow in this embodiment according to the above steps, taking a point cloud data set and a standard point cloud data set of the real estate building as input of an ICP matching algorithm, obtaining a matching result of the point cloud data set and the standard point cloud data set of the real estate building according to a specific flow of the ICP algorithm, and after obtaining the matching result, the matching result includes a data point with a coincident spatial position and a data point with a non-coincident spatial position, the ICP algorithm is the prior art, and redundant description is not repeated here; for the data points with coincident spatial positions, taking actual position information corresponding to the standard point cloud data as position information of point cloud data on the real estate building; and for the data points with different spatial positions, calculating the difference between the data points and the matched standard point cloud data on each coordinate axis, and acquiring the position information of the point cloud data on the real estate building by using the actual position information and the difference corresponding to the standard point cloud data. And after the position information of all data points is obtained, acquiring required mapping data according to the statistical requirement of the real estate building, and uploading the mapping data to the cloud for storage.
And then, finishing the acquisition of the position information of the point cloud data of the real estate building.
The method includes the steps that firstly, a point cloud data set of the real estate building is obtained, the characteristic stability of each data point in the point cloud data set of the real estate building is determined, the distribution condition of the data points in each acquisition image is considered by the characteristic stability, the problem that when a scanner scans an area with severe surface change of the real estate building, the local density of adjacent point cloud data points is reduced due to scanning angles, the influence on point cloud data matching is caused by the increase of coordinate errors of the data points, the characteristic stability is used for representing the stability degree of characteristics of the data points on the corresponding acquisition images in the point cloud data set, the larger the characteristic stability is, the more uniform the distribution of the corresponding data points and the surrounding data points in the space is shown, the higher the similarity of the corresponding data points and key points is, the more obvious the characteristics are, the more possible the corresponding data points are normal data points, the characteristic points are screened out from the point cloud data set of the real estate building based on the characteristic points in the point cloud data set of the real estate building subsequently, the interference of abnormal points is eliminated, and the reliability of subsequent matching results can be improved; according to the Euclidean distance between each feature point and the feature point in the preset neighborhood, the point cloud descriptor of each feature point, the point cloud descriptor of the feature point in the preset neighborhood of each feature point, the structural diagram corresponding to each feature point and the structural diagram corresponding to the feature point in the preset neighborhood of each feature point, the feature descriptor of each feature point is constructed, the influence of the distance is considered by the feature descriptor, data points with different distances are analyzed, the problem that data points with longer distances are ignored when the descriptor is constructed in the prior art is solved, the structural similarity between the feature point and the data points in the neighborhood is also considered by the feature descriptor, accurate feature information can be obtained in a local area with a complex structural line, the influence of isolated interference points is avoided, and the matching precision between subsequent data points is improved; in the embodiment, the point cloud data is matched based on the feature descriptors of the feature points in the point cloud data set of the real estate building and the feature descriptors of the data points in the standard point cloud data set, so that the position information of the point cloud data of the real estate building is determined, the interference of abnormal points on the matching result is eliminated, and the accuracy and the reliability of the acquisition of the geographic information mapping data of the real estate building are improved.
Claims (8)
1. A real estate geographic information mapping data acquisition method is characterized by comprising the following steps:
acquiring a point cloud data set of a real estate building;
acquiring each acquisition image based on the spatial coordinates of each data point in the point cloud data set; obtaining the feature stability of each data point in the point cloud data set according to the Thiessen polygon corresponding to each data point in each acquisition image, the normal vector of the key point in the acquisition image where each data point is located and the normal vector of the acquisition image where each data point is located; screening feature points based on the feature stability;
constructing a structure chart corresponding to each feature point based on each feature point and the feature points in a preset neighborhood, and constructing a point cloud descriptor of each feature point based on a HOG operator corresponding to each feature point in each acquisition graph, a normal vector of the feature point in the acquisition graph where each feature point is located and the gray value of each feature point; obtaining a feature descriptor of each feature point according to the Euclidean distance between each feature point and the feature point in the preset neighborhood of the feature point, the point cloud descriptor of each feature point, the point cloud descriptor of the feature point in the preset neighborhood of each feature point, and the structural similarity between the structural diagram corresponding to each feature point and the structural diagram corresponding to the feature point in the preset neighborhood of each feature point;
matching the characteristic points in the point cloud data set of the real estate building with the data points in the standard point cloud data set based on the characteristic descriptors of the characteristic points in the point cloud data set of the real estate building and the characteristic descriptors of the data points in the standard point cloud data set, and determining the position information of the point cloud data of the real estate building based on the matching result.
2. The real estate geographic information mapping data collection method of claim 1 wherein said obtaining each collection map based on spatial coordinates of each data point in said point cloud data set comprises:
the acquisition diagrams comprise a transverse acquisition diagram and a longitudinal acquisition diagram;
the original point of a space coordinate system is an O point, three coordinate axes of the coordinate system are respectively an X axis, a Y axis and a Z axis, each plane parallel to the plane YOZ is marked as a first plane, each plane parallel to the plane XOZ is marked as a second plane, the first plane of the abscissa of any data point in the point cloud data set of the real estate building is used as a transverse acquisition map, and the second plane of the ordinate of any data point in the point cloud data set of the real estate building is used as a longitudinal acquisition map.
3. The real estate geographical information mapping data collection method of claim 2, wherein obtaining the feature stability of each data point in the point cloud data set according to the Thiessen polygon corresponding to each data point in each collection map, the normal vector of the key point in the collection map where each data point is located, and the normal vector of the collection map where each data point is located comprises:
for the a-th data point in the point cloud dataset:
according to the Thiessen polygons corresponding to the a-th data point in each acquisition map and the Thiessen polygons corresponding to the data points in the preset neighborhood of the a-th data point in each acquisition map, obtaining the transverse angle difference and the longitudinal angle difference of the a-th data point corresponding to the data points in the preset neighborhood of the a-th data point;
calculating the area difference of the Thiessen polygon corresponding to the a-th data point in the transverse acquisition map and the Thiessen polygon corresponding to each data point in the preset neighborhood of the a-th data point in the transverse acquisition map, and recording the area difference as the transverse area difference; calculating the area difference of the Thiessen polygon corresponding to the a-th data point in the longitudinal acquisition diagram and the Thiessen polygon corresponding to each data point in the preset neighborhood of the a-th data point in the longitudinal acquisition diagram, and recording the area difference as the longitudinal area difference; calculating the spatial dispersion of the a-th data point based on the transverse area difference, the longitudinal area difference, the transverse angle difference and the longitudinal angle difference, wherein the transverse area difference, the longitudinal area difference, the transverse angle difference and the longitudinal angle difference are in positive correlation with the spatial dispersion;
respectively calculating cosine similarity of a normal vector of each key point in each acquisition graph where the a-th data point is located and a normal vector of each acquisition graph where the a-th data point is located, and obtaining structural stability of the a-th data point based on the cosine similarity, wherein the cosine similarity and the structural stability are in positive correlation;
and taking the ratio of the structural stability to the spatial dispersion as the characteristic stability of the a-th data point.
4. The real estate geographical information mapping data collection method according to claim 3, wherein the method for obtaining the lateral angle difference of the a-th data point and each data point in the preset neighborhood thereof comprises:
respectively calculating the absolute value of the difference value between the maximum internal angle of the a-th data point in the corresponding Thiessen polygon in the transverse acquisition diagram and the maximum internal angle of the ith data point in the corresponding Thiessen polygon in the transverse acquisition diagram, and recording the absolute value as the difference of the maximum internal angles; respectively calculating the absolute value of the difference value between the minimum internal angle of the a-th data point in the corresponding Thiessen polygon in the transverse acquisition diagram and the minimum internal angle of the ith data point in the corresponding Thiessen polygon in the transverse acquisition diagram, and recording the absolute value as the difference of the minimum internal angles; the ith data point is a data point in a preset neighborhood of the a-th data point; and recording the sum of the difference of the maximum internal angle and the difference of the minimum internal angle as the transverse angle difference corresponding to the ith data point in the preset neighborhood of the ith data point.
5. The real estate geographical information mapping data collection method of claim 1, wherein said screening feature points based on said feature stability comprises: and acquiring the minimum value of the characteristic stability of all data points in the standard point cloud data set, taking the minimum value as a screening threshold value, and taking the data points with the characteristic stability of the point cloud data set of the real estate building greater than or equal to the screening threshold value as the characteristic points.
6. The method for collecting real estate geographic information mapping data according to claim 2, wherein constructing a point cloud descriptor for each feature point based on a HOG operator corresponding to each feature point in each collection map, a normal vector of the feature point in the collection map where each feature point is located, and a gray value of each feature point comprises:
for the qth feature point:
calculating the sum value of the HOG operator in the transverse acquisition graph where the q-th characteristic point is located and the HOG operator in the longitudinal acquisition graph where the q-th characteristic point is located;
calculating the normal vector of the q-th feature point in the transverse acquisition graph where the q-th feature point is located and the variance of the normal vectors of all feature points in the preset neighborhood of the q-th feature point, and recording the variance as a first variance; calculating the normal vector of the q-th feature point in the longitudinal acquisition image where the q-th feature point is located and the variance of the normal vectors of all feature points in a preset neighborhood of the q-th feature point, recording the variance as a second variance, and taking the sum of the first variance and the second variance as the normal vector distribution variance of the q-th feature point;
and combining the sum, the normal vector distribution variance and the gray value of the qth characteristic point together to be used as a point cloud descriptor of the qth characteristic point.
7. The real estate geographical information mapping data collection method of claim 1 wherein the feature descriptors for each feature point are calculated using the following formula:
wherein ,
for the feature descriptor of the qth feature point, be->
For the distance weighting of the kth characteristic point in the predetermined neighborhood of the qth characteristic point, ->
Is the Euclidean distance between the qth characteristic point and the kth characteristic point in the preset neighborhood, and>
for the number of feature points in a predetermined neighborhood of the qth feature point, < >>
Is the radius of the preset neighborhood, is>
Is the mean value of Euclidean distances between the qth characteristic point and all characteristic points in the preset neighborhood, and/or the corresponding value>
For a point cloud descriptor of the qth feature point>
For the point cloud descriptor of the kth characteristic point in the preset neighborhood of the qth characteristic point, the method determines whether the point cloud descriptor is valid or invalid>
Is a natural constant->
For presetting the adjusting parameter, is selected>
For a structure map corresponding to the qth feature point>
Is a structural diagram corresponding to the kth characteristic point in the preset neighborhood of the qth characteristic point,
is a structure diagram>
And a structure diagram>
Structural similarity of (a), in>
The absolute value sign is taken.
8. The method according to claim 1, wherein the constructing a structure diagram corresponding to each feature point based on the feature points in the feature points and the preset neighborhood comprises: and respectively connecting each feature point with each feature point in a preset neighborhood by taking each feature point as a center to obtain a structure chart corresponding to each feature point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310231199.8A CN115953604B (en) | 2023-03-13 | 2023-03-13 | Real estate geographic information mapping data acquisition method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310231199.8A CN115953604B (en) | 2023-03-13 | 2023-03-13 | Real estate geographic information mapping data acquisition method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115953604A true CN115953604A (en) | 2023-04-11 |
| CN115953604B CN115953604B (en) | 2023-05-30 |
Family
ID=85891296
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310231199.8A Active CN115953604B (en) | 2023-03-13 | 2023-03-13 | Real estate geographic information mapping data acquisition method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115953604B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116954342A (en) * | 2023-09-18 | 2023-10-27 | 惠州市鑫晖源科技有限公司 | Method for monitoring operation of power supply of omnibearing efficient heat dissipation host based on Internet of things |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107123161A (en) * | 2017-06-14 | 2017-09-01 | 西南交通大学 | A kind of the whole network three-dimensional rebuilding method of contact net zero based on NARF and FPFH |
| CN109241978A (en) * | 2018-08-23 | 2019-01-18 | 中科光绘（上海）科技有限公司 | The rapid extracting method of planar chip in ground three-dimensional laser point cloud |
| CN109253717A (en) * | 2018-10-09 | 2019-01-22 | 安徽大学 | A kind of mining area surface sedimentation 3 D laser scanning surface subsidence monitoring sets station method |
| US20200410690A1 (en) * | 2018-08-27 | 2020-12-31 | Tencent Technology (Shenzhen) Company Limited | Method and apparatus for segmenting point cloud data, storage medium, and electronic device |
| CN112596063A (en) * | 2020-11-27 | 2021-04-02 | 北京迈格威科技有限公司 | Point cloud descriptor construction method and device and closed loop detection method and device |
| CN114972459A (en) * | 2022-05-31 | 2022-08-30 | 哈尔滨理工大学 | Point cloud registration method based on low-dimensional point cloud local feature descriptor |
-
2023
- 2023-03-13 CN CN202310231199.8A patent/CN115953604B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107123161A (en) * | 2017-06-14 | 2017-09-01 | 西南交通大学 | A kind of the whole network three-dimensional rebuilding method of contact net zero based on NARF and FPFH |
| CN109241978A (en) * | 2018-08-23 | 2019-01-18 | 中科光绘（上海）科技有限公司 | The rapid extracting method of planar chip in ground three-dimensional laser point cloud |
| US20200410690A1 (en) * | 2018-08-27 | 2020-12-31 | Tencent Technology (Shenzhen) Company Limited | Method and apparatus for segmenting point cloud data, storage medium, and electronic device |
| CN109253717A (en) * | 2018-10-09 | 2019-01-22 | 安徽大学 | A kind of mining area surface sedimentation 3 D laser scanning surface subsidence monitoring sets station method |
| CN112596063A (en) * | 2020-11-27 | 2021-04-02 | 北京迈格威科技有限公司 | Point cloud descriptor construction method and device and closed loop detection method and device |
| CN114972459A (en) * | 2022-05-31 | 2022-08-30 | 哈尔滨理工大学 | Point cloud registration method based on low-dimensional point cloud local feature descriptor |
Non-Patent Citations (2)
| Title |
|---|
| 丁一: "基于改进SIFT 的离线签名鉴别方 法研究", 《中国优秀硕士学位论文全文数据库》 * |
| 袁钰杰: "增强现实环境中的三维重建技术研究", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116954342A (en) * | 2023-09-18 | 2023-10-27 | 惠州市鑫晖源科技有限公司 | Method for monitoring operation of power supply of omnibearing efficient heat dissipation host based on Internet of things |
| CN116954342B (en) * | 2023-09-18 | 2023-12-15 | 惠州市鑫晖源科技有限公司 | Method for monitoring operation of power supply of omnibearing efficient heat dissipation host based on Internet of things |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115953604B (en) | 2023-05-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11049267B2 (en) | Apparatus, method, and system for alignment of 3D datasets | |
| Kang et al. | Automatic targetless camera–lidar calibration by aligning edge with gaussian mixture model | |
| Tazir et al. | CICP: Cluster Iterative Closest Point for sparse–dense point cloud registration | |
| CN112446844B (en) | Point cloud feature extraction and registration fusion method | |
| Hu et al. | An automatic 3D registration method for rock mass point clouds based on plane detection and polygon matching | |
| CN115953604A (en) | Real estate geographic information mapping data acquisition method | |
| CN115267724B (en) | Position re-identification method of mobile robot capable of estimating pose based on laser radar | |
| CN115761303A (en) | Ground object classification method based on airborne laser radar point cloud and remote sensing image data | |
| Toth | Strip adjustment and registration | |
| Dold | Extended Gaussian images for the registration of terrestrial scan data | |
| Widyaningrum et al. | Comprehensive comparison of two image-based point clouds from aerial photos with airborne LiDAR for large-scale mapping | |
| CN114998395A (en) | Effective embankment three-dimensional data change detection method and system | |
| Ghorbani et al. | A novel 3-D local DAISY-style descriptor to reduce the effect of point displacement error in point cloud registration | |
| Chen et al. | Multi-stage matching approach for mobile platform visual imagery | |
| CN113532424A (en) | Integrated equipment for acquiring multidimensional information and cooperative measurement method | |
| CN117058008A (en) | Remote sensing image geometry and radiation integrated correction method, device, equipment and medium | |
| CN110969650B (en) | Intensity image and texture sequence registration method based on central projection | |
| Honti et al. | Automation of cylinder segmentation from point cloud data | |
| Kim et al. | New approach for planar patch segmentation using airborne laser data | |
| Panday | Fitting of parametric building models to oblique aerial images | |
| CN115147471A (en) | Laser point cloud automatic registration method based on curvature density characteristics | |
| CN118036899B (en) | Building decoration intelligent management system based on BIM | |
| CN117422746B (en) | Partition nonlinear geographic registration method, device, equipment and storage medium | |
| CN117076704B (en) | Image quality analysis and evaluation method and system | |
| Toth et al. | Strip adjustment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |