CN113537324A - Family type space matching method and device based on thin geometric plane spline interpolation - Google Patents
Family type space matching method and device based on thin geometric plane spline interpolation Download PDFInfo
- Publication number
- CN113537324A CN113537324A CN202110756389.2A CN202110756389A CN113537324A CN 113537324 A CN113537324 A CN 113537324A CN 202110756389 A CN202110756389 A CN 202110756389A CN 113537324 A CN113537324 A CN 113537324A
- Authority
- CN
- China
- Prior art keywords
- house type
- type space
- clustering
- space
- matching
- 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 44
- 238000013507 mapping Methods 0.000 claims abstract description 83
- 239000011159 matrix material Substances 0.000 claims abstract description 53
- 238000012545 processing Methods 0.000 claims abstract description 15
- 230000006870 function Effects 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 11
- 238000013506 data mapping Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000008859 change Effects 0.000 claims description 4
- 238000013075 data extraction Methods 0.000 claims description 3
- 238000013461 design Methods 0.000 abstract description 15
- 238000005034 decoration Methods 0.000 abstract description 11
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013434 data augmentation Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/53—Querying
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Probability & Statistics with Applications (AREA)
- Databases & Information Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a family space matching method based on thin geometric plane spline interpolation. The method comprises the steps of receiving an input house type graph, dividing a house type area and obtaining a house type space; extracting all feature data in the house type space and carrying out unified feature mapping processing on the feature data of the wall, the door and the window; matching the house type space feature mapping matrix with the clustering centers of different levels of a house type space matching database according to the function area to which the house type space belongs to obtain the maximum value of the overlapping degree of the clustering centers of the house type space and each level of clustering, and determining the clustering category to which the house type space belongs according to the maximum value of the overlapping degree; and returning the house type space matching result according to the cluster type of the house type space. The invention can solve the problem that the traditional house type matching cannot be matched under the conditions of rotation and mirror image, and takes the relative positions of the door and the window in the house type space into consideration, thereby improving the reliability of successful matching and greatly improving the design efficiency of home decoration design.
Description
Technical Field
The invention relates to the field of home decoration design, in particular to a family type space matching method and device based on thin geometric plane spline interpolation.
Background
With the development of the real estate industry, the more demands are made on the house decoration design for purchasing the real estate, and people hope to refer to the decoration design of the same type of house before formal decoration, and enter the design decoration link of the real estate after the design which is the same type of house and liked is selected and improved. However, the mainstream method for house type matching at present is based on context image analysis, the similarity of house types is calculated through a convex geometric polygonal model for matching, but the method is easily influenced by the conditions of house type image scaling, rotation and the like, so that the corresponding house type space cannot be matched, and the method is not beneficial to fast screening out the proper house type by demand personnel as a reference, so that the design efficiency of the whole house decoration is low, besides contour information is removed when matching a single house type space, the information of doors and windows is often more important than wall information, the moving line of the house type space and the design of the house decoration are directly influenced, and the positions and the sizes of the doors and the windows are required to be considered when matching the house type space.
Disclosure of Invention
In view of the above disadvantages of the prior art, the present invention is directed to a method and an apparatus for matching a family space based on thin geometry planar spline interpolation, which are used to solve the above problems in the prior art.
In order to achieve the above objects and other related objects, the present invention provides a family space matching method and device based on thin geometric plane spline interpolation, the method comprising: receiving an input house type graph, dividing a house type area and acquiring a house type space; extracting all feature data in the house type space and carrying out unified feature mapping processing on the feature data of the wall, the door and the window; matching the house type space feature mapping matrix with the clustering centers of different levels of a house type space matching database according to the function area to which the house type space belongs to obtain the maximum value of the overlapping degree of the clustering centers of the house type space and each level of clustering, and determining the clustering category to which the house type space belongs according to the maximum value of the overlapping degree; and returning the house type space matching result according to the cluster type of the house type space.
In an embodiment of the present invention, the method further includes: pre-establishing the house type space matching database, wherein the house type space matching database is established by the following steps: the method comprises the steps that a house type space is obtained by dividing a house type area of a house type sample by an example, feature data of the house type space are extracted, and unified feature mapping processing is carried out on a wall body, a door and a window in the feature data; calculating the overlapping degree of the house type spaces of the same functional area according to the characteristic mapping matrix, adjusting the data weight of walls, doors and windows to perform different-level distance clustering on the house type spaces and extracting distance clustering centers; carrying out hierarchical clustering on the house type space according to the distance between the first-layer distance clustering centers and extracting hierarchical clustering centers; and setting the cluster classification field, storing corresponding data and establishing the house type space matching database.
In an embodiment of the present invention, the method further includes identifying a space identifier of the dwelling space; extracting wall, door and window characteristic data of the house type space; and calculating the intersection point of the house type space wall body according to the feature data of the house type space wall body, drawing the wall body line according to the intersection point, and verifying the accuracy of the wall body line by calculating the overlapping degree of the drawn wall body line and the original house type space wall body.
In an embodiment of the present invention, the performing unified feature mapping processing on the user-type spatial feature data in the method includes: setting m feature mapping control points through thin geometric plane spline interpolation, wherein m is a preset value; setting the size of a feature mapping matrix, wherein the size of the matrix can be represented by h x h, and h is determined according to the sizes of all house type space samples; constructing a feature mapping loss function, which can be expressed as:
in an embodiment of the present invention, the method further includes: adjusting the weights of wall, door and window data in the house type space sample feature mapping matrix to be j, k and l, performing first-layer distance clustering on the house type space, and extracting a first-layer distance clustering center, wherein: j > k > l; adjusting the weight of wall, door and window data in the house type space sample feature mapping matrix to be j1, k1 and l1, clustering the second layer distance of the house type space, and extracting the second layer distance clustering center, wherein: k1> j1> l1, and the second-layer distance clustering of the house space is carried out based on the first-layer distance clustering; adjusting the weights of wall, door and window data in the house type space sample feature mapping matrix to be j2, k2 and l2, performing third-layer distance clustering on the house type space, and extracting the third-layer distance clustering center, wherein: l2> -k 2> -j 2, and the third-layer distance clustering of the house-type space is carried out on the basis of the second-layer distance clustering; extracting first-layer distance clustering centers, calculating the overlapping degree of the first-layer distance clustering centers, and carrying out hierarchical clustering according to the overlapping degree.
In an embodiment of the present invention, the method further includes: randomly traversing p% of feature data mapping data of the household type space samples belonging to the same functional area, and selecting an initial clustering center in a K-Means + + mode, wherein: p is a preset value, and the selected initial clustering centers are not in the same cluster; and calculating the overlap degree of the house type space according to the characteristic mapping matrixes of all the house type space samples in the same cluster and the initial clustering center, and calculating a new clustering center by taking the overlap degree as a distance function, wherein the new clustering center can be represented by a probability distribution map.
In an embodiment of the present invention, the method further includes: the calculated expression of the degree of overlap is
Wherein: TP represents the coincidence data quantity of the house type space characteristic mapping matrix and the cluster center of the house type space matching database, FP represents the non-coincidence data quantity of the house type space characteristic mapping matrix, and FN represents the non-coincidence data quantity of the cluster center of the house type space matching database; comparing the overlapping degree of the house type space characteristic mapping matrix and all the clustering centers of each layer of clusters to obtain the maximum value of the overlapping degree; and storing the characteristic mapping matrix of the house type space into the clustering category database according to the clustering category to which the house type space belongs.
To achieve the above and other related objects, the present invention provides a family space matching apparatus based on thin geometry planar spline interpolation, the apparatus comprising: the receiving module is used for receiving the floor plan; the division module is used for dividing the house type area in the house type graph into single house type spaces; the characteristic data extraction module is used for extracting all characteristic data of the house type space; the characteristic data mapping module is used for uniformly mapping all characteristic data of the house type space; the matching module is used for matching the feature mapping matrix of the house type space with the clustering centers of different levels of the house type space matching database according to the functional area to which the house type space belongs and obtaining a matching result; and the output module is used for outputting the house type space matching result according to the clustering category matching result.
To achieve the above and other related objects, the present invention provides a computer-readable storage medium, wherein a computer program is stored, and when the computer program is loaded and executed by a processor, the method for matching a house type space based on thin geometric planar spline interpolation is implemented.
To achieve the above and other related objects, the present invention provides an electronic device, comprising: a processor, a memory; wherein the memory is for storing a computer program; the processor is used for loading and executing the computer program to enable the electronic equipment to execute the family space matching method based on the thin geometric plane spline interpolation.
As described above, according to the house type space matching method and device based on thin geometric plane spline interpolation provided by the present invention, all the information of wall lines, doors, windows, etc. of a single space are extracted from a complete house type diagram, and feature matching is performed with a single space of other house types, so as to find the most similar single space. The technology solves the problem that the traditional house type matching cannot be matched under the conditions of rotation and mirror image, and takes the relative positions of a door and a window into consideration, thereby improving the reliability of successful matching, greatly improving the design efficiency of home decoration design, reducing the design cost, and being easier for users to accept.
Drawings
Fig. 1 is a flowchart of a family space matching method based on thin geometric planar spline interpolation according to an embodiment of the present invention.
Fig. 2 is a flow chart illustrating the building of the family-space matching database according to an embodiment of the present invention.
Fig. 3 is a schematic diagram illustrating a subscriber-type space matching case in an embodiment of the invention.
Fig. 4 is a schematic diagram of a subscriber space matching apparatus module according to an embodiment of the invention.
Fig. 5 is a schematic diagram of an electronic device according to an embodiment of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The method is based on the thin geometric plane spline interpolation model, and maps the information of a single space to a characteristic space, so that the matching standard degree of the house type space is detected. Based on this, the invention provides a family space matching method and device based on thin geometric plane spline interpolation, which are described in detail through embodiments below.
Referring to fig. 1, the present embodiment provides a family space matching method based on thin geometric plane spline interpolation, including the following steps:
s11: and receiving the input house type graph, dividing the house type area and acquiring the house type space.
Specifically, a house type graph input by a user is received, and the house type areas in the house type graph are divided into independent house type spaces on the basis of not damaging wall lines and door and window information of the original space by carrying out example division on the house type areas in the house type graph.
S12: and extracting all characteristic data in the house type space and carrying out unified characteristic mapping processing on the wall body, the door and the window characteristic data.
Specifically, the space identification of each of the residential spaces in the residential area is recognized through the OCR technology so as to confirm the functional area to which each of the residential spaces belongs, such as a restaurant, a living room, a main bed, a sub bed, a balcony, a kitchen, etc. Identifying each house type space characteristic data through a target detection network, wherein the characteristic data comprises wall, door and window data, and the method comprises the following steps: the wall data comprises coordinate data, wall thickness, bearing walls or non-heavy walls, outer walls and inner walls; the door data comprises the position of the door and index data of the wall surface to which the door belongs; the window data includes the position of the window and index data of the wall to which the window belongs, and orientation data of the window. Identifying the type and attributes of doors and windows in the dwelling space through a convolutional neural network, such as: the door types comprise safety doors, sliding doors and the like; the door attribute includes attribute information such as a door opening direction, a door size, a door height, and transparency.
Before mapping processing is carried out on the house type space characteristic data, a transition point of each wall is calculated according to the acquired house type space wall data through a singular value of second-order difference of horizontal and vertical coordinates to serve as a wall vertex, a wall line is drawn according to the vertex data through OpenCV, and the accuracy of the wall line is verified through calculating the overlapping degree of the drawn wall line and the original house type space wall. And if the accuracy rate meets the standard, carrying out unified feature mapping processing on the feature data.
Furthermore, m feature mapping control points are set through Thin geometric plane spline interpolation (namely Thin plate spline) to perform feature mapping on wall, door and window data in the family type feature data so as to eliminate errors of a family type graph space region caused by distortion, rotation, scaling and the like, wherein m is a preset value. And uniformly mapping the wall, door and window data to a uniform h x h feature matrix through an energy function and different weighting coefficients, wherein h is determined according to the sample size of all the house type spaces, and the size of the feature matrix can be set to 100 x 100, for example.
Further, when feature data mapping is performed, excessive loss of the house type space feature data is avoided by constructing a feature mapping loss function to ensure that a geometric figure represented by the mapped house type space feature data is smoother, and the loss function can adopt
Preferably, by
Mapping the house type space characteristic data as an optimal mapping function, wherein z represents a characteristic vector of 1 x h, d represents a characteristic matrix of h x h mapping planes, c represents the change of a mapping matrix m x h from the house type y to the house type f,
representing a feature vector of 1 m, the expression of which can be used
And (4) showing.
The characteristic data of the house type space can be represented by a characteristic mapping chart after being uniformly mapped to a matrix with the same size.
S13: and matching the house type space feature mapping matrix with the clustering centers of different levels of a house type space matching database according to the function area to which the house type space belongs, acquiring the maximum value of the overlapping degree of the clustering centers of the house type space and each level of clustering, and determining the clustering category to which the house type space belongs according to the maximum value of the overlapping degree.
Specifically, matching a house type space feature mapping matrix with a house type space matching database and different clustering centers in a first-layer distance cluster of the same functional area of the house type space, calculating the overlapping degree of the house type space and the clustering centers of different categories in the first-layer distance cluster, obtaining the maximum value of the overlapping degree, and determining the distance cluster category of the house type space in the first-layer distance cluster; then, matching the feature mapping matrix of the house type space with different clustering centers in a second layer of distance clustering under the category, calculating the overlapping degree of the house type space and the centers of the different categories in the layer of distance clustering, obtaining the maximum value of the overlapping degree, and determining the distance clustering category to which the house type space belongs in the second layer of distance clustering; finally, matching the house type spatial feature mapping matrix with different clustering centers in a third-layer distance cluster under the category according to the second-layer distance cluster category of the spatial house type, calculating the overlapping degree of the house type space and the clustering centers of different categories in the third-layer distance cluster, obtaining the maximum value of the overlapping degree, and determining the distance cluster category of the house type space in the third-layer distance cluster; and finally, determining the house type space with the highest overlapping degree and the most similar degree to the house type space according to the determined category of each layer of cluster center. In general, the first-level distance cluster is represented as the geometric shape of a house-type space region, namely the geometric shape of a closed region formed by walls, doors and windows; the second layer distance cluster is expressed as the category, the position and the like of the door in the house type space; the third layer of distance clustering is expressed as the category, position and the like of a window in the house type space; the matching of the house type space is completed by determining the geometric shape and the door and window category of the house type space.
It should be noted that, when the family space matching database is established, after hierarchical clustering is performed on the basis of the first-level distance classification, matching needs to be performed with a clustering center in the hierarchical clustering in advance after feature mapping is performed on the family space.
Furthermore, the calculation mode of the overlap degree (interaction over Unit) of the house type space characteristic mapping matrix and the clustering centers of different levels in the house type space matching database is
Wherein: TP represents the coincidence data quantity of the house type space characteristic mapping matrix and the cluster center of the house type space matching database, FP represents the non-coincidence data quantity of the house type space characteristic mapping matrix, and FN represents the non-coincidence data quantity of the cluster center of the house type space matching database.
It should be noted that before the house type matching is performed, a house type space matching database needs to be established, and the flow of the establishing step is shown in fig. 2:
s21: and the house type space is obtained by dividing the house type area of the house type sample by the example, the characteristic data of the house type space is extracted, and the unified characteristic mapping processing is carried out on the wall body, the door and the window in the characteristic data.
Specifically, the processing method of the step after data augmentation for the collected house type pattern book and the house type pattern sample is the same as that in steps S11 and S12, and will not be described in detail again.
S22: and calculating the overlapping degree of the house type spaces of the same functional area according to the characteristic mapping matrix, adjusting the data weight of the wall, the door and the window to perform different-level distance clustering on the house type spaces, and extracting a clustering center.
Specifically, p% of characteristic mapping matrixes of the house type space samples belonging to the same functional area are traversed randomly, one house type sample is selected as an initial clustering center in a K-Means + + mode, and the mode is used for ensuring that the selected initial clustering centers are not in the same cluster. Wherein: p is a preset value and is adjusted according to the total number of the house type pattern books. And calculating the overlapping degree of the house type spaces according to the characteristic mapping matrixes of all the house type space samples in the same cluster and the initial clustering center, calculating the best matching house type space as a new clustering center by taking the overlapping degree as a distance function, and iterating the clustering centers according to the increase of the number of the samples. Preferably, the present invention employs the probability distribution map as a cluster center for the heart.
Specifically, the calculation formula for the overlapping degree of the two mapped house type space samples is calculated as
Wherein: TP represents the coincidence data quantity of the feature mapping matrixes of two user type space samples, FP represents the non-coincidence data quantity of the feature mapping matrix of one user type space sample, and FN represents the non-coincidence data quantity of the feature mapping matrix of the other user type space sample. It should be noted that, since mapping the house type space samples to the feature space of the fixed size matrix may cause different scaling ratios, the wall line width needs to be calculated when calculating the overlapping degree.
Further, on the basis of the distance clustering method, the weights of the wall, the door and the window are adjusted for clustering. According to the method, the weight of wall, door and window data in the house type space sample feature mapping matrix is adjusted to be j > k > l according to the first-layer distance clustering side emphasis, first-layer distance clustering is carried out on the house type space, and the first-layer distance clustering center is extracted.
Further, after the first-layer distance clustering is completed, the weight of wall, door and window data in the family space sample feature mapping matrix is adjusted to be k1> j1> l1 according to the second-layer distance clustering side emphasis in each clustering category of the first layer, the second-layer distance clustering of the family space is carried out, the second-layer distance clustering center is extracted, and the second-layer distance clustering of the family space is carried out on the basis of the first-layer distance clustering.
Further, after the distance clustering of the second layer is completed, the weight of wall, door and window data in the house type space sample feature mapping matrix is adjusted to be l2> -k 2> -j 2 according to the side emphasis of distance clustering of the third layer in each cluster category of the second layer, the distance clustering of the third layer of the house type space is performed, and the distance clustering center of the third layer is extracted.
Preferably, the overlapping degree between the first-layer distance clustering centers is calculated according to the extracted first-layer distance clustering centers, hierarchical clustering is carried out according to the overlapping degree, clustering centers of different classes of hierarchical clustering are extracted, and hierarchical classification identification is carried out on a sample space.
S23: and setting the clustering type field, storing corresponding data and establishing the house type space matching database.
Specifically, the characteristic mapping matrix of the clustered house type space samples is stored in the corresponding clustering category field, and the house type space matching database is completed.
S14: and returning the house type space matching result according to the cluster type of the house type space.
Specifically, the user type space feature mapping matrix is stored in the clustering category database according to the clustering category to which the user type space belongs, and the user type space most similar to a single space region in a user type graph input by the user is returned to the user, so that the user can design the home decoration according to the favorite style.
In summary, the house type space matching process of the present invention is roughly described by the case schematic as shown in fig. 3: carrying out example segmentation on the sub-bedroom 1 in the family pattern, carrying out feature mapping on the segmented family space according to the acquired feature data, and matching the bedroom feature mapping chart with a clustering center probability chart of the first-layer distance clustering to acquire the most similar first-layer distance clustering category; matching according to the bedroom door feature mapping graph and a clustering center probability graph of the second-layer distance cluster to obtain the most similar second-layer distance cluster type; matching the bedroom window characteristic mapping chart with a clustering center probability chart of the third-layer distance clustering to obtain the most similar third-layer distance clustering category; and outputting the house type space most similar to the bedroom 1 in the house type to the user through the determined distance clustering categories of the first layer, the second layer and the third layer.
All or part of the steps for implementing the above method embodiments may be performed by hardware associated with a computer program. Based upon such an understanding, the present invention also provides a computer program product comprising one or more computer instructions. The computer instructions may be stored in a computer readable storage medium. The computer-readable storage medium can be any available medium that a computer can store or a data storage device, such as a server, a data center, etc., that is integrated with one or more available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.
Referring to fig. 4, the present embodiment provides a family space matching apparatus 40 based on thin geometric plane spline interpolation, and since the technical principle of the present embodiment is similar to that of the foregoing method embodiment, repeated details of the same technical details are not repeated. The apparatus 40 of the present embodiment includes the following modules:
a receiving module 41, configured to receive the user profile.
And a dividing module 42, configured to divide the house type area in the house type map into a single house type space.
And a feature data extraction module 43, configured to extract all feature data of the user type space.
And the feature data mapping module 44 is configured to perform unified mapping processing on all feature data of the user type space.
And the matching module 45 is used for matching the family type space characteristic mapping matrix with the clustering centers of different levels of the family type space matching database according to the functional area to which the family type space belongs and obtaining a matching result.
And the output module 46 is used for outputting the house type space matching result according to the clustering class matching result.
Those skilled in the art should understand that the division of the modules in the embodiment of fig. 4 is only a logical division, and the actual implementation can be fully or partially integrated into one or more physical entities. And the modules can be realized in a form that all software is called by the processing element, or in a form that all the modules are realized in a form that all the modules are called by the processing element, or in a form that part of the modules are called by the hardware.
Referring to fig. 5, the embodiment provides an electronic device, which may be a desktop device, a portable computer, a smart phone, and the like. In detail, the electronic device comprises at least, connected by a bus: the system comprises a memory and a processor, wherein the memory is used for storing computer programs, and the processor is used for executing the computer programs stored by the memory so as to execute all or part of the steps in the method embodiment.
The above-mentioned system bus may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The system bus may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus. The Memory may include a Random Access Memory (RAM), and may further include a non-volatile Memory (non-volatile Memory), such as at least one disk Memory.
The Processor may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; the device can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic device, discrete hardware component.
In summary, the house type space matching method and device based on the thin geometric plane spline interpolation solve the problem that the traditional house type matching cannot be matched under the conditions of house type rotation and mirror image, and also consider the relative positions of the door and the window, so that the matching reliability is improved, the design efficiency and the reference value of home decoration design are greatly improved, and various defects in the prior art are effectively overcome, and the house type space matching method and device have high industrial utilization value.
The foregoing embodiments are merely illustrative of the principles and utilities of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or change the above-mentioned embodiments without departing from the spirit and scope of the present invention. Accordingly, it is intended that all equivalent modifications or changes which can be made by those skilled in the art without departing from the spirit and technical spirit of the present invention be covered by the claims of the present invention.
Claims (10)
1. A house type space matching method based on thin geometric plane spline interpolation is characterized by comprising the following steps:
receiving an input house type graph, dividing a house type area and acquiring a house type space;
extracting all feature data in the house type space and carrying out unified feature mapping processing on the feature data of the wall, the door and the window;
matching the house type space feature mapping matrix with the clustering centers of different levels of a house type space matching database according to the function area to which the house type space belongs to obtain the maximum value of the overlapping degree of the clustering centers of the house type space and each level of clustering, and determining the clustering category to which the house type space belongs according to the maximum value of the overlapping degree;
and returning the house type space matching result according to the cluster type of the house type space.
2. The method according to claim 1, further comprising pre-establishing the family-type space matching database, wherein the step of establishing the family-type space matching database is:
the method comprises the steps that a house type space is obtained by dividing a house type area of a house type sample by an example, feature data of the house type space are extracted, and unified feature mapping processing is carried out on a wall body, a door and a window in the feature data;
calculating the overlapping degree of the house type spaces of the same functional area according to the characteristic mapping matrix, adjusting the data weight of walls, doors and windows to perform different-level distance clustering on the house type spaces and extracting distance clustering centers;
carrying out hierarchical clustering on the house type space according to the distance between the first-layer distance clustering centers and extracting hierarchical clustering centers;
and setting the cluster classification field, storing corresponding data and establishing the house type space matching database.
3. The method of claim 1 or 2, further comprising:
identifying a space identifier of the house type space;
extracting wall, door and window characteristic data of the house type space;
and calculating the intersection point of the house type space wall body according to the feature data of the house type space wall body, drawing the wall body line according to the intersection point, and verifying the accuracy of the wall body line by calculating the overlapping degree of the drawn wall body line and the original house type space wall body.
4. The method according to claim 1 or 2, wherein the unified feature mapping processing on the user type space feature data comprises:
setting m feature mapping control points through thin geometric plane spline interpolation, wherein m is a preset value;
setting the size of a feature mapping matrix, wherein the size of the matrix can be represented by h x h, and h is determined according to the sizes of all house type space samples;
constructing a feature mapping loss function, which can be expressed as:
a feature mapping matrix is calculated, and the calculation function can be expressed as:
5. the method of claim 2, further comprising:
adjusting the weights of wall, door and window data in the house type space sample feature mapping matrix to be j, k and l, performing first-layer distance clustering on the house type space, and extracting a first-layer distance clustering center, wherein: j > k > l;
adjusting the weights of wall, door and window data in the house type space sample feature mapping matrix to be j1, k1 and l1, clustering the second layer distance of the house type space, and extracting the second layer distance clustering center, wherein: k1> j1> l1, and the second-layer distance clustering of the house space is carried out based on the first-layer distance clustering;
adjusting the weights of wall, door and window data in the house type space sample feature mapping matrix to be j2, k2 and l2, performing third-layer distance clustering on the house type space, and extracting the third-layer distance clustering center, wherein: l2> -k 2> -j 2, and the third-layer distance clustering of the house-type space is carried out on the basis of the second-layer distance clustering;
and calculating the overlapping degree between the first-layer distance clustering centers according to the extracted first-layer distance clustering centers, carrying out hierarchical clustering on the basis of the overlapping degree and extracting the centers of the hierarchical clustering.
6. The method of claim 2 or 5, further comprising:
randomly traversing p% of feature data mapping data of the household type space samples belonging to the same functional area, and selecting an initial clustering center in a K-Means + + mode, wherein: p is a preset value, and the selected initial clustering centers are not in the same cluster;
and calculating the overlap degree of the house type space according to the characteristic mapping matrixes of all the house type space samples in the same cluster and the initial clustering center, and calculating a new clustering center by taking the overlap degree as a distance function, wherein the new clustering center can be represented by a probability distribution map.
7. The method of claim 1, further comprising:
the calculated expression of the degree of overlap is
Wherein: TP represents the coincidence data quantity of the house type space characteristic mapping matrix and the cluster center of the house type space matching database, FP represents the non-coincidence data quantity of the house type space characteristic mapping matrix, and FN represents the non-coincidence data quantity of the cluster center of the house type space matching database;
comparing the overlapping degree of the house type space characteristic mapping matrix and all the clustering centers of each layer of clusters to obtain the maximum value of the overlapping degree;
and storing the characteristic mapping matrix of the house type space into the clustering category database according to the clustering category to which the house type space belongs.
8. A family space matching apparatus based on thin geometry planar spline interpolation, the apparatus comprising:
the receiving module is used for receiving the floor plan;
the division module is used for dividing the house type area in the house type graph into a single house type space;
the characteristic data extraction module is used for extracting all characteristic data of the house type space;
the characteristic data mapping module is used for uniformly mapping all characteristic data of the house type space;
the matching module is used for matching the feature mapping matrix of the house type space with the clustering centers of different levels of the house type space matching database according to the functional area to which the house type space belongs and obtaining a matching result;
and the output module is used for outputting the house type space matching result according to the clustering category matching result.
9. A computer-readable storage medium, in which a computer program is stored which, when loaded and executed by a processor, implements the family space matching method based on thin geometric planar spline interpolation of any of claims 1 to 7.
10. An electronic device, comprising: a processor, a memory; wherein,
the memory is used for storing a computer program;
the processor is configured to load and execute the computer program to cause the electronic device to perform the thin-geometry-planar-spline-interpolation-based family space matching method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110756389.2A CN113537324B (en) | 2021-07-05 | 2021-07-05 | House type space matching method and device based on thin geometric plane spline interpolation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110756389.2A CN113537324B (en) | 2021-07-05 | 2021-07-05 | House type space matching method and device based on thin geometric plane spline interpolation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113537324A true CN113537324A (en) | 2021-10-22 |
| CN113537324B CN113537324B (en) | 2023-09-12 |
Family
ID=78097755
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110756389.2A Active CN113537324B (en) | 2021-07-05 | 2021-07-05 | House type space matching method and device based on thin geometric plane spline interpolation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113537324B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113946900A (en) * | 2021-12-21 | 2022-01-18 | 深圳小库科技有限公司 | Method for quickly recommending similar house types based on house type profiles and distribution characteristics |
| CN114255267A (en) * | 2021-12-20 | 2022-03-29 | 北京房江湖科技有限公司 | Method and apparatus for registering a family plot |
| CN116740276A (en) * | 2023-06-09 | 2023-09-12 | 北京优贝卡科技有限公司 | House type diagram generation method, device, equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109933840A (en) * | 2019-01-18 | 2019-06-25 | 江苏艾佳家居用品有限公司 | A kind of region Auto-matching algorithm and system based on house type geometrical characteristic |
| CN110197225A (en) * | 2019-05-28 | 2019-09-03 | 广东三维家信息科技有限公司 | House type spatial match method and system based on deep learning |
| KR20190121275A (en) * | 2019-10-07 | 2019-10-25 | 엘지전자 주식회사 | System, apparatus and method for indoor positioning |
-
2021
- 2021-07-05 CN CN202110756389.2A patent/CN113537324B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109933840A (en) * | 2019-01-18 | 2019-06-25 | 江苏艾佳家居用品有限公司 | A kind of region Auto-matching algorithm and system based on house type geometrical characteristic |
| CN110197225A (en) * | 2019-05-28 | 2019-09-03 | 广东三维家信息科技有限公司 | House type spatial match method and system based on deep learning |
| KR20190121275A (en) * | 2019-10-07 | 2019-10-25 | 엘지전자 주식회사 | System, apparatus and method for indoor positioning |
Non-Patent Citations (2)
| Title |
|---|
| ALEX X. LEE ET AL.: "A Non-Rigid Point and Normal Registration Algorithm with Applications to Learning from Demonstrations", 《IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION》, pages 935 - 942 * |
| 李伟等: "基于户型图实例分割的室内场景建模研究", 《浙江工业大学学报》, vol. 49, no. 2, pages 1 - 3 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114255267A (en) * | 2021-12-20 | 2022-03-29 | 北京房江湖科技有限公司 | Method and apparatus for registering a family plot |
| CN113946900A (en) * | 2021-12-21 | 2022-01-18 | 深圳小库科技有限公司 | Method for quickly recommending similar house types based on house type profiles and distribution characteristics |
| CN113946900B (en) * | 2021-12-21 | 2022-03-29 | 深圳小库科技有限公司 | Method for quickly recommending similar house types based on house type profiles and distribution characteristics |
| CN116740276A (en) * | 2023-06-09 | 2023-09-12 | 北京优贝卡科技有限公司 | House type diagram generation method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113537324B (en) | 2023-09-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113537324B (en) | House type space matching method and device based on thin geometric plane spline interpolation | |
| US20180225593A1 (en) | Transforming property data into sufficiently sized, relatively homogeneous data segments for configuring automated modeling systems | |
| CN110414550B (en) | Training method, device and system of face recognition model and computer readable medium | |
| Leng et al. | A multi‐scale plane‐detection method based on the Hough transform and region growing | |
| El‐Sayed et al. | Plane detection in 3D point cloud using octree‐balanced density down‐sampling and iterative adaptive plane extraction | |
| CN117078048B (en) | Digital twinning-based intelligent city resource management method and system | |
| CN114565807B (en) | Method and device for training target image retrieval model | |
| CN111178196B (en) | Cell classification method, device and equipment | |
| CN111339960B (en) | Face recognition method based on discrimination low-rank regression model | |
| CN116385707A (en) | Deep learning scene recognition method based on multi-scale features and feature enhancement | |
| CN112818162A (en) | Image retrieval method, image retrieval device, storage medium and electronic equipment | |
| JP2023510945A (en) | Scene identification method and apparatus, intelligent device, storage medium and computer program | |
| CN111597921B (en) | Scene recognition method, device, computer equipment and storage medium | |
| CN108197795A (en) | The account recognition methods of malice group, device, terminal and storage medium | |
| CN111738831A (en) | Service processing method, device and system | |
| CN114283332A (en) | Fuzzy clustering remote sensing image segmentation method, system, terminal and storage medium | |
| Lee et al. | Imat: The iterative medial axis transform | |
| CN117035837A (en) | Method for predicting electricity purchasing demand of power consumer and customizing retail contract | |
| CN113435479A (en) | Feature point matching method and system based on regional feature expression constraint | |
| CN111680346B (en) | House type diagram complement method and device, computer readable storage medium and electronic equipment | |
| WO2020199483A1 (en) | Image processing method and apparatus for financial data, and device and computer-readable storage medium | |
| CN113919449B (en) | Resident electric power data clustering method and device based on precise fuzzy clustering algorithm | |
| CN112507137B (en) | Small sample relation extraction method based on granularity perception in open environment and application | |
| CN112668590B (en) | Visual phrase construction method and device based on image feature space and airspace space | |
| CN114612923A (en) | House type graph wall processing method, system, medium and equipment based on target detection |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20230809 Address after: No. 344, Sanlin Road, Pudong New Area, Shanghai Applicant after: B&T Home Network Technology (Shanghai) Co.,Ltd. Address before: 200120 B & Q Pudong business building, No. 393 Yinxiao Road, Pudong New Area, Shanghai Applicant before: Baianju information technology (Shanghai) Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |