CN112613528B - Point cloud simplifying method and device based on significance variation and storage medium - Google Patents
Point cloud simplifying method and device based on significance variation and storage medium Download PDFInfo
- Publication number
- CN112613528B CN112613528B CN202011619176.7A CN202011619176A CN112613528B CN 112613528 B CN112613528 B CN 112613528B CN 202011619176 A CN202011619176 A CN 202011619176A CN 112613528 B CN112613528 B CN 112613528B
- Authority
- CN
- China
- Prior art keywords
- point cloud
- salient region
- salient
- unsig
- original point
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000000605 extraction Methods 0.000 claims abstract description 8
- 239000011159 matrix material Substances 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 10
- 238000000354 decomposition reaction Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000000513 principal component analysis Methods 0.000 claims description 6
- 230000004927 fusion Effects 0.000 claims description 5
- 101100391172 Dictyostelium discoideum forA gene Proteins 0.000 claims description 3
- 238000005056 compaction Methods 0.000 claims description 3
- 101100391182 Dictyostelium discoideum forI gene Proteins 0.000 claims 1
- 230000000694 effects Effects 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
-
- 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
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Length Measuring Devices With Unspecified Measuring Means (AREA)
- Image Generation (AREA)
Abstract
The application discloses a point cloud simplifying method based on significance variation, which comprises the following steps: step 110, acquiring an original point cloud, and establishing a kd-tree of the original point cloud; step 120, establishing a curve Bian Bianfen index of all data points of the original point cloud according to the kd-tree; step 130, dividing all data points of the original point cloud into a salient region and a non-salient region according to a salient division principle; step 140, extracting the salient region and storing the salient region; step 150, simplifying the non-salient region by a space dichotomy to obtain a simplified non-salient region; and step 160, fusing the dominant region and the simplified non-significant region to obtain a final simplified result. The method not only maintains the original advantages of the space dichotomy, but also greatly improves the extraction capacity of the characteristic edge line, has good robustness, and is suitable for most point cloud data.
Description
Technical Field
The disclosure relates to the technical field of computer vision, in particular to a point cloud simplifying method and device based on significance variation and a storage medium.
Background
With the continuous maturity of the point cloud data technology, the application of the point cloud data technology in different fields, such as industry, medical treatment, military and the like, is also becoming wider and wider. The point cloud data is mainly obtained through contact type measurement optical three-dimensional scanning equipment, the obtained data has high precision, but the most original point cloud data has high redundancy, so that the obtained point cloud data needs to be simplified.
Traditional point cloud reduction algorithms include curvature methods, bounding box methods, spatial dichotomy, and the like. The space dichotomy has the advantages of high compaction speed, high compaction ratio, good conformality and the like. However, the spatial dichotomy has some problems, such as unobvious feature edge extraction, partial hollowness of the reduced point cloud, and the like.
Disclosure of Invention
The present disclosure aims to solve at least one of the above problems, and provides a point cloud reduction method, device and storage medium based on saliency variation.
In order to achieve the above object, according to an aspect of the present disclosure, there is provided a point cloud reduction method based on a saliency variation, the point cloud reduction method based on the saliency variation including the steps of:
step 110, acquiring an original point cloud, and establishing a kd-tree of the original point cloud;
step 120, establishing a curve Bian Bianfen index of all data points of the original point cloud according to the kd-tree;
step 130, dividing all data points of the original point cloud into a salient region and a non-salient region according to a salient division principle;
step 140, extracting the salient region and storing the salient region;
step 150, simplifying the non-salient region by a space dichotomy to obtain a simplified non-salient region;
and step 160, fusing the salient regions and the simplified non-salient regions to obtain a final simplified result.
Further, in the step 120, the curved edge variation index of all the data points of the original point cloud is established according to the kd-Tree, specifically includes,
step 121, defining the acquired original point cloud as U, wherein all data points in U are U i ＝[x i ,y i ,z i ]Searching u through the kd-Tree index i K near points of (2);
step 122, each u i Is regarded as a local surface P i ；
Step 123, calculating P i Average point of (2)
Step 124, according to P i Andcalculation matrix->
Step 125, build P i Corresponding covariance matrix
Step 126, for the covariance matrix M i Singular value decomposition is carried out to obtain singular value lambda i1 ,λ i2 ,λ i3 Wherein lambda is i1 ＞λ i2 ＞λ i3 ；
Step 127, according to the singular value λ i1 ,λ i2 ,λ i3 Calculate each point u i Is a curved surface variation delta i ，
Step 128, each point u i According to its curved surface variation delta i And (5) sorting the sizes of the objects to obtain a sorting result.
Further, in the step 130, all data points of the original point cloud are divided into a salient region and a non-salient region according to the principle of saliency division, specifically including the following,
step 131, define salient region threshold, ρ
Wherein N is the number of point clouds of the original point cloud U, alpha is an adjusting factor, delta (U) i ) Data point u i Is a curved surface variation of (2);
further, α is specifically any one of 2, 5, and 7.
Step 132, dividing the original point cloud U into a salient region U according to ρ sig And non-salient region U unsig ，
Further, in the step 150, the non-significant area is reduced by spatial dichotomy to obtain a reduced non-significant area, which specifically includes the following steps,
step 151, calculate U unsig Average point of (2)
Step 152, establishing non-salient region U according to principal component analysis unsig Corresponding covariance matrix M' 3×3 As will be described below,
wherein n is U unsig The number of data points;
step 153, calculating M 'according to singular value decomposition' 3×3 Singular value λ' 1 ,λ' 2 ,λ' 3 (λ' 1 ＞λ' 2 ＞λ' 3 ) And its corresponding feature vector xi 1 ',ξ 2 ',ξ 3 '；
Step 154, define f i Is U (U) unsig The distance of each data point in (c) from the long axis normal plane,
step 155, according to f i U is determined by positive and negative unsig Divided into subsets U 1 ,U 2 ：
Step 156, introducing a point threshold n max And a surface variation threshold delta max Continuously pair subset U 1 ,U 2 Bisecting until the final subset U k ，0<k<n satisfies U k The number of the middle points is less than n max Or curve variation delta (U) k ) Below delta max Stopping bisection to obtain all final subsets U k ，0<k<n；
Step 157, for each U k Calculating the representative point p as the final reduced result,
step 158, integrate all U' s k Is represented by (A) is non-salient region U unsig The binary reduced result of (1) is marked as U J ；
The long axis normal plane in step 154 is specifically,
to be used forAs the center point, xi 3 ' establishing a fitting plane for the normal vector->ξ 1 The' direction is the fitting plane +.>Long axis direction, ζ 2 The' direction is the fitting plane +.>Is arranged in the short axis direction of the lens.
The application also provides a point cloud simplifying device based on the significance variation, which comprises:
the original point cloud acquisition module is used for acquiring an original point cloud and establishing a kd-tree of the original point cloud;
the curve edge variation index building module is used for building a curve Bian Bianfen index of all data points of the original point cloud according to the kd-Tree;
the region dividing module is used for dividing all data points of the original point cloud into a salient region and a non-salient region according to a salient dividing principle;
the salient region extraction module is used for extracting the salient region and storing the salient region;
the non-salient region simplifying module is used for simplifying the non-salient region by a space dichotomy to obtain a simplified non-salient region;
and the fusion module is used for fusing the salient region and the simplified non-salient region to obtain a final simplified result.
The application proposes a computer readable storage medium storing a computer program which, when executed by a processor, implements the steps of a saliency-based point cloud reduction method as described.
The beneficial effects of the present disclosure are: according to the application, the original point cloud data is divided into a salient region and a non-salient region according to a salient variation principle, and the salient region is reserved; then, the non-salient region is reduced by a space dichotomy method, and a non-salient region reduction result is obtained; and finally, the significant area and the non-significant area are integrated to obtain a final simplified result, so that the original advantages of the space dichotomy are maintained, the extraction capacity of the characteristic edge is greatly improved, the robustness is good, and the method is suitable for most point cloud data.
Drawings
The above and other features of the present disclosure will become more apparent from the detailed description of the embodiments illustrated in the accompanying drawings, in which like reference numerals designate like or similar elements, and which, as will be apparent to those of ordinary skill in the art, are merely some examples of the present disclosure, from which other drawings may be made without inventive effort, wherein:
FIG. 1 is a flow chart of a point cloud reduction method based on saliency variation;
FIG. 2 is a flowchart of a related algorithm of a point cloud reduction method based on saliency variation;
FIG. 3 is a schematic view of an origin cloud according to a first embodiment of the present application;
FIG. 4 is a schematic diagram of an original point cloud according to the first embodiment after being simplified by a conventional dichotomy;
fig. 5 is a schematic diagram of a point cloud after the original point cloud of the first embodiment is processed and simplified by a point cloud reduction method based on saliency variation according to the present application.
Detailed Description
The conception, specific structure, and technical effects produced by the present disclosure will be clearly and completely described below in connection with the embodiments and the drawings to fully understand the objects, aspects, and effects of the present disclosure. It should be noted that, without conflict, the embodiments of the present application and features of the embodiments may be combined with each other.
Referring to fig. 1 and 2, the present disclosure proposes a point cloud reduction method based on a saliency variation, which includes the following steps:
step 110, acquiring an original point cloud, and establishing a kd-tree of the original point cloud;
step 120, establishing a curve Bian Bianfen index of all data points of the original point cloud according to the kd-tree;
step 130, dividing all data points of the original point cloud into a salient region and a non-salient region according to a salient division principle;
step 140, extracting the salient region and storing the salient region;
step 150, simplifying the non-salient region by a space dichotomy to obtain a simplified non-salient region;
and step 160, fusing the salient regions and the simplified non-salient regions to obtain a final simplified result.
In the step 120, the curved edge variation indexes of all the data points of the original point cloud are established according to the kd-Tree, which is a preferred embodiment of the present application,
step 121, defining the obtained original point cloudU, all data points in U are U i ＝[x i ,y i ,z i ]Searching u through the kd-Tree index i K near points of (2);
step 122, each u i Is regarded as a local surface P i ；
Step 123, calculating P i Average point of (2)
Step 124, according to P i Andcalculation matrix->
Step 125, build P i Corresponding covariance matrix
Step 126, for the covariance matrix M i Singular value decomposition is carried out to obtain singular value lambda i1 ,λ i2 ,λ i3 Wherein lambda is i1 ＞λ i2 ＞λ i3 ；
Step 127, according to the singular value λ i1 ,λ i2 ,λ i3 Calculate each point u i Is a curved surface variation delta i ，
Step 128, each point u i According to its curved surface variation delta i And (5) sorting the sizes of the objects to obtain a sorting result.
As a preferred embodiment of the present application, in the above step 130, all data points of the original point cloud are divided into a salient region and a non-salient region according to a salient division principle, specifically including the following,
step 131, define salient region threshold, ρ
Wherein N is the number of point clouds of the original point cloud U, alpha is an adjusting factor, delta (U) i ) Data point u i Is a curved surface variation of (2);
in a preferred embodiment of the present application, α is specifically any one of 2, 5, and 7, and is generally 2, 5, or 7, although it is needless to say that the value may be adjusted according to the actual situation.
Step 132, dividing the original point cloud U into a salient region U according to ρ sig And non-salient region U unsig ，
As a preferred embodiment of the present application, in the step 150, the non-significant area is reduced by spatial dichotomy to obtain a reduced non-significant area, which specifically includes,
step 151, calculate U unsig Average point of (2)
Step 152, establishing non-salient region U according to principal component analysis unsig Corresponding covariance matrix M' 3×3 As will be described below,
wherein n is U unsig The number of data points;
step 153, calculating M 'according to singular value decomposition' 3×3 Singular value λ' 1 ,λ' 2 ,λ' 3 (λ' 1 ＞λ' 2 ＞λ' 3 ) And its corresponding feature vector xi 1 ',ξ 2 ',ξ 3 '；
Step 154, define f i Is U (U) unsig The distance of each data point in (c) from the long axis normal plane,
step 155, according to f i U is determined by positive and negative unsig Divided into subsets U 1 ,U 2 ：
Step 156, introducing a point threshold n max And a surface variation threshold delta max Continuously pair subset U 1 ,U 2 Bisecting until the final subset U k ，0<k<n satisfies U k The number of the middle points is less than n max Or curve variation delta (U) k ) Below delta max Stopping bisection to obtain all final subsets U k ，0<k<n；
Step 157, for each U k Calculating the representative point p as the final reduced result,
step 158, integrate all U' s k Is represented by (A) is non-salient region U unsig The binary reduced result of (1) is marked as U J ；
The long axis normal plane in step 154 is specifically,
to be used forAs the center point, xi 3 ' establishing a fitting plane for the normal vector->ξ 1 The' direction is the fitting plane +.>Long axis direction, ζ 2 The' direction is the fitting plane +.>Is arranged in the short axis direction of the lens.
In the preferred embodiment, only the final fusion U is required sig And U J And obtaining a final simplifying result.
Referring to fig. 3, fig. 4 and fig. 5, for the first embodiment of the present application, it can be seen that the point cloud significantly reduced by the conventional dichotomy is not as much as the point cloud significantly reduced by the present method, and the following advantages can be found by comparison:
the algorithm optimizes the traditional spatial dichotomy based on the principle of significance variation, so that the advantages of the original method are maintained, including high speed, high reduction ratio and good shape retention;
the extraction proportion of the salient region is controllable and can be adjusted according to actual needs;
the feature extraction effect is greatly enhanced, and a new idea is provided for point cloud simplification.
The preferred embodiments of the present application are now integrated to yield the following process steps,
1. acquiring an original point cloud U through a scanning instrument;
2. and establishing a kd-Tree index of the original point cloud.
3. For each point U in U i ＝[x i ,y i ,z i ]K adjacent points (K points in total) are found through the kd-Tree index.
4. Each point u i K proximities (K points in total) of (a) are regarded as a local surface P i ，
5. Calculation of P i Average point of (2)
6. According to P i Andcalculation matrix->
7. Establishing P i Corresponding covariance matrix
8. For covariance matrix M i Singular value decomposition (Singular Value Decompositon, SVD) is performed to obtain singular value lambda i1 ,λ i2 ,λ i3 (λ i1 ＞λ i2 ＞λ i3 )。
9. According to singular value lambda i1 ,λ i2 ,λ i3 Approximate calculation of each Point u i Is a curved surface variation delta i ：
10. Each point u i According to its curved surface variation delta i Is a size ordering of (2).
11. Defining a salient region threshold ρ:
wherein N is the number of point clouds of the original point cloud U, alpha is an adjusting factor, and the values are generally 2, 5 and 7.
12. Dividing the original point cloud U into a significant region U according to ρ sig And non-salient region U unsig ：
13. Extracting significant region U sig And stored.
14. For non-salient region U unsig Performing spatial dichotomy (Binary Space Partitioning, BSP) is reduced.
The flow is as follows (15-22):
15. calculation U unsig Average point of (2)
16. Establishing non-salient regions U from Principal Component Analysis (PCA) unsig Corresponding covariance matrix (n points):
17. calculation of M 'from singular value decomposition' 3×3 Singular value λ' 1 ,λ' 2 ,λ' 3 And corresponding feature vector ζ 1 ,ξ 2 ,ξ 3 。
18. Definition f i Is U (U) unsig Distance of each point of (2) from the long axis normal plane:
19. according to f i U is determined by positive and negative unsig Divided into subsets U 1 ,U 2 ：
20. Introducing a point threshold n max And a surface variation threshold delta max Continue to subset U 1 ,U 2 Bisecting until the final subset U k (0<k<n) one of two iteration termination conditions is satisfied:
(1)U k the number of medium points cannot exceed n max ；
(2) Curved surface variation delta (U) k ) Cannot exceed delta max 。
21. For each U k Calculating the representative point p as a final reduction result:
22. integrate all U' s k Is represented by (A) is non-salient region U unsig The binary reduced result of (1) is marked as U J 。
23. Fusion U sig And U J And the final simplifying result is obtained.
In addition, a significant variation is explained, wherein the significant variation is that the original point cloud data is divided into a significant region and a non-significant region according to a surface variation (surface variation) of the point cloud data. The salient region is the characteristic region that we want.
The application also provides a point cloud simplifying device based on the significance variation, which comprises:
the original point cloud acquisition module is used for acquiring an original point cloud and establishing a kd-tree of the original point cloud;
the curve edge variation index building module is used for building a curve Bian Bianfen index of all data points of the original point cloud according to the kd-Tree;
the region dividing module is used for dividing all data points of the original point cloud into a salient region and a non-salient region according to a salient dividing principle;
the salient region extraction module is used for extracting the salient region and storing the salient region;
the non-salient region simplifying module is used for simplifying the non-salient region by a space dichotomy to obtain a simplified non-salient region;
and the fusion module is used for fusing the salient region and the simplified non-salient region to obtain a final simplified result.
The application proposes a computer readable storage medium storing a computer program which, when executed by a processor, implements the steps of a saliency-based point cloud reduction method as described.
The modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical modules, i.e., may be located in one place, or may be distributed over a plurality of network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist alone physically, or two or more modules may be integrated into one module. The integrated modules may be implemented in hardware or in software functional modules.
The integrated modules, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on this understanding, the present application may implement all or part of the flow of the method of the above embodiment, or may be implemented by a computer program to instruct related hardware, where the computer program may be stored in a computer readable storage medium, and when the computer program is executed by a processor, the computer program may implement the steps of each of the method embodiments described above. Wherein the computer program comprises computer program code which may be in source code form, object code form, executable file or some intermediate form etc. The computer readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), an electrical carrier signal, a telecommunications signal, a software distribution medium, and so forth. It should be noted that the computer readable medium may include content that is subject to appropriate increases and decreases as required by jurisdictions in which such content is subject to legislation and patent practice, such as in certain jurisdictions in which such content is not included as electrical carrier signals and telecommunication signals.
While the present application has been described in considerable detail and with particularity with respect to several described embodiments, it is not intended to be limited to any such detail or embodiments or any particular embodiment, but is to be construed as providing broad interpretation of such claims by reference to the appended claims in view of the prior art so as to effectively encompass the intended scope of the application. Furthermore, the foregoing description of the application has been presented in its embodiments contemplated by the inventors for the purpose of providing a useful description, and for the purposes of providing a non-essential modification of the application that may not be presently contemplated, may represent an equivalent modification of the application.
The present application is not limited to the above embodiments, but is merely preferred embodiments of the present application, and the present application should be construed as being limited to the above embodiments as long as the technical effects of the present application are achieved by the same means. Various modifications and variations are possible in the technical solution and/or in the embodiments within the scope of the application.
Claims (6)
1. The point cloud simplifying method based on the saliency variation is characterized by comprising the following steps of:
step 110, acquiring an original point cloud, and establishing a kd-tree of the original point cloud;
step 120, establishing a curve Bian Bianfen index of all data points of the original point cloud according to the kd-tree;
step 130, dividing all data points of the original point cloud into a salient region and a non-salient region according to a salient division principle;
step 140, extracting the salient region and storing the salient region;
step 150, simplifying the non-salient region by a space dichotomy to obtain a simplified non-salient region;
step 160, fusing the significant areas and the simplified non-significant areas to obtain a final simplified result;
in the step 150, the non-significant area is reduced by spatial dichotomy to obtain a reduced non-significant area, which specifically includes the following steps,
step 151, calculate U unsig Average point of (2)
Step 152, establishing non-salient region U according to principal component analysis unsig Corresponding covariance matrix M' 3×3 As will be described below,
wherein n is U unsig The number of data points;
step 153, calculating M 'according to singular value decomposition' 3×3 Singular value λ' 1 ，λ′ 2 ，λ′ 3 And its corresponding feature vector xi' 1 ,ξ′ 2 ,ξ′ 3 Wherein lambda' 1 ＞λ′ 2 ＞λ′ 3 ；
Step 154, define f i Is U (U) unsig The distance of each data point in (c) from the long axis normal plane,
step 155, according to f i U is determined by positive and negative unsig Divided into subsets U 1 ,U 2 ：
Step 156, introducing a point threshold n max And a surface variation threshold delta max Continuously pair subset U 1 ,U 2 Bisecting until the final subset U k ，0<k<n satisfies U k The number of the middle points is less than n max Or curve variation delta (U) k ) Below delta max Stopping bisection to obtain all final subsets U k ，0<k<n；
Step 157 ofEach U k Calculating the representative point p as the final reduced result,
step 158, integrate all U' s k Is represented by (A) is non-salient region U unsig The binary reduced result of (1) is marked as U J ；
The long axis normal plane in step 154 is specifically,
to be used forIs taken as a center point, xi' 3 Establishing a fitting plane for the normal vector>ξ′ 1 The direction is fitting plane +.>In the long axis direction, ζ' 2 The direction is fitting plane +.>Is arranged in the short axis direction of the lens.
2. The method for point cloud compaction based on saliency variation according to claim 1, wherein the creating of the curved edge variation index of all data points of the original point cloud according to kd-Tree in the step 120 comprises the steps of,
step 121, defining the acquired original point cloud as U, wherein all data points in U are U i ＝[x i ,y i ,z i ]Finding u through kd-Tree index i K near points of (2);
step 122, each u i Is regarded as a local surface P i ；
Step 123, calculating P i Average point of (2)
Step 124, according to P i Andcalculation matrix->
Step 125, build P i Corresponding covariance matrix
Step 126, for the covariance matrix M i Singular value decomposition is carried out to obtain singular value lambda i1 ,λ i2 ,λ i3 Wherein lambda is i1 ＞λ i2 ＞λ i3 ；
Step 127, according to the singular value λ i1 ,λ i2 ,λ i3 Calculate each point u i Is a curved surface variation delta i ，
Step 128, each point u i According to its curved surface variation delta i And (5) sorting the sizes of the objects to obtain a sorting result.
3. The method according to claim 2, wherein in the step 130, all data points of the original point cloud are divided into salient regions and non-salient regions according to a salient division principle, specifically including,
step 131, define salient region threshold, ρ
Wherein N is the number of point clouds of the original point cloud U, alpha is an adjusting factor, delta (U) i ) Data point u i Is a curved surface variation of (2);
step 132, dividing the original point cloud U into a salient region U according to ρ sig And non-salient region U unsig ，
4. A point cloud reduction method based on saliency variation according to claim 3, wherein α is specifically any one of values 2, 5, and 7.
5. The utility model provides a point cloud retrencies device based on saliency variation which characterized in that includes:
the original point cloud acquisition module is used for acquiring an original point cloud and establishing a kd-tree of the original point cloud;
the curve edge variation index building module is used for building a curve Bian Bianfen index of all data points of the original point cloud according to the kd-Tree;
the region dividing module is used for dividing all data points of the original point cloud into a salient region and a non-salient region according to a salient dividing principle;
the salient region extraction module is used for extracting the salient region and storing the salient region;
the non-salient region simplifying module is used for simplifying the non-salient region by a space dichotomy to obtain a simplified non-salient region;
the fusion module is used for fusing the salient region and the simplified non-salient region to obtain a final simplified result;
wherein the non-salient region is reduced by a spatial dichotomy to obtain a reduced non-salient region, which comprises the following steps,
step 151, calculate U unsig Average point of (2)
Step 152, establishing non-salient region U according to principal component analysis unsig Corresponding covariance matrix M' 3×3 As will be described below,
wherein n is U unsig The number of data points;
step 153, calculating M 'according to singular value decomposition' 3×3 Singular value λ' 1 ，λ′ 2 ，λ′ 3 And its corresponding feature vector xi' 1 ,ξ′ 2 ,ξ′ 3 Wherein lambda' 1 ＞λ′ 2 ＞λ′ 3 ；
Step 154, define f i Is U (U) unsig The distance of each data point in (c) from the long axis normal plane,
step 155, according to f i U is determined by positive and negative unsig Divided into subsets U 1 ,U 2 ：
Step 156, introducing a point threshold n max And a surface variation threshold delta max Continuously pair subset U 1 ,U 2 Bisecting until the final subset U k ，0<k<n satisfies U k The number of the middle points is less than n max Or curve variation delta (U) k ) Below delta max Stopping bisection to obtain all final subsets U k ，0<k<n；
Step 157, for eachU (U) k Calculating the representative point p as the final reduced result,
step 158, integrate all U' s k Is represented by (A) is non-salient region U unsig The binary reduced result of (1) is marked as U J ；
The long axis normal plane in step 154 is specifically,
to be used forAs the center point, xi 3 ' establishing a fitting plane for the normal vector->ξ 1 The' direction is the fitting plane +.>Long axis direction, ζ 2 The' direction is the fitting plane +.>Is arranged in the short axis direction of the lens.
6. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program which, when executed by a processor, implements the steps of a saliency-based point cloud reduction method according to any of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011619176.7A CN112613528B (en) | 2020-12-31 | 2020-12-31 | Point cloud simplifying method and device based on significance variation and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011619176.7A CN112613528B (en) | 2020-12-31 | 2020-12-31 | Point cloud simplifying method and device based on significance variation and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112613528A CN112613528A (en) | 2021-04-06 |
| CN112613528B true CN112613528B (en) | 2023-11-03 |
Family
ID=75249694
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011619176.7A Active CN112613528B (en) | 2020-12-31 | 2020-12-31 | Point cloud simplifying method and device based on significance variation and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112613528B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115002470A (en) * | 2022-05-27 | 2022-09-02 | 腾讯科技（深圳）有限公司 | Media data processing method, device, equipment and readable storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107341825A (en) * | 2017-07-06 | 2017-11-10 | 西南科技大学 | A kind of method for simplifying for large scene high-precision three-dimensional laser measurement cloud data |
| CN108830931A (en) * | 2018-05-23 | 2018-11-16 | 上海电力学院 | A kind of laser point cloud compressing method based on dynamic grid k neighborhood search |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8175376B2 (en) * | 2009-03-09 | 2012-05-08 | Xerox Corporation | Framework for image thumbnailing based on visual similarity |
| CN106295542A (en) * | 2016-08-03 | 2017-01-04 | 江苏大学 | A kind of road target extracting method of based on significance in night vision infrared image |
| US20180053040A1 (en) * | 2016-08-19 | 2018-02-22 | Umm Al-Qura University | System and method for 3d local surface matching |
-
2020
- 2020-12-31 CN CN202011619176.7A patent/CN112613528B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107341825A (en) * | 2017-07-06 | 2017-11-10 | 西南科技大学 | A kind of method for simplifying for large scene high-precision three-dimensional laser measurement cloud data |
| CN108830931A (en) * | 2018-05-23 | 2018-11-16 | 上海电力学院 | A kind of laser point cloud compressing method based on dynamic grid k neighborhood search |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112613528A (en) | 2021-04-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109887015B (en) | Point cloud automatic registration method based on local curved surface feature histogram | |
| CN109166145B (en) | Fruit tree leaf growth parameter extraction method and system based on cluster segmentation | |
| CN110032925B (en) | Gesture image segmentation and recognition method based on improved capsule network and algorithm | |
| CN113077476B (en) | Height measurement method, terminal device and computer storage medium | |
| CN113628263A (en) | Point cloud registration method based on local curvature and neighbor characteristics thereof | |
| CN112613528B (en) | Point cloud simplifying method and device based on significance variation and storage medium | |
| CN111145129A (en) | Point cloud denoising method based on hyper-voxels | |
| CN111652855A (en) | Point cloud simplification method based on survival probability | |
| CN111340862A (en) | Point cloud registration method and device based on multi-feature fusion and storage medium | |
| CN115049925A (en) | Method for extracting field ridge, electronic device and storage medium | |
| CN114549525B (en) | Industrial image detection method based on improved canny algorithm | |
| KR101296734B1 (en) | Method for object segmentation by combination of gaussian mixture model and rgb clustering | |
| CN115131424A (en) | Clustering and boundary feature-preserving automobile trunk point cloud volume calculation method | |
| CN117132630A (en) | Point cloud registration method based on second-order spatial compatibility measurement | |
| CN115937151A (en) | Method for judging curling degree of crop leaves | |
| CN106683105A (en) | Image segmentation method and image segmentation device | |
| CN112270746B (en) | Aluminum alloy 3D printing point cloud simplifying algorithm based on neighborhood covariance characteristic parameter threshold | |
| CN113963010A (en) | Object contour line extraction method and system | |
| CN110334226B (en) | Depth image retrieval method fusing feature distribution entropy | |
| CN112329662A (en) | Multi-view saliency estimation method based on unsupervised learning | |
| CN116486371A (en) | Obstacle detection method and obstacle detection device based on laser point cloud | |
| CN109840924B (en) | Product image rapid generation method based on series countermeasure network | |
| CN112102376B (en) | Multi-view cloud registration method, device and storage medium of hybrid sparse ICP | |
| CN116977593A (en) | Single wood segmentation method based on super-voxel concave-convex segmentation and color region growth | |
| CN114299586A (en) | Intelligent deep learning system based on convolutional neural network |
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 |