CN103714555A - Four-dimensional motion point cloud segmentation and reconstruction method based on motion track - Google Patents
Four-dimensional motion point cloud segmentation and reconstruction method based on motion track Download PDFInfo
- Publication number
- CN103714555A CN103714555A CN201310684707.4A CN201310684707A CN103714555A CN 103714555 A CN103714555 A CN 103714555A CN 201310684707 A CN201310684707 A CN 201310684707A CN 103714555 A CN103714555 A CN 103714555A
- Authority
- CN
- China
- Prior art keywords
- data
- carried out
- frame
- motion
- movement locus
- 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
- 230000033001 locomotion Effects 0.000 title claims abstract description 70
- 238000000034 method Methods 0.000 title claims abstract description 47
- 230000011218 segmentation Effects 0.000 title abstract 2
- 238000000605 extraction Methods 0.000 claims abstract description 17
- 238000007621 cluster analysis Methods 0.000 claims description 22
- 238000013075 data extraction Methods 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 5
- 230000008878 coupling Effects 0.000 claims description 3
- 238000010168 coupling process Methods 0.000 claims description 3
- 238000005859 coupling reaction Methods 0.000 claims description 3
- 230000003595 spectral effect Effects 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Images
Landscapes
- Length Measuring Devices By Optical Means (AREA)
- Length Measuring Devices With Unspecified Measuring Means (AREA)
Abstract
The invention relates to the field of computer graphics and provides a four-dimensional motion point cloud segmentation and reconstruction method based on a motion track. Skeleton extraction of a motion track is realized. The method comprises the steps of acquiring motion data, carrying out registration on adjacent data, extracting motion track data, carrying out clustering analysis on the motion track data and carrying out consistent skeleton extraction on the clustering-analyzed motion data. A template does not need to be defined in advance, skeleton extraction does not need to be carried out on data of each frame, and the requirement on the quality of point cloud is low.
Description
Technical field
The present invention relates to area of computer graphics, particularly relate to a kind of four-dimensional movement point cloud based on movement locus and cut apart and method for reconstructing.
Background technology
Existingly motion-capturedly need masterplate or higher to a cloud quality requirements with cutting apart, or extract skeleton from single frames point cloud, carry out consistance optimization and the processing of multiframe skeleton.
Prior art is cut apart great majority based on a cloud geometric properties, the continuity in shortage time and space and consistance to object; The point cloud skeleton method for reconstructing of existing pursuit space-time consistency, higher to the quality requirements of input point cloud, because they need to extract coarse skeleton from the sequential point cloud of input, then carry out the consistance optimization in time and space.
Summary of the invention
The present invention adopts a kind of four-dimensional movement point cloud based on movement locus to cut apart and method for reconstructing, realized the skeletal extraction of movement locus, the present invention does not need predefined masterplate, does not need the data of each frame first to carry out skeletal extraction yet, lower to a cloud quality requirements.The present invention adopts following scheme:
Four-dimensional movement point cloud based on movement locus is cut apart and a method for reconstructing, comprising:
S1, obtain exercise data;
S2, described exercise data is carried out to consecutive frame Registration of Measuring Data;
S3, to carrying out motion trace data extraction by the data after step S2 registration;
S4, the motion trace data described in step S3 is carried out to cluster analysis;
S5, the data that obtain after cluster analysis described in step S4 are carried out to consistance skeletal extraction.
Preferably, described in, obtain exercise data, the exercise data catching method of employing based on laser scanner.
Preferably, laser scanner is with continuous the carrying out cloud data scanning and obtain moving object of certain frame per second; Data after obtaining are stored in computer with the form of a file of every frame.
Preferably, described exercise data is carried out to the method for consecutive frame Registration of Measuring Data for adopting the method for non-rigid coupling to realize, described non-rigid matching process is that each point in each frame is found corresponding point in next frame.
Preferably, to carry out the method for motion trace data extraction by the data after step S2 registration, be, the method that adopts depth-first is the similarly movement locus growth that is connected between adjacent two frames of adjacent direction, thereby obtains between multiframe movement locus between points.
Preferably, the method for the motion trace data described in step S3 being carried out to cluster analysis is that the range formula based on movement locus, adopts spectral clustering or kmean clustering procedure, according to the similarity of motion trace data, moving object is divided into different motion parts.
Preferably, it is characterized in that, the data that obtain are carried out to consistance skeletal extraction also comprise that the part to having cut apart calculates the step of conforming skeleton according to syntople after cluster analysis described in step S4.
Preferably, the method that calculates conforming skeleton according to syntople is, the node to skeleton, from adjacent node and adjacent node, to the corresponding point of next frame, find corresponding relation, thereby the key node of skeleton is transformed into next frame, and iteration is carried out, and can be transformed in N frame data.
A kind of four-dimensional movement point cloud based on movement locus disclosed by the invention is cut apart and method for reconstructing, by obtaining exercise data, carry out adjacent data registration, motion trace data is extracted, motion trace data cluster analysis, and the exercise data after cluster analysis carries out the extraction of consistance share price, realized the skeletal extraction of movement locus, the present invention does not need predefined masterplate, does not need the data of each frame first to carry out skeletal extraction yet, lower to a cloud quality requirements.
Accompanying drawing explanation
Fig. 1 is that 1 one kinds of the embodiment of the present invention four-dimensional movement point cloud based on movement locus is cut apart and method for reconstructing process flow diagram;
Fig. 2 is that the embodiment of the present invention 1 four frame point clouds are as example;
Fig. 3 is the example after the embodiment of the present invention 1 consecutive frame Registration of Measuring Data;
Fig. 4 be 1 of the embodiment of the present invention with aim at after the line relation of point;
Fig. 5 is movement locus between points between the embodiment of the present invention 1 multiframe;
Fig. 6 is the distance definition exploded view directly perceived of the embodiment of the present invention 1 different tracks;
Fig. 7 is the some movement locus before the embodiment of the present invention 1 cluster analysis;
Fig. 8 is the movement locus after the embodiment of the present invention 1 cluster analysis.
Embodiment
In order to make object of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.
The embodiment of the present invention provides a kind of four-dimensional movement point cloud based on movement locus to cut apart and method for reconstructing, comprising:
S1, obtain exercise data;
S2, described exercise data is carried out to consecutive frame Registration of Measuring Data;
S3, to carrying out motion trace data extraction by the data after step S2 registration;
S4, the motion trace data described in step S3 is carried out to cluster analysis.
S5, the data that obtain after cluster analysis described in step S4 are carried out to consistance skeletal extraction.
The embodiment of the present invention is by obtaining exercise data, carry out adjacent data registration, motion trace data is extracted, motion trace data cluster analysis, exercise data after cluster analysis carries out the extraction of consistance share price, has realized the skeletal extraction of movement locus, and the present invention does not need predefined masterplate, do not need the data of each frame first to carry out skeletal extraction yet, lower to a cloud quality requirements.Below the present invention will be described in detail.
embodiment 1:
Refer to shown in Fig. 1, for a kind of four-dimensional movement point cloud based on movement locus of the present invention is cut apart and method for reconstructing process flow diagram.The method comprises the steps:
S1, obtain exercise data.
The present embodiment provides four frame point clouds as example, as Fig. 2, adopts the exercise data catching method based on laser scanner, and laser scanner is with continuous the carrying out cloud data scanning and obtain moving object of certain frame per second; Data after obtaining are stored in computer with the form of a file of every frame.
S2, described exercise data is carried out to consecutive frame Registration of Measuring Data.
Exercise data is carried out to consecutive frame Registration of Measuring Data, adopt the method for non-rigid coupling to realize, matching process is that each point in each frame is found corresponding point in next frame, as Fig. 3, in every frame, right-hand point is the point after registration, point with aim at after the line relation of point as Fig. 4.
S3, to carrying out motion trace data extraction by the data after step S2 registration.
To carrying out motion trace data extraction by the data after step S2 registration, the method that adopts depth-first is the similarly movement locus growth that is connected between adjacent two frames of adjacent direction, thereby obtains between multiframe movement locus between points, as Fig. 5.
S4, the motion trace data described in step S3 is carried out to cluster analysis.
Motion trace data described in step S3 is carried out to cluster analysis, and the range formula based on movement locus, adopts spectral clustering or kmean clustering procedure, according to the similarity of motion trace data, moving object is divided into different motion parts.
Similarity between motion trace data is calculated and is adopted Euclidean distance algorithm, the Euclidean distance between consideration track and the relation between direction of motion.During cluster analysis, choose at random N bar track as cluster centre, then similar track data is constantly added in the middle of nearest cluster, upgrade cluster centre simultaneously, till this step iteration proceeds to all tracks and met other conditions of user's appointment by cluster work.The definition of different distance is calculated and is made with the following method:
D1=Euclidean distance;
D2=angle α;
D=normalization (d1)+normalization (d2);
The distance definition of different tracks is intuitively shown as Fig. 6.
Some movement locus before cluster analysis are as Fig. 7, and the movement locus after cluster analysis, as Fig. 8.
S5, the data that obtain after cluster analysis described in step S4 are carried out to consistance skeletal extraction.
According to syntople, calculate conforming skeleton, comprise the node to skeleton, from adjacent node and adjacent node, to the corresponding point of next frame, find corresponding relation, thereby the key node of skeleton is transformed into next frame, iteration is carried out, and can be transformed in N frame data.
The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any modifications of doing within the spirit and principles in the present invention, be equal to and replace and improvement etc., within all should being included in protection scope of the present invention.
Claims (8)
1. the four-dimensional movement point cloud based on movement locus is cut apart and a method for reconstructing, it is characterized in that, comprising:
S1, obtain exercise data;
S2, described exercise data is carried out to consecutive frame Registration of Measuring Data;
S3, to carrying out motion trace data extraction by the data after step S2 registration;
S4, the motion trace data described in step S3 is carried out to cluster analysis;
S5, the data that obtain after cluster analysis described in step S4 are carried out to consistance skeletal extraction.
2. method according to claim 1, is characterized in that, described in obtain exercise data, adopt the exercise data catching method based on laser scanner.
3. method according to claim 2, is characterized in that, laser scanner is with continuous the carrying out cloud data scanning and obtain moving object of certain frame per second; Data after obtaining are stored in computer with the form of a file of every frame.
4. method according to claim 1, it is characterized in that, described exercise data is carried out to the method for consecutive frame Registration of Measuring Data for adopting the method for non-rigid coupling to realize, described non-rigid matching process is that each point in each frame is found corresponding point in next frame.
5. method according to claim 1, it is characterized in that, to carry out the method for motion trace data extraction by the data after step S2 registration, be, the method that adopts depth-first is the similarly movement locus growth that is connected between adjacent two frames of adjacent direction, thereby obtains between multiframe movement locus between points.
6. method according to claim 1, it is characterized in that, the method of the motion trace data described in step S3 being carried out to cluster analysis is, range formula based on movement locus, adopt spectral clustering or kmean clustering procedure, according to the similarity of motion trace data, moving object is divided into different motion parts.
7. method according to claim 1, is characterized in that, the data that obtain is carried out to consistance skeletal extraction also comprise that the part to having cut apart calculates the step of conforming skeleton according to syntople after cluster analysis described in step S4.
8. method according to claim 7, it is characterized in that, the method that calculates conforming skeleton according to syntople is, node to skeleton, from adjacent node and adjacent node, to the corresponding point of next frame, find corresponding relation, thereby the key node of skeleton is transformed into next frame, and iteration is carried out, and can be transformed in N frame data.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310684707.4A CN103714555B (en) | 2013-12-13 | 2013-12-13 | A kind of four-dimensional motion point cloud segmentation and method for reconstructing based on movement locus |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310684707.4A CN103714555B (en) | 2013-12-13 | 2013-12-13 | A kind of four-dimensional motion point cloud segmentation and method for reconstructing based on movement locus |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103714555A true CN103714555A (en) | 2014-04-09 |
CN103714555B CN103714555B (en) | 2017-06-13 |
Family
ID=50407498
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310684707.4A Active CN103714555B (en) | 2013-12-13 | 2013-12-13 | A kind of four-dimensional motion point cloud segmentation and method for reconstructing based on movement locus |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103714555B (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104504897A (en) * | 2014-09-28 | 2015-04-08 | 北京工业大学 | Intersection traffic flow characteristic analysis and vehicle moving prediction method based on trajectory data |
CN105006007A (en) * | 2015-07-23 | 2015-10-28 | 北京理工大学 | Data-driven action reconstruction method |
CN105912983A (en) * | 2016-04-01 | 2016-08-31 | 北京理工大学 | Motion reconstruction method based on sparse input signals |
CN106204635A (en) * | 2016-06-27 | 2016-12-07 | 北京工业大学 | Based on L0the human body successive frame skeleton optimization method minimized |
CN110378904A (en) * | 2018-07-09 | 2019-10-25 | 北京京东尚科信息技术有限公司 | The method and apparatus that point cloud data is split |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102467753A (en) * | 2010-11-04 | 2012-05-23 | 中国科学院深圳先进技术研究院 | Method and system for reconstructing time-varying point cloud based on framework registration |
CN103218817A (en) * | 2013-04-19 | 2013-07-24 | 深圳先进技术研究院 | Partition method and partition system of plant organ point clouds |
CN103268631A (en) * | 2013-05-23 | 2013-08-28 | 中国科学院深圳先进技术研究院 | Method and device for extracting point cloud framework |
-
2013
- 2013-12-13 CN CN201310684707.4A patent/CN103714555B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102467753A (en) * | 2010-11-04 | 2012-05-23 | 中国科学院深圳先进技术研究院 | Method and system for reconstructing time-varying point cloud based on framework registration |
CN103218817A (en) * | 2013-04-19 | 2013-07-24 | 深圳先进技术研究院 | Partition method and partition system of plant organ point clouds |
CN103268631A (en) * | 2013-05-23 | 2013-08-28 | 中国科学院深圳先进技术研究院 | Method and device for extracting point cloud framework |
Non-Patent Citations (1)
Title |
---|
袁冠等: "《基于结构相似度的轨迹聚类算法》", 《通 信 学 报》 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104504897A (en) * | 2014-09-28 | 2015-04-08 | 北京工业大学 | Intersection traffic flow characteristic analysis and vehicle moving prediction method based on trajectory data |
CN105006007A (en) * | 2015-07-23 | 2015-10-28 | 北京理工大学 | Data-driven action reconstruction method |
CN105006007B (en) * | 2015-07-23 | 2018-06-29 | 北京理工大学 | A kind of action method for reconstructing of data-driven |
CN105912983A (en) * | 2016-04-01 | 2016-08-31 | 北京理工大学 | Motion reconstruction method based on sparse input signals |
CN105912983B (en) * | 2016-04-01 | 2019-05-24 | 北京理工大学 | Movement method for reconstructing based on sparse input signal |
CN106204635A (en) * | 2016-06-27 | 2016-12-07 | 北京工业大学 | Based on L0the human body successive frame skeleton optimization method minimized |
CN106204635B (en) * | 2016-06-27 | 2018-11-30 | 北京工业大学 | Based on L0The human body successive frame bone optimization method of minimum |
CN110378904A (en) * | 2018-07-09 | 2019-10-25 | 北京京东尚科信息技术有限公司 | The method and apparatus that point cloud data is split |
CN110378904B (en) * | 2018-07-09 | 2021-10-01 | 北京京东尚科信息技术有限公司 | Method and device for segmenting point cloud data |
Also Published As
Publication number | Publication date |
---|---|
CN103714555B (en) | 2017-06-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111192292B (en) | Target tracking method and related equipment based on attention mechanism and twin network | |
CN104376596B (en) | A kind of three-dimensional scene structure modeling and register method based on single image | |
EP3234806B1 (en) | Scalable 3d mapping system | |
CN103714555A (en) | Four-dimensional motion point cloud segmentation and reconstruction method based on motion track | |
US9697609B2 (en) | Method and apparatus for estimating pose | |
CN104317391B (en) | A kind of three-dimensional palm gesture recognition exchange method and system based on stereoscopic vision | |
WO2017214595A1 (en) | Systems and methods for performing three-dimensional semantic parsing of indoor spaces | |
CN110163239B (en) | Weak supervision image semantic segmentation method based on super-pixel and conditional random field | |
CN109829356B (en) | Neural network training method and pedestrian attribute identification method based on neural network | |
CN105493078B (en) | Colored sketches picture search | |
CN110555412B (en) | End-to-end human body gesture recognition method based on combination of RGB and point cloud | |
CN103957397B (en) | A kind of low resolution depth image top sampling method based on characteristics of image | |
CN103218827B (en) | The contour tracing method of segmentation and figure matching and correlation is combined in Shape-based interpolation transmission | |
CN104063723A (en) | Stroke reduction method of offline handwritten Chinese character and device thereof | |
Liu et al. | Automatic motion capture data denoising via filtered subspace clustering and low rank matrix approximation | |
CN103186787A (en) | Low-quality Chinese character primary skeleton extraction algorithm based on point cloud model | |
CN105046689A (en) | Method for fast segmenting interactive stereo image based on multilayer graph structure | |
CN103942778A (en) | Fast video key frame extraction method of principal component characteristic curve analysis | |
CN104091336A (en) | Stereoscopic image synchronous segmentation method based on dense disparity map | |
Wang et al. | Tc-sfm: Robust track-community-based structure-from-motion | |
CN103247052A (en) | Image segmentation algorithm for local region characteristics through nonsubsampled contourlet transform | |
CN103871060A (en) | Smooth direction wave domain probability graph model-based image segmentation method | |
CN104657944A (en) | Compressed sensing remote sensing image reconstruction method based on reference image texture constraints | |
CN102270338B (en) | Method for effectively segmenting repeated object based on image representation improvement | |
CN101510317A (en) | Method and apparatus for generating three-dimensional cartoon human face |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
EE01 | Entry into force of recordation of patent licensing contract | ||
EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20140409 Assignee: SHENZHEN YIHUO TECHNOLOGY CO.,LTD. Assignor: SHENZHEN INSTITUTES OF ADVANCED TECHNOLOGY CHINESE ACADEMY OF SCIENCES Contract record no.: X2023980043401 Denomination of invention: A four-dimensional motion point cloud segmentation and reconstruction method based on motion trajectory Granted publication date: 20170613 License type: Common License Record date: 20231013 |