CN100337250C - Method for detecting collision between body stand and flexible object - Google Patents

Method for detecting collision between body stand and flexible object Download PDF

Info

Publication number
CN100337250C
CN100337250C CNB2004100865597A CN200410086559A CN100337250C CN 100337250 C CN100337250 C CN 100337250C CN B2004100865597 A CNB2004100865597 A CN B2004100865597A CN 200410086559 A CN200410086559 A CN 200410086559A CN 100337250 C CN100337250 C CN 100337250C
Authority
CN
China
Prior art keywords
control shaft
local distance
manikin
point
summit
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.)
Expired - Fee Related
Application number
CNB2004100865597A
Other languages
Chinese (zh)
Other versions
CN1766931A (en
Inventor
毛天露
王兆其
夏时洪
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Institute of Computing Technology of CAS
Original Assignee
Institute of Computing Technology of CAS
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Institute of Computing Technology of CAS filed Critical Institute of Computing Technology of CAS
Priority to CNB2004100865597A priority Critical patent/CN100337250C/en
Publication of CN1766931A publication Critical patent/CN1766931A/en
Application granted granted Critical
Publication of CN100337250C publication Critical patent/CN100337250C/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Image Analysis (AREA)

Abstract

The invention relates to a collision detecting method between a human body model and a flexible object, which comprises the following steps: 1, the preprocessing stage is to establish a local distance table on the surface of a human body model and record the distance information between each area of the surface of the human body model and the relative bones on the bottom layer according to certain order; 2, each top point/ surface area of a flexible object is detected, the vertical point position of the current top point/surface area on each control shaft 1 is obtained, and through the discussion of the vertical point positions and the search of the local distance table, a user can judge weather the current top point/ surface area has position collisions with the surface of the human body model at present. The present has the advantages that the requirements of a computer generating space to reality are satisfied, the requirements of a practical application system to speed is satisfied, the time and the complexity of the collision detection are only correlative to the number of the top points on the flexible object model, and the time and the complexity of the calculation of the collision detection are invariant under the condition that the human body model moves acutely.

Description

Collision detection method between a kind of manikin and flexible article
Technical field
The present invention relates to field of Computer Graphics, specifically, relate to the method for collision detection between a kind of manikin and flexible article.
Background technology
In real world, human body always contacts with various flexible articles, as the dress ornament of body worn, and the mattress quilt that sleep is used etc.If will in the computing machine span, really describe the objective world, comprise various personages, article and their motion in the objective world, need to manikin with its on every side relevant flexible article carry out motion simulation or animation simulation.Wherein one of key issue is in emulation or animation simulation process manikin and flexible article to be carried out the collision detection of locus.
In the computing machine span, flexible article and manikin are represented with polygonal mesh usually.In order to obtain the stronger sense of reality, a simple flexible article, as the model of skirt, its summit number is usually at several thousand orders of magnitude, and the dough sheet number of common manikin is usually at several ten thousand orders of magnitude.Polygonal mesh is the geometric model of no thickness, any trickle verisimilitude that all can influence visual effect that penetrates.Carry out the method for collision detection in the prior art, speed have levels faster bounding box method, volume elements method, but the computation complexity of these methods is still higher, is difficult to requirement of real time.For some real-time application, how guaranteeing under the real-time prerequisite that it is one of the research difficult point in current graphics field that thousands of summit and dough sheet are carried out the collision detection of accurate locus.
Summary of the invention
The objective of the invention is to provides a kind of locus fast, accurately collision detection method for manikin and flexible article in the computing machine span.
To achieve these goals, the collision detection method between manikin provided by the invention and flexible article comprises the steps:
1) detects pretreatment stage before, set up the local distance table on the manikin surface, the local distance table has write down the range information between each zone of human body surface and the bottom associated bone in certain sequence, this bottom associated bone is by Control Shaft 1 expression, and Control Shaft 1 is the line segment in join dependency bone two adjacent joints;
21) each summit on the traversal flexible article for each position of traversal human body, current summit, is made vertical line from current summit to working as front Control Shaft 1 place straight line, and is tried to achieve vertical point;
22) according to the position of point of hanging down, and search the local distance table, judge whether occurrence positions conflicts when the front with manikin on flexible article fooled preceding summit.
Described step 1) comprises following substep:
11) according to the characteristics of current manikin, model is carried out the position divide, draw the Control Shaft 1 at each position;
12) travel through each position, the geometric jacquard patterning unit surface when the front is carried out area dividing;
13) each regional area is called local distance to the distance of its position, place Control Shaft 1 with this distance on the computational geometry surface;
141) all local distances are deposited in certain sequence in the local distance table at this position;
142) will work as the maximal value of local distance on each section of front, deposit in the local distance table by the section order;
143) will select that the maximal value of all local distances deposits in the local distance table in the position;
144) in the beginning record position of local distance table sequence number, toponym, position section number, the number of rays of each section.
Also comprise following substep in the described step 12):
121), equally spaced make the section 9 vertical with Control Shaft 1 along this position Control Shaft 1 for working as the front;
122) traveling through each section, for current section, is the center with the intersection point of this section and its Control Shaft 1, as start ray 10, is at interval with the equal angular with the fixed-direction of agreement, outwards makes ray equably;
123) with step 122) in the intersection point on the ray of doing and surface, position, place characterize each regional area.
Described step 22) also comprise following substep:
221) in step 21) in tried to achieve the position of the point that hangs down, if the point that hangs down is positioned at beyond the Control Shaft, then can judge current summit with work as the front not occurrence positions conflict; Point is positioned on the Control Shaft if hang down, and then tries to achieve the sequence number of two sections adjacent with vertical line according to the vertical point and the distance at Control Shaft two ends;
222) search when front local distance table, if current summit to the distance of the point that hangs down greater than local distance maximum on adjacent two sections, then can judge current summit with work as the front not occurrence positions conflict; Otherwise,, try to achieve the sequence number of four rays adjacent with vertical line again according to the initial direction of making ray under vertical line direction, Control Shaft position and the current motion state when the front;
223) search when front local distance table, if current summit to the distance of the point that hangs down local distance, then can judge greater than adjacent four rays current summit with work as the front not occurrence positions conflict; Otherwise, judge current summit and taken place when the front or be about to occurrence positions to conflict.
Collision detection method between manikin provided by the invention and flexible article, its advantage is:
(a) satisfy the requirement of the computing machine span to verisimilitude, collision detection method energy provided by the invention load comprises the manikin and the flexible article model of a large amount of dough sheets.
(b) satisfy the requirement of real application systems to speed, the dough sheet number of establishing manikin is m, and the summit number on the flexible article model is n, and then the time complexity of collision detection is O (n), and is only relevant with summit number n on the flexible article model.
(c) be under the situation of strenuous exercise at manikin, the time complexity that collision detection is calculated is constant.
Under human motion state, owing to need not to upgrade the local distance table at each position, consistent under the method for therefore carrying out collision detection and the human body stationary state, algorithm complex still is O (n).And the summit number that the complexity of algorithm is conflicted with actual generation locus is irrelevant, and therefore, in the practical application that a large amount of conflicts take place, for example virtual costume is performed, and the complexity of its algorithm still remains on O (n) order of magnitude.Manikin described in the invention and the collision detection method between flexible article, can the accurate detection manikin under various states (static, motion), and the position conflict situations between the flexible article (as clothes, curtain etc.) on every side satisfies the needs of various real-time application.
Description of drawings
Fig. 1 is the synoptic diagram of a certain position of human body and associated bone thereof;
Fig. 2 is that the position between a certain position of human body and its Control Shaft concerns synoptic diagram in the motion process;
Fig. 3 is a position view of asking local distance process midship section and ray;
Fig. 4 is that several possible position at summit and a certain position of human body concerns synoptic diagram on the flexible article;
Fig. 5 is the example of manikin and flexible article in the computing machine span;
Fig. 6 is for generating the process flow diagram of local distance table among the present invention;
Fig. 7 is the process flow diagram of concrete testing process once among the present invention.
Embodiment
Generally speaking, in the computing machine span, each moment is owing to reasons such as human motion, air flow, spatial relation between manikin and the flexible article can change, and therefore need the locus conflict situations between them be detected again.
Because the characteristics of the physiological structure of human body, can be divided into several positions to manikin, wherein all there is an associated bone in each position, it serves as the axle rotation with an end points of this bone that associated bone can drive its position, place, associated bone is called Control Shaft, its end points as turning axle is called the father joint of this Control Shaft.As shown in Figure 1, be example with the position right leg 5 of manikin, its Control Shaft 1 is the line of right knee joint 2 to right ankle-joint 3, right knee joint 2 is the father joint of Control Shaft 1.Fig. 2 is that the position between a certain position of human body and its Control Shaft concerns synoptic diagram in the motion process, and wherein 1 is Control Shaft, and 4 is the father joint, and 6 is a surperficial dough sheet at this position, and 7 is the original position at this position, 8 be this position around the father joint 4 postrotational positions.As can be seen, whole position produces identical displacement and rotation together with its Control Shaft in motion process, and these characteristics have determined the distance between manikin arbitrary zone, surface and its Control Shaft can suppose to remain unchanged.
Utilize this hypothesis just, the present invention has introduced this notion of local distance table, and the local distance table has write down the range information between the relevant bottom bone with it of each regional area of human body surface in certain sequence.Because the distance between hypothesis manikin arbitrary zone, surface and its Control Shaft remains unchanged in motion process, no matter which kind of athletic posture human body is in, and the local distance information in each position local distance table all remains unchanged.
In specific implementation process, method of the present invention comprises following two steps: at first, the pretreatment stage before detecting beginning, surperficial at manikin for setting up the local distance table in each position.Secondly, in testing process,,, judge the position conflict situations between this moment flexible article and the manikin fast by searching the local distance table according to the distance between current time flexible article and the human skeleton.The present invention will be further described below in conjunction with accompanying drawing and embodiment.
1, the foundation of local distance table
The foundation of local distance table is to carry out at pretreatment stage, does not take the time overhead in the testing process.Set up the local distance table and may further comprise the steps (as shown in Figure 6):
Step 101: manikin is carried out the position divide, for each position, the definition associated bone, this associated bone is represented with the line segment that connects two adjacent segment point, is called Control Shaft.
Step 102 a: position in the selected manikin, (4 is the father joint among the figure) as shown in Figure 3, equally spaced make the section 9 vertical along this position Control Shaft 1 with Control Shaft 1, this size at interval can be arranged along the axial length of control to accuracy requirement with when the front according to application system.Generally speaking, the section number at a position is between 6~12.
Step 103: on each section, being the center with this section and Control Shaft 1 intersection point, as start ray 10, is at interval with the fixed-direction of agreement with the equal angular, and ray outwards is set uniformly.Try to achieve the intersection point of all rays and human body surface, see Fig. 3.Angle intervals between the ray also can be arranged according to the accuracy requirement of application system.Generally speaking, the number of rays on each section is between 6~18.
Step 104: calculate all rays of selected position from starting point to this ray and the distance the human body surface intersection point, this distance is become local distance.All local distances at selected position are deposited in the local distance table at this position in certain sequence.Table 1 is the example of a local distance table.
Step 105: will select the maximal value of local distance on each section at position, and deposit in the local distance table by the section order.
Step 106: will select that the maximal value of all local distances deposits in the local distance table in the position.
Step 107: the local distance that judges whether in the local distance table, to have deposited in all sites in the manikin.If be judged as be, enter step 108,, get back to step 102 if be judged as not, selected next position, repeated execution of steps 102 is to step 106.
Step 108: the local distance of all sites in the manikin has all deposited the local distance table in, and the local distance table is set up and finished.
Be noted that in the local distance table, depositing in proper order of local distance information can be carried out artificial agreement according to the subsequent query needs.Beginning general record position sequence number, toponym, the position section number of table, the number of rays of each section.
Table 1
The position sequence number 2
Toponym Right leg
The section number 12
Single section number of rays 10
The local distance of No. 1 ray on No. 1 section 0.1105
The local distance of No. 2 rays on No. 1 section 0.1110
... 0.1116
... 0.1158
... 0.1178
... 0.1201
... 0.1183
... 0.1165
... 0.1134
The local distance of No. 10 rays on No. 1 section 0.1112
Maximum local distance on No. 1 section 0.1201
... ...
... ...
The maximum local distance in position 0.1339
2, the detection of locus conflict situations
Concrete each time testing process may further comprise the steps (as shown in Figure 7):
Step 200: each summit on the traversal flexible article, for each position of traversal human body, current summit.
Step 201: make vertical line from current summit to working as Control Shaft place, front straight line, and try to achieve vertical point.Fig. 4 is that several possible position at summit and a certain position of human body concerns synoptic diagram on the flexible article, and wherein 1 is Control Shaft, P1, P2, P3 is respectively three summits on the flexible article, and regional A, B represent respectively corresponding to summit P2, the human body surface regional area of P3.Cross P1 respectively, P2, P3 makes the vertical line of Control Shaft 1, the some P1 ' that obtains hanging down, P2 ', P3 '.
Step 202: if the point that hangs down is positioned at beyond the Control Shaft, then can judge current summit with when the front not occurrence positions conflict (being the P1 situation among Fig. 4), get back to step 201 and detect next and organize data; Point is arranged in (being Fig. 4 P2 or P3 situation) on the Control Shaft if hang down, and enters step 203.
Step 203: calculating the distance at hang down point and Control Shaft two ends, searching local distance table, trying to achieve the sequence number of two sections adjacent with vertical line when the front.
Step 204: search when front local distance table, if current summit to the point that hangs down apart from d greater than local distance (the P2 situation among Fig. 4) maximum on adjacent two sections, then can judge current summit with when the front not occurrence positions conflict, get back to step 201 and detect next group data; Otherwise, enter step 205.
Step 205: according under vertical line direction and the current motion state when the ray initial direction of front, and the sequence number of adjacent two sections of having tried to achieve is tried to achieve the sequence number of four rays adjacent with vertical line.In step 204, obtained two sections adjacent, searched, can obtain four adjacent on these two sections rays (two adjacent rays are wherein arranged on each section) respectively with current summit when front local distance table with current summit.Be noted that under motion state, may take place with the Control Shaft when the front is rotatablely moving of rotating shaft, the ray initial direction of therefore working as the front should be that the acquiescence initial direction in the local distance table adds the angle when the front rotation.
Step 206: search when front local distance table, if current summit to the distance of the point that hangs down maximal value, then can judge greater than adjacent four ray local distances current summit with when the front not occurrence positions conflict, get back to step 201 and detect next and organize data; Otherwise, judge current summit and taken place when the front or be about to occurrence positions to conflict, enter step 207.
Step 207: judging whether to travel through each summit on the flexible article and the combination at each position of human body, is to enter step 208 if be judged as; If be judged as not, get back to step 201 and detect next group data.
Step 208: obtain testing result, the collision detection process finishes.
From above narration as can be seen, the collision detection between manikin and the flexible article can be decomposed into the collision detection between the single summit and the single position of manikin on the flexible article.If the dough sheet number of manikin is m, summit number on the flexible article model is n, generally m>>n, the position number of human body is generally a less constant (generally speaking less than 15), therefore (time that is collision detection is only relevant with the summit number n on the flexible article model at O (n) order of magnitude to utilize the present invention to carry out the time complexity of collision detection between manikin and the flexible article, irrelevant with the size of the dough sheet number m of manikin), be significantly less than present existed algorithms.
Under human motion state, to show apart from height owing to need not to upgrade the part at each position, consistent under the method for therefore carrying out collision detection and the human body stationary state, algorithm complex still is O (n).And the summit number that the complexity of algorithm is conflicted with actual generation locus is irrelevant, and therefore, in the practical application that a large amount of conflicts take place, for example virtual costume is performed, and the complexity of its algorithm still remains on O (n) order of magnitude.
Manikin described in the invention and the collision detection method between flexible article, can the accurate detection manikin under various states (static, motion), and the position conflict situations between the flexible article (as clothes, curtain etc.) on every side satisfies the needs of various real-time application.

Claims (5)

1, the collision detection method between a kind of manikin and flexible article is characterized in that, comprises the steps:
1) detects pretreatment stage before, set up the local distance table on the manikin surface, the local distance table has write down the range information between each zone of human body surface and the bottom associated bone in certain sequence, this bottom associated bone is by Control Shaft (1) expression, and Control Shaft (1) is the line segment in join dependency bone two adjacent joints;
21) each summit on the traversal flexible article for each position of traversal human body, current summit, is made vertical line from current summit to working as front Control Shaft (1) place straight line, and is tried to achieve vertical point;
22) according to the position of point of hanging down, and search the local distance table, judge whether occurrence positions conflicts when the front with manikin on flexible article fooled preceding summit.
2, by the collision detection method between described manikin of claim 1 and flexible article, described step 1) is characterized in that, comprises following substep:
11) according to the characteristics of current manikin, model is carried out the position divide, draw the Control Shaft (1) at each position;
12) travel through each position, the geometric jacquard patterning unit surface when the front is carried out area dividing;
13) each regional area is called local distance to the distance of its position, place Control Shaft (1) with this distance on the computational geometry surface;
141) all local distances are deposited in certain sequence in the local distance table at this position;
142) will work as the maximal value of local distance on each section of front, deposit in the local distance table by the section order;
143) will select that the maximal value of all local distances deposits in the local distance table in the position.
3, by the collision detection method between described manikin of claim 2 and flexible article, described step 12) is characterized in that, also comprises following substep:
121) for working as the front, equally spaced make the section (9) vertical with Control Shaft (1) along this position Control Shaft (1);
122) traveling through each section, for current section, is the center with the intersection point of this section and its Control Shaft (1), as start ray (10), is at interval with the equal angular with the fixed-direction of agreement, outwards makes ray equably;
123) with step 122) in the intersection point on the ray of doing and surface, position, place characterize each regional area.
4, by the collision detection method between described manikin of claim 2 and flexible article, finish described step 143) after, also have following substep:
144) in the beginning record position of local distance table sequence number, toponym, position section number, the number of rays of each section.
5, by the collision detection method between the described manikin of claim 1 and flexible article, described step 22) also comprise following substep:
221) in step 21) in tried to achieve the position of the point that hangs down, if the point that hangs down is positioned at beyond the Control Shaft, then can judge current summit with work as the front not occurrence positions conflict; Point is positioned on the Control Shaft if hang down, and then tries to achieve the sequence number of two sections adjacent with vertical line according to the vertical point and the distance at Control Shaft two ends;
222) search when front local distance table, if current summit to the distance of the point that hangs down greater than local distance maximum on adjacent two sections, then can judge current summit with work as the front not occurrence positions conflict; Otherwise,, try to achieve the sequence number of four rays adjacent with vertical line again according to the initial direction of making ray under vertical line direction, Control Shaft position and the current motion state when the front;
223) search when front local distance table, if current summit to the distance of the point that hangs down local distance, then can judge greater than adjacent four rays current summit with work as the front not occurrence positions conflict; Otherwise, judge current summit and taken place when the front or be about to occurrence positions to conflict.
CNB2004100865597A 2004-10-26 2004-10-26 Method for detecting collision between body stand and flexible object Expired - Fee Related CN100337250C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CNB2004100865597A CN100337250C (en) 2004-10-26 2004-10-26 Method for detecting collision between body stand and flexible object

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CNB2004100865597A CN100337250C (en) 2004-10-26 2004-10-26 Method for detecting collision between body stand and flexible object

Publications (2)

Publication Number Publication Date
CN1766931A CN1766931A (en) 2006-05-03
CN100337250C true CN100337250C (en) 2007-09-12

Family

ID=36742797

Family Applications (1)

Application Number Title Priority Date Filing Date
CNB2004100865597A Expired - Fee Related CN100337250C (en) 2004-10-26 2004-10-26 Method for detecting collision between body stand and flexible object

Country Status (1)

Country Link
CN (1) CN100337250C (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101140663B (en) * 2007-10-16 2010-06-02 中国科学院计算技术研究所 Clothing cartoon computation method
CN100592342C (en) * 2008-03-21 2010-02-24 中国科学院计算技术研究所 A creation method of candidate set for physical body-clothe confliction detection

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1166220A (en) * 1995-10-09 1997-11-26 任天堂株式会社 Stereo image processing system
CN1407511A (en) * 2001-09-08 2003-04-02 袁铭辉 Tree-dimensional garment developing system
WO2003031005A2 (en) * 2001-10-09 2003-04-17 Massachusetts Institute Of Technology Methods and apparatus for detecting and correcting penetration between objects
US6714213B1 (en) * 1999-10-08 2004-03-30 General Electric Company System and method for providing interactive haptic collision detection

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1166220A (en) * 1995-10-09 1997-11-26 任天堂株式会社 Stereo image processing system
US6714213B1 (en) * 1999-10-08 2004-03-30 General Electric Company System and method for providing interactive haptic collision detection
CN1407511A (en) * 2001-09-08 2003-04-02 袁铭辉 Tree-dimensional garment developing system
WO2003031005A2 (en) * 2001-10-09 2003-04-17 Massachusetts Institute Of Technology Methods and apparatus for detecting and correcting penetration between objects

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
计算机仿真 束林扬,姜建国,第123页至第124页,织物仿真中的冲突检测和处理 2003 *

Also Published As

Publication number Publication date
CN1766931A (en) 2006-05-03

Similar Documents

Publication Publication Date Title
CN102490909B (en) Multi-body separating analogy method of aircraft
CN106845515A (en) Robot target identification and pose reconstructing method based on virtual sample deep learning
Schvartzman et al. Star-contours for efficient hierarchical self-collision detection
CN109101932B (en) Multi-task and proximity information fusion deep learning method based on target detection
JP6320016B2 (en) Object detection apparatus, object detection method and program
Hutter et al. Optimized continuous collision detection for deformable triangle meshes
CN1797474A (en) Fast method for posting players to electronic game
CN110990913A (en) Side span shape finding method adopting horizontal plane internal force control
CN107479693A (en) Real-time hand recognition methods based on RGB information, storage medium, electronic equipment
CN100337250C (en) Method for detecting collision between body stand and flexible object
CN104391482B (en) Longitudinal tensile loading trajectory design method based on mold surface extension
CN101739713A (en) Method for detecting parallel collision based on subdivision
CN110634149B (en) Non-rigid target characteristic point matching method for optical motion capture system
CN110349176A (en) Method for tracking target and system based on triple convolutional networks and perception interference in learning
CN106023317A (en) Weighted Voronoi diagram generation method used for big data test
CN1567383A (en) Method for three-dimensional animation production using literal script
CN111523576B (en) Density peak clustering outlier detection method suitable for electron quality detection
CN103235862B (en) Select disassembly sequence planning method and device
CN104298345A (en) Control method for man-machine interaction system
Najmaei et al. A new sensory system for modeling and tracking humans within industrial work cells
CN105303602A (en) Automatic 3D human body model bone binding process and method
CN105014966A (en) Hanging pre-detection method and system for 3D printing model
CN104408685A (en) Collision elimination method of large-scale object group mixing CPU (central processing unit) and GPU (graphics processing unit)
CN104992161B (en) A kind of Hanzi component segmentation and structural determination method based on part identification
CN102663674B (en) Parallelization real-time unmarked human body motion tracking method

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20070912

Termination date: 20161026

CF01 Termination of patent right due to non-payment of annual fee