CN103886115B - The foundation of a kind of three-dimensional virtual human platform based on different building shape and call method thereof - Google Patents
The foundation of a kind of three-dimensional virtual human platform based on different building shape and call method thereof Download PDFInfo
- Publication number
- CN103886115B CN103886115B CN201210556599.8A CN201210556599A CN103886115B CN 103886115 B CN103886115 B CN 103886115B CN 201210556599 A CN201210556599 A CN 201210556599A CN 103886115 B CN103886115 B CN 103886115B
- Authority
- CN
- China
- Prior art keywords
- angle
- buttocks
- chest
- central point
- sides
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 38
- 210000000038 chest Anatomy 0.000 claims abstract description 118
- 210000000481 breast Anatomy 0.000 claims abstract description 51
- 238000005259 measurement Methods 0.000 claims abstract description 38
- 230000000877 morphologic effect Effects 0.000 claims abstract description 16
- 230000007935 neutral effect Effects 0.000 claims abstract description 12
- 230000000392 somatic effect Effects 0.000 claims abstract description 10
- 238000007619 statistical method Methods 0.000 claims abstract description 7
- 210000001217 buttock Anatomy 0.000 claims description 106
- 230000003187 abdominal effect Effects 0.000 claims description 37
- 238000013528 artificial neural network Methods 0.000 claims description 21
- 210000001015 abdomen Anatomy 0.000 claims description 20
- 238000012549 training Methods 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 6
- 210000003141 lower extremity Anatomy 0.000 claims description 4
- 238000012546 transfer Methods 0.000 claims description 4
- 230000009466 transformation Effects 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 3
- 241001269238 Data Species 0.000 claims 1
- 230000011218 segmentation Effects 0.000 claims 1
- 238000013461 design Methods 0.000 abstract description 8
- 238000006243 chemical reaction Methods 0.000 abstract description 4
- 238000004519 manufacturing process Methods 0.000 abstract description 4
- 230000008929 regeneration Effects 0.000 abstract description 3
- 238000011069 regeneration method Methods 0.000 abstract description 3
- 230000006870 function Effects 0.000 description 6
- 239000011159 matrix material Substances 0.000 description 5
- 241000208340 Araliaceae Species 0.000 description 2
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 2
- 235000003140 Panax quinquefolius Nutrition 0.000 description 2
- 235000008434 ginseng Nutrition 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses foundation and the call method thereof of a kind of three-dimensional virtual human platform based on different building shape.Described method includes following four steps: 1, classify the buman body type obtained with height/chest measurement, side form, breast waist difference for index, and uses the method averaged to obtain the somatic data of each build intermediate;2, BP neutral net is built, it is achieved anthropometric data is to the conversion of build modeling parameters;3, the build interface document of each build intermediate is write, it is achieved the virtual reappearance of each build intermediate;4, by statistical analysis, height/chest measurement index, people's body side surface morphological index and the discrimination formula of breast waist difference index are set up respectively;Differentiated by formula, it is achieved respective virtual people's platform is called.The present invention can realize the regeneration of dummy body form and call respective virtual people tableland, can realize being carried out the conjunction Dam Configuration Design of clothing by dummy body form, provides a kind of new method for meeting clothing MTM production, can be widely applied to custom made clothing platform.
Description
Technical field
The present invention relates to foundation and the call method thereof of a kind of three-dimensional virtual human platform based on different building shape, belong to clothing and produce
Industry digitizing technique field.
Background technology
Along with the raising of people's living standard, the clothing MTM mode of production based on clothing personalized designs is arisen at the historic moment.Clothes
In the personalized designs of dress, the zoariumization design of garment dimension based on client's build is a very important aspect.Therefore,
As emerging garment fit evaluation means, three-dimensional based on customization virtual human body and customization virtual costume tries
Clothing, of increased attention.How by several body dimension simple and easy to get, generate the virtual of respective client build
People's platform, thus carrying out the conjunction Dam Configuration Design of clothing by this people's platform is the big study hotspot in this area one.
Summary of the invention
It is an object of the invention to provide foundation and the call method thereof of a kind of three-dimensional virtual human platform based on different building shape, with
Realize being carried out by dummy body form the conjunction Dam Configuration Design of clothing, produce a kind of new method of offer for meeting clothing MTM.
For achieving the above object, the technical solution used in the present invention is as follows:
The foundation of a kind of three-dimensional virtual human platform based on different building shape and call method thereof, comprise the steps:
The first step, for index, the buman body type obtained is classified with height/chest measurement, side form, breast waist difference, and transport
The somatic data of each build intermediate is obtained by the method averaged;
The Changing Pattern of the virtual human body in second step, analyzing three-dimensional fitting software, with physical characteristic parameter for input ginseng
Number, be applied to three-dimensional modeling shape parameter as output parameter, build BP neutral net, it is achieved anthropometric data arrives
The conversion of build modeling parameters;
3rd step, analyze the coded system of virtual human body interface document in this three-dimensional fitting software, write in the middle of each build
The build interface document of body, it is achieved the virtual reappearance of each build intermediate;
4th step, by statistical analysis, set up height/chest measurement index, people's body side surface morphological index and breast waist difference respectively and refer to
Target discrimination formula;Differentiated by formula, it is achieved respective virtual people's platform is called.
As a kind of preferred version, described side morphological index refers to that people's body side surface abdomen salient point is to the line of breast salient point and water
The angle of the angle of plane and people's body side surface buttocks salient point line to back of the body salient point and horizontal plane.
As further preferred scheme, the acquisition of described side morphological index is at Matlab (matrix labotstory (matrix
Laboratory) under environment, use Roberts edge detection operator (one of operator), find the edge of people's body side surface photo,
By matrix operations identification matrix of edge salient point, by salient point coordinate, it is thus achieved that described side morphological index.
As further preferred version, before identifying matrix of edge salient point, first image is divided into two by horizontal 1:1,
It is divided into two by longitudinal 3:4.
As a kind of preferred version, the operation of the first step includes:
A) use Scanworx spatial digitizer (one of scanner in high precision), automatically obtain whole somatometries
According to, and manually measure and the calculating of derived variable increases by 36 measure the items below human body: front fillet, rear fillet, average
Fillet, the distance/front length of side neck point to breast point, BP dot spacing/chest measurement, abdomen coign, the angle of Attacking Midfielder's central point to chest both sides,
Side waist central point to the angle of chest both sides, Attacking Midfielder's central point to chest both sides angle/side waist central point to chest both sides
Angle, the vertical angle of Attacking Midfielder to breast, the angle of Attacking Midfielder's central point to buttocks both sides, the folder of side waist central point to buttocks both sides
Angle, the angle/angle of side waist central point to buttocks both sides of Attacking Midfielder's central point to buttocks both sides, Attacking Midfielder's central point are to buttocks both sides
The angle/angle of Attacking Midfielder's central point chest both sides, side waist central point to buttocks both sides angle/side waist central point to chest two
The angle of side, low back to buttocks vertical angle, the back of the body length/height, the vertical angle of side knee center point to waist central point, side
Waist central point hits exactly angle and low back to collare angle before the angle of buttocks central point, lower body shaft angle, upper half health shaft angle, upper body;
B) with height/chest measurement for index human body is divided into height, in, short three classes, the most a height of relative chest measurement is high, in be relative
In chest measurement, short short for relative chest measurement;With people's body side surface morphological index: the line of abdomen bump to breast salient point and the angle of horizontal plane,
Human body, to the angle of line and the horizontal plane of back of the body salient point, is divided into AC, AB, AA, BC, BB, BA, CC, CB, CA nine class by buttocks salient point;
Wherein: AC refers to that the buttocks of people's body side surface protrudes than back, chest protrudes than abdominal part;AB refers to that the buttocks of people's body side surface is more convex than back
Go out, chest and abdominal part approximate on a vertical vertical line;AA refers to that the buttocks of people's body side surface is more convex than chest than back protrusion, abdominal part
Go out;BC refers to that the buttocks of people's body side surface and back approximate on a vertical vertical line, chest protrudes than abdominal part;BB refers to people's body side surface
Buttocks and back approximate on a vertical vertical line, chest and abdominal part approximate on a vertical line;BA refers to the buttocks of people's body side surface
Approximate on a vertical vertical line with back, chest protrudes than abdominal part;CC refers to that the back of people's body side surface is than buttocks protrusion, chest ratio
Abdominal part protrudes;CB refers to that the back of people's body side surface approximates on a vertical vertical line than buttocks protrusion, chest and abdominal part;CA refers to human body
The back of side is protruded than buttocks, abdominal part protrudes than chest;It addition, by national standard with breast waist difference for index human body is divided into Y,
A, B, C tetra-class;There are 108 class builds;
C) somatic data of 108 class builds is carried out statistical analysis, after seeking its meansigma methods, i.e. obtain each build intermediate
Somatic data.
As a kind of preferred version, the operation of second step includes:
D) control some the shape adjustment Parameters variation of virtual human body in three-dimensional fitting software, observe this Parameters variation
Time, the Changing Pattern of virtual human body build, it is thus achieved that the physical characteristic parameter bigger on body surface impact;
E) select 36 idiotype characteristic parameters as the input parameter of BP neutral net, select 11 and be applied to three-dimensional
The shape parameter of modeling is output parameter, builds the initial model of BP network under Matlab environment;
F) by network training, minimax method, normalized square mean method and the most right without processing three kinds of methods are analyzed
Data be standardized processing after on the impact of neural network forecast ability, analyze the different network numbers of plies and arrange neural network forecast ability
Impact, analyze different training function setup on the impact of neural network forecast ability, analyze different node transfer function and arrange
On the impact of neural network forecast ability, analyze different the number of hidden nodes and the impact on neural network forecast ability is set, analyzes different net
Network input and output parameter arrange the impact on neural network forecast ability, obtain optimal network structure, set up physical characteristic parameter
With the shape adjustment nonlinearity in parameters mapping relations in three-dimensional fitting software.
As further preferred scheme, 36 described idiotype characteristic parameters are: front fillet, rear fillet, average shoulder angle, side
Neck point is to the distance of breast point, the distance/front length of side neck point to breast point, BP dot spacing, BP dot spacing/chest measurement, abdomen coign, Qian Yaozhong
Heart point is to the angle of chest both sides, the angle of side waist central point to chest both sides, the angle/side of Attacking Midfielder's central point to chest both sides
Waist central point is in the angle of chest both sides, the vertical angle of Attacking Midfielder to breast, the angle of Attacking Midfielder's central point to buttocks both sides, side waist
Heart point is to the angle of buttocks both sides, the angle of the angle/side waist central point to buttocks both sides of Attacking Midfielder's central point to buttocks both sides, front
Waist central point to the angle/angle of Attacking Midfielder's central point chest both sides of buttocks both sides, side waist central point to buttocks both sides angle/
Side waist central point is long to the angle of chest both sides, the vertical angle of low back to buttocks, the back of the body, back of the body length/height, side knee center point arrive
Before the vertical angle of waist central point, the angle of side waist central point to buttocks central point, lower body shaft angle, upper half health shaft angle, upper body
Angle, center, low back to collare angle, highly, enclose under neck circumference, shoulder breadth, chest measurement, breast, waistline, lower hip circumference, upper hip circumference, outer lower limb length and interior
Long.
As further preferred scheme, described 11 are applied to the shape parameter of three-dimensional modeling and are: shoulder parameter, breast
Portion's location parameter, breast peak distance parameter, abdomen parameter, health width parameter, buttocks width parameter, upper half height parameter, posture are joined
Number, lower part of the body pose parameter, bottom position parameter and upper body pose parameter.
As a kind of preferred version, the operation of the 3rd step includes:
G) coded system of the interface document of human parameters modeling in analyzing three-dimensional fitting software;
H) setting up third party software platform, built-in neural network structure, by by defeated for corresponding physical characteristic data file
Enter, through the Parameters Transformation of neutral net, automatically generate and be applied to people's platform of various build intermediate in three-dimensional fitting software and connect
Mouth file.
As a kind of preferred version, the operation of the 4th step includes:
I) under Matlab environment, the automatic identification algorithm of people's body side surface morphological index is set up: changed by side colour picture
For gray-scale map, then by Robort operator, find the edge of people's body side surface photo, read human body edge image by Matlab
The part having edge lines be 1, remaining is 0;Dividing the image into into 4 parts, the most horizontal ratio is 1:1, longitudinal
The ration of division is 3:4;Utilize the determinant of digital picture, find first method that 1 ranks occur and find every image
Salient point;Bump location after being merged by determinant, obtains the coordinate figure of each salient point;It is calculated human body by coordinate figure
Side abdomen salient point is to breast salient point line and horizontal angle, people's body side surface buttocks salient point to back of the body salient point line and horizontal folder
Angle;
J) human body classifying height/chest measurement, side morphological index, breast waist difference index, carries out each classification side respectively
The statistics of the boundary value of the classification human body that FAXIA obtains, and set up corresponding build discrimination formula;By build discrimination formula pair
Build belongs to class and differentiates, calls corresponding interface document, generates corresponding intermediate people's platform.
Compared with prior art, the present invention has following significance progress and a beneficial effect:
The present invention carries out K-Mean average classification by the somatic data obtaining three-dimensional human body measurement, and adds up corresponding
The data at each position of human body intermediate value;Adjust to the build being applied to existing three-dimensional fitting software by building characteristics of human body's parameter
Four layers of BP model of whole parameter and the coded system of the interface document of virtual human body, it is achieved that the regeneration of dummy body form;By building
The fuzzy discrimination formula of vertical each human body classification indicators, it is achieved buman body type differentiates, it is achieved that call respective virtual people tableland;This
Invention can carry out the conjunction Dam Configuration Design of clothing by dummy body form, provides a kind of new method for meeting clothing MTM production, can be wide
General it is applied to custom made clothing platform.
Detailed description of the invention
Below in conjunction with embodiment, the present invention is illustrated in further detail:
Three-dimensional fitting software described in the present embodiment is the V-stitcher software of Gerber company of the U.S., described
Matlab is Matlab7.1 version.
The foundation of a kind of based on different building shape the three-dimensional virtual human platform that the present invention provides and call method thereof, including as follows
Step:
The first step, for index, the buman body type obtained is classified with height/chest measurement, side form, breast waist difference, and transport
The somatic data of each build intermediate is obtained by the method averaged;
The Changing Pattern of the virtual human body in second step, analyzing three-dimensional fitting software, with physical characteristic parameter for input ginseng
Number, be applied to three-dimensional modeling shape parameter as output parameter, build BP neutral net, it is achieved anthropometric data arrives
The conversion of build modeling parameters;
3rd step, analyze the coded system of virtual human body interface document in this three-dimensional fitting software, write in the middle of each build
The build interface document of body, it is achieved the virtual reappearance of each build intermediate;
4th step, by statistical analysis, set up height/chest measurement index, people's body side surface morphological index and breast waist difference respectively and refer to
Target discrimination formula;Differentiated by formula, it is achieved respective virtual people's platform is called.
As a kind of preferred version, described method includes operating as follows:
The first step:
A) use Scanworx spatial digitizer (one of scanner in high precision), automatically obtain whole somatometries
According to, and manually measure and the calculating of derived variable increases by 36 measure the items below human body: front fillet, rear fillet, average
Fillet, the distance/front length of side neck point to breast point, BP dot spacing/chest measurement, abdomen coign, the angle of Attacking Midfielder's central point to chest both sides,
Side waist central point to the angle of chest both sides, Attacking Midfielder's central point to chest both sides angle/side waist central point to chest both sides
Angle, the vertical angle of Attacking Midfielder to breast, the angle of Attacking Midfielder's central point to buttocks both sides, the folder of side waist central point to buttocks both sides
Angle, the angle/angle of side waist central point to buttocks both sides of Attacking Midfielder's central point to buttocks both sides, Attacking Midfielder's central point are to buttocks both sides
The angle/angle of Attacking Midfielder's central point chest both sides, side waist central point to buttocks both sides angle/side waist central point to chest two
The angle of side, low back to buttocks vertical angle, the back of the body length/height, the vertical angle of side knee center point to waist central point, side
Waist central point hits exactly angle and low back to collare angle before the angle of buttocks central point, lower body shaft angle, upper half health shaft angle, upper body;
B) with height/chest measurement for index human body is divided into height, in, short three classes, the most a height of relative chest measurement is high, in be relative
In chest measurement, short short for relative chest measurement;With people's body side surface morphological index: the line of abdomen bump to breast salient point and the angle of horizontal plane,
Human body, to the angle of line and the horizontal plane of back of the body salient point, is divided into AC, AB, AA, BC, BB, BA, CC, CB, CA nine class by buttocks salient point;
Wherein: AC refers to that the buttocks of people's body side surface protrudes than back, chest protrudes than abdominal part;AB refers to that the buttocks of people's body side surface is more convex than back
Go out, chest and abdominal part approximate on a vertical vertical line;AA refers to that the buttocks of people's body side surface is more convex than chest than back protrusion, abdominal part
Go out;BC refers to that the buttocks of people's body side surface and back approximate on a vertical vertical line, chest protrudes than abdominal part;BB refers to people's body side surface
Buttocks and back approximate on a vertical vertical line, chest and abdominal part approximate on a vertical line;BA refers to the buttocks of people's body side surface
Approximate on a vertical vertical line with back, chest protrudes than abdominal part;CC refers to that the back of people's body side surface is than buttocks protrusion, chest ratio
Abdominal part protrudes;CB refers to that the back of people's body side surface approximates on a vertical vertical line than buttocks protrusion, chest and abdominal part;CA refers to human body
The back of side is protruded than buttocks, abdominal part protrudes than chest;It addition, by national standard with breast waist difference for index human body is divided into Y,
A, B, C tetra-class;There are 108 class builds;
C) somatic data of 108 class builds is carried out statistical analysis, after seeking its meansigma methods, i.e. obtain each build intermediate
Somatic data.
Second step:
D) control some the shape adjustment Parameters variation of virtual human body in three-dimensional fitting software, observe this Parameters variation
Time, the Changing Pattern of virtual human body build, it is thus achieved that the physical characteristic parameter bigger on body surface impact;
E) select 36 idiotype characteristic parameters as the input parameter of BP neutral net, select 11 and be applied to three-dimensional
The shape parameter of modeling is output parameter, builds the initial model of BP network under Matlab environment;
F) by network training, minimax method, normalized square mean method and the most right without processing three kinds of methods are analyzed
Data be standardized processing after on the impact of neural network forecast ability, analyze the different network numbers of plies and arrange neural network forecast ability
Impact, analyze different training function setup on the impact of neural network forecast ability, analyze different node transfer function and arrange
On the impact of neural network forecast ability, analyze different the number of hidden nodes and the impact on neural network forecast ability is set, analyzes different net
Network input and output parameter arrange the impact on neural network forecast ability, obtain optimal network structure, set up physical characteristic parameter
With the shape adjustment nonlinearity in parameters mapping relations in three-dimensional fitting software;
36 described idiotype characteristic parameters are: front fillet, rear fillet, average shoulder angle, the distance of side neck point to breast point, side
Neck point is to the distance/front length of breast point, BP dot spacing, BP dot spacing/chest measurement, abdomen coign, the folder of Attacking Midfielder's central point to chest both sides
Angle, the angle of side waist central point to chest both sides, Attacking Midfielder's central point to chest both sides angle/side waist central point to chest both sides
Angle, the vertical angle of Attacking Midfielder to breast, the angle of Attacking Midfielder's central point to buttocks both sides, the folder of side waist central point to buttocks both sides
Angle, the angle/angle of side waist central point to buttocks both sides of Attacking Midfielder's central point to buttocks both sides, Attacking Midfielder's central point are to buttocks both sides
The angle/angle of Attacking Midfielder's central point chest both sides, side waist central point to buttocks both sides angle/side waist central point to chest two
The angle of side, low back are long to vertical angle, the back of the body of buttocks, the vertical folder of back of the body length/height, side knee center point to waist central point
Angle, low back is hit exactly to collare before angle, the angle of side waist central point to buttocks central point, lower body shaft angle, upper half health shaft angle, upper body
Angle, highly, enclose under neck circumference, shoulder breadth, chest measurement, breast, waistline, lower hip circumference, upper hip circumference, outer lower limb length and interior length;
Described 11 are applied to the shape parameter of three-dimensional modeling: shoulder parameter, chest locations parameter, breast peak distance
Parameter, abdomen parameter, health width parameter, buttocks width parameter, upper half height parameter, pose parameter, lower part of the body pose parameter, bottom
Location parameter and upper body pose parameter.
3rd step:
G) coded system of the interface document of human parameters modeling in analyzing three-dimensional fitting software;
H) setting up third party software platform, built-in neural network structure, by by defeated for corresponding physical characteristic data file
Enter, through the Parameters Transformation of neutral net, automatically generate and be applied to people's platform of various build intermediate in three-dimensional fitting software and connect
Mouth file.
4th step:
I) under Matlab environment, the automatic identification algorithm of people's body side surface morphological index is set up: changed by side colour picture
For gray-scale map, then by Robort operator, find the edge of people's body side surface photo, read human body edge image by Matlab
The part having edge lines be 1, remaining is 0;Dividing the image into into 4 parts, the most horizontal ratio is 1:1, longitudinal
The ration of division is 3:4;Utilize the determinant of digital picture, find first method that 1 ranks occur and find every image
Salient point;Bump location after being merged by determinant, obtains the coordinate figure of each salient point;It is calculated human body by coordinate figure
Side abdomen salient point is to breast salient point line and horizontal angle, people's body side surface buttocks salient point to back of the body salient point line and horizontal folder
Angle;
J) human body classifying height/chest measurement, side morphological index, breast waist difference index, carries out each classification side respectively
The statistics of the boundary value of the classification human body that FAXIA obtains, and set up corresponding build discrimination formula;By build discrimination formula pair
Build belongs to class and differentiates, calls corresponding interface document, generates corresponding intermediate people's platform.
Embodiment
With height/chest measurement for index human body is divided into height, in, short three classes, the most a height of relative chest measurement is high, in be relative breast
In enclosing, short short for relative chest measurement;With people's body side surface index: the line of abdomen bump to breast salient point and the angle of horizontal plane, buttocks salient point
Line and the angle of horizontal plane to back of the body salient point;Human body is divided into AC, AB, AA, BC, BB, BA, CC, CB, CA nine class, wherein: AC
The buttocks referring to people's body side surface protrudes than back, chest protrudes than abdominal part, AB refer to the buttocks of people's body side surface protrude than back, chest and
Abdominal part approximates on a vertical vertical line, and AA refers to that the buttocks of people's body side surface protrudes than back, abdominal part protrudes than chest, and BC refers to human body
The buttocks of side and back approximate on a vertical vertical line, chest protrudes than abdominal part, and BB refers to buttocks and the back of people's body side surface
Approximation on a vertical vertical line, chest and abdominal part approximate on a vertical line, BA refers to buttocks and the back approximation of people's body side surface
On a vertical vertical line, chest protrude than abdominal part, CC refers to that the back of people's body side surface is protruded than buttocks, chest protrudes than abdominal part,
CB refers to that the back of people's body side surface approximates on a vertical vertical line than buttocks protrusion, chest and abdominal part, and CA refers to the back of the body of people's body side surface
Portion protrudes than buttocks, abdominal part protrudes than chest.With breast waist difference, human body is divided into Y, A, B, C tetra-class by national standard.There are 108
Class build.
The build data of 108 class builds are carried out statistical analysis, after seeking its meansigma methods, obtains corresponding intermediate data.
With shoulder parameter, chest locations parameter, breast peak distance parameter, abdomen parameter, health width parameter, buttocks width parameter,
Upper half height parameter, pose parameter, lower part of the body pose parameter, bottom position parameter, upper body pose parameter are 11 neutral nets
Output parameter, with 36 idiotype characteristic parameters: front fillet, rear fillet, average shoulder angle, the distance of side neck point to breast point, side neck point
To the distance/front length of breast point, BP dot spacing, BP dot spacing/chest measurement, abdomen coign, the angle of Attacking Midfielder's central point to chest both sides, side
Waist central point is to the angle of chest both sides, the folder of the angle/side waist central point to chest both sides of Attacking Midfielder's central point to chest both sides
Angle, the vertical angle of Attacking Midfielder to breast, the angle of Attacking Midfielder's central point to buttocks both sides, the angle of side waist central point to buttocks both sides,
Attacking Midfielder's central point is to the angle/angle of side waist central point to buttocks both sides of buttocks both sides, Attacking Midfielder's central point to buttocks both sides
Angle/the angle of Attacking Midfielder's central point chest both sides, side waist central point to buttocks both sides angle/side waist central point to chest both sides
Angle, the vertical angle of low back to buttocks, the back of the body be long, back of the body length/height, the vertical angle of side knee center point to waist central point,
Side waist central point to hit exactly before the angle of buttocks central point, lower body shaft angle, upper half health shaft angle, upper body angle, low back to collare angle,
Highly, enclose under neck circumference, shoulder breadth, chest measurement, breast, waistline, lower hip circumference, upper hip circumference, outer lower limb length, interior a length of input parameter,
Build four layers of BP neutral net in the environment of Matlab7.1, take minimax method that input data are normalized, net
The training function setup of network be trainlm}, and the node transfer function of network be set to logsig, purelin, logsig},
Epoch is set to 1000, and learning rate is set to 0.01, it is achieved from characteristics of human body's parameter, the build to application three-dimensional fitting software is adjusted
The conversion of whole parameter.
In analyzing three-dimensional fitting software, the coded system of the interface document of human parameters modeling, sets up third party software and puts down
Platform, built-in neural network structure, by corresponding physical characteristic data file is inputted, through the Parameters Transformation of neutral net,
Automatically generate the people's platform interface document being applied to various build intermediate in three-dimensional fitting software.
The automatic identification algorithm of people's body side surface index is set up: first by side colour picture in the environment of Matlab7.1
Be converted to gray-scale map, then by Robort operator, find the edge of people's body side surface photo, so read human body by Matlab
The part having edge lines of edge image is 1, and remaining is 0.Convex in order to quickly calculate acquisition the chest salient point of human body, back
Point, buttocks salient point and abdominal part salient point, divide the image into into 4 parts, and wherein grid scale is 1:1, and longitudinal ration of division is
3:4.Utilize the determinant of digital picture, find first method that 1 ranks occur and find the salient point of every image.Pass through
Bump location after determinant merging, obtains the coordinate figure of each salient point.It is calculated by coordinate figure: people's body side surface abdomen salient point
To breast salient point line and horizontal angle, people's body side surface buttocks salient point to back of the body salient point line and horizontal angle.
Then the human body classified height/chest measurement, side morphological index, breast waist difference index, carries out each classification respectively
The statistics of the boundary value of the classification human body obtained under method, and set up corresponding build discrimination formula.By build discrimination formula
Build is belonged to class differentiate, call corresponding interface document, generate corresponding intermediate people's platform.
The most visible: the present invention carries out K-Mean average by the somatic data obtaining three-dimensional human body measurement and divides
Class, and add up the data at the corresponding each position of human body intermediate value;By building characteristics of human body's parameter to being applied to existing three-dimensional
Four layers of BP model of the shape adjustment parameter of fitting software and the coded system of the interface document of virtual human body, it is achieved that visual human
The regeneration of platform;By setting up the fuzzy discrimination formula of each human body classification indicators, it is achieved buman body type differentiates, it is achieved that to corresponding empty
Anthropomorphic tableland is called;The present invention can carry out the conjunction Dam Configuration Design of clothing by dummy body form, provides for meeting clothing MTM production
A kind of new method, can be widely applied to custom made clothing platform.
Finally be necessary it is pointed out here that, described above be served only for technical scheme make the most specifically
Bright, it is impossible to be interpreted as limiting the scope of the invention, those skilled in the art makes according to the foregoing of the present invention
Some nonessential improvement and adjustment belong to protection scope of the present invention.
Claims (1)
1. the foundation of a three-dimensional virtual human platform based on different building shape and call method thereof, it is characterised in that include walking as follows
Rapid:
A) use Scanworx spatial digitizer, automatically obtain whole anthropometric datas, and manually measure and derive from
The calculating of variable increases by 36 measure the items of below human body: front fillet, rear fillet, average shoulder angle, side neck point to breast point distance/
Front length, BP dot spacing/chest measurement, abdomen coign, Attacking Midfielder's central point are to the angle of chest both sides, the folder of side waist central point to chest both sides
Angle, the angle/angle of side waist central point to chest both sides of Attacking Midfielder's central point to chest both sides, the vertical angle of Attacking Midfielder to breast,
Attacking Midfielder's central point is to the angle of buttocks both sides, the angle of side waist central point to buttocks both sides, Attacking Midfielder's central point to buttocks both sides
Angle/the angle of side waist central point to buttocks both sides, angle/Attacking Midfielder's central point chest both sides of Attacking Midfielder's central point to buttocks both sides
Angle, the angle/angle of side waist central point to chest both sides of side waist central point to buttocks both sides, the vertical folder of low back to buttocks
Angle, the back of the body length/height, the vertical angle of side knee center point to waist central point, the angle of side waist central point to buttocks central point,
Angle and low back is hit exactly to collare angle before lower body shaft angle, upper half health shaft angle, upper body;
B) with height/chest measurement for index human body is divided into height, in, short three classes, the most a height of relative chest measurement is high, in be relative chest measurement
In, short short for relative chest measurement;With people's body side surface morphological index: the line of abdomen bump to breast salient point and the angle of horizontal plane, buttocks are convex
Human body, to the angle of line and the horizontal plane of back of the body salient point, is divided into AC, AB, AA, BC, BB, BA, CC, CB, CA nine class by point;Wherein:
AC refers to that the buttocks of people's body side surface protrudes than back, chest protrudes than abdominal part;AB refers to that the buttocks of people's body side surface is than back protrusion, chest
Approximate on a vertical vertical line with abdominal part;AA refers to that the buttocks of people's body side surface protrudes than back, abdominal part protrudes than chest;BC refers to people
The buttocks of body side surface and back approximate on a vertical vertical line, chest protrudes than abdominal part;BB refers to buttocks and the back of the body of people's body side surface
Portion approximates on a vertical vertical line, chest and abdominal part approximate on a vertical line;BA refers to that the buttocks of people's body side surface and back are near
Like on a vertical vertical line, chest than abdominal part protrude;CC refers to that the back of people's body side surface is more convex than abdominal part than buttocks protrusion, chest
Go out;CB refers to that the back of people's body side surface approximates on a vertical vertical line than buttocks protrusion, chest and abdominal part;CA refers to people's body side surface
Back is protruded than buttocks, abdominal part protrudes than chest;It addition, human body is divided into Y, A, B, C with breast waist difference for index by national standard
Four classes;There are 108 class builds;
C) somatic data of 108 class builds is carried out statistical analysis, after seeking its meansigma methods, i.e. obtain the human body of each build intermediate
Data;
D) some the shape adjustment Parameters variation of virtual human body in three-dimensional fitting software is controlled, when observing this Parameters variation, empty
Intend the Changing Pattern of buman body type, it is thus achieved that the physical characteristic parameter bigger on body surface impact;
E) 36 idiotype characteristic parameters are selected: front fillet, rear fillet, average shoulder angle, the distance of side neck point to breast point, side neck point arrive
The distance of breast point/front length, BP dot spacing, BP dot spacing/chest measurement, abdomen coign, the angle of Attacking Midfielder's central point to chest both sides, side waist
Central point to the angle of chest both sides, the angle of the angle/side waist central point to chest both sides of Attacking Midfielder's central point to chest both sides,
Attacking Midfielder is to the vertical angle of breast, the angle of Attacking Midfielder's central point to buttocks both sides, the angle of side waist central point to buttocks both sides, Attacking Midfielder
Central point to the angle/angle of side waist central point to buttocks both sides of buttocks both sides, Attacking Midfielder's central point to buttocks both sides angle/
The angle of Attacking Midfielder's central point chest both sides, the folder of angle/side waist central point to chest both sides of side waist central point to buttocks both sides
Angle, the vertical angle of low back to buttocks, back of the body length, the back of the body length/height, the vertical angle of side knee center point to waist central point, side
Waist central point hits exactly angle, low back to collare angle, height before the angle of buttocks central point, lower body shaft angle, upper half health shaft angle, upper body
Enclose under degree, neck circumference, shoulder breadth, chest measurement, breast, waistline, lower hip circumference, upper hip circumference, outer lower limb length and interior length is as the input of BP neutral net
Parameter, selects 11 shape parameters being applied to three-dimensional modeling: take on parameter, chest locations parameter, breast peak distance parameter, abdomen
Parameter, health width parameter, buttocks width parameter, upper half height parameter, pose parameter, lower part of the body pose parameter, bottom position are joined
Number and upper body pose parameter are output parameter, build the initial model of BP network under Matlab environment;
F) by network training, minimax method, normalized square mean method and without processing three kinds of methods respectively to data are analyzed
On the impact of neural network forecast ability after being standardized processing, analyze the different network numbers of plies shadow to neural network forecast ability is set
Ring, analyze different training function setup on the impact of neural network forecast ability, analyze different node transfer function and arrange net
The impact of network predictive ability, analyze that different the number of hidden nodes arranges the impact on neural network forecast ability, to analyze heterogeneous networks defeated
Enter and the impact on neural network forecast ability is set with output parameter, obtain optimal network structure, set up physical characteristic parameter and use
Shape adjustment nonlinearity in parameters mapping relations in three-dimensional fitting software;
G) coded system of the interface document of human parameters modeling in analyzing three-dimensional fitting software;
H) setting up third party software platform, built-in neural network structure, by by the input of corresponding physical characteristic data file, warp
Cross the Parameters Transformation of neutral net, automatically generate the people's platform interface literary composition being applied to various build intermediate in three-dimensional fitting software
Part;
I) under Matlab environment, the automatic identification algorithm of people's body side surface morphological index is set up: side colour picture is converted to ash
Degree figure, then by Robort operator, find the edge of people's body side surface photo, read having of human body edge image by Matlab
The part of edge lines is 1, and remaining is 0;Dividing the image into into 4 parts, the most horizontal ratio is 1:1, longitudinal segmentation
Ratio is 3:4;Utilize the determinant of digital picture, find first method that 1 ranks occur and find the most convex of every image
Point;Bump location after being merged by determinant, obtains the coordinate figure of each salient point;It is calculated people's body side surface by coordinate figure
Abdomen salient point is to breast salient point line and horizontal angle, people's body side surface buttocks salient point to back of the body salient point line and horizontal angle;
J) human body classifying height/chest measurement, side morphological index, breast waist difference index, is carried out under each sorting technique respectively
The statistics of the boundary value of the classification human body obtained, and set up corresponding build discrimination formula;By build discrimination formula to build
Belong to class to differentiate, call corresponding interface document, generate corresponding intermediate people's platform.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210556599.8A CN103886115B (en) | 2012-12-20 | 2012-12-20 | The foundation of a kind of three-dimensional virtual human platform based on different building shape and call method thereof |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210556599.8A CN103886115B (en) | 2012-12-20 | 2012-12-20 | The foundation of a kind of three-dimensional virtual human platform based on different building shape and call method thereof |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103886115A CN103886115A (en) | 2014-06-25 |
CN103886115B true CN103886115B (en) | 2016-12-28 |
Family
ID=50955007
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210556599.8A Expired - Fee Related CN103886115B (en) | 2012-12-20 | 2012-12-20 | The foundation of a kind of three-dimensional virtual human platform based on different building shape and call method thereof |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103886115B (en) |
Families Citing this family (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104881892B (en) * | 2015-06-04 | 2017-10-27 | 厦门唯尔酷信息技术有限公司 | Model deformation method and its system |
CN106418848B (en) * | 2016-08-04 | 2018-09-18 | 上海工程技术大学 | A kind of somatotype method customized for men's shirt |
CN107895145A (en) * | 2017-10-31 | 2018-04-10 | 南京信息工程大学 | Method based on convolutional neural networks combination super-Gaussian denoising estimation finger stress |
CN108009577A (en) * | 2017-11-29 | 2018-05-08 | 南京工业大学 | Method for realizing virtual fitting mirror |
CN108985657A (en) * | 2018-08-06 | 2018-12-11 | 百度在线网络技术(北京)有限公司 | Evaluation method, device and the server of personage's clothing collocation |
CN109636907A (en) * | 2018-12-13 | 2019-04-16 | 谷东科技有限公司 | A kind of terrain reconstruction method and system based on AR glasses |
CN110135078B (en) * | 2019-05-17 | 2023-03-14 | 浙江凌迪数字科技有限公司 | Human body parameter automatic generation method based on machine learning |
CN110533767A (en) * | 2019-08-15 | 2019-12-03 | 东华大学 | Bulging body men's shirt template parameter optimization method based on three-dimensional garment design |
CN111783182B (en) * | 2020-07-07 | 2024-03-01 | 恒信东方文化股份有限公司 | Modeling method and system of three-dimensional virtual mannequin |
CN112598788A (en) * | 2020-12-30 | 2021-04-02 | 代少稳 | Human body image drawing method |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10628729B2 (en) * | 2010-06-08 | 2020-04-21 | Styku, LLC | System and method for body scanning and avatar creation |
-
2012
- 2012-12-20 CN CN201210556599.8A patent/CN103886115B/en not_active Expired - Fee Related
Non-Patent Citations (2)
Title |
---|
<基于BP神经网络的服装号型推荐方法研究>;郑爱花;《中国优秀硕士学位论文全文数据库 工程科技I辑》;20110615(第6期);摘要、第1.2.2节和第4.5节、第3.1-3.3节、第4.2节、第4.3节和图4.2、表4.1-4.2 * |
N我国服装人台研究现状概述;黄英;《浙江纺织服装职业技术学院学报》;20110930(第3期);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN103886115A (en) | 2014-06-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103886115B (en) | The foundation of a kind of three-dimensional virtual human platform based on different building shape and call method thereof | |
CN106250874B (en) | Recognition methods and the device of a kind of dress ornament and carry-on articles | |
CN103886117B (en) | A kind of improve the method for virtual human body modeling accuracy in three-dimensional fitting software | |
CN106250867B (en) | A kind of implementation method of the skeleton tracking system based on depth data | |
CN103106586A (en) | Three-dimensional virtual fitting system | |
CN102609684B (en) | Human body posture detection method and device | |
CN108765430A (en) | A kind of heart left chamber region segmentation method based on cardiac CT image and machine learning | |
CN103337072A (en) | Texture and geometric attribute combined model based indoor target analytic method | |
CN108009577A (en) | Method for realizing virtual fitting mirror | |
CN109087317A (en) | A kind of Lung neoplasm image partition method | |
CN106327536A (en) | Neck line measuring method based on sectional point clouds | |
CN107680130A (en) | One kind is based on the somatometric method and system of image | |
CN111264951B (en) | Non-contact cheongsam customized human body three-dimensional size measurement method based on deep learning | |
CN105138975A (en) | Human body complexion area segmentation method based on deep belief network | |
CN105488303B (en) | A kind of waist abdomen stern somatotype method and its measuring device based on characteristic distance collection | |
CN115272579A (en) | Single-image three-dimensional garment reconstruction method based on multi-feature fusion | |
CN108829958A (en) | A kind of clothes automatic plate making method and its system | |
CN107038703A (en) | A kind of goods distance measurement method based on binocular vision | |
CN106419922B (en) | A kind of somatic data classification method | |
CN108564586A (en) | A kind of body curve's measurement method and system based on deep learning | |
Petrak et al. | Research of 3D body models computer adjustment based on anthropometric data determined by laser 3D scanner | |
CN109446622A (en) | A kind of modeling method of dress ornament version shape | |
Griffin et al. | Method to capture and analyze the waist-hip-thigh body region of seated-standing 3D scans | |
CN109645607A (en) | A kind of novel the shapes of the clothes construction design method | |
KR20140140997A (en) | Method and apparatus for designing body-type customized compression wear by using standard body shape |
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: 20161228 |
|
CF01 | Termination of patent right due to non-payment of annual fee |