CN102609697B - Vine model creating method for finger vein tridimensional characteristic identification - Google Patents
Vine model creating method for finger vein tridimensional characteristic identification Download PDFInfo
- Publication number
- CN102609697B CN102609697B CN 201210109091 CN201210109091A CN102609697B CN 102609697 B CN102609697 B CN 102609697B CN 201210109091 CN201210109091 CN 201210109091 CN 201210109091 A CN201210109091 A CN 201210109091A CN 102609697 B CN102609697 B CN 102609697B
- Authority
- CN
- China
- Prior art keywords
- node
- model
- tendril
- vein
- stem
- 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
- 210000003462 vein Anatomy 0.000 title claims abstract description 66
- 238000000034 method Methods 0.000 title claims abstract description 19
- 241000345998 Calamus manan Species 0.000 claims abstract description 14
- 235000012950 rattan cane Nutrition 0.000 claims abstract description 14
- 238000005192 partition Methods 0.000 claims description 14
- 238000013459 approach Methods 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 8
- 239000011159 matrix material Substances 0.000 claims description 6
- 241000233788 Arecaceae Species 0.000 claims description 4
- 238000003384 imaging method Methods 0.000 claims description 4
- 230000009466 transformation Effects 0.000 claims description 3
- 238000013519 translation Methods 0.000 claims description 3
- 230000033228 biological regulation Effects 0.000 claims description 2
- 239000007787 solid Substances 0.000 claims description 2
- 238000000605 extraction Methods 0.000 abstract description 4
- 230000008520 organization Effects 0.000 abstract description 4
- 238000010586 diagram Methods 0.000 description 8
- 230000000694 effects Effects 0.000 description 7
- 238000011160 research Methods 0.000 description 4
- 210000001519 tissue Anatomy 0.000 description 3
- 125000000174 L-prolyl group Chemical group [H]N1C([H])([H])C([H])([H])C([H])([H])[C@@]1([H])C(*)=O 0.000 description 2
- 210000000988 bone and bone Anatomy 0.000 description 2
- 238000002224 dissection Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000008439 repair process Effects 0.000 description 2
- 210000003484 anatomy Anatomy 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000008468 bone growth Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 210000001503 joint Anatomy 0.000 description 1
- 210000003205 muscle Anatomy 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 230000007480 spreading Effects 0.000 description 1
- 230000002277 temperature effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 230000003313 weakening effect Effects 0.000 description 1
- 238000004804 winding Methods 0.000 description 1
Images
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention relates to a vine model creating method for finger vein tridimensional characteristic identification. The vine model creating method the following steps of: firstly, simplifying and abstracting a finger vein spatial structure according to principles of uniform pipe diameter, graded rattan, node subdivision and extension constraint; and then, providing elements, attributes and mathematical description of a vein model; and finally, drawing a vein stereopicture according to the data structure of the vein model. The finger vein vine model is used for guiding the tridimensional reconstruction to reduce the influences of certain factors, for example, repairing broken veins, eliminating burrs and noise and fitting vein spatial curve; and the method disclosed by the invention can abstract and simplify the complex tridimensional vein spatial structure to facilitate the data organization and characteristic extraction.
Description
Technical field
The present invention relates generally to a kind of tendril model modelling approach for the identification of finger vena three-dimensional feature, belongs to security protection biometrics identification technology field.
Background technology
Finger vena identification is the study hotspot in current living things feature recognition field, and its main advantage is that vein conceals in finger interior, belongs to the live body feature, is difficult for being forged, stealing, and gathers convenient and safe.The aspects such as figure image intensifying, image segmentation, feature extracting and matching have been contained for the research of finger vena identification at present, but above-mentioned research is all only studied for finger vena equatorial projection image spread, Main Problems is: the three-dimensional distribution of (1) finger vena has uniqueness, but may there be same or analogous equatorial projection in the three-dimensional different finger vena that distributes, and this security to vein identification has brought hidden danger; (2) often there are the posture changings such as rotation in finger in registration and verification process, causes easily same finger vena to have the equatorial projection of larger difference, thereby causes the decline of discrimination.
The three-dimensional identification of research finger vena is the effective way that solves above-mentioned two problems, also is the development trend of living things feature recognition, and achievement in research is not also arranged before the party's appearance.Finger vena tendril model of the present invention is the prerequisite of carrying out the three-dimensional Study of recognition of finger vena, and its meaning is mainly reflected in two aspects:
(1) instructs the finger vena three-dimensional reconstruction.Vein is concealed in finger interior, the locus that can't adopt the means such as laser acquisition to obtain vein, and adopting the three-dimensional rebuilding method based on stereoscopic vision is the effective way of finger vena three-dimensional reconstruction.Current vein image obtain extensive employing infrared transmitting or the reflection mode is carried out, owing to the interference of muscle, skeletal tissue is difficult to obtain clearly finger venous image, cause the vein streakline after cutting apart fracture and burr interference phenomenon often to occur; And finger venous image is also pointed the various factors of self, when for example pointing pressurized, gather the image medium sized vein and fracture and distortion can occur, under same light source, finger thickness difference or environment temperature are different, gather image medium sized vein sharpness and have larger difference.Above-mentioned factor brings very large difficulty can for the vein three-dimensional reconstruction, and adopts finger vena tendril model to instruct three-dimensional reconstruction, can weaken the impact of these factors, for example: repair vein, rejecting burr and the noise of fracture, match vein space curve.
(2) instruct the finger vena three-dimensional feature to extract.Finger vena branch is numerous, and space structure is complicated, the extraction of three-dimensional feature and array organization's difficulty, and adopt finger vena tendril model with three-dimensional vein space structure abstract and the simplification of complexity, also to be convenient to the tissue of data and the extraction of feature.
Summary of the invention
Technical matters to be solved by this invention is to provide a kind of tendril model modelling approach for the identification of finger vena three-dimensional feature.
For realizing above-mentioned goal of the invention, the present invention adopts following technical scheme:
At first by caliber homogeneous, rattan classification, node partition, spread four principles of constraint and simplify and abstract finger vena space structure; Then provide element, attribute and the mathematical description of tendril model; Utilize at last the data structure of tendril model to draw the vein stereo-picture.
1, finger vena space structure short-cut method
Fig. 1 is the dissection synoptic diagram of certain finger, can find out vein as tendril, twines bone growth, refer to the back of the body, refer to that side, each vein of fingers and palms cross mutually, and duplexure is many, and general structure is complicated, is not easy to Organization of Data and description.The tendril model that the present invention proposes at first by caliber homogeneous, rattan classification, node partition, spread four principles of constraint venous structures be reduced to binary tree structure.Clear in order to describe, set forth the vein simplification principle below in conjunction with the digital palmar veins shown in Fig. 1 (c), Fig. 2 has shown Fig. 1 (c) medium sized vein structural representation.
1.1 caliber homogeneous principle
Anatomy finds that finger mainline (finger vena identification mainly selects middle joint to refer to) caliber is generally 0.7~1.0mm, consider the physical resolution of existing video camera, caliber difference is less during the mainline imaging, and the vein caliber is subjected to finger pressure, ambient temperature effect larger during imaging, the caliber feature does not have conspicuousness and robustness, therefore the caliber of supposition finger mainline is identical, and the venous structures after the simplification as shown in Figure 3.
1.2 rattan principle of grading
For the ease of distinguishing different venous tributaries, give different grades with each venous tributary, and with different color marks.The rattan principle of grading is: by from referring to that root is to finger tip, fingers and palms to referring to that the back of the body, counterclockwise direction carry out step by step classification (initial progression is 1) to venous tributary, each bar venous tributary for each node (joint of each bar venous tributary) connection, classified venous tributary progression is n, and venous tributary progression to be fractionated is n+1.Fig. 4 has shown the initial classification synoptic diagram of each bar vein among Fig. 3.
1.3 node partition principle
Generally speaking, two thin veins and a thick vein intersect, and corresponding node is tee joint, is convenient to describe with binary tree structure.If node does not satisfy the tee joint characteristic, then need carry out node partition.Fig. 5 provides one five subdivision synoptic diagram that is communicated with node, and 3 principles of node partition are described in conjunction with this figure.
Principle 1: if same other venous tributary of level that node is connected is greater than two, then these venous tributaries are sorted by counter clockwise direction, if there be the 3rd vein, then increase by 1 node at the 2nd vein, connect the 3rd vein, similarly, increase by 1 node at the 3rd vein, be used for connecting the 4th vein, the rest may be inferred.Among Fig. 5, node P
1Same other venous tributary of level that is connected has 4, is { L by counterclockwise ordering
2, L
3, L
4, L
5.L
3Upper increase node P
2, be used for connecting L
4Then L
4Upper increase node P
3, be used for connecting L
5Node P behind the subdivision
1, P
2, P
3All satisfy the tee joint characteristic.
Principle 2: newly-generated node and the air line distance of last node equal the vein caliber.
Principle 3: the translation transformation on each venous tributary only has living space before and after the subdivision, other space attributes constant (such as length, curvature etc.).
1.4 spread the constraint principle
Vein spreads in the process at the winding bone, different venous tributaries cross mutually, cause occurring the ring texture that not easy-to-use binary tree is described, for this reason, proposition spreads the constraint principle: venous tributary is (also being in the rattan classification process) in the process that spreads around bone, if run into classified node, then continue to spread, until arrive the node that does not have classification.Spreading constraint just changes the connectedness of vein node and merges venous tributary for the ease of Organization of Data, not have obliterated data (binding effect is just spread and the discarded part divided data in order showing in Fig. 6~9), therefore not affect the significant characteristics of venous structures.
2, finger vena tendril model description method
For the binary tree venous structures after simplifying, set up finger vena tendril model, its synoptic diagram as shown in figure 10:
The chief component unit of tendril model have: stem, node.
Stem: all vein limbs are referred to as stem in the model.Among Figure 10, L
1~L
7It all is stem.
Node: the two-end-point of all vein limbs is referred to as node.Among Figure 10, P
0~P
7It all is node.
The main attribute of tendril model has: the locus of node, the Betti number of node; The length of stem, curvature, around rate.Be described in detail as follows:
The Betti number of node: classification levels is that the Betti number of the start node of that stem of 1 is 1 in the definition tendril model, and the Betti number of the terminal node of all stems that classification levels is the highest is 2, and the Betti number of other nodes is 3.
Character 1: nodes=stem number+1.
Character 2: in each tendril model, the Betti number of node satisfies following formula:
N
2=N
1+N
3
N
1For Betti number is 1 nodes, N
2For Betti number is 2 nodes, N
3For Betti number is 3 nodes, and N
1=1.
The length of stem, curvature, around rate: in three dimensions, the length of stem, curvature, around rate refer to the length, curvature of the axis space curve of stem, around rate.
Tendril model L can be unique definite by one group of point set, is described below:
L={R
k|k=1,2,...,N-1}
Wherein, N is the sum of node in the tendril model, R
kThe expression Betti number is described below greater than the attribute of 1 node k:
R
k={C,M}
Wherein, that stem take node k as terminating point is designated as l
k, C represents the position code value of node k in the tendril model, its number of significant digit represents l
kClassification levels; M represents l
kThe space curve parameter.
(1) position code value
Position code value C is used for organising data and calculate classification levels.The regulation Betti number is that the code value of 1 node is 0, and the code value of subsequent node is encoded according to the organizational form of binary tree, and coding rule is:
If present node P
kCode value be X, the node P that is generated by this node
tCode value be Y, then
Y=X×2+b
Wherein, if the node that present node generates has 2, and P
tAt P
kThe right side, b=0 then, otherwise b=1.
If stem l
kClassification levels be n, then n satisfies
2
n-1≤Y<2
n
Table 1 has shown position code value and the classification levels of model shown in Figure 10.
Table 1 position code value table
Nodename | Code value C | Binary coding | Classification levels n |
P 1 | 1 | 1 | 1 |
|
2 | 10 | 2 |
P 3 | 3 | 11 | 2 |
P 4 | 4 | 100 | 3 |
P 5 | 5 | 101 | 3 |
P 6 | 6 | 110 | 3 |
P 7 | 7 | 111 | 3 |
(2) space curve parameter
Stem l
kMiddle axial curve be a space curve, can obtain by solid matching method the volume coordinate of its part sampled point, then adopt this space curve of B spline-fitting, the reference mark matrix of M this moment when representing curve.
M=(E
TE)
-1E
TP
Wherein, E represents B spline base function matrix of coefficients, and P represents the sample point coordinate matrix.
3, the finger vena stereo-picture is drawn
After adopting tendril model description finger vena of the present invention, according to parameter M and classification levels n, utilize Pro/E software can draw the three dimensions hierarchy of each stem.Figure 11 has shown the three-dimensional structure drafting effect of the model of tendril shown in the table 2, and wherein color index is as shown in table 3.
Table 2 tendril model data
C | M (unit: mm) |
1 | 130,0,0,140,30,0,150,70,0,150,110,0 |
2 | 150,110,0,150,130,0,150,160,0,150,190,0 |
3 | 150,110,0,120,105,5,90,100,15,70,100,20 |
4 | 150,190,0,150,240,0,100,290,0,60,330,0 |
5 | 150,190,0,90,140,20,50,90,30,20,20,20 |
10 | 20,20,20,20,40,20,20,60,20,15,80,15 |
11 | 20,20,20,50,60,-60,100,90,-40,140,120,-20 |
20 | 15,80,15,20,130,15,15,180,15,15,230,15 |
21 | 15,80,15,0,130,-20,60,190,-60,100,300,0 |
Table 3 color index table
n | color(red,green,blue) |
1 | 255,0,0 |
2 | 251,142,191 |
3 | 255,0,246 |
4 | 168,0,255 |
5 | 78,3,250 |
Adopt finger vena tendril model to instruct three-dimensional reconstruction, can weaken the impact of some factors, for example: repair vein, rejecting burr and the noise of fracture, match vein space curve; Can with three-dimensional vein space structure abstract and the simplification of complexity, be convenient to the tissue of data and the extraction of feature.
Description of drawings
Fig. 1 points the dissection synoptic diagram;
Fig. 2 finger vena structure;
Venous structures after Fig. 3 caliber homogeneous is simplified;
The initial classification synoptic diagram of Fig. 4 finger vena;
Fig. 5 node partition synoptic diagram;
Spread binding effect during the classification of Fig. 6 3rd level;
Spread binding effect during the 6th grade of classification of Fig. 7;
Spread binding effect during the 7th grade of classification of Fig. 8;
Fig. 9 spreads binding effect after the loop cuts off;
Figure 10 tendril model synoptic diagram;
Figure 11 tendril modeling rendering effect.
Embodiment
The present invention is a kind of tendril model modelling approach for finger vena three-dimensional feature identification, at first by caliber homogeneous, rattan classification, node partition, spread four principles of constraint and simplify and abstract finger vena space structure; Then provide element, attribute and the mathematical description of tendril model; Utilize at last the data structure of tendril model to draw the vein stereo-picture.Detailed process is as follows:
One, finger vena space structure short-cut method
1.1 caliber homogeneous principle
Caliber difference is less during the mainline imaging, supposes that the caliber of finger mainline is identical.
1.2 rattan principle of grading
The rattan principle of grading is: by from referring to that root is to finger tip, fingers and palms to referring to that the back of the body, counterclockwise direction carry out step by step classification to venous tributary, each bar venous tributary for each node connection, classified venous tributary progression is n, and venous tributary progression to be fractionated is n+1.
1.3 node partition principle
Node partition comprises 3 principles, principle 1: if same other venous tributary of level that node is connected is greater than two, then these venous tributaries are sorted by counter clockwise direction, if there be the 3rd vein, then increase by 1 node at the 2nd vein, connect the 3rd vein, if there be the 4th vein, increase by 1 node at the 3rd vein, be used for connecting the 4th vein, the rest may be inferred; Principle 2: newly-generated node and the air line distance of last node equal the vein caliber; Principle 3: the translation transformation on each venous tributary only has living space before and after the subdivision, other space attributes are constant.
1.4 spread the constraint principle
Venous tributary if run into classified node, then continues to spread in the rattan classification process, until arrive the node that does not have classification.
Two, finger vena tendril model description method
For the binary tree venous structures after simplifying, set up finger vena tendril model:
The chief component unit of tendril model have: stem, node; Stem: all vein limbs are referred to as stem in the model; Node: the two-end-point of all vein limbs is referred to as node;
The main attribute of tendril model has: the locus of node, the Betti number of node; The length of stem, curvature, around rate.
Tendril model L can be unique definite by one group of point set, is described below:
L={R
k|k=1,2,...,N-1}
Wherein, N is the sum of node in the tendril model, R
kThe expression Betti number is greater than the attribute of 1 node k,
R
k={C,M}
That stem take node k as terminating point is designated as l
k, C represents the position code value of node k in the tendril model, its number of significant digit represents l
kClassification levels; M represents l
kThe space curve parameter.
Three, the finger vena stereo-picture is drawn
After adopting tendril model description finger vena, according to parameter M and classification levels n, utilize the three dimensions hierarchy of each stem of Pro/E Software on Drawing.
Claims (4)
1. a tendril model modelling approach that is used for the identification of finger vena three-dimensional feature is characterized in that, at first by caliber homogeneous, rattan classification, node partition, spread four principles of constraint and simplify and abstract finger vena space structure; Then provide element, attribute and the mathematical description of tendril model; Detailed process is as follows:
One, the finger vena space structure is simplified
1.1 caliber homogeneous principle
Caliber difference is less during the mainline imaging, supposes that the caliber of finger mainline is identical;
1.2 rattan principle of grading
The rattan principle of grading is: by from referring to root to finger tip, fingers and palms to referring to that the back of the body, counterclockwise direction carry out step by step classification to venous tributary, for each bar venous tributary that each node connects, classified venous tributary progression is
n, venous tributary progression to be fractionated is
n+ 1;
1.3 node partition principle
Node partition comprises 3 principles, principle 1: if same other venous tributary of level that node is connected is greater than two, then these venous tributaries are sorted by counter clockwise direction, if there be the 3rd vein, then increase by 1 node at the 2nd vein, connect the 3rd vein, if there be the 4th vein, increase by 1 node at the 3rd vein, be used for connecting the 4th vein, the rest may be inferred; Principle 2: newly-generated node and the air line distance of last node equal the vein caliber; Principle 3: the translation transformation on each venous tributary only has living space before and after the subdivision, other space attributes are constant;
1.4 spread the constraint principle
Venous tributary if run into classified node, then continues to spread in the rattan classification process, until arrive the node that does not have classification;
Two, finger vena tendril model description
For the binary tree venous structures after simplifying, set up finger vena tendril model:
The component of tendril model comprises: stem, node; Stem: all vein limbs are referred to as stem in the model; Node: the two-end-point of all vein limbs is referred to as node;
The attribute of tendril model comprises: the locus of node, the Betti number of node; The length of stem, curvature, around rate;
The tendril model
LCan be unique definite by one group of point set, be described below:
Wherein,
NBe the sum of node in the tendril model,
R k The expression Betti number is greater than 1 node
kAttribute,
With node
kFor that stem of terminating point is designated as
l k ,
CThe expression node
kPosition code value in the tendril model, its number of significant digit represents
l k Classification levels;
MExpression
l k The space curve parameter;
In the tendril model, the position code value
CBe used for organising data and calculate classification levels, the regulation Betti number is that the code value of 1 node is 0, and the code value of subsequent node is encoded according to the organizational form of binary tree, and coding rule is:
If present node
P k Code value be X, the node that is generated by this node
P t Code value be Y, then
Wherein, if the node that present node generates has 2, and
P t P k The right side, then
b=0, otherwise
b=1,
If stem
l k Classification levels be
n, then
nSatisfy
2. a kind of tendril model modelling approach for finger vena three-dimensional feature identification according to claim 1, it is characterized in that, in the tendril model, the Betti number of node: classification levels is that the Betti number of the start node of that stem of 1 is 1 in the definition tendril model, the Betti number of the terminal node of all stems that classification levels is the highest is 2, and the Betti number of other nodes is 3;
Character 1: nodes=stem number+1;
Character 2: in each tendril model, the Betti number of node satisfies following formula:
N 1For Betti number is 1 nodes,
N 2For Betti number is 2 nodes,
N 3For Betti number is 3 nodes, and
N 1=1.
3. a kind of tendril model modelling approach for finger vena three-dimensional feature identification according to claim 1 is characterized in that, in the tendril model, the length of stem, curvature, around rate refer to the length, curvature of the axis space curve of stem, around rate.
4. a kind of tendril model modelling approach for the identification of finger vena three-dimensional feature according to claim 1 is characterized in that, in the tendril model, and stem
l k Middle axial curve be a space curve, obtain the volume coordinate of its part sampled point by solid matching method, then adopt this space curve of B spline-fitting, at this moment
MReference mark matrix during the expression curve,
Wherein,
EExpression B spline base function matrix of coefficients,
PExpression sample point coordinate matrix.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201210109091 CN102609697B (en) | 2012-04-14 | 2012-04-14 | Vine model creating method for finger vein tridimensional characteristic identification |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201210109091 CN102609697B (en) | 2012-04-14 | 2012-04-14 | Vine model creating method for finger vein tridimensional characteristic identification |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102609697A CN102609697A (en) | 2012-07-25 |
CN102609697B true CN102609697B (en) | 2013-04-17 |
Family
ID=46527056
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN 201210109091 Expired - Fee Related CN102609697B (en) | 2012-04-14 | 2012-04-14 | Vine model creating method for finger vein tridimensional characteristic identification |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102609697B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105975950A (en) * | 2016-05-27 | 2016-09-28 | 国创科视科技股份有限公司 | Rapid comparison method for three-dimensional finger vein similarity |
CN106919941B (en) * | 2017-04-26 | 2018-10-09 | 华南理工大学 | A kind of three-dimensional finger vein identification method and system |
CN111160247B (en) * | 2019-12-28 | 2023-05-12 | 智冠一掌通科技(深圳)有限公司 | Method for three-dimensional modeling and identification by scanning palm vein |
CN112508064A (en) * | 2020-11-24 | 2021-03-16 | 广州广电运通金融电子股份有限公司 | Finger vein identity recognition method and device, computer equipment and storage medium |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
RU2192175C2 (en) * | 1999-09-03 | 2002-11-10 | Научно-исследовательский центр Татарстана "Восстановительная травматология и ортопедия" | Method for cutaneous plasty at ringtype scalping the skin of the fourth finger |
CN101980243B (en) * | 2010-10-15 | 2012-11-07 | 中国人民解放军国防科学技术大学 | Binocular vision-based finger vein three-dimensional identification system |
CN102222229B (en) * | 2011-07-28 | 2015-12-02 | 陈庆武 | Method for preprocessing finger vein images |
-
2012
- 2012-04-14 CN CN 201210109091 patent/CN102609697B/en not_active Expired - Fee Related
Also Published As
Publication number | Publication date |
---|---|
CN102609697A (en) | 2012-07-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102609697B (en) | Vine model creating method for finger vein tridimensional characteristic identification | |
CN103065352B (en) | Plant three-dimensional reconstruction method based on image and scanning data | |
US7917346B2 (en) | Geospatial modeling system providing simulated tree trunks and branches for groups of tree crown vegetation points and related methods | |
CN107423678A (en) | A kind of training method and face identification method of the convolutional neural networks for extracting feature | |
CN109903257A (en) | A kind of virtual hair-dyeing method based on image, semantic segmentation | |
CN106648103A (en) | Gesture tracking method for VR headset device and VR headset device | |
CN103106684A (en) | Method and system of three-dimensional reconstruction of shape structure of trees in leaved state | |
US20140354630A1 (en) | Tree model and forest model generating method and apparatus | |
CN106056564B (en) | Edge clear image interfusion method based on joint sparse model | |
CN105405162B (en) | Tree point cloud three-dimensional rebuilding method based on partial structurtes and directional perception | |
CN106778785A (en) | Build the method for image characteristics extraction model and method, the device of image recognition | |
CN109741438B (en) | Three-dimensional face modeling method, device, equipment and medium | |
CN107507171A (en) | A kind of lung CT image air flue three-dimensional framework tree extraction and labeling method | |
CN110827295A (en) | Three-dimensional semantic segmentation method based on coupling of voxel model and color information | |
CN108460398A (en) | Image processing method, device, cloud processing equipment and computer program product | |
CN109360175A (en) | A kind of infrared image interfusion method with visible light | |
CN111476241B (en) | Character clothing conversion method and system | |
CN103646421B (en) | Trees lightweight 3D method for reconstructing based on enhancement mode PyrLK optical flow method | |
CN101866495A (en) | Tree modeling method based on skeleton point cloud | |
CN112967178A (en) | Image conversion method, device, equipment and storage medium | |
CN107993280A (en) | Beauty method and system based on threedimensional model | |
CN101847267A (en) | Tree modeling method based on depth search | |
CN110163796A (en) | A kind of image generating method and frame that unsupervised multi-modal confrontation encodes certainly | |
CN103839289B (en) | A kind of true tree Interactive Modeling method based on sparse graph picture | |
US20130271460A1 (en) | Apparatus and method for generating 3d tree images |
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 | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20130417 |