CN105608737A - Human foot three-dimensional reconstruction method based on machine learning - Google Patents
Human foot three-dimensional reconstruction method based on machine learning Download PDFInfo
- Publication number
- CN105608737A CN105608737A CN201610069095.1A CN201610069095A CN105608737A CN 105608737 A CN105608737 A CN 105608737A CN 201610069095 A CN201610069095 A CN 201610069095A CN 105608737 A CN105608737 A CN 105608737A
- Authority
- CN
- China
- Prior art keywords
- foot
- dimensional
- machine learning
- model
- method based
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a three-dimensional modeling technology, specifically relates to a human foot three-dimensional reconstruction method based on machine learning, and belongs to the field of application of a computer vision technology. The invention discloses a human foot three-dimensional reconstruction method based on machine learning, for solving the problems of low measurement speed, low precision and tedious operation of a conventional contact type measurement scheme and the problems of tedious operation and high cost of a non-contact type measurement scheme despite high measurement precision. According to the invention, foot images of several positions are obtained randomly by use of an image obtaining apparatus, then, by use of a machine learning method, key points of feet are obtained by use of a foot model trained in advance, a standard foot three-dimensional model is driven for approximation by use of the key points, finally, a foot three-dimensional model is obtained through iteration, and parameters of the feet can be calculated according to reconstructed three-dimensional model.
Description
Technical field
The present invention relates to dimensional Modeling Technology, be specifically related to a kind of human foot three-dimensional rebuilding method based on machine learning, belong toComputer vision technique application.
Background technology
Three-dimensional reconstruction is to utilize two-dimensional projection to recover mathematical procedure and the computer technology of object dimensional information (shape etc.). It is according to trueThe data reconstruction of real field scape goes out to have the threedimensional model of accurate geometry information and photorealistic.
Human foot is carried out to three-dimensional reconstruction, can more improve and digitlization foot physiological characteristic effectively, simple and expedientlyMeasure its relevant morphological feature parameter, have extensively in fields such as Foot-biomechanics research, medical science, physical culture, shoemaking, 3D printingsGeneral application.
At present foot three-dimensional reconstruction scheme is had to contact and contactless two kinds:
1) contact measurement method uses probe contact foot surface, to a certain extent foot is pushed and causes foot tableFace deforms, the foot parameter inaccuracy obtaining like this; The measuring speed of contact type measurement mode is slow in addition, if measureWhole foot profile, measuring speed is also restraining factors of considering in foot measurement, contact type measurement mode operates ratio simultaneouslyTrouble, wastes time and energy, and will inevitably be replaced by Contactless foot measuring method fast.
2) contactless measuring method is in conjunction with photoelectric technology, computer technology and electronic scanning device, in the process of foot measurementIn do not need directly to contact foot surface and just can recover the three-dimensional appearance of foot, and foot tri-dimensional facial type data that obtain and connecingThe foot relevant parameter that touch method obtains is compared, and possesses the advantage that data volume is many, measuring speed is fast, precision is high, is measuringIn journey, having broken away from heavy manual labor, is the prefered method in foot measurement field.
The method of foot non-contact measurement reconstruction of three-dimensional model has at present: laser scanning, structured light or binocular triangulation etc.,Although its certainty of measurement is high, measuring system and measuring principle complexity, cost are high, complex operation, only have professional person's abilityComplete.
Summary of the invention
Technical problem to be solved by this invention is: propose a kind of human foot three-dimensional rebuilding method based on machine learning, solveIn conventional art, contact type measurement scheme measuring speed is slow, precision is low, troublesome poeration, although and non-contact measurement scheme is measuredPrecision is high, but still complex operation, problem that cost is high.
The present invention solves the problems of the technologies described above adopted scheme:
A human foot three-dimensional rebuilding method based on machine learning, comprises the following steps:
A, choose the article with standard two dimension yardstick and be placed near foot as object of reference, and guarantee fully contacts with foot;
B, utilize image acquiring device to obtain at random the above images of two of foots;
Foot characteristic point on the image obtaining in the foot characteristic point Model Identification that C, utilization have trained and annotation step B;
D, determine the attitude parameter of foot according to the foot characteristic point of obtaining in step C, and logical according to described attitude parameter adjustmentBy the attitude of three-dimensional foot model;
The corresponding points of each foot characteristic point of obtaining in E, determining step C on general three-dimensional foot model, obtain corresponding portionThe foot point cloud dividing;
F, the foot point cloud that step e is obtained are normalized and obtain the cloud data that comprises true coordinate;
G, a have cloud is mated to aligning, obtain the cloud data of whole foot;
H, the cloud data of whole foot is carried out to interpolation, obtain the intensive cloud data of whole foot, thereby obtain foot3D model;
I, the foot 3D model obtaining is carried out to texture mapping, obtain the final 3D foot of rebuilding.
As further optimization, in steps A, selection standard A4 paper, as object of reference, is stepped on the foot of 3D model to be reconstructedOn described standard A 4 paper; Here be not limited in A4 paper, but any object that standard two dimension yardstick is provided, such as bodyPart card waits also can.
As further optimization, in step B, in the time utilizing image acquiring device to obtain at random foot image, ensure to get footAt least each image in the portion left and right sides.
As further optimization, in step C, described characteristic point, for characterizing key point and the profile of foot, comprising: sufficient heelPoint, the longest toe cusp, l articulationes metatarsophalangeae bump, the 5th articulationes metatarsophalangeae bump, internal malleolus cusp, external malleolus cusp, navicular bone areMillet cake is selected, supported to Gao Dian, inner side arch of foot peak, big toe side salient point, toe front end peak, l articulationes metatarsophalangeae upper limb.
The invention has the beneficial effects as follows: the present invention, by the foot picture of single image acquisition device random shooting, utilizes engineeringThe method of practising is carried out 3D reconstruction to foot, without multiple sensors, also without demarcation; Because the method is simple, be easy to realize,Precision is high, workable.
Brief description of the drawings
Fig. 1 is foot characteristic point schematic diagram.
Detailed description of the invention
The present invention is intended to propose a kind of method of utilizing machine learning principle to carry out foot three-dimensional reconstruction, can realize foot phaseThe measurement of related parameter, with more perfect, digitized forms is described human foot physiological phenomenon accurately.
The method utilizes image acquiring device to obtain at random the foot picture of several positions, then adopts the method for machine learning, profitWith the foot model training in advance, obtain the key point of foot, then utilize key point to drive the foot threedimensional model of standard to enterRow approaches, and finally obtains the threedimensional model of foot by iteration, and can calculate according to the threedimensional model of rebuilding the parameter of foot.
Embodiment:
Human foot three-dimensional rebuilding method in this example comprises following performing step:
(1) near the foot of 3D model to be reconstructed, place article with standard two dimension (long and wide) yardstick as referenceThing, these article have sufficient contact (as: pin is stepped on standard A 4 paper) with foot;
(2) utilize image acquiring device random (without accurately knowing angle) to obtain two of foot with the epigraph (left and right sidesAt least each one);
(3) the foot characteristic point of the image obtaining in the foot characteristic point Model Identification that employing has trained and mark (2),Characteristic point is for characterizing key point and the profile of foot, and as shown in Figure 1, it comprises: sufficient heel point 1, the longest toe cusp 2, theL articulationes metatarsophalangeae bump the 3, the 5th articulationes metatarsophalangeae bump 4, internal malleolus cusp 5, external malleolus cusp 6, navicular bone peak 7, interior parapodumBow peak 8, big toe side salient point 9, toe front end peak 10, l articulationes metatarsophalangeae upper limb are selected 11, are supported millet cake 12.
(4) utilize 3) in characteristic point determine the attitude parameter of foot, and adjust general three-dimensional foot model according to attitude parameterAttitude; The general three-dimensional foot model is here a standard 3D model, and the object of this step is by the attitude of universal modelBe adjusted into the same visual angle of foot image to be reconstructed.
(5) determine characteristic point corresponding points on general three-dimensional foot model in (4) in (3), obtain the foot of corresponding partPortion's point cloud;
(6) utilize object of reference in (1), the foot point cloud obtaining is normalized and obtains the cloud data that comprises true coordinate;
(7) a have cloud is mated to aligning, obtain the cloud data of whole foot;
(8) above-mentioned cloud data is carried out to interpolation, obtain the intensive cloud data of foot, thereby obtain the 3D model of foot;
(9) the foot model in (8) is carried out to texture mapping, obtain the final 3D foot of rebuilding.
Claims (4)
1. the human foot three-dimensional rebuilding method based on machine learning, is characterized in that, comprises the following steps:
A, choose the article with standard two dimension yardstick and be placed near foot as object of reference, and guarantee fully contacts with foot;
B, utilize image acquiring device to obtain at random the above images of two of foots;
Foot characteristic point on the image obtaining in the foot characteristic point Model Identification that C, reference have trained and annotation step B;
D, determine the attitude parameter of foot according to the foot characteristic point of obtaining in step C, and logical according to described attitude parameter adjustmentBy the attitude of three-dimensional foot model;
The corresponding points of each foot characteristic point of obtaining in E, determining step C on general three-dimensional foot model, obtain corresponding portionThe foot point cloud dividing;
F, the foot point cloud that step e is obtained are normalized and obtain the cloud data that comprises true coordinate;
G, a have cloud is mated to aligning, obtain the cloud data of whole foot;
H, the cloud data of whole foot is carried out to interpolation, obtain the intensive cloud data of whole foot, thereby obtain foot3D model;
I, the foot 3D model obtaining is carried out to texture mapping, obtain the final 3D foot of rebuilding.
2. a kind of human foot three-dimensional rebuilding method based on machine learning as claimed in claim 1, is characterized in that stepIn A, selection standard A4 paper, as object of reference, steps down in the foot of 3D model to be reconstructed on described standard A 4 paper.
3. a kind of human foot three-dimensional rebuilding method based on machine learning as claimed in claim 1, is characterized in that stepIn B, in the time utilizing image acquiring device to obtain at random foot image, ensure to get at least each image in the foot left and right sides.
4. a kind of human foot three-dimensional rebuilding method based on machine learning as claimed in claim 1, is characterized in that stepIn C, described characteristic point, for characterizing key point and the profile of foot, comprising: sufficient heel point, the longest toe cusp, l sole of the foot toeAR, the 5th articulationes metatarsophalangeae bump, internal malleolus cusp, external malleolus cusp, navicular bone peak, inner side arch of foot peak, thumbMillet cake is selected, supported to toe side salient point, toe front end peak, l articulationes metatarsophalangeae upper limb.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610069095.1A CN105608737B (en) | 2016-02-01 | 2016-02-01 | A kind of human foot three-dimensional rebuilding method based on machine learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610069095.1A CN105608737B (en) | 2016-02-01 | 2016-02-01 | A kind of human foot three-dimensional rebuilding method based on machine learning |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105608737A true CN105608737A (en) | 2016-05-25 |
CN105608737B CN105608737B (en) | 2019-01-08 |
Family
ID=55988648
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610069095.1A Active CN105608737B (en) | 2016-02-01 | 2016-02-01 | A kind of human foot three-dimensional rebuilding method based on machine learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105608737B (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106815428A (en) * | 2017-01-13 | 2017-06-09 | 中国空气动力研究与发展中心高速空气动力研究所 | A kind of wind tunnel balance calibration data processing method based on intelligent optimization algorithm |
TWI607412B (en) * | 2016-09-10 | 2017-12-01 | 財團法人工業技術研究院 | Measurement systems and methods for measuring multi-dimensions |
CN109330106A (en) * | 2018-11-01 | 2019-02-15 | 成都牛晶科技有限公司 | A kind of subscript dimension measurement method based on mobile phone photograph |
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 |
CN109815830A (en) * | 2018-12-28 | 2019-05-28 | 梦多科技有限公司 | A method of obtaining foot information in the slave photo based on machine learning |
CN109840592A (en) * | 2018-12-24 | 2019-06-04 | 梦多科技有限公司 | A kind of method of Fast Labeling training data in machine learning |
CN109887077A (en) * | 2019-03-07 | 2019-06-14 | 百度在线网络技术(北京)有限公司 | Method and apparatus for generating threedimensional model |
CN110379001A (en) * | 2019-07-04 | 2019-10-25 | 新拓三维技术(深圳)有限公司 | A kind of foot product method for customizing, device, terminal device and computer readable storage medium |
CN112257582A (en) * | 2020-10-21 | 2021-01-22 | 北京字跳网络技术有限公司 | Foot posture determination method, device, equipment and computer readable medium |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
IT202100023615A1 (en) * | 2021-09-14 | 2023-03-14 | Trya S R L | METHOD AND SYSTEM FOR MEASURING A FOOT AND GENERATING A THREE-DIMENSIONAL MODEL OF A FOOT |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090306801A1 (en) * | 2006-11-27 | 2009-12-10 | Northeastern University | Patient specific ankle-foot orthotic device |
CN101658347A (en) * | 2009-09-24 | 2010-03-03 | 浙江大学 | Method for obtaining dynamic shape of foot model |
CN104021589A (en) * | 2014-06-27 | 2014-09-03 | 江苏中佑石油机械科技有限责任公司 | Three-dimensional fitting simulating method |
-
2016
- 2016-02-01 CN CN201610069095.1A patent/CN105608737B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090306801A1 (en) * | 2006-11-27 | 2009-12-10 | Northeastern University | Patient specific ankle-foot orthotic device |
CN101658347A (en) * | 2009-09-24 | 2010-03-03 | 浙江大学 | Method for obtaining dynamic shape of foot model |
CN104021589A (en) * | 2014-06-27 | 2014-09-03 | 江苏中佑石油机械科技有限责任公司 | Three-dimensional fitting simulating method |
Non-Patent Citations (1)
Title |
---|
袁振宇: "基于线结构光扫描的脚型重构测量研究", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI607412B (en) * | 2016-09-10 | 2017-12-01 | 財團法人工業技術研究院 | Measurement systems and methods for measuring multi-dimensions |
US10008005B2 (en) | 2016-09-10 | 2018-06-26 | Industrial Technology Research Institute | Measurement system and method for measuring multi-dimensions |
CN106815428B (en) * | 2017-01-13 | 2020-05-19 | 中国空气动力研究与发展中心高速空气动力研究所 | Wind tunnel balance calibration data processing method based on intelligent optimization algorithm |
CN106815428A (en) * | 2017-01-13 | 2017-06-09 | 中国空气动力研究与发展中心高速空气动力研究所 | A kind of wind tunnel balance calibration data processing method based on intelligent optimization algorithm |
CN109330106A (en) * | 2018-11-01 | 2019-02-15 | 成都牛晶科技有限公司 | A kind of subscript dimension measurement method based on mobile phone photograph |
CN109330106B (en) * | 2018-11-01 | 2021-08-24 | 成都牛晶科技有限公司 | Foot code size measuring method based on mobile phone photographing |
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 |
CN109840592A (en) * | 2018-12-24 | 2019-06-04 | 梦多科技有限公司 | A kind of method of Fast Labeling training data in machine learning |
CN109840592B (en) * | 2018-12-24 | 2019-10-18 | 梦多科技有限公司 | A kind of method of Fast Labeling training data in machine learning |
CN109815830A (en) * | 2018-12-28 | 2019-05-28 | 梦多科技有限公司 | A method of obtaining foot information in the slave photo based on machine learning |
CN109887077A (en) * | 2019-03-07 | 2019-06-14 | 百度在线网络技术(北京)有限公司 | Method and apparatus for generating threedimensional model |
WO2020177348A1 (en) * | 2019-03-07 | 2020-09-10 | 百度在线网络技术(北京)有限公司 | Method and apparatus for generating three-dimensional model |
CN109887077B (en) * | 2019-03-07 | 2022-06-03 | 百度在线网络技术(北京)有限公司 | Method and apparatus for generating three-dimensional model |
CN110379001A (en) * | 2019-07-04 | 2019-10-25 | 新拓三维技术(深圳)有限公司 | A kind of foot product method for customizing, device, terminal device and computer readable storage medium |
CN112257582A (en) * | 2020-10-21 | 2021-01-22 | 北京字跳网络技术有限公司 | Foot posture determination method, device, equipment and computer readable medium |
Also Published As
Publication number | Publication date |
---|---|
CN105608737B (en) | 2019-01-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105608737A (en) | Human foot three-dimensional reconstruction method based on machine learning | |
CN104126989B (en) | A kind of based on the foot surfaces 3 D information obtaining method under multiple stage RGB-D pick up camera | |
CN103971409A (en) | Measuring method for foot three-dimensional foot-type information and three-dimensional reconstruction model by means of RGB-D camera | |
US10460517B2 (en) | Mobile device human body scanning and 3D model creation and analysis | |
CN107111833B (en) | Fast 3D model adaptation and anthropometry | |
CN109816724A (en) | Three-dimensional feature extracting method and device based on machine vision | |
CN107016697B (en) | A kind of height measurement method and device | |
CN104408762A (en) | Method for obtaining object image information and three-dimensional model by using monocular unit and two-dimensional platform | |
CN105701820A (en) | Point cloud registration method based on matching area | |
CN102779354B (en) | Three-dimensional reconstruction method for traditional Chinese medicine inspection information surface based on photometric stereo technology | |
CN102763923B (en) | Human body lower limb automatic measuring device and method | |
CN103948196A (en) | Human body data measuring method | |
CN103822581B (en) | A kind of irregularly shaped object volume measuring method based on compressed sensing | |
CN103535960A (en) | Human body three-dimensional measurement method based on digital images | |
CN110633005A (en) | Optical unmarked three-dimensional human body motion capture method | |
CN105212436B (en) | The measuring system and measuring method of non-contact 3-D laser foot type | |
CN103431582A (en) | Method for designing child shoe tree ground pattern based on dynamic and static foot models | |
CN104331924A (en) | Three-dimensional reconstruction method based on single camera SFS algorithm | |
CN108564653A (en) | Human skeleton tracing system and method based on more Kinect | |
CN104665107B (en) | A kind of vola three dimensional data collection processing system and data acquisition treatment method | |
CN107305699A (en) | Based on two-dimensional image information foot three-dimensional rebuilding method and its foot-footwear matching system | |
Sansoni et al. | 3-D optical measurements in the field of cultural heritage: the case of the Vittoria Alata of Brescia | |
CN103075977A (en) | Automatic combining algorithm for point cloud data in binocular stereoscopic vision system | |
CN113487726B (en) | Motion capture system and method | |
CN106767433A (en) | A kind of method and system for measuring foot sizing |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |