CN102783743A - Method for automatically detecting human body lower limb feature points - Google Patents
Method for automatically detecting human body lower limb feature points Download PDFInfo
- Publication number
- CN102783743A CN102783743A CN2012102393218A CN201210239321A CN102783743A CN 102783743 A CN102783743 A CN 102783743A CN 2012102393218 A CN2012102393218 A CN 2012102393218A CN 201210239321 A CN201210239321 A CN 201210239321A CN 102783743 A CN102783743 A CN 102783743A
- Authority
- CN
- China
- Prior art keywords
- human body
- point
- lower limb
- fibula
- detection
- 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
- 238000000034 method Methods 0.000 title claims abstract description 33
- 210000003141 lower extremity Anatomy 0.000 title claims abstract description 19
- 210000002414 leg Anatomy 0.000 claims abstract description 28
- 210000002082 fibula Anatomy 0.000 claims abstract description 27
- 210000003127 knee Anatomy 0.000 claims abstract description 23
- 238000001514 detection method Methods 0.000 claims abstract description 21
- 244000309466 calf Species 0.000 claims abstract description 16
- 210000001699 lower leg Anatomy 0.000 claims description 18
- 210000000689 upper leg Anatomy 0.000 claims description 14
- 239000000284 extract Substances 0.000 claims description 5
- 230000003467 diminishing effect Effects 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract 1
- 230000006835 compression Effects 0.000 abstract 1
- 238000007906 compression Methods 0.000 abstract 1
- 238000007670 refining Methods 0.000 abstract 1
- 238000005259 measurement Methods 0.000 description 13
- 230000002459 sustained effect Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 238000007598 dipping method Methods 0.000 description 3
- 238000003384 imaging method Methods 0.000 description 3
- 206010046996 Varicose vein Diseases 0.000 description 2
- 210000003423 ankle Anatomy 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 2
- 238000003825 pressing Methods 0.000 description 2
- 208000027185 varicose disease Diseases 0.000 description 2
- 210000003462 vein Anatomy 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 210000003205 muscle Anatomy 0.000 description 1
- 230000007170 pathology Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 230000000392 somatic effect Effects 0.000 description 1
- 230000000153 supplemental effect Effects 0.000 description 1
- 239000004753 textile Substances 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention relates to a method for automatically detecting human body lower limb feature points. The method comprises the following steps of 1) image collection: a testee wears a pair of special measuring trousers, an image collecting device is utilized for collecting a front side image and a lateral side image of the testee, and the leg outline is extracted; 2) detection in a position being 5cm from the crotch: the judgment is carried out according to the edges of the special measuring trousers, i.e., the edges are the position being 5cm from the crotch; 3) detection in a thickest position of the calf: a method of gradually reducing the detection range and carrying out refining layer by layer is adopted for detecting the thickest position of the calf; and 4) detection in a position of the fibula point: the detection is carried out in positions above the thickest position of the calf and below the position being 5cm from the crotch, the front side most concave position and the back side most concave position of the knee are determined, and the average value of the two positions is the fibula point. The method has the beneficial effects that through the automatic identification on each feature point of the lower limb of the human body, each dimension required by a medical compression stocking can be simply, conveniently and fully automatically obtained, a semi-automatic measuring method requiring label pasting or mouse clicking in the prior art is improved, and the method provided by the invention is more convenient, faster and more accurate.
Description
Technical field
The present invention relates to the clothes field of measuring technique, especially a kind of method of automatic human body lower limb characteristic point.
Background technology
The medical socks that reduce pressure in proper order pressurize to lumen of vessels through shrinking leg muscle by the slow principle of pressing, and promote the venous backflow heart, reduce veins of lower extremity extravasated blood, guarantee the veins of lower extremity good circulation.These socks can not only the supplemental treatment varix of lower limb, can also effectively prevent varix of lower limb, has important use and is worth and vast market prospect.The curative effect that the medical pressure socks want to obtain must customize corresponding socks like height, thigh circumference degree, calf circumference degree according to patient's leg size, and suitable pressure could in use be provided, and satisfies slow effect of pressing treatment.If socks are too loose, the pressure of generation is too small, can not effectively exert pressure; If opposite socks too tighten less than normal, will produce bigger pressure, let patient feel to feel bad, uncomfortable, and be unwilling to use.Therefore it still is important measuring seeming of patient's leg size.The method of traditional measurement human body is a manual measurement method, is main tool with tape, anthropometer, goniometer, type variable human body section gauge appearance etc., can directly measure partes corporis humani position size, and tool using is simple.But different measuring person's experience is different or measure gimmick standardized degree difference, causes measurement result to differ widely, and exists than mistake, and wastes time and energy; With contacting of measured, also can let the measured feel poverty-stricken or tired sometimes.People seek new measuring method and replace the traditional-handwork mensuration.It is the basis that automatic body is measured with the contemporary optics; Melt technology such as photoelectronics, computer graphics, information processing, computer vision and be one; The image information of measurand is used as the carrier that detects and transmit to be used; From image, extract useful letter, carry out processed, generate needed size.The automatic body measuring technique has remedied the deficiency that traditional-handwork is measured, and can easily and fast, accurately obtain somatic data.In recent years, the automatic body measurement has had many achievements.The Vitus Smart 3D laser scanning equipment of scanning systems such as the WBX of U.S. Cyberware company, WB4, French Lactra company, their essence is triangulation, the utilization similar triangle theory; With a plurality of laser range finders the measured who stand in the measure case is measured from a plurality of orientation; Calculate the three-dimensional coordinate of human body surface each point,, form human 3d model through Computer Processing; Then model is carried out three-dimensional dimension and measure, obtain point-device human dimension.This method certainty of measurement is high, but complex equipments is expensive, and floor space is big, is unfavorable for promoting the use of.People such as Meados proposed More's consistency profiles in 1970, were that the shade (projection) of application grating and the Moire fringe of grating formation are caught human body.Proposed image Moire technique, projection Moire technique, scanning Moire technique on this basis and improved one's methods.Basic principle be with light source irradiation to the key light grid that place on the testee, through the record distortion of grating on object the grating picture, obtain the depth information of body surface each point, thereby obtain body surface information.The Moire technique principle is simple, and precision is higher, but the data acquisition speed is slow, and poor stability is responsive to noise ratio.U.S. textile garment technology company's T C2 selects for use white light layering contour measurement to obtain the human body three-dimensional data message, detects the pattern profile that produces with six video cameras, synthesizes a complete model then.Can avoid the measured to take off based on the measuring method of red place line sensing technology and change one's clothes, and obtain the human body dead size; Body Line scanning system like the Hamamatsu company of Japan adopts red place line emitting diode to obtain scan-data, just do not need mark can extract three-dimensional data basically, and the mistakes and omissions data are less.
Generally there is following problem in these non-contact human measurement devices: (1) error is bigger.Because the said equipment is external basically, its model all is based upon on themselves basis, national demographic data storehouse, and when using in China, error is generally higher.According to some article, chest measurement, waistline that TC2 measures all have the 4-5cm error; The error of Canadian BOSS-21 also has about 5cm.(2) these equipment great majority are to use the 3D anthropometric scanning technology to gather initial data and image; The informix that is obtained is handled the three-dimensional dimension that obtains the partes corporis humani position; Carrying out human body rebuilds; General development system complex structure, bulky, cost is high, cost an arm and a leg, a set of equipment is at least more than 500,000.(3) these equipment are not to measure development to human body lower limbs, and each measurement all can produce huge data, can not produce the needed size of medical socks automatically, need the manual click feature point of people and search requiredly in the data to generating, and use very inconvenience.
Summary of the invention
The present invention will solve the shortcoming of above-mentioned prior art, provides a kind of realization full automatic lower limb dimensional measurement, simple, convenient, the automatic method of human body lower limb characteristic point efficiently.
The present invention solves the technical scheme that its technical problem adopts: the method for this automatic human body lower limb characteristic point; May further comprise the steps: 1) IMAQ: the measured wears specialty and measures trousers; Utilize image collecting device, gather measured front, two images in side and extract leg profile; 2) detection at the following 5 centimeters places of thigh root: the edge of measuring trousers according to specialty judges that promptly the edge is exactly the following 5 centimeters places of thigh root; 3) the shank thickness detects: adopt and progressively dwindle detection range, the method for layering refinement detects the thickest place of shank; 4) detection at fibula point place: more than shank is the thickest with under the thigh root, detect below 5 centimeters, confirm the most recessed and the most recessed position of rear side, knee front side, the mean value of two positions is the fibula point.
As preferably, getting the position that leg widths begins to diminish in the step 1) is the pants border height.
The detection range S1 that the shank thickness is set as preferably, step 2) is 20cm-46cm, and every 1cm carries out width and adds up; When accumulated value begins when diminishing greatly; Be exactly the zone at calf place, selected this place 2cm up and down is a region S 2, and detection range narrows down to S2 like this; Carry out every 2mm width again in the S2 zone and add up and compare, the value maximum is the shank thickness.
As preferably; Detecting below 5 centimeters more than the shank thickness with under the thigh root in the step 3); Find the minimum row-coordinate of leg widths in this scope earlier, detect in the 10cm on this journey the knee front side the most recessed with the recess of rear side, this two places average height is the height of fibula point.
Inventing useful effect is: the present invention is through discern each characteristic point of human body lower limbs automatically; Can simple and conveniently automatically carry out obtaining of each required size of medical pressure socks; Improved the semi-automatic measuring method that in the past needs labelling or click, more convenient, accurate.
The specific embodiment
Be described further in the face of the present invention down:
Embodiment: the measured wears specialty and measures trousers; Utilize image collecting device, gather measured front, two images in side and extract leg profile, wherein the measuring table assigned address at people station; This position is apart from dead ahead camera 180cm; Apart from side camera 178cm, image resolution ratio is 2048x1536, and two cameras overhead highly are 45cm.
The leg profile image of frontal automatically detects the left trouser legs edge that trousers are measured in identification, and the design of these measurements trousers is that the edge by left trouser legs is that crotch is done for following 5 centimeters, so the left trouser legs edge of measurement trousers is exactly following 5 centimeters of a thigh root.Because leg is three-dimensional, and image imaging is a two dimensional surface, human body is imaged as not the oblique line at sustained height on one's body on two-dimension picture at the trouser legs edge of sustained height.Can know from the image imaging principle, the object of sustained height, near more from camera, residing position is high more in picture, only could in picture, present sustained height with camera distance object identical and that oneself height is also identical.Left trouser legs edge in the image, distance (leg widths) is increasing between its left and right edges from top to bottom, the present invention gets leg widths, and to begin from the position that diminishes greatly be the pants border height.
To the leg profile image of side, detect the thickest position of identification shank automatically and be the calf position.At first to confirm the scope of shank; Know that according to the human dimension regularity of distribution ankle generally is high about 12cm, knee height is 0.26 of a human body height; Can know according to " GB10000-88 Chinese adult human dimension " standard; For the crowd of height from 140cm to 190cm, 99% people's knee height scope is 20cm-46cm at 36.5cm-46.9cm so the detection range S1 of shank thickness is set; This is the thickest detection range (actual size of later indication can convert the value of image with pixel unit into according to the image imaging principle), further dwindles detection range again, and per 20 line width of advancing add up.Can know that from the health inherent feature from being the trend that broadens gradually to calf more than the ankle, it is curved after arriving maximum, will slowly to be reduced to knee.When accumulated value begins when diminishing greatly, be exactly the zone at calf place, selected should the place 30 behavior region S 2 up and down, detection range narrows down to S2 like this; Carry out per two line width again in the S2 zone and add up and compare, value is maximum is the shank thickness.From S1 to S2, progressively dwindle detection range like this, carry out the layering refinement and detect, the erroneous judgement of having avoided human body lower limbs difference to bring.
To the leg profile image of side, detect identification fibula point position automatically.Can know from the human body characteristic: the fibula point is positioned at the following 2-4 centimetre of knee, and the front side of knee is protruding and rear side is recessed almost at sustained height.Through a large amount of experiments (manual markings fibula point; In the side image of gathering; The relation of manual detection knee position and fibula point position) analyzes and to know: for normal development; Do not receive the leg of damage and pathology, fibula point position necessarily is higher than the narrowest place more than the calf, near knee, leg profile near the recess of left hand edge.But different stances has a considerable influence to the left and right edges of knee is recessed; The recessed reduction of left hand edge during like little leaning forward, and knee is protruding not obvious, the recessed rising of left hand edge during hypsokinesis slightly; The recessed point of left hand edge is also different with the relation of fibula point like this, is higher than fibula point sometimes and is lower than the fibula point sometimes.Because the knee front side is protruding and rear side is recessed almost at sustained height in theory, by the recessed variation that comes the balance stance of knee rear side.Stance is different, and the recessed degree of knee rear side is also different, if leg is very thin and stand erectly and littlely forward incline; The state that then is in line, recessed point will be put much higher than fibula, simultaneously before recess before the knee projection can down reduce; Be lower than the fibula point, the fibula point is between the recess of knee front and back.Consider no matter how to do and measure the posture requirement; People's stance is certain to discrepant; For reducing the influence that stance detects fibula point as far as possible, the present invention takes all factors into consideration the height that calculates fibula point through the recess B point three place's height of the recess A point of the narrowest basePos of place more than the calf, knee front side and knee rear side.Be specially:
Detect the narrowest place more than the calf, be called for short basePos.The offside profile image; Begin toward following 5 centimeters of thigh root from calf, the width of adding up every capable leg is a shank left and right edges coordinate difference, for avoiding error effect; Use accumulation method; To the width summation that adds up of per 4 row, and all and in find minimum of a value, think to be that calf is to the narrowest basePos of place between the thigh root here.
Begin from basePos, up search the recess A point in knee front side for 10 centimeters.Image origin is in the upper left corner, and certain some horizontal direction value is X (x >=0) in the image, and the vertical direction value is Y (y >=0).If the vertical coordinate of leg profile left hand edge each point is y, then the y value characteristics at A place are: rate of change is maximum and be that rate of change is to reduce from increasing to change.Can find A by these characteristics.
Begin from basePos, up search the recess B point in knee front side for 10 centimeters.Method is the same, and the y value characteristics at B place are: rate of change is maximum and be that rate of change is from being reduced to increase.Can find B by these characteristics.
Fibula point highly is an A, some B height flat average (highA+highB)/2.
Because fibula point location difficulty is very big, even manual work can not find out at a glance that need go for to shank with hand, the error of different people hand dipping is also bigger.Method of the present invention possibly not accomplish accurately to locate the fibula point, but but can in the fair scope of certain error, find the fibula point.
The single measurement sample results is following:
Method | 5 centimeters of thigh roots | Calf | The fibula point |
Measured value of the present invention | 72.3cm | 31.1cm | 40.7cm |
The hand dipping value | 72.5cm | 31.5cm | 41.2cm |
Error | 0.2cm | 0.4cm | 0.5cm |
Hand dipping and the present invention's measurement through more than 200 human bodies compare, and average result is as shown in the table:
? | 5 centimeters of thigh roots | Calf | The fibula point |
Mean error | ?0.57cm | 0.43cm | 0.62cm |
Except that the foregoing description, the present invention can also have other embodiments.All employings are equal to the technical scheme of replacement or equivalent transformation formation, all drop on the protection domain of requirement of the present invention.
Claims (4)
1. the method for an automatic human body lower limb characteristic point may further comprise the steps:
1) IMAQ: the measured wears specialty and measures trousers, utilizes image collecting device, gathers measured front, two images in side and extracts leg profile;
2) detection at the following 5 centimeters places of thigh root: the edge of measuring trousers according to specialty judges that promptly the edge is exactly the following 5 centimeters places of thigh root;
3) the shank thickness detects: adopt and progressively dwindle detection range, the method for layering refinement detects the thickest place of shank;
4) detection at fibula point place: more than shank is the thickest with under the thigh root, detect below 5 centimeters, confirm the most recessed and the most recessed position of rear side, knee front side, the mean value of two positions is the fibula point.
2. the method for automatic human body lower limb characteristic point according to claim 1 is characterized in that: getting the position that leg widths begins to diminish in the step 1) is the pants border height.
3. the method for automatic human body lower limb characteristic point according to claim 1; It is characterized in that: step 2) in the shank thickness is set detection range S1 be 20cm-46cm; Per 20 line width of advancing add up, and when accumulated value begins when diminishing greatly, are exactly the zone at calf place; 30 behavior region S 2 about selected should the locating, detection range narrows down to S2 like this; Carry out per 2 line width again in the S2 zone and add up and compare, the value maximum is the shank thickness.
4. the method for automatic human body lower limb characteristic point according to claim 1; It is characterized in that: detecting below 5 centimeters more than the shank thickness with under the thigh root in the step 3); Find the minimum row-coordinate of leg widths in this scope earlier; Detect in the 10cm on this journey the knee front side the most recessed with the recess of rear side, this two places average height is the height of fibula point.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210239321.8A CN102783743B (en) | 2012-07-11 | 2012-07-11 | Method for automatically detecting human body lower limb feature points |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210239321.8A CN102783743B (en) | 2012-07-11 | 2012-07-11 | Method for automatically detecting human body lower limb feature points |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102783743A true CN102783743A (en) | 2012-11-21 |
CN102783743B CN102783743B (en) | 2014-07-23 |
Family
ID=47149503
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210239321.8A Expired - Fee Related CN102783743B (en) | 2012-07-11 | 2012-07-11 | Method for automatically detecting human body lower limb feature points |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102783743B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109190519A (en) * | 2018-08-15 | 2019-01-11 | 上海师范大学 | A kind of human body image crotch detection method |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5956525A (en) * | 1997-08-11 | 1999-09-21 | Minsky; Jacob | Method of measuring body measurements for custom apparel manufacturing |
CN101228973A (en) * | 2007-01-22 | 2008-07-30 | 殷实 | Non-contact measurement method and system for human outside measurement |
CN101467798A (en) * | 2007-12-27 | 2009-07-01 | 香港理工大学 | Method for producing intelligent pressure coat |
-
2012
- 2012-07-11 CN CN201210239321.8A patent/CN102783743B/en not_active Expired - Fee Related
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5956525A (en) * | 1997-08-11 | 1999-09-21 | Minsky; Jacob | Method of measuring body measurements for custom apparel manufacturing |
CN101228973A (en) * | 2007-01-22 | 2008-07-30 | 殷实 | Non-contact measurement method and system for human outside measurement |
CN101467798A (en) * | 2007-12-27 | 2009-07-01 | 香港理工大学 | Method for producing intelligent pressure coat |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109190519A (en) * | 2018-08-15 | 2019-01-11 | 上海师范大学 | A kind of human body image crotch detection method |
CN109190519B (en) * | 2018-08-15 | 2021-07-16 | 上海师范大学 | Human body image crotch detection method |
Also Published As
Publication number | Publication date |
---|---|
CN102783743B (en) | 2014-07-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102763923B (en) | Human body lower limb automatic measuring device and method | |
CN105608737B (en) | A kind of human foot three-dimensional rebuilding method based on machine learning | |
Kouchi et al. | Errors in landmarking and the evaluation of the accuracy of traditional and 3D anthropometry | |
CN202396601U (en) | Body-measuring and size-selecting device | |
Han et al. | Automatic body landmark identification for various body figures | |
CN102406280A (en) | Device for selecting size by body measurement | |
Chi et al. | Body scanning of dynamic posture | |
CN104366906A (en) | Arch contour detection method and system | |
CN204070693U (en) | A kind of three-dimensional human body measurement device of structure based light technology | |
CN112182682B (en) | Sportswear model generation method and sportswear model generation system | |
KR101557492B1 (en) | Apparatus and Method for generating user's three dimensional body model based on depth information | |
CN102783743B (en) | Method for automatically detecting human body lower limb feature points | |
KR101499699B1 (en) | Apparatus and Method for generating user's three dimensional body model based on depth information | |
CN105105416A (en) | Foot shape data collector and foot shape measuring method | |
CN105011447B (en) | Human body measuring reference point obtaining method for clothes | |
Kuzmichev et al. | Sizing and fit for pressure garments | |
CN202819751U (en) | Human body lower limb automatic measuring device | |
CN107729712B (en) | Gait stability evaluation method based on kinematic data | |
CN109801326A (en) | It is a kind of for obtaining the image measuring method of human somatotype data | |
CN109288176A (en) | A kind of method of cutting the garment according to the figure based on millimeter wave | |
CN207428505U (en) | A kind of clothing measurement on human body | |
CN207428504U (en) | A kind of amount body device of dress designing | |
CN206137260U (en) | Three -dimensional body trousers of surveying of women | |
Wen et al. | Study on segmentation of 3d human body based on point cloud data | |
CN206137243U (en) | Three -dimensional steel ring shape of surveying body brassiere of women |
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: 20140723 Termination date: 20150711 |
|
EXPY | Termination of patent right or utility model |