CN109801329A - Human somatotype data measuring method based on multi-cam - Google Patents
Human somatotype data measuring method based on multi-cam Download PDFInfo
- Publication number
- CN109801329A CN109801329A CN201910071198.5A CN201910071198A CN109801329A CN 109801329 A CN109801329 A CN 109801329A CN 201910071198 A CN201910071198 A CN 201910071198A CN 109801329 A CN109801329 A CN 109801329A
- Authority
- CN
- China
- Prior art keywords
- skeleton
- cam
- human
- data
- measuring method
- 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.)
- Pending
Links
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses the human somatotype data measuring methods based on multi-cam, the following steps are included: A, acquiring respective image by multi-cam equipment and being pre-processed, the pre-treatment step are as follows: generate human ROI region in described image, the dummy skeleton at human body position in human ROI region is extracted by morphology Skeleton Extract；The coordinate of the dummy skeleton point is calculated according to multi-view geometry principle, and the estimated value of the length of the dummy skeleton is calculated by the coordinate；B, it is filtered to remove according to the estimated value of the dummy skeleton length for pre-processing and obtaining and make an uproar, and obtain non-noise data, again to the skeleton data for arriving finally after the non-noise data weighted average, human somatotype data are constituted finally by the skeleton data.
Description
Technical field
The present invention relates to multi-cams, human somatotype DATA REASONING, and in particular to the human somatotype number based on multi-cam
According to measurement method.
Background technique
When needing to measure human body stature, figure data, conventional method is often responsible for scene amount body by special messenger, and our
Case propose method be then using multi-cam equipment (any mould group equipped with multi-cam, mobile phone or other hardware devices) into
The self-service measurement of row, may not need other people and assists completing.
Summary of the invention
The present invention provides to solve the problems, such as that the prior art can not effectively acquire human somatotype data based on multi-cam
Human somatotype data measuring method, take the photograph equipment based on double and human body extracted by ROI, morphology Skeleton Extract etc.
Figure data.
The present invention is achieved through the following technical solutions:
Human somatotype data measuring method based on multi-cam, comprising the following steps:
A, respective image is acquired by multi-cam equipment and is pre-processed, the pre-treatment step are as follows:
Human ROI region is generated in described image, and people in human ROI region is extracted by morphology Skeleton Extract
The dummy skeleton of body body region；
The coordinate of the dummy skeleton point is calculated according to multi-view geometry principle, and is calculated by the coordinate
The estimated value of the length of the dummy skeleton；
B, it is filtered to remove according to the estimated value of the dummy skeleton length for pre-processing and obtaining and make an uproar, and obtain non-make an uproar
Point data, then to the skeleton data for arriving finally after the non-noise data weighted average, finally by the skeleton data structure
Adult body figure data.
Design principle of the invention is that the area for belonging to human body in picture is determined using the mode that ROI (area-of-interest) extracts
Domain, and required dummy skeleton is extracted by image processing algorithm, then be based on multi-view geometry principle, dummy skeleton can be found out
Geometry numerical value (length of relative position and skeleton line including middle skeleton point), and by the geometry numerical value of dummy skeleton by removing
Make an uproar and weighted average processing after, obtain final figure data；Wherein when extracting dummy skeleton, it is extracted description human body first
The basic skeleton of grown form, then extension skeleton is extracted based on this, more detailed portray is carried out to human body key position.
Further, described image is video or picture.
Further, the dummy skeleton includes basic skeleton and the extension skeleton that is expanded based on this basic skeleton,
Middle skeleton includes skeletal point and the skeleton line that is connected and composed by skeletal point.
Further, the human body position in described image is divided into head, trunk, four limbs, the head, trunk, four
Skeletal point and skeleton line can be generated in limb.
Further, the multi-cam equipment can acquire the image of human body or object different location, direction.
Compared with prior art, the present invention having the following advantages and benefits:
1, the present invention is based on the human somatotype data measuring methods of multi-cam, efficiently and complete by computerized algorithm
The figure data for acquiring human body reduce manual measurement trouble, assist without other people, can independently complete entire step completely；
2, that the present invention is based on the human somatotype data measuring method steps of multi-cam is simple, the speed of service is fast；
3, the present invention is based on the human somatotype data measuring method of multi-cam, measurement data is accurate, perfect.
Detailed description of the invention
Attached drawing described herein is used to provide to further understand the embodiment of the present invention, constitutes one of the application
Point, do not constitute the restriction to the embodiment of the present invention.In the accompanying drawings:
Fig. 1 is overall procedure schematic diagram of the present invention；
Fig. 2 is pretreatment process schematic diagram；
Fig. 3 is extension skeleton structure schematic diagram.
Specific embodiment
To make the objectives, technical solutions, and advantages of the present invention clearer, below with reference to embodiment and attached drawing, to this
Invention is described in further detail, and exemplary embodiment of the invention and its explanation for explaining only the invention, are not made
For limitation of the invention.
Embodiment
As shown in Figure 1, by it is double take the photograph mobile phone for, in Image Acquisition and pretreatment stage, adjust the position of dual camera first
It sets, it is ensured that the position can take human body.And to guarantee measurement effect, the person of being taken only wears personal thin clothing as far as possible
Object, start shooting when, people be moved to can by dual camera shoot position, at this time Image Pretreatment Algorithm carry out human testing with
Whether position identification, judgement currently take the position of human body and captured human body.User under the guidance of software, into
Whole body shooting, arm shooting, the shooting of bust waistline hip circumference, leg shooting etc. are completed in the adjustment that line position is set.Image Acquisition is completed
Afterwards, the video data of two cameras of algorithm process calculates human parameters based on image segmentation result and binocular ranging
And generate threedimensional model.
As shown in Fig. 2, being recorded a video when Image Acquisition by software, while image preprocessing is carried out, the step is specifically such as
Under: firstly, carrying out ROI (Region of interest) detection in video, ROI region is to belong to the area of human body in image
Domain；Physical feeling identification is carried out in ROI region, what the mode that physical feeling is identified by extraction human skeleton was completed.This
In skeletal extraction refer on morphology, human body can be divided into several connected parts such as trunk, four limbs in the picture,
Each part is substituted with an imaginary skeleton；Skeletal extraction includes skeletal point and skeleton line.Skeleton line is by part skeletal point
It being formed by connecting two-by-two, the quantity of skeletal point requires different according to measurement accuracy, and application scenarios difference has 9 points, 13 points, and 16
Point, 18 points are even more.
Here it is illustrated for figure data measuring method based on the example of 13 points to skeletal point and skeleton line with one
Processing mode.Manikin based on 13 each skeletal points as shown in Figure 3 can be on the image using deep learning method
The position for extracting these points, such as point red in figure.After obtaining skeletal point, to four limbs on image, torso portion is connected adjacent
The available description trunk of skeletal point, the skeleton line of four limbs, such as line segment blue in figure.Obtained skeletal point is operated by such
Become basic skeleton point and basic skeleton line with skeleton line.Thus can detect that whether present image includes human body, and current
Human body is mainly which position is photographed.
It on the basis of obtaining basic skeleton line, then is extended, be expanded skeletal point and extension skeleton line.They
Generating mode is as shown below.On the image, to take some position of this basic skeleton line (such as shown in right human hand in scheming
Midpoint), make vertical line, two intersection points can be obtained with the boundary (i.e. the contour edge of human region on image) of ROI on image.This
Dotted line is to extend skeleton line, and as shown in phantom in FIG., intersection point is to extend skeletal point, as shown in figure intermediate cam point.
According to different application scenarios and required precision, multiple groups can differently be expanded for a basic skeleton line
Extend skeletal point and skeleton line.
For two camera institute acquired images of dual camera, while the extraction of skeletal point and skeleton line is carried out, remembered
Record the position of the skeletal point and skeleton line detected in every frame image on the image plane.And judgement extracts those can be same
When the skeletal point that is observed by two cameras.According to the principle of binocular ranging, can calculate under Current camera coordinate system, this
The three-dimensional coordinate of a little skeletal points may further obtain the length of corresponding skeleton line, and record and obtain this skeleton line away from camera
Distance.
Finally after the completion of progress Data Post, all image procossings, for each skeleton line, one can be obtained
The length estimated value of series, and when obtaining the estimated value, distance of the skeleton line away from camera.According to these numerical value to every section of skeleton
Line carries out data filtering, noise is removed, and be weighted and averaged to non-noise data, it is hereby achieved that the length of every section of skeleton line
Degree.These numerical value constitute human somatotype data.
Above-described specific embodiment has carried out further the purpose of the present invention, technical scheme and beneficial effects
It is described in detail, it should be understood that being not intended to limit the present invention the foregoing is merely a specific embodiment of the invention
Protection scope, all within the spirits and principles of the present invention, any modification, equivalent substitution, improvement and etc. done should all include
Within protection scope of the present invention.
Claims (8)
1. the human somatotype data measuring method based on multi-cam, which comprises the following steps:
A, respective image is acquired by multi-cam equipment and is pre-processed, the pre-treatment step are as follows:
ROI is generated in described image, and the dummy skeleton at human body position in ROI is extracted by image processing algorithm；
The estimated value of the dummy skeleton length is calculated according to multi-view geometry principle；
B, it is filtered to remove according to the estimated value of the dummy skeleton length for pre-processing and obtaining and make an uproar, and obtain non-noise number
According to, then to the skeleton data for arriving finally after the non-noise data weighted average, people is constituted finally by the skeleton data
Body figure data.
2. the human somatotype data measuring method according to claim 1 based on multi-cam, which is characterized in that the figure
As being video or picture.
3. the human somatotype data measuring method according to claim 1 based on multi-cam, which is characterized in that the void
Quasi- skeleton includes skeletal point, and connects the skeleton line formed by the skeletal point.
4. the human somatotype data measuring method according to claim 3 based on multi-cam, which is characterized in that the bone
Frame point includes basic skeleton point and extension skeletal point, and the skeleton line includes basic skeleton line and extension skeleton line.
5. the human somatotype data measuring method according to claim 3 based on multi-cam, which is characterized in that the figure
Human body position as in is divided into head, trunk, four limbs, and skeletal point and skeleton can be generated in the head, trunk, four limbs
Line.
6. the human somatotype data measuring method according to claim 3 based on multi-cam, which is characterized in that the bone
Stringing is made of being connected two-by-two as skeletal point.
7. the human somatotype data measuring method according to claim 1 based on multi-cam, which is characterized in that described more
Camera device can acquire the image of human body or object different location, direction.
8. the human somatotype data measuring method according to claim 1 based on multi-cam, which is characterized in that the people
Body figure data are threedimensional model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910071198.5A CN109801329A (en) | 2019-01-25 | 2019-01-25 | Human somatotype data measuring method based on multi-cam |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910071198.5A CN109801329A (en) | 2019-01-25 | 2019-01-25 | Human somatotype data measuring method based on multi-cam |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109801329A true CN109801329A (en) | 2019-05-24 |
Family
ID=66558841
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910071198.5A Pending CN109801329A (en) | 2019-01-25 | 2019-01-25 | Human somatotype data measuring method based on multi-cam |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109801329A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105006014A (en) * | 2015-02-12 | 2015-10-28 | 上海交通大学 | Method and system for realizing fast fitting simulation of virtual clothing |
| KR20160035497A (en) * | 2014-09-23 | 2016-03-31 | (주)이튜 | Body analysis system based on motion analysis using skeleton information |
| CN107256565A (en) * | 2017-05-19 | 2017-10-17 | 安徽信息工程学院 | The measuring method and system of human body predominant body types parameter based on Kinect |
| CN107270829A (en) * | 2017-06-08 | 2017-10-20 | 南京华捷艾米软件科技有限公司 | A kind of human body measurements of the chest, waist and hips measuring method based on depth image |
-
2019
- 2019-01-25 CN CN201910071198.5A patent/CN109801329A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20160035497A (en) * | 2014-09-23 | 2016-03-31 | (주)이튜 | Body analysis system based on motion analysis using skeleton information |
| CN105006014A (en) * | 2015-02-12 | 2015-10-28 | 上海交通大学 | Method and system for realizing fast fitting simulation of virtual clothing |
| CN107256565A (en) * | 2017-05-19 | 2017-10-17 | 安徽信息工程学院 | The measuring method and system of human body predominant body types parameter based on Kinect |
| CN107270829A (en) * | 2017-06-08 | 2017-10-20 | 南京华捷艾米软件科技有限公司 | A kind of human body measurements of the chest, waist and hips measuring method based on depth image |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109299659A (en) | A kind of human posture recognition method and system based on RGB camera and deep learning | |
| US8116517B2 (en) | Action analysis apparatus | |
| CN104881881B (en) | Moving Objects method for expressing and its device | |
| US7162082B2 (en) | Method and apparatus for performing a clean background subtraction | |
| CN109784130B (en) | Pedestrian re-identification method, device and equipment thereof | |
| CN112487921B (en) | Face image preprocessing method and system for living body detection | |
| CN109919977A (en) | A kind of video motion personage tracking and personal identification method based on temporal characteristics | |
| CN105869166B (en) | A kind of human motion recognition method and system based on binocular vision | |
| CN109938737A (en) | A kind of human body body type measurement method and device based on deep learning critical point detection | |
| CN109271950A (en) | A kind of human face in-vivo detection method based on mobile phone forward sight camera | |
| CN104167016A (en) | Three-dimensional motion reconstruction method based on RGB color and depth image | |
| CN105869217B (en) | A kind of virtual real fit method | |
| CN109815816A (en) | A kind of examinee examination hall abnormal behaviour analysis method based on deep learning | |
| KR102202490B1 (en) | Device and method for measuring three-dimensional body model | |
| CN115035546B (en) | Three-dimensional human body posture detection method and device and electronic equipment | |
| CN111985579A (en) | Double-person diving synchronism analysis method based on camera cooperation and three-dimensional skeleton estimation | |
| CN115462783A (en) | Infant crawling posture analysis system based on skeleton key point detection | |
| Zhang et al. | Data-driven flower petal modeling with botany priors | |
| CN110580681B (en) | High-resolution cardiac motion pattern analysis device and method | |
| CN108197549A (en) | Face identification method and terminal based on 3D imagings | |
| CN114120068A (en) | Image processing method, image processing device, electronic equipment, storage medium and computer product | |
| CN112861588B (en) | Living body detection method and device | |
| CN116205952B (en) | Face recognition and tracking method and device, electronic equipment and storage medium | |
| CN109801329A (en) | Human somatotype data measuring method based on multi-cam | |
| CN115953838A (en) | Gait image tracking and identifying system based on MLP-Yolov5 network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190524 |
|
| RJ01 | Rejection of invention patent application after publication |