CN101187546A - Personnel space orientation automatic measuring method and system - Google Patents
Personnel space orientation automatic measuring method and system Download PDFInfo
- Publication number
- CN101187546A CN101187546A CNA200610118372XA CN200610118372A CN101187546A CN 101187546 A CN101187546 A CN 101187546A CN A200610118372X A CNA200610118372X A CN A200610118372XA CN 200610118372 A CN200610118372 A CN 200610118372A CN 101187546 A CN101187546 A CN 101187546A
- Authority
- CN
- China
- Prior art keywords
- image
- point
- personnel
- space orientation
- automatic measurement
- 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
Images
Landscapes
- Length Measuring Devices By Optical Means (AREA)
- Image Processing (AREA)
Abstract
The invention relates to a method and a system for automatic measuring personnel dimensional orientation, wherein a mathematical model of a real dimensional orientation coordinate and an image point dimensional coordinate is established through measuring the real dimensional orientation coordinate and the image point dimensional coordinate of a reference object, then an image of a target personnel is uptaken, the image point dimensional coordinate of a target point is detected after the target point of the target personnel in the image is detected and positioned, and then the real dimensional orientation coordinate of the target point is calculated according to the mathematical model. The invention can particularly realize automatic measuring to the dimensional orientation of left and right eyes of the personnel, the function of high measurement accuracy, quick measuring speed, and real-time measurement can be realized, simultaneously a dimensional orientation coordinate of the left and right eyes is further utilized to gain a posture orientation of human face or even a direction of a visual line, thereby a premise condition for personnel controlling, remeassuring and reprocessing is provided.
Description
Technical field
The present invention relates to a kind of method for automatic measurement and system of personnel space orientation.
Background technology
Regular meeting need measure the dimensional orientation of human eye in the reality; for example supervisory system need be obtained the positional information of people's face; perhaps certain control system need be measured people's position and carry out corresponding actions, perhaps has system to follow the tracks of or the like its sight line according to the orientation of human eye.
Therefore, how personnel's dimensional orientation is carried out fast and automatically measurement, and can guarantee that the precision of measuring becomes the technical task that industry needs to be resolved hurrily.
Summary of the invention
The object of the present invention is to provide a kind of method for automatic measurement of personnel space orientation,, realize that measuring accuracy height, measuring speed reach the function that measurement is carried out in real time soon to realize the automatic measurement of personnel's stereo omnibearing.
The object of the present invention is to provide a kind of automatic measurement system of personnel space orientation, the attitude that obtains people's face with the dimensional orientation coordinate of realizing further utilizing right and left eyes towards, or even the function pointed to of sight line, and then with the control that is embodied as personnel, measure and handle again the function that precondition is provided again.
In order to achieve the above object, the invention provides a kind of method for automatic measurement of personnel space orientation, it mainly may further comprise the steps: 1) the picture point volume coordinate of the described object of reference image of taking simultaneously according to the real space azimuthal coordinates of object of reference and at least two video cameras is set up the mathematical model in picture point and real space orientation; 2) described at least two video cameras while photographic subjects personnel's image; 3) detect the described target person image of each video camera picked-up respectively to identify the impact point of the target person in each width of cloth image; 4) locate impact point that is identified and the picture point volume coordinate of measuring described impact point; 5) according to the picture point volume coordinate of described impact point and the real space azimuthal coordinates of described calculated with mathematical model respective objects point, to determine the target person dimensional orientation.
Wherein, comprised step before described step 1): (1) measures the real space azimuthal coordinates of object of reference; (2) adopt at least two camera heads to take the image of described object of reference simultaneously; (3) the picture point volume coordinate of the reference point of the described object of reference of mensuration, described step 1) adopts photogrammetric collinearity equation to set up described mathematical model, impact point described in the described step 3) is a left and right sides eyes, described step 3) adopts eye socket centralized positioning method to locate described eyes, described step 3) adopts pupil center's localization method to locate described eyes, and described step 3) adopts the described eyes of canthus line centralized positioning.
The present invention also provides a kind of automatic measurement system of personnel space orientation, it mainly comprises: at least two camera heads that are used to take the image of object of reference and target person, be used to gather the captured object of reference and the image of target person, and detect the image detection module of the image of the captured target person of each camera head with the impact point in the recognition image, be used for measuring the picture point volume coordinate of reference point of the object of reference of the image of being gathered, and set up the mathematical model in picture point and real space orientation, and be used to measure the picture point volume coordinate of identified impact point and according to the image measurement module of the real space azimuthal coordinates of the described impact point of picture point spatial coordinates calculation of described mathematical model and impact point according to the picture point volume coordinate and the real space azimuthal coordinates thereof of described reference point.
Wherein, described at least two camera heads keep the parallel and predetermined at interval distance of optical axis.
In sum, the method for automatic measurement of personnel space orientation of the present invention and system are by setting up the mathematical model of real space azimuthal coordinates and picture point volume coordinate in advance, photographic subjects personnel's image and measure the picture point volume coordinate of the impact point of target person then, and go out respective objects personnel's dimensional orientation coordinate according to the calculated with mathematical model of being set up, realized the automatic measurement of personnel's stereo omnibearings, realization measuring accuracy height, measuring speed reaches the function that measurement is carried out in real time soon, further utilize attitude that the dimensional orientation coordinate of right and left eyes can obtain people's face towards, or even sight line points to, and then can be personnel's control, measure again and handle again precondition is provided.
Description of drawings
Fig. 1 is the basic framework synoptic diagram of the automatic measurement system of personnel space orientation of the present invention.
Fig. 2 is the practical structures synoptic diagram of the automatic measurement system of personnel space orientation of the present invention.
Fig. 3 is the operating process synoptic diagram of the method for automatic measurement of personnel space orientation of the present invention.
Embodiment
See also Fig. 1 and Fig. 2, the automatic measurement system of personnel space orientation of the present invention mainly comprises: at least two camera heads 11,12, image detection module 13 and image measurement module 14 below will be described in detail aforementioned each module.
Described at least two camera heads 11,12 are used for taking simultaneously the image of object of reference and target person, in the present embodiment, adopt the video camera of 2 suitable commercial measurements, be fixed and be installed on the metal support of size extensibility stable (swelling property is little), draw back certain distance (such as 1 meter) between two video cameras, and make two optical axises parallel and vertical with the metal support maintenance, wherein, distance between optical axes is called base length, when taking, two video cameras carry out simultaneously, are noted that, the distance between two video cameras can be determined according to actual conditions, but not exceed with present embodiment, described object of reference also can be chosen according to actual conditions, for example, selects certain special sign on the buildings.
Described image detection module 13 is used to gather the captured object of reference and the image of target person, and the image that detects the captured target person of each camera head is with the impact point in the recognition image, it can adopt the image pick-up card of high speed, high resolution to finish the high-definition picture of two video cameras of synchronous acquisition, and detect captured image in real time by CPU (central processing unit) (CPU), also can adopt digital signal processor (DSP) to finish image acquisition and testing, with from two camera heads 11,12 two captured width of cloth image recognitions go out identical point, for example, identify the impact point of target person, in the present embodiment, a plurality of reference points of object of reference also can adopt described image detection module 13 to discern automatically by manually discerning certainly.
Described image measurement module 14 is used for measuring the picture point volume coordinate of reference point of the object of reference of the image of being gathered, and the mathematical model of setting up picture point and real space orientation according to the picture point volume coordinate and the real space azimuthal coordinates thereof of described reference point, and be used to measure the picture point volume coordinate of identified impact point and according to the real space azimuthal coordinates of the described impact point of picture point spatial coordinates calculation of described mathematical model and impact point, in the present embodiment, described image measurement module 14 has been measured the picpointed coordinate of a plurality of reference points for accurately setting up described mathematical model, reduce the error of described mathematical model, it can adopt CPU (central processing unit) (CPU) to measure the picture point volume coordinate of impact point and calculate three-dimensional coordinate fast, also can adopt DSP to calculate the real space coordinate of impact point.
See also Fig. 2 again, it is the operating process synoptic diagram of the method for automatic measurement of personnel space orientation of the present invention, in the method for automatic measurement of described personnel space orientation, execution in step S10 at first measures the real space azimuthal coordinates of object of reference, for example adopts measuring method accurately to measure the real space azimuthal coordinates of described object of reference, because of measuring method is a prior art, and known to those skilled in the art knowing no longer given unnecessary details at this, then execution in step S11.
In step S11, adopt at least two camera heads to take the image of described object of reference simultaneously, in the present embodiment, adopt two video cameras that are fixed on the metal support to take the image of object of reference, then execution in step S12.
In step S12, measure the picture point volume coordinate of the reference point of described object of reference, normally be its picture point volume coordinate with respect to the distance of the initial point in picture point space by measuring described reference point, described initial point generally selects the point of fixity in the upper left corner of entire image as the coordinate starting point, in addition, also can adopt the theoretical picture point volume coordinate of measuring described reference point of collimation mark, be not described in detail in this, in the present embodiment, can measure the picture point volume coordinate of described object of reference by CPU, in addition, be noted that, can adopt manual type also can adopt automatic recognition method to the identification of the reference point in the captured image, then execution in step S13.
In step S13, the picture point volume coordinate of the described object of reference image of taking simultaneously according to the real space azimuthal coordinates of object of reference and at least two video cameras is set up the mathematical model in picture point and real space orientation, in the present embodiment, adopt photogrammetric collinearity equation to set up described mathematical model, accordingly, real space azimuthal coordinates and corresponding picture point volume coordinate according to object of reference can calculate elements of interior orientation and elements of exterior orientation, can set up corresponding collinearity equation mathematical model in view of the above, collinearity equation is disclosed by existing photogrammetry, no longer given unnecessary details at this, in addition, for reducing the error of the mathematical model of being set up, can select each the picture point volume coordinates of a plurality of reference points and real space coordinate as a reference, the mathematical model of being set up with continuous correction, improve the precision of described mathematical model, then execution in step S14.
In step S14, described at least two video cameras absorb the image of target person simultaneously, for example absorb the personnel's that are monitored image, then execution in step S15.
In step S15, the described target person image that detects the picked-up of each video camera respectively with identify target person in each width of cloth image impact point, described impact point is a left and right sides eyes, because of the position relation of human eye is determined, be that left eye always keeps left, right eye is always kept right, therefore in the image of two shot by camera, can discern human eye in two width of cloth images by image detecting technique, and the eye image that is positioned at the left side in two width of cloth images is thought same place, the eye image that is positioned at the right side in two width of cloth images is thought another same place, in addition, described impact point is not to exceed with eyes, for example, also can select nose for use, then execution in step S16.
In step S16, locate impact point in the image of described target person and measure the picture point volume coordinate of described impact point, in the present embodiment, can adopt eye socket centralized positioning method to locate described eyes, also can adopt pupil center's localization method to locate described eyes, eye socket centralized positioning method and pupil center's localization method are prior art, are not described in detail in this, can obtain left eye and right eye two same place two-dimensional image point coordinate data thus, then execution in step S17.
In described step S17, according to the picture point volume coordinate of described impact point and the real space azimuthal coordinates of described calculated with mathematical model respective objects point, computation process is the aforementioned inverse process of setting up mathematical model, therefore no longer repeats.
In sum, the actual measurement precision of the method for automatic measurement of personnel space orientation of the present invention and system directly is subject to the video camera quality, the levels of precision of camera position, the influence of factors such as measurement parameter demarcation and human eye two-dimensional localization precision, because video camera of the present invention is fixed on the metal support, the position is comparatively accurate, therefore, according to relevant system's practical experience, the human eye bearing accuracy can reach 1 pixel in the image, final space measurement precision reaches as high as picture side's 0.1 pixel, on processing speed, influence factor comprises image acquisition, people's face detects and the needed time of human eye localization process, usually human eye is located can in 40 milliseconds, finishing the image of 352 * 288 sizes on the CPU of P43.0G, if on CPU, realize the present invention, then might accomplish the real-time eye space measurement of coordinates of per second 25 frames.
Have again, according to eye socket central point position method can calculate further facial towards, add the sensing that pupil center's location rule can reflect actual sight line more accurately.
Claims (9)
1. the method for automatic measurement of a personnel space orientation is characterized in that may further comprise the steps:
1) the picture point volume coordinate of the described object of reference image of taking simultaneously according to the real space azimuthal coordinates of object of reference and at least two video cameras is set up the mathematical model in picture point and real space orientation;
2) described at least two video cameras absorb the image of target person simultaneously;
3) detect the described target person image of each video camera picked-up respectively to identify the impact point of the target person in each width of cloth image;
4) locate impact point that is identified and the picture point volume coordinate of measuring described impact point;
5) according to the picture point volume coordinate of described impact point and the real space azimuthal coordinates of described calculated with mathematical model respective objects point.
2. the method for automatic measurement of personnel space orientation as claimed in claim 1 is characterized in that: comprise step before the described step 1):
(1) the real space azimuthal coordinates of measurement object of reference;
(2) adopt at least two camera heads to take the image of described object of reference simultaneously;
(3) the picture point volume coordinate of the reference point of the described object of reference of mensuration.
3. the method for automatic measurement of personnel space orientation as claimed in claim 1, it is characterized in that: described step 1) adopts photogrammetric collinearity equation to set up described mathematical model.
4. the method for automatic measurement of personnel space orientation as claimed in claim 1 is characterized in that: the left and right sides eyes of the face of impact point behaviour described step 2).
5. the method for automatic measurement of personnel space orientation as claimed in claim 4 is characterized in that: described step 3) adopts eye socket centralized positioning method to locate described eyes.
6. the method for automatic measurement of personnel space orientation as claimed in claim 4 is characterized in that: described step 3) adopts pupil center's localization method to locate described eyes.
7. the method for automatic measurement of personnel space orientation as claimed in claim 4 is characterized in that: the described eyes of described step 3) employing canthus line centralized positioning.
8. the automatic measurement system of a personnel space orientation is characterized in that comprising:
At least two camera heads are used to take the image of object of reference and target person;
Image detection module is used to gather the captured object of reference and the image of target person, and the image that detects the captured target person of each camera head is with the impact point in the recognition image;
The image measurement module, be used for measuring the picture point volume coordinate of reference point of the object of reference of the image of being gathered, and set up the mathematical model in picture point and real space orientation, and be used to measure the picture point volume coordinate of identified impact point and according to the real space azimuthal coordinates of the described impact point of picture point spatial coordinates calculation of described mathematical model and impact point according to the picture point volume coordinate and the real space azimuthal coordinates thereof of described reference point.
9. the automatic measurement system of personnel space orientation as claimed in claim 8 is characterized in that: the parallel and predetermined at interval distance of described two camera heads maintenance optical axis at least.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB200610118372XA CN100491900C (en) | 2006-11-15 | 2006-11-15 | Personnel space orientation automatic measuring method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB200610118372XA CN100491900C (en) | 2006-11-15 | 2006-11-15 | Personnel space orientation automatic measuring method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101187546A true CN101187546A (en) | 2008-05-28 |
| CN100491900C CN100491900C (en) | 2009-05-27 |
Family
ID=39480039
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB200610118372XA Expired - Fee Related CN100491900C (en) | 2006-11-15 | 2006-11-15 | Personnel space orientation automatic measuring method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100491900C (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102577337A (en) * | 2009-09-16 | 2012-07-11 | 日本电气株式会社 | Mobile information device and display control method |
| CN103186233A (en) * | 2011-12-31 | 2013-07-03 | 上海飞来飞去新媒体展示设计有限公司 | Panoramic interaction control method for eye location |
| CN113091673A (en) * | 2021-03-23 | 2021-07-09 | 深圳市华星光电半导体显示技术有限公司 | Line width measuring method and line width measuring device |
-
2006
- 2006-11-15 CN CNB200610118372XA patent/CN100491900C/en not_active Expired - Fee Related
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102577337A (en) * | 2009-09-16 | 2012-07-11 | 日本电气株式会社 | Mobile information device and display control method |
| US8786581B2 (en) | 2009-09-16 | 2014-07-22 | Nec Corporation | Mobile information apparatus and display control method |
| CN102577337B (en) * | 2009-09-16 | 2015-08-05 | 日本电气株式会社 | Mobile information apparatus and display control method |
| US9390683B2 (en) | 2009-09-16 | 2016-07-12 | Nec Corporation | Mobile information apparatus and display control method |
| CN103186233A (en) * | 2011-12-31 | 2013-07-03 | 上海飞来飞去新媒体展示设计有限公司 | Panoramic interaction control method for eye location |
| CN103186233B (en) * | 2011-12-31 | 2016-03-09 | 上海飞来飞去新媒体展示设计有限公司 | Panoramic interaction control method for eye location |
| CN113091673A (en) * | 2021-03-23 | 2021-07-09 | 深圳市华星光电半导体显示技术有限公司 | Line width measuring method and line width measuring device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100491900C (en) | 2009-05-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7925049B2 (en) | Stereo-based visual odometry method and system | |
| US8823779B2 (en) | Information processing apparatus and control method thereof | |
| EP2745504B1 (en) | Image projector, image processing method, computer program and recording medium | |
| US9109889B2 (en) | Determining tilt angle and tilt direction using image processing | |
| EP2825841B1 (en) | Method, device and computer programme for extracting information about a staircase | |
| EP3067861A2 (en) | Determination of a coordinate conversion parameter | |
| CN100365654C (en) | Image processing device | |
| JP6981531B2 (en) | Object identification device, object identification system, object identification method and computer program | |
| CN113188484B (en) | Method for detecting outline area of head of hot-rolled coil | |
| JPWO2014002725A1 (en) | Three-dimensional measurement method, apparatus and system, and image processing apparatus | |
| CN112232279A (en) | Personnel spacing detection method and device | |
| CN108932732B (en) | Method and device for acquiring data information of monitored object | |
| KR20140055898A (en) | Digital map generating system for performing spatial modelling through a distortion correction of image | |
| KR101925289B1 (en) | Method and apparatus for identifying location/angle of terminal | |
| CN100491900C (en) | Personnel space orientation automatic measuring method and system | |
| KR20130034528A (en) | Position measuring method for street facility | |
| CN115371673A (en) | Binocular camera target positioning method based on Bundle Adjustment in unknown environment | |
| KR20140114594A (en) | Auto-Camera Calibration Method Based on Human Object Tracking | |
| CN111105467A (en) | Image calibration method and device and electronic equipment | |
| KR101234917B1 (en) | Method and system for generating 3-dimensional space data using image correction by threshold applied a weight information | |
| KR20190104663A (en) | Apparatus for measuring body size using drone | |
| KR101879858B1 (en) | Spatial modelling system for modelling spatial data by extracting common feature using comparison process between coordinate information of target object extracted from corrected image and coordinate information of arial photograph | |
| KR101332093B1 (en) | Spatial image processing system for obtaining 3-dimensional space data using coordination information of target object obtained by a plurality of sensors | |
| CN110178157A (en) | Self-position estimation device, self-position estimation method, program and image processing apparatus | |
| JP4674316B2 (en) | Position detection apparatus, position detection method, and position detection program |
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: 20090527 Termination date: 20151115 |
|
| EXPY | Termination of patent right or utility model |