CN102306285B - Human shape recognition method - Google Patents
Human shape recognition method Download PDFInfo
- Publication number
- CN102306285B CN102306285B CN201110232631.2A CN201110232631A CN102306285B CN 102306285 B CN102306285 B CN 102306285B CN 201110232631 A CN201110232631 A CN 201110232631A CN 102306285 B CN102306285 B CN 102306285B
- Authority
- CN
- China
- Prior art keywords
- height
- analyzing spot
- data
- distance
- scan
- 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.)
- Active
Links
Images
Landscapes
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention relates to a human shape recognition method, which comprises the following steps: scanning a scanning target in a scene to acquire scanning data; acquiring the height and distance of each scanning point in the scanning data; extracting scanning points with the height larger than a height threshold value; establishing a cone, wherein the cone comprises all scanning points with the height larger than a height threshold value; and if the height and the radius of the conical body accord with preset height and radius, judging that the scanning target in the scene is humanoid. According to the human shape identification method, the height and the distance of each scanning point are obtained by obtaining the scanning data of the scanning target in the scene, the cone is established by utilizing the scanning points with the height larger than the height threshold value, and the height and the radius of the cone are compared with the preset height and radius, so that the method for accurately judging whether the scanning target in the scene is the human shape is realized.
Description
Technical field
The present invention relates to humanoid recognition methods, relate in particular to and utilize 3 D laser scanning to detect the humanoid method of passenger flow.
Background technology
Along with social economy's prosperity and development, the increasing substantially of people's living standard, social public activities such as trip, tourism, party, shopping, amusement have been indispensable contents in people's daily life.Crowded places such as various public places, commercial street, tourist attractions usually are that bustling crowd is in an endless stream.Be example with the subway, more than 400 of the total gateways of Beijing Metro, 16 of transfer stops day send 4,000,000 person-times of passengers.Crowd massing has brought huge potential safety hazard for these places, has proposed higher management expectancy also for administrative authoritys at different levels.According to relevant regulations, surpass certain limit in crowd's quantity, administrative authority need take the restriction crowd to enter and the crowd such as dredges at the prediction scheme measure.Under the overall background that entire society more and more payes attention to safety problem, in time obtain information such as stream of people's speed, flow of the people, understand and grasp passenger flow ruuning situation in real time, to improving passenger flow dispatch control, emergency disposal and the public service level of going on a journey, set up safety, passenger flow organizes transportation system to have a very important role efficiently.
Two-dimensional laser or the infrared rays of adopting carry out coarse scan to the target in the scanning scene, the volume of the flow of passengers in the statistics scanning scene more at present.And the shape of the target that two-dimensional laser or infrared ray can't accurately scan, can only obtain scanning the rough shape of target, thereby other target misattribution that often will scan is humanoid, so adopt two-dimensional laser or infrared ray statistics volume of the flow of passengers degree of accuracy not high, can't obtain volume of the flow of passengers information accurately.
Summary of the invention
The objective of the invention is the defective at prior art, a kind of humanoid recognition methods is proposed, obtain the height and distance of each analyzing spot of scene interscan target by scanning, utilize height to set up cone-shaped body greater than the analyzing spot of height threshold, the height of cone-shaped body is compared with radius with default height with radius, thereby judged whether the scanning target in the scene is humanoid.
For achieving the above object, the invention provides a kind of humanoid recognition methods, described method comprises the steps:
Scanning target in the scene is scanned, obtain scan-data;
Obtain the height and distance of each analyzing spot in the described scan-data;
Extract height greater than the analyzing spot of height threshold;
Set up a cone-shaped body, described cone-shaped body comprises that whole height are greater than the analyzing spot of height threshold;
If the height of described cone-shaped body and radius meet default height and radius, judge that then the scanning target in the described scene is humanoid.
The humanoid recognition methods that the present invention proposes, obtain the height and distance of each analyzing spot by the scan-data that obtains scene interscan target, utilize height to set up cone-shaped body greater than the analyzing spot of height threshold, the height of cone-shaped body is compared with radius with default height with radius, thereby realized whether the scanning target in a kind of accurate judgement scene is humanoid method.
Description of drawings
Fig. 1 is the process flow diagram of the humanoid recognition methods of the embodiment of the invention;
Fig. 2 is that the position of the humanoid recognition methods of the embodiment of the invention concerns synoptic diagram.
Embodiment
Below by drawings and Examples, technical scheme of the present invention is described in further detail.
The humanoid recognition methods that the present invention proposes, three-dimensional laser scanner is fixed on aerial certain position as measurement point, utilize three-dimensional laser that the scanning target in the scene is scanned, obtain scan-data, obtain the height and distance (analyzing spot refers to the point of the scanning target different parts that three-dimensional laser scanner scans) of each analyzing spot in the scan-data, select the height of each analyzing spot in the scan-data greater than the analyzing spot of predefined height threshold; Set up a cone-shaped body, the analyzing spot of all height greater than predefined height threshold all included; With the height of cone-shaped body and radius and default height and radius ratio, if the height of cone-shaped body and radius meet default height and radius, judge that then the scanning target is humanoid, thereby realize whether a kind of accurate judgement scanning target is humanoid method.
Fig. 1 is the process flow diagram of the humanoid recognition methods of the embodiment of the invention, and as shown in Figure 1, the concrete steps of the humanoid recognition methods of the embodiment of the invention are:
Three-dimensional laser scanner is fixed on the surveyed area sky apart from ground L place as measurement point, three-dimensional laser scanner vertically downward, position centered by the straight line that three-dimensional laser scanner extends vertically downward, the three-dimensional laser scanner sweep limit for the vertical direction angle be in the scope of α, the α value is between 0 °～90 °, utilizes three-dimensional laser scanner that surveyed area is scanned.
When having the scanning target to enter scanning area in the scene, three-dimensional laser scanner scans the scanning target, obtains scan-data, and three-dimensional laser scanner utilizes laser distance measuring principle that scanning target different parts is scanned, and obtains scan-data.The scan-data that obtains for the analyzing spot of scanning target different parts to the range data of measurement point and detect the laser that scans target different parts analyzing spot and the angle data between the center.Analyzing spot refers to the point of the scanning target different parts that three-dimensional laser scanner scans.
Fig. 2 is the synoptic diagram that the position of the humanoid recognition methods of the embodiment of the invention concerns, as shown in Figure 2, P is the position of measurement point (three-dimensional laser scanner), Q is one of them analyzing spot, the position of corresponding ground when O point three-dimensional laser scanner scans vertically downward, position centered by the straight line that PO is linked to be.
Concrete, when detecting the scanning target, records three-dimensional laser scanner each analyzing spot to the distance L of three-dimensional laser scanner ' and detect the laser of analyzing spot and the angle α ' of vertical direction, as the distance L of analyzing spot A to three-dimensional laser scanner
a, detecting scan laser that A orders and the angle of vertical direction is a, then the A information of ordering is recorded as A (L
a, a), the information of other each analyzing spots is B (L by that analogy
b, b), C (L
c, c) ....
The height of each analyzing spot refers to each analyzing spot apart from the height on ground in the scan-data, and the distance of each analyzing spot refers to each analyzing spot to the distance at center in the scan-data.
The scan-data (surface sweeping is put the distance of measurement point and detected the laser of analyzing spot and the angle of vertical direction) that obtains according to step 101 draws each analyzing spot apart from the height on ground;
Again as shown in Figure 2, analyzing spot is figure middle distance H apart from the height on ground.Obtain analyzing spot to the distance of measurement point and detect the laser of analyzing spot and the angle information of vertical direction after, can be according to formula:
H＝L-L′cosα′(1)
Obtain each analyzing spot apart from the height on ground, wherein L be measurement point apart from the height on ground, L ' is that each analyzing spot is to the position of measurement point in the scan-data, α ' is for detecting the laser of each analyzing spot in the scan-data and the angle of vertical direction.
The information of each analyzing spot is brought into the height H that formula 1 can obtain each analyzing spot
A, H
B, H
C....
The scan-data (surface sweeping is put the distance of measurement point and detected the laser of analyzing spot and the angle of vertical direction) that obtains according to step 101 draws the distance of each analyzing spot decentering position.
Again as shown in Figure 2, the distance of analyzing spot decentering position i.e. figure middle distance M, in obtaining scan-data each analyzing spot to the distance of measurement point and detect the laser of analyzing spot and the angle information of vertical direction after, can be according to formula:
M＝L′sinα′(2)
Obtain analyzing spot decentering position apart from M, wherein L ' is that each analyzing spot is to the position of measurement point in the scan-data, α ' is for detecting the laser of each analyzing spot in the scan-data and the angle of vertical direction.
Each analyzing spot that step 102 is obtained is analyzed apart from the elevation information on ground, found out local extreme point, extreme point herein can be point highly the highest in the scan-data.Find out point around it according to extreme point, height in these analyzing spots greater than height threshold H
0Point pick out and form the some set.Height threshold H herein
0Decide height threshold H according to actual
0Can be to the arbitrary value ordinary people's height from ground, be preferably between 1m～1.2m, because at 1m to being normal person's waist or chest locations approximately between the 1.2m, in detection, get more than 1m～1.2m be human body waist or the analyzing spot more than the chest be big for fear of intensity of passenger flow in mistake analyzing spots of other scanning targets are counted in this scanning target, and be that the above analyzing spot of human body waist or chest seldom has other analyzing spot error statistics that scan targets in this scanning target more than 1m～1.2m.
To the set of select height greater than the analyzing spot of height threshold, adopt the mode of level and smooth sampling, tentatively carry out highl stratification.Set up model according to select height greater than the set of the analyzing spot of height threshold.The model of setting up can be cone-shaped, and the cone-shaped body model is included in the set of select height greater than the analyzing spot of height threshold in the model, and the area of qualified cone-shaped body is the smaller the better, height N and the radius R of the cone-shaped body of record area minimum.
Behind the height N and radius R of the cone-shaped body that obtains the area minimum, according to formula:
H′＝N-N′(3)
Obtain the relative height H ' from height N ' to the cone-shaped body top, wherein N is the height of cone-shaped body, and N ' value is between 1m～1.2m.
According to formula:
R′＝R?H′/N(4)
Obtain N ' locate the radius R of corresponding cone-shaped body '.
With H ', R ' and h given in advance
0, r
0Compare h
0Value is between 40cm～120cm, is that ordinary people's waist or chest are to the distance on the crown; r
0Between 20cm～30cm, it is the radius of the circumference of chest or waist.If the height H of cone-shaped body ' and radius R ' all meet h given in advance
0, r
0, judge that then the scanning target is humanoid, and do mark so that statistics; If the height H of cone ' and radius R ' have only one of them to meet or all do not meet h given in advance
0, r
0, judge that then the scanning target is not humanoid, then data are cast out, do not make marks.
The present invention utilizes three-dimensional laser that the scanning target in the scene is scanned, obtain the height and distance of each analyzing spot in the scan-data, and therefrom select height and set up a cone-shaped body greater than the set of the analyzing spot of preset height threshold value, by with the height of cone-shaped body and radius and default height and radius ratio, judge whether the scanning target in the scene is humanoid method, this humanoid recognition methods can judge fast and accurately whether the scanning target is humanoid by the target that real time scan enters in the scene, is conducive to add up accurately the volume of the flow of passengers.
Above-described embodiment; purpose of the present invention, technical scheme and beneficial effect are further described; institute is understood that; the above only is the specific embodiment of the present invention; and be not intended to limit the scope of the invention; within the spirit and principles in the present invention all, any modification of making, be equal to replacement, improvement etc., all should be included within protection scope of the present invention.
Claims (2)
1. a humanoid recognition methods is characterized in that, described method comprises the steps:
Scanning target in the scene is scanned, obtain the scan-data of the different parts of described scanning target;
Obtain the height and distance of each analyzing spot in the described scan-data, described analyzing spot is the point of the different parts of described scanning target, the height of described each analyzing spot be described each analyzing spot apart from the height on ground, the distance of described each analyzing spot is the distance of described each analyzing spot distance center position; Described center is the straight line that measurement point extends vertically downward;
Extract height greater than the analyzing spot of height threshold, described height threshold is arbitrary value between 1m～1.2m;
Described height is carried out highl stratification greater than the mode of the level and smooth sampling of analyzing spot employing of height threshold, set up a cone-shaped body, described cone-shaped body is to comprise that whole height are greater than the cone-shaped body of the area minimum of the analyzing spot of height threshold;
If the height of described cone-shaped body and radius meet default height and radius, judge that then the scanning target in the described scene is humanoid;
The described height that obtains each analyzing spot in the described scan-data is specially, according to formula:
H=L-L′cosα′
Obtain the height of each analyzing spot in the described scan-data, wherein L be measurement point to the distance on ground, L ' is that each analyzing spot is to the distance of measurement point in the described scan-data, α ' is for detecting the laser of each analyzing spot in the described scan-data and the angle of vertical direction;
The described distance of obtaining each analyzing spot in the described scan-data is specially, according to formula:
M=L′sinα′
Obtain the distance of each analyzing spot in the described scan-data, wherein L ' is that each analyzing spot is to the distance of measurement point in the described scan-data, and α ' is for detecting the laser of each analyzing spot in the described scan-data and the angle of vertical direction.
2. humanoid recognition methods according to claim 1 is characterized in that, described scanning target in the scene is scanned is specially, and adopts three-dimensional laser that the scanning target in the scene is scanned.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110232631.2A CN102306285B (en) | 2011-08-15 | 2011-08-15 | Human shape recognition method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110232631.2A CN102306285B (en) | 2011-08-15 | 2011-08-15 | Human shape recognition method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102306285A CN102306285A (en) | 2012-01-04 |
| CN102306285B true CN102306285B (en) | 2013-07-17 |
Family
ID=45380145
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110232631.2A Active CN102306285B (en) | 2011-08-15 | 2011-08-15 | Human shape recognition method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102306285B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111308463B (en) * | 2020-01-20 | 2022-06-07 | 京东方科技集团股份有限公司 | Human body detection method and device, terminal equipment, storage medium and electronic equipment |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201465103U (en) * | 2009-08-06 | 2010-05-12 | 北京北大千方科技有限公司 | Passenger flow detection and analysis device |
| CN101644717B (en) * | 2009-08-31 | 2011-06-08 | 北京大学 | Pedestrian advancing direction judging method adopting laser scanning |
| CN101777127B (en) * | 2010-03-11 | 2011-11-23 | 成都三泰电子实业股份有限公司 | Human body head detection method |
-
2011
- 2011-08-15 CN CN201110232631.2A patent/CN102306285B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN102306285A (en) | 2012-01-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104567708B (en) | Full section of tunnel high speed dynamical health detection means and method based on active panoramic vision | |
| Ai et al. | Critical assessment of an enhanced traffic sign detection method using mobile LiDAR and INS technologies | |
| CN103456172B (en) | A kind of traffic parameter measuring method based on video | |
| WO2018122819A1 (en) | Binocular image analysis-based asphalt road surface damage detection system | |
| CN106049210A (en) | Intelligent rail state detection platform | |
| McElhinney et al. | Initial results from European Road Safety Inspection (EURSI) mobile mapping project | |
| CN106428558B (en) | A kind of track synthesis method for inspecting based on the dual-purpose unmanned plane of sky-rail | |
| CN106980113A (en) | Article detection device and object detecting method | |
| CN103605978A (en) | Urban illegal building identification system and method based on three-dimensional live-action data | |
| CN108974044A (en) | Railroad track assets survey system | |
| CN106741890B (en) | A kind of high-speed railway safety detecting system based on the dual-purpose unmanned plane of empty rail | |
| CN104777522B (en) | Foreign matter detection system and detection method between a kind of Train door and platform screen door based on compound detection | |
| JP2017224935A (en) | Object detection/evaluation system and object detection/evaluation method | |
| CN114845924A (en) | Method for on-site and real-time collection and processing of geometrical parameters of a railway line | |
| CN103164958B (en) | Method and system for vehicle monitoring | |
| CN103699677A (en) | Criminal track map drawing system and method based on face recognition | |
| CN203310400U (en) | Limit detection system | |
| CN106494611B (en) | A kind of dual-purpose patrol unmanned machine of sky rail | |
| Lehtomäki et al. | Performance analysis of a pole and tree trunk detection method for mobile laser scanning data | |
| CN105844995A (en) | Railway line operation and maintenance measuring method based on vehicle-mounted LiDAR technology | |
| KR101237876B1 (en) | Methods and Systems which can classify vehicles class through the magnetic characteristic analysis of vehicles | |
| Hsu et al. | Design and implementation of an intelligent road detection system with multisensor integration | |
| CN204178554U (en) | A kind of laser scanning model recognition system freely flowing charge | |
| CN113295094B (en) | Pantograph dynamic envelope intrusion detection method, device and system | |
| CN102306285B (en) | Human shape recognition method |
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 | ||
| CP02 | Change in the address of a patent holder | ||
| CP02 | Change in the address of a patent holder |
Address after: 100193 2 / F, block B, qianfang building, 27 Zhongguancun Software Park, 8 Dongbeiwang West Road, Haidian District, Beijing Patentee after: CHINA TRANSINFO TECHNOLOGY Corp. Address before: 100191 Beijing Haidian District 39 Xueyuan Road 8 real estate building. Patentee before: CHINA TRANSINFO TECHNOLOGY Corp. |