CN103838438B - A kind of infrared multipoint positioning method - Google Patents
A kind of infrared multipoint positioning method Download PDFInfo
- Publication number
- CN103838438B CN103838438B CN201410114588.3A CN201410114588A CN103838438B CN 103838438 B CN103838438 B CN 103838438B CN 201410114588 A CN201410114588 A CN 201410114588A CN 103838438 B CN103838438 B CN 103838438B
- Authority
- CN
- China
- Prior art keywords
- point
- visual angle
- distance
- true
- region
- 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.)
- Expired - Fee Related
Links
Landscapes
- Image Analysis (AREA)
Abstract
The present invention relates to a kind of infrared multipoint positioning method, described localization method comprises the following steps: preset window distance threshold t1；Each field of view information is added up at the traversal each visual angle of horizontal direction, it is assumed that the areal obtained is respectively K1, K2..., KN；The traversal each visual angle of vertical direction, adds up each field of view information, obtains areal and be respectively P1, P2..., PM；Calculate that candidate point concentrates the standard scores of each point；Statistics candidate point concentrates the true integration of each point；The judgement of true and false point.The invention have the benefit that the concept proposing various visual angles；Using two most for the areal that is blocked visual angles as benchmark visual angle, it is utilized to carry out the calculating of candidate touch point；By the region that is blocked to distance on the basis of the minimum range of point, other distance thoughts that herewith distance compares；Utilize the thought that auxiliary view is integrated；The most important thing is above 4 points.
Description
Technical field
The present invention relates to a kind of infrared multipoint positioning method.
Background technology
Existing infrared multiple spot identification technology has logical judgment method and image treating two kinds, and both of which is based on one
To many hardware scanning modes, ask for obstacles borders according to light circuit-switched data；Ask for standard according to border (to wait
Choosing) touch point set, and set true touch point threshold value；Travel through each candidate touch point, it is determined that current
Point is the credibility of true touch point.Credibility judges as follows, the region formed according to this border,
Through the light path set in this region, and should judge what this light path collective entity was blocked successively in anti-dianoetic
Number.Number is blocked divided by theory, if more than threshold value, being then true point, otherwise with the actual number that blocks
It is false point, removes it；Follow the tracks of and export, the touch point centre of form and exact boundary can be utilized respectively as sentencing
Determining region, during in touch point farther out, effect is preferable.But when touch point is nearer, can be owing to hiding mutually
Keep off and cause asking for mistake.It addition, when border, touch point interacts, border, touch point can be caused to become
Change, and then ask for region, touch point (determinating area) mistake, so since, the judgement of touch point verity
Lost efficacy.
The most basic thought of image treating is: true touch point relatively sizing grid substantially wants big, corrosion (filter
Ripple) dynamics controls to be removed by grid noise, retains true touch point simultaneously and be preferred.Therefore image method will
The actual touch object that None-identified is identical or close with grid noise size, though be now clearly present by
It is also such for blocking light path；It addition, time difference during high-speed motion can cause corrosion after lose situation a little.Send out
A person of good sense finds that above-mentioned multiple spot is asked for scheme and be there is significant deficiency, because logical approach only make use of the local of light path
Characteristic, and cannot be in view of the overall permanence of light path and the relation that influences each other of true touch point；And image
Method only make use of global property, it is impossible to identifies the actual touch object suitable with noise, and image method meter
Calculation amount is big, memory consumption big, very sensitive to delay problem, easily loses true touch point.Logical approach has
The advantage that computing is simple, treating capacity is little, the logical approach of a kind of improvement of the present invention so that it is can correctly process
With the situation of delay data, the impact that when reducing and eliminate multiple spot, touch point is blocked mutually, more correctly
Ground identifies all true touch points, preferably rejects ghost point and propose.
Summary of the invention
For weak point present in the problems referred to above, the present invention provides a kind of infrared multipoint positioning method.
For achieving the above object, the present invention provides a kind of infrared multipoint positioning method, described localization method bag
Include following steps:
Step 1: preset window distance threshold t1, difference distance threshold t2, true some decision threshold t3；
Step 2: determine level and the benchmark visual angle in vertical direction respectively；
Step 3: the candidate's point set drawn by horizontal reference visual angle and vertical reference visual angle calculates the most every
The standard scores of individual point；
Step 4: the candidate's point set drawn by horizontal reference visual angle and vertical reference visual angle is each in calculating
The true integration of point；
Step 5: the judgement of true and false point.
Further, in step 1, the region that is blocked just is integrated to the distance put less than this threshold value t1,
The most not integration, prevents the erroneous judgement that wire jumper and light path instability cause；With point and the region narrow spacing that is blocked
On the basis of from, if other points and this region distance difference with minimum range that is blocked are less than this threshold value t2, then
These points are also carried out integration；Threshold value t3 is the percentage ratio of the decimal in 0-1.
Further, in step 2, travel through each visual angle of horizontal direction, add up each field of view information,
The areal obtained is respectively K1, K2..., KN；The traversal each visual angle of vertical direction, adds up each
Field of view information, obtains areal and is respectively P1, P2..., PM；In horizontal direction, number of regions
Measure most i-th visual angle Ni(1≤i≤N) is considered as horizontal reference visual angle；In vertical direction, number of regions
Measure most jth visual angle Mj(1≤j≤M) is considered as vertical reference visual angle.
Further, in step 3, utilize two benchmark field of view orthogonal calculation respectively, obtain benchmark
Scene, i.e. intersection point set, as the intersection point in region, (point herein refers to by level the intersection point of centrage
The benchmark scene that benchmark field of view and vertical reference field of view obtain jointly, i.e. touch area), these
Intersection point is referred to as candidate's point set, and the intersection point of zone boundary may determine that the size of contact, according to the some position obtained
Put with putting size that (point herein refers to by horizontal reference field of view and vertical reference field of view common
The benchmark scene obtained, i.e. touch area), and light path visual angle slope (visual angle of light path refer to by
Block just to line and inclined line, just to line correspondence benchmark visual angle, inclined line correspondence auxiliary view), counter push away through
The visual angle number of specified point (specified point refers to touch point, is i.e. blocked a little), this number is as this point
Standard scores, if now calculate be 0, then force be set to 1.
Further, in step 4, the benchmark visual angle in removing both horizontally and vertically, by other visual angles
It is considered as auxiliary view, calculates in auxiliary view each region that is blocked respectively (by horizontal reference field of view
The benchmark scene jointly obtained with vertical reference field of view, be i.e. blocked region) and candidate point centrostigma
Distance；For each region that is blocked, it is sorted from small to large with the distance of all candidate points, with
Minimum range is as reference range, if reference range d < t1, then relevant to d dot product 1；If other distances
DOther and poor dOff < t2, the wherein dOff=dOther-d of minimum range d, then relevant to this distance point
Long-pending 1 point, the most not integration.
Further, in step 5, candidate point is asked to concentrate ratio r of each true integration of point and standard scores,
If r > t3, this point is true, is otherwise false, carries out smooth output according to true point and historical record.
The invention have the benefit that the concept proposing various visual angles；By most for the areal that is blocked two
Individual visual angle, as benchmark visual angle (resolution is the highest), utilizes it to carry out the calculating of candidate touch point；To be hidden
Gear region is to distance on the basis of the minimum range of point, and other (disappear apart from thoughts that herewith distance compares
The error brought except time delay (the most each fluorescent tube light on and off time irreversibility and contact rapid movement))；Profit
The thought being integrated by auxiliary view (other visual angles)；The most important thing is above 4 points, touch completing
After touching dot information extraction, residue ghost point can be removed further according to the restricting relation of each touch point, this
Step not necessarily, but can make algorithm more perfect.In scheme, a lot of steps may be by other
Some algorithms substitute, and therefore should protect emphatically this roadmap and solution principle thought.
Accompanying drawing explanation
Fig. 1 is the perspective definition figure of a kind of infrared multipoint positioning method of the present invention；
Fig. 2 is the area schematic that is blocked of a kind of infrared multipoint positioning method of the present invention；
Fig. 3 is be blocked region and the some distance schematic diagram of a kind of infrared multipoint positioning method of the present invention.
Detailed description of the invention
As Figure 1-3, the infrared multipoint positioning method of one described in the embodiment of the present invention, described location
Method comprises the following steps:
Step 1: preset window distance threshold t1, difference distance threshold t2, true some decision threshold t3；
Step 2: determine level and the benchmark visual angle in vertical direction respectively；
Step 3: the candidate's point set drawn by horizontal reference visual angle and vertical reference visual angle calculates the most every
The standard scores of individual point；
Step 4: the candidate's point set drawn by horizontal reference visual angle and vertical reference visual angle is each in calculating
The true integration of point；
Step 5: the judgement of true and false point.
Further, in step 1, the region that is blocked just is integrated to the distance put less than this threshold value t1,
The most not integration, prevents the erroneous judgement that wire jumper and light path instability cause；With point and the region narrow spacing that is blocked
On the basis of from, if other points and this region distance difference with minimum range that is blocked are less than this threshold value t2, then
These points are also carried out integration；Threshold value t3 is the percentage ratio of the decimal in 0-1.
Further, in step 2, travel through each visual angle of horizontal direction, add up each field of view information,
The areal obtained is respectively K1, K2..., KN；The traversal each visual angle of vertical direction, adds up each
Field of view information, obtains areal and is respectively P1, P2..., PM；In horizontal direction, number of regions
Measure most i-th visual angle Ni(1≤i≤N) is considered as horizontal reference visual angle；In vertical direction, number of regions
Measure most jth visual angle Mj(1≤j≤M) is considered as vertical reference visual angle.
In step 3, utilize two benchmark field of view orthogonal calculation respectively, obtain benchmark scene, i.e. hand over
Point set, the intersection point of centrage is as the intersection point in region, and these intersection points are referred to as candidate's point set, zone boundary
Intersection point may determine that the size of contact, according to the some position obtained and some size, and the visual angle of light path
Slope, counter pushes away the visual angle number through specified point, and the standard scores that this number is put as this, if now calculating
It is 0, then forces to be set to 1.
Further, in step 4, the benchmark visual angle in removing both horizontally and vertically, by other visual angles
It is considered as auxiliary view, calculates each region distance with candidate point centrostigma that is blocked in auxiliary view respectively；
For each region that is blocked, it is sorted from small to large with the distance of all candidate points, with minimum range
As reference range, if reference range d < t1, then relevant to d dot product 1；If other distances dOther
With the poor dOff < t2 of minimum range d, wherein dOff=dOther-d, then to this relevant dot product 1 point of distance,
The most not integration.
Further, in step 5, candidate point is asked to concentrate ratio r of each true integration of point and standard scores,
If r > t3, this point is true, is otherwise false, carries out smooth output according to true point and historical record.
The computing formula of the application:
Data:
Algorithm inputs: AD gathers data, and example is as follows:
Algorithm exports:
Calculated xy coordinate, data type is float.
Only as described above, only presently preferred embodiments of the present invention, be such as familiar with the special of this skill
Industry personage.After understanding the technological means of the present invention, natural energy is according to actual needs, in the present invention
Teaching under changed.The most all equal changes made according to scope of the present invention patent and modification,
Once should still remain within the scope of the patent.
Claims (2)
1. an infrared multipoint positioning method, it is characterised in that: described localization method comprises the following steps:
Step 1: preset window distance threshold t1, difference distance threshold t2, true some decision threshold t3；
The region distance to point that is blocked just is integrated less than this threshold value t1, and the most not integration prevents jumping
The erroneous judgement that line and light path instability cause；On the basis of point with the region minimum range that is blocked, if other points
With this region distance difference with minimum range that is blocked less than this threshold value t2, then these points are also carried out integration；
Threshold value t3 is the percentage ratio of the decimal in 0-1；
Step 2: determine level and the benchmark visual angle in vertical direction respectively；
The traversal each visual angle of horizontal direction, adds up each field of view information, and the areal obtained is respectively
For K1, K2..., KN；The traversal each visual angle of vertical direction, adds up each field of view information, obtains
Areal is respectively P1, P2..., PM；In horizontal direction, the i-th visual angle that region quantity is most
Ni(1≤i≤N) is considered as horizontal reference visual angle；In vertical direction, the jth visual angle that region quantity is most
Mj(1≤j≤M) is considered as vertical reference visual angle；
Step 3: the candidate's point set drawn by horizontal reference visual angle and vertical reference visual angle calculates the most every
The standard scores of individual point；
Utilize two benchmark field of view orthogonal calculations respectively, obtain benchmark scene, i.e. intersection point set, in
The intersection point of heart line is as the intersection point in region, and these intersection points are referred to as candidate's point set, and the intersection point of zone boundary is permissible
Determine the size of contact, according to the some position obtained and some size, and the visual angle slope of light path, counter push away
Through the visual angle number of specified point, the standard scores that this number is put as this, if now calculating is 0, then force
It is set to 1；
Step 4: the candidate's point set drawn by horizontal reference visual angle and vertical reference visual angle is each in calculating
The true integration of point；
Other visual angles are considered as auxiliary view, count respectively by the benchmark visual angle in removing both horizontally and vertically
Calculate each region distance with candidate point centrostigma that is blocked in auxiliary view；For each region that is blocked,
It is sorted from small to large with the distance of all candidate points, with minimum range as reference range, if benchmark
Distance d < t1, then relevant to d dot product 1；If other distances dOther and the poor dOff < t2 of minimum range d,
Wherein dOff=dOther-d, then relevant to this distance dot product 1 point, the most not integration；
Step 5: the judgement of true and false point.
Infrared multipoint positioning method the most according to claim 1, it is characterised in that: in step 5,
Asking candidate point to concentrate ratio r of each true integration of point and standard scores, if r > t3, this point is true, is otherwise false,
Smooth output is carried out according to true point and historical record.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410114588.3A CN103838438B (en) | 2014-03-25 | 2014-03-25 | A kind of infrared multipoint positioning method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410114588.3A CN103838438B (en) | 2014-03-25 | 2014-03-25 | A kind of infrared multipoint positioning method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103838438A CN103838438A (en) | 2014-06-04 |
| CN103838438B true CN103838438B (en) | 2016-12-07 |
Family
ID=50802014
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410114588.3A Expired - Fee Related CN103838438B (en) | 2014-03-25 | 2014-03-25 | A kind of infrared multipoint positioning method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103838438B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105335020B (en) * | 2014-08-15 | 2018-09-25 | 青岛海信电器股份有限公司 | A kind of touch point recognition methods and device |
| CN105373262B (en) * | 2014-09-02 | 2018-09-25 | 青岛海信电器股份有限公司 | A kind of method and device of the identification invalid light path of infrared touch panel |
| TWI529583B (en) * | 2014-12-02 | 2016-04-11 | 友達光電股份有限公司 | Touch system and touch detection method |
| CN105404433B (en) * | 2015-12-04 | 2019-06-07 | 青岛海信电器股份有限公司 | A kind of touch control identification method and display device based on infrared touch panel |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101387931A (en) * | 2008-10-14 | 2009-03-18 | 贺伟 | Infrared touch screen multi-point recognizing method |
| CN102270063A (en) * | 2010-06-03 | 2011-12-07 | 上海优熠电子科技有限公司 | Infrared true multi-point touch screen |
| CN102339170A (en) * | 2011-05-31 | 2012-02-01 | 广州视睿电子科技有限公司 | Signal scanning calculation method and system of infrared touch system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101697133B1 (en) * | 2009-09-30 | 2017-02-01 | 베이징 아이어터치 시스템 코퍼레이션 리미티드 | Touch screen, touch system and method for positioning a touch object in a touch system |
-
2014
- 2014-03-25 CN CN201410114588.3A patent/CN103838438B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101387931A (en) * | 2008-10-14 | 2009-03-18 | 贺伟 | Infrared touch screen multi-point recognizing method |
| CN102270063A (en) * | 2010-06-03 | 2011-12-07 | 上海优熠电子科技有限公司 | Infrared true multi-point touch screen |
| CN102339170A (en) * | 2011-05-31 | 2012-02-01 | 广州视睿电子科技有限公司 | Signal scanning calculation method and system of infrared touch system |
Non-Patent Citations (1)
| Title |
|---|
| "基于MCU的红外多点触摸屏设计";李均，谷灵康;《电脑知识与技术》;20120731;第8卷(第19期);第4701-4704页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103838438A (en) | 2014-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103838438B (en) | A kind of infrared multipoint positioning method | |
| CN103699908B (en) | Video multi-target tracking based on associating reasoning | |
| Dubois et al. | Human activities recognition with RGB-Depth camera using HMM | |
| CN102510506B (en) | Virtual and real occlusion handling method based on binocular image and range information | |
| CN106780557A (en) | A kind of motion target tracking method based on optical flow method and crucial point feature | |
| CN103164711A (en) | Regional people stream density estimation method based on pixels and support vector machine (SVM) | |
| CN103489175A (en) | Road surface detecting method and device | |
| CN105005999A (en) | Obstacle detection method for blind guiding instrument based on computer stereo vision | |
| Kuk et al. | Fast lane detection & tracking based on Hough transform with reduced memory requirement | |
| CN101763636A (en) | Method for tracing position and pose of 3D human face in video sequence | |
| CN105426858A (en) | Vision and vibration information fusion based ground type identification method | |
| US20190362163A1 (en) | Method for validation of obstacle candidate | |
| CN104183142A (en) | Traffic flow statistics method based on image visual processing technology | |
| CN104392239A (en) | License plate identification method and system | |
| Li et al. | SVM-based information fusion for weld deviation extraction and weld groove state identification in rotating arc narrow gap MAG welding | |
| CN104537342B (en) | A kind of express lane line detecting method of combination ridge border detection and Hough transformation | |
| CN103135136A (en) | Automatic fault interpretation device for three-dimensional seismic data body | |
| CN105844328A (en) | Method applied to automatic commissioning personnel counting system and automatic commissioning personnel counting system | |
| CN104574441B (en) | A kind of tumble real-time detection method based on GMM and temporal model | |
| CN112085675B (en) | Depth image denoising method, foreground segmentation method and human motion monitoring method | |
| CN109584265A (en) | A kind of method for tracking target and device | |
| CN104751146A (en) | Indoor human body detection method based on 3D (three-dimensional) point cloud image | |
| CN102750522B (en) | A kind of method of target following | |
| CN105096292A (en) | Object quantity estimation method and device | |
| CN105741326B (en) | A kind of method for tracking target of the video sequence based on Cluster-Fusion |
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 | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20171213 Address after: 100085 Beijing city Haidian District on the 28 Street Hospital No. 2 Building 2 layer 205 Patentee after: BEIJING HUAYUAN SHENGYA SCIENCE AND TECHNOLOGY CO.,LTD. Address before: 100176 days in Beijing City, park of Daxing District economic and Technological Development Zone in the two district two building 19 layer F1B9 Patentee before: QLTOUCH TECH Co.,Ltd. |
|
| TR01 | Transfer of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20161207 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |