CN106767502A - A kind of circle codification index point with start information - Google Patents
A kind of circle codification index point with start information Download PDFInfo
- Publication number
- CN106767502A CN106767502A CN201611108392.9A CN201611108392A CN106767502A CN 106767502 A CN106767502 A CN 106767502A CN 201611108392 A CN201611108392 A CN 201611108392A CN 106767502 A CN106767502 A CN 106767502A
- Authority
- CN
- China
- Prior art keywords
- round dot
- coding
- layer
- witness marker
- index point
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/24—Measuring arrangements characterised by the use of optical techniques for measuring contours or curvatures
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of circle codification index point with start information, suitable for technical field of image processing, more particularly to the system of the index point coding and decoding in stereoscopic vision, the three-decker that coded target is made up of coding maker layer, witness marker layer and beginning flag layer, compared with existing other encoding schemes, this programme uses array circle rather than annulus, and group/cording quantity is more, can also have good stability to the image of affine transformation, it is adaptable to which the multi-view stereo vision under big visual field is demarcated.
Description
Technical field
The present invention relates to image processing field, specifically a kind of circle codification index point with start information.
Background technology
Object dimensional surface measurement technology is in automatic detection, profiling in kind, automated production, CAD, system
Make and the field such as quality control has a wide range of applications.Dimensional visual measurement method with its have it is untouchable, quick, to ring
Border requires first-selection of the advantages of low and measurement range is big as scheme.
In computer vision and three-dimensional measurement, in the measurement of large-sized three-dimensional body, multiple figures are in general needed
Complete three-dimensional appearance can be just obtained as sensor carries out data acquisition from multiple angles to three-dimensional body.And in such many biographies
In sensor network, preferably using the global registration method of global controlling network, by the depth data transform of different visual angles to system
The method of one reference frame completes the matching of many visual fields, then the accuracy of matching is to improve three-dimensional data splicing accuracy
Important step.
The establishment (corresponding point matching) of point correspondence is the Three-dimensional Gravity based on stereoscopic vision between the image of different visual fields
The basis built.But common non-coding index point is a round dot, its imaging be usually it is oval, each other cannot be
Made a distinction in form, for the stereo visual system without any priori (without demarcating), it is impossible to realize non-coding mark
The Corresponding matching of will point.Therefore need to design the distinguishing mark point of outward appearance --- coded target, by outward appearance for each is marked
Will point establishes different encoded radios, makes each coded target have unique identity information to establish the correspondence between encoded point
Relation.Since eighties of last century, coded target is widely used in digital close shot industrial photogrammetry.
The number of patent application of Zhang Chunsen and Feng Pengfei designs is that the coded target in 201310294713.9 is annular volume
The additional starting point of code mark point；The profile coding maker identifying system based on coordinate of Dong Ming et al. designs, patent
Application No. 201420686452.5, circular index point is encoded plus scatterplot when being in its design；Deng get Yan and Zhu Tao
The number of patent application of design be 201410348728.3 in coded target be made up of centre spot and encoded point；Meng Xiangli etc.
The number of patent application of people's design is the starting point and encoded point of the coded target designed in 201510815768.9 in same layer
In space, and starting point and central point are equal in magnitude.
In the prior art, annular coded target extensive use in large-scale workpiece measurement, but actually annular coding
There is problems with index point：
1st, the width of coding-belt is different：When coding-belt is more compared with long codes quantity, coding accurately identifies rate can be reduced；
2nd, without start code point, code capacity is small；In general, the coded target meeting without beginning flag point
Choose possible coding and get only minimum value as unique beginning flag position；During decoding, the method can cycle calculations, the increasing of calculating time
Plus simultaneously, group/cording quantity is also reduced.
3rd, coding-belt decodes stability decline when inclination angle is larger, is unfavorable for recognizing and matches.
The content of the invention
It is an object of the invention to provide a kind of circle codification index point with start information, to solve prior art presence
Problem.
In order to achieve the above object, the technical solution adopted in the present invention is：
, a kind of circle codification index point with start information, it is characterised in that：Witness marker layer including innermost layer,
The beginning flag layer of secondary internal layer, outermost coding maker layer, wherein witness marker layer are a witness marker round dot, the positioning
Mark round dot is filled circles, and beginning flag layer is the beginning flag round dot between witness marker layer and coding maker layer, should
Beginning flag round dot is also filled circles, the coded sequences that coding maker layer is constituted for multiple coding round dots, multiple coding round dots with
Witness marker round dot is uniformly looped around around witness marker round dot for the center of circle, and each coding round dot is filled circles or open circles, institute
Stating witness marker round dot, beginning flag round dot and coding round dot can mutually be distinguished by diameter.
A kind of described circle codification index point with start information, it is characterised in that：The diameter of beginning flag round dot
It is minimum.
A kind of described circle codification index point with start information, it is characterised in that：Constitute the volume of coding maker layer
Code round dot has 12.
A kind of described circle codification index point with start information, it is characterised in that：Witness marker round dot and surrounding
The position relationship of round dot can obtain coded sequence and beginning flag round dot, and beginning flag round dot determines the start bit of coded sequence
Put, coded sequence obtains 12 2 binary values being made up of 12 coding round dots, and then obtain by counterclockwise decoding
The numbering of coded target.
A kind of coding/decoding method of coded target, it is characterised in that：Comprise the following steps：
Step A, the coded target to being attached on object is photographed, and is extracted and is believed comprising coded target and surrounding
The initial pictures of breath, binary image is obtained for the binary conversion treatment that described initial pictures carry out self adaptation, is then passed through
And position adjacent pixel identical to pixel value scans for obtaining different connected domains, the connected region image needed for generation.
Step B, judges the circularity of connected domain, filters out each information for meeting coded target, then connect by each
The size relation in domain judges to obtain the coordinate information of witness marker round dot.
Step C, coded sequence and the beginning flag point of surrounding are found by the position relationship with witness marker round dot, decoding
Sequence obtains the numbering of decimal coded index point.
The present invention has identification accurate, and group/cording quantity is influenceed more and not by shooting angle.With before other people design
Circle codification index point compare, the present invention have group/cording quantity it is many, coding circle identification need not be differentiated by angle, stabilization
14 it is solid or open circles can increase the advantages such as the accuracy rate of identification.
Compared with prior art, advantage of the invention is also resided in：
1st, all it is made up of solid or open circles, index point is easily recognized and judged；
2nd, with start code index point, code capacity is greatly increased；
3rd, the coding circle of this programme is made up of solid and open circles, and identification is simple efficiently, and identifying schemes stabilization, than according to
Angle searching coded target is compared and is more not likely to produce error hiding.
Brief description of the drawings
Fig. 1 is coded target schematic diagram of the present invention, and wherein coding maker layer is by 10 open circles and 2 filled circles structures
Into.
Fig. 2 is lab diagram shooting figure of the present invention.
Fig. 3 is the OTSU Threshold segmentation figures of experiment.
Fig. 4 is the anchor point identification figure for encoding sequence number.
Fig. 5 a are the schematic diagrames for encoding sequence number.
Fig. 5 b are the lab diagrams after coding and decoding.
Specific embodiment
A kind of circle codification index point with start information, including innermost layer witness marker layer A, the starting of secondary internal layer
Reference lamina B, outermost coding maker layer C, wherein witness marker layer A is a witness marker round dot 1, the witness marker round dot
1 is filled circles, and beginning flag layer B is the beginning flag round dot 2 between witness marker layer and coding maker layer, starting mark
Will round dot 2 is also filled circles, and coding maker layer C is the coded sequence that multiple coding round dots 3 are constituted, multiple coding round dot 3
With witness marker round dot 1 for the center of circle is uniformly looped around around witness marker round dot 1, each coding round dot 3 is filled circles or hollow
Circle, witness marker round dot 1, beginning flag round dot 2 and coding round dot 3 can mutually be distinguished by diameter.
The diameter of beginning flag round dot 2 is minimum.
The coding round dot 3 for constituting coding maker layer has 12.
Witness marker round dot 1 can obtain coded sequence and beginning flag round dot 2 with the position relationship of surrounding round dot, starting
Mark round dot 2 determines the original position of coded sequence, and coded sequence is obtained by 12 coding round dots by counterclockwise decoding
12 2 binary values of 3 compositions, and then obtain the numbering of coded target.
A kind of coding/decoding method for coded target of the invention is comprised the following steps that：
Step A, the coded target to being attached on object is photographed, and is extracted and is believed comprising coded target and surrounding
The initial pictures of breath, binary image is obtained for the binary conversion treatment that described initial pictures carry out self adaptation, is then passed through
And position adjacent pixel identical to pixel value scans for obtaining different connected domains, the connected region image needed for generation.
Step B, judges the circularity of connected domain, filters out each information for meeting coded target, then connect by each
The size relation in domain judges to obtain the coordinate information of witness marker round dot.
Step C, coded sequence and the beginning flag point of surrounding are found by the position relationship with witness marker round dot, decoding
Sequence obtains the numbering of decimal coded index point.
Further, according to step A, first, target image is acquired using double camera, obtains the figure of left and right camera
2, then use the thresholding method of self adaptation -- otsu algorithms, for image I (x, y), prospect (i.e. target) and background point
Cut threshold value and be denoted as T, the pixel number for belonging to prospect accounts for the ratio of entire image and is designated as ω 0, its average gray μ 0；Background pixel point
The ratio that number accounts for entire image is ω 1, and its average gray is μ 1.The overall average gray scale of image is designated as μ, and inter-class variance is designated as g.It is false
If the background of image is dark, and the size of image is M × N, and number of pixels of the gray value of pixel less than threshold value T is remembered in image
Make N0, number of pixels of the pixel grey scale more than threshold value T is denoted as N1, then has:
ω1+ω0=1 (9)
N0+N1=M × N (10)
μ=ω0×μ0+ω1×μ1 (11)
G=ω0(μ0-μ)2+ω1(μ1-μ)2 (12)
Image after binaryzation carries out the judgement of connected domain, and 1 is determined whether to 8 fields that each pixel is broken:If
1, then with above inside a connected domain；If being not 1, a connected domain, image such as Fig. 3 are re-established
Further, according to step B, circularity judgesP represents girth, A represent area, connected domain obtain after energy
The girth and area of connected domain are calculated, in general, circular C values are 1, in experiment, due to the pass for having affine transformation
System, circle may occur in picture as oval, therefore the value of C is 1.2, can so be screened on the basis of connected domain
Go out is circle and oval connected domain, such as Fig. 4.
Further, it is big by area because the area of index point center circle is around justify more than 2 times according to step B
The small ratio with surrounding area, can rapidly determine center circle and starting point connects domain information experimental results such as Fig. 4.
According to step C, in the case that starting point and central point are obtained, K mean cluster scheme is used, have found surrounding
The information of circle is encoded according to K mean cluster method:K object is first randomly selected as initial cluster centre.Then calculate every
Each object, is distributed to the cluster centre nearest apart from it by the distance between individual object and each seed cluster centre.Cluster
Center and distribute to their object and just represent a cluster.Once whole objects are all assigned, the cluster of each cluster
Center can be recalculated according to existing object in cluster.This process will be repeated constantly until meeting certain end condition.
End condition can be it is following any one:
1) (or minimal amount) object is not had to be reassigned to different clusters.
2) (or minimal amount) cluster centre is not had to change again.
3) error sum of squares Local Minimum.
Further, according to step C, coding circle is decoded according to counter clockwise direction, and corresponding with initial point position is first
The information of individual value, 12 binary numerical value altogether, being converted into decimal system scheme can be as shown in figure 5a and 5b.
Claims (5)
1. a kind of circle codification index point with start information, it is characterised in that：Including innermost layer witness marker layer, it is secondary in
The beginning flag layer of layer, outermost coding maker layer, wherein witness marker layer are a witness marker round dot, the witness marker
Round dot is filled circles, and beginning flag layer is the beginning flag round dot between witness marker layer and coding maker layer, the starting
Mark round dot is also filled circles, and coding maker layer is the coded sequence that multiple coding round dots are constituted, and multiple encodes round dot to position
Mark round dot is uniformly looped around around witness marker round dot for the center of circle, and each encodes round dot for filled circles or open circles, described fixed
Bit flag round dot, beginning flag round dot and coding round dot can mutually be distinguished by diameter.
2. a kind of circle codification index point with start information according to claim 1, it is characterised in that：Beginning flag
The diameter of round dot is minimum.
3. a kind of circle codification index point with start information according to claim 1, it is characterised in that：Constitute coding
The coding round dot of reference lamina has 12.
4. a kind of circle codification index point with start information according to claim 1, it is characterised in that：Constitute coding
The coding round dot of reference lamina has 12.
5. a kind of coding/decoding method for coded target described in claim 1, it is characterised in that：Comprise the following steps：
Step A, the coded target to being attached on object is photographed, and is extracted comprising coded target and peripheral information
Initial pictures, obtain binary image, then by picture for the binary conversion treatment that described initial pictures carry out self adaptation
The element pixel that value is identical and position is adjacent scans for obtaining different connected domains, the connected region image needed for generation.
Step B, judges the circularity of connected domain, filters out each information for meeting coded target, then by each connected domain
Size relation judges to obtain the coordinate information of witness marker round dot.
Step C, coded sequence and the beginning flag point of surrounding, decoding sequence are found by the position relationship with witness marker round dot
Obtain the numbering of decimal coded index point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611108392.9A CN106767502A (en) | 2016-12-06 | 2016-12-06 | A kind of circle codification index point with start information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611108392.9A CN106767502A (en) | 2016-12-06 | 2016-12-06 | A kind of circle codification index point with start information |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106767502A true CN106767502A (en) | 2017-05-31 |
Family
ID=58878220
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611108392.9A Pending CN106767502A (en) | 2016-12-06 | 2016-12-06 | A kind of circle codification index point with start information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106767502A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107314771A (en) * | 2017-07-04 | 2017-11-03 | 合肥工业大学 | Unmanned plane positioning and attitude angle measuring method based on coded target |
| CN108520258A (en) * | 2018-04-04 | 2018-09-11 | 湖南科技大学 | Character code mark |
| CN109271823A (en) * | 2018-08-08 | 2019-01-25 | 佛山市顺德区中山大学研究院 | A kind of method for rapidly positioning of customized two dimensional code |
| CN109470142A (en) * | 2018-11-22 | 2019-03-15 | 北京航天计量测试技术研究所 | A kind of monocular pose measuring method based on circular index point |
| CN109579729A (en) * | 2018-11-12 | 2019-04-05 | 岭南师范学院 | A kind of annular encoded point and its coding/decoding method with start bit |
| CN109920010A (en) * | 2019-03-19 | 2019-06-21 | 易思维（杭州）科技有限公司 | The method that a kind of pair of polyphaser visual detection equipment is demarcated |
| CN110009692A (en) * | 2019-03-28 | 2019-07-12 | 渤海大学 | For the large-scale controlling filed artificial target of camera calibration and its coding method |
| CN111457914A (en) * | 2020-01-22 | 2020-07-28 | 南京航空航天大学 | Multi-rudder wing dynamic swing angle measuring method based on encodable visual cooperation target |
| CN113129394A (en) * | 2020-12-23 | 2021-07-16 | 合肥工业大学 | Parallelogram coding mark based on region division coding and coding method thereof |
| CN113240744A (en) * | 2020-01-23 | 2021-08-10 | 华为技术有限公司 | Image processing method and device |
| CN114994082A (en) * | 2022-06-28 | 2022-09-02 | 深圳市合力泰光电有限公司 | Double-precision mark point pattern structure and identification method |
-
2016
- 2016-12-06 CN CN201611108392.9A patent/CN106767502A/en active Pending
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107314771B (en) * | 2017-07-04 | 2020-04-21 | 合肥工业大学 | Unmanned aerial vehicle positioning and attitude angle measuring method based on coding mark points |
| CN107314771A (en) * | 2017-07-04 | 2017-11-03 | 合肥工业大学 | Unmanned plane positioning and attitude angle measuring method based on coded target |
| CN108520258A (en) * | 2018-04-04 | 2018-09-11 | 湖南科技大学 | Character code mark |
| CN109271823A (en) * | 2018-08-08 | 2019-01-25 | 佛山市顺德区中山大学研究院 | A kind of method for rapidly positioning of customized two dimensional code |
| CN109579729A (en) * | 2018-11-12 | 2019-04-05 | 岭南师范学院 | A kind of annular encoded point and its coding/decoding method with start bit |
| CN109579729B (en) * | 2018-11-12 | 2020-11-13 | 岭南师范学院 | Annular coding point with start bit and decoding method thereof |
| CN109470142A (en) * | 2018-11-22 | 2019-03-15 | 北京航天计量测试技术研究所 | A kind of monocular pose measuring method based on circular index point |
| CN109920010A (en) * | 2019-03-19 | 2019-06-21 | 易思维（杭州）科技有限公司 | The method that a kind of pair of polyphaser visual detection equipment is demarcated |
| CN110009692A (en) * | 2019-03-28 | 2019-07-12 | 渤海大学 | For the large-scale controlling filed artificial target of camera calibration and its coding method |
| CN111457914A (en) * | 2020-01-22 | 2020-07-28 | 南京航空航天大学 | Multi-rudder wing dynamic swing angle measuring method based on encodable visual cooperation target |
| CN113240744A (en) * | 2020-01-23 | 2021-08-10 | 华为技术有限公司 | Image processing method and device |
| CN113129394A (en) * | 2020-12-23 | 2021-07-16 | 合肥工业大学 | Parallelogram coding mark based on region division coding and coding method thereof |
| CN113129394B (en) * | 2020-12-23 | 2022-09-06 | 合肥工业大学 | Parallelogram coding mark based on region division coding and coding method thereof |
| CN114994082A (en) * | 2022-06-28 | 2022-09-02 | 深圳市合力泰光电有限公司 | Double-precision mark point pattern structure and identification method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106767502A (en) | A kind of circle codification index point with start information | |
| CN109285198B (en) | Method for coding and identifying annular coding mark points | |
| CN104662589B (en) | For the parallax detection in the image using array camera seizure and the system and method for correction | |
| US20160239975A1 (en) | Highly robust mark point decoding method and system | |
| CN103530889A (en) | Method for decoding coding mark points of measuring stick based on skeleton extraction | |
| CN106485183A (en) | A kind of Quick Response Code localization method and system | |
| CN107545277B (en) | Model training, identity verification method and device, storage medium and computer equipment | |
| CN104915949A (en) | Image matching algorithm of bonding point characteristic and line characteristic | |
| CN106651882B (en) | A kind of bird's nest impurities identification and detection method and device based on machine vision | |
| CN108573511B (en) | Point-distributed cooperative coding mark and identification and positioning method thereof | |
| CN110580481B (en) | Light field image key position detection method based on EPI | |
| CN101079102A (en) | Fingerprint identification method based on statistic method | |
| JP2007003233A (en) | Code target, cord detection system, and three-dimensional information acquisition system | |
| CN108550166B (en) | Spatial target image matching method | |
| CN110490913A (en) | Feature based on angle point and the marshalling of single line section describes operator and carries out image matching method | |
| CN104408772A (en) | Grid projection-based three-dimensional reconstructing method for free-form surface | |
| CN109215016A (en) | A kind of recognition positioning method of coding maker | |
| CN109190742A (en) | A kind of coding/decoding method of the coding characteristic point based on gray feature | |
| CN102446356A (en) | Parallel and adaptive matching method for acquiring remote sensing images with homogeneously-distributed matched points | |
| CN106203373B (en) | A kind of human face in-vivo detection method based on deep vision bag of words | |
| CN112364881B (en) | Advanced sampling consistency image matching method | |
| CN115345870A (en) | Method for realizing large-scene precise deformation monitoring based on monocular camera and self-luminous target with code | |
| CN115082888A (en) | Lane line detection method and device | |
| CN109034328A (en) | A kind of random anti-counterfeit marker characteristic extracts recognition methods, device and server | |
| Jin et al. | Depth-map driven planar surfaces detection |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170531 |