CN105701496A - Go board surface identification method based on artificial intelligence technology - Google Patents
Go board surface identification method based on artificial intelligence technology Download PDFInfo
- Publication number
- CN105701496A CN105701496A CN201610016730.XA CN201610016730A CN105701496A CN 105701496 A CN105701496 A CN 105701496A CN 201610016730 A CN201610016730 A CN 201610016730A CN 105701496 A CN105701496 A CN 105701496A
- Authority
- CN
- China
- Prior art keywords
- chessboard
- line
- determined
- card
- view data
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/48—Extraction of image or video features by mapping characteristic values of the pattern into a parameter space, e.g. Hough transformation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
Abstract
The present invention provides a go board surface identification method and device. The method comprises the steps of pre-processing a frame of obtained go board surface image to generate the pre-processed image data; carrying out the clustering processing on the image data via a clustering method to generate a clustering result, and dividing the clustering result into three kinds of results to represent the white points, the black points and the blank points in a checkerboard respectively; carrying out the image identification on the pre-processed image data, and determining all identifiable line generation sample lines in the frame of image; carrying out the fitting processing on the sample lines to determine the checkerboard lines in two directions of the checkerboard; intersecting the determined checkerboard lines in the two directions, and determining the intersection point coordinates of the checkerboard lines on the checkerboard; determining the current go board surface state according to the determined clustering results and the intersection point coordinates of the checkerboard lines on the checkerboard. According to the present invention, the tedious problem that the major of notation modes at present need the artificial input is solved, a lot of cumbersome equipment investment is saved, and the tedious workload at the notation is reduced.
Description
Technical field
The present invention relates to image recognition technology, be a kind of go card recognition methods and device concretely。
Background technology
Electric Weiqi of the prior art note spectrum equipment is generally all made up of chessboard, outut device etc., when playing chess, chess player normally plays chess on equipment, data are transmitted automatically to display platform, various functions is realized by software, therefore this note spectrum discrimination method is used, must purchasing various electronics note spectrum equipment, and the time occasion competed is all very restricted, the quality of product is also uneven；And more existing go note spectrum software is also required to be manually entered existing chess manual, and, both the above method all cannot identify existing chess manual automatically。
Additionally, the software of recognizable chess manuals more of the prior art, although having accomplished to utilize procedure identification to have this function of chess manual, but require to shoot in advance the picture of a frame sky chessboard, make comparisons with empty chessboard after next piece of chess piece, record sub-point, repeat this process afterwards, the method requires chessboard in the process played chess, the position imaging first-class equipment cannot be moved, therefore, the method is to context, the aspects such as the angle of shooting require very high, and, its maximum problem is must at same position and angle photographs blank chessboard before note spectrum, this makes to use the method when the demand having note to compose in advance, if chess under carry out note to a certain extent again and compose and had no idea, namely the time not only note composed, the place that chessboard is put, outside the aspects such as chessboard is fixing with the relative position of photographic head are restricted, also effective only for the user just having note spectrum demand before playing chess。
Summary of the invention
For meeting, the chess manual of generation in process of arbitrarily playing chess is carried out note spectrum demand so that note spectrum, not by the restriction of the factors such as any time, place and shooting angle, embodiments provides a kind of go card recognition methods, including:
One two field picture of the go card obtained is carried out the pretreatment pretreated view data of generation；
The clustering processing of described view data is generated cluster result by the method adopting cluster, and described cluster result, for class value to be divided three classes, represents the white point in chessboard, black color dots and blank spot respectively；
Described pretreated view data is carried out image recognition, it is determined that in this two field picture, all discernible lines generate sample line；
It is fitted described sample line processing the chessboard line determined in chessboard both direction；
Chessboard line in the both direction determined is intersected, it is determined that the intersecting point coordinate of chessboard line on chessboard；
Current I-go board surface state is determined according to the intersecting point coordinate of chessboard line on the cluster result determined and chessboard。
In the embodiment of the present invention, a two field picture of the go card of described acquisition is carried out pretreatment and includes:
One two field picture of the go card obtained is carried out color enhancement, ashing process, Laplace transform and standardization processing。
In the embodiment of the present invention, the pretreated view data of described generation includes: generate the picture element matrix of this two field picture。
In the embodiment of the present invention, described carries out image recognition to described pretreated view data, it is determined that in this two field picture, all discernible lines generation sample line includes:
Utilize the Hough transformation in image recognition that view data is carried out image recognition。
In the embodiment of the present invention, sample line is fitted processes the chessboard line determined in chessboard both direction and includes:
Described sample line is sequentially carried out direct matching, weighted fitting and ransac matching and determines the chessboard line in a direction；
From described sample line, delete the chessboard line determined, remaining model line is carried out again according to the order of sequence direct matching, weighted fitting and ransac matching and determines the chessboard line of other direction。
Meanwhile, the present invention also provides for a kind of go card identification device, and device includes:
Pretreatment module, generates pretreated view data for a two field picture of the go card obtained is carried out pretreatment；
Clustering processing module, for adopting the method for cluster that the clustering processing of described view data is generated cluster result, described cluster result, for class value to be divided three classes, represents the white point in chessboard, black color dots and blank spot respectively；
Sample line generation module, for carrying out image recognition to described pretreated view data, it is determined that in this two field picture, all discernible lines generate sample line；
Fitting module, processes the chessboard line determined in chessboard both direction for described sample line is fitted；
Module determined by chessboard, for being intersected by the chessboard line in the both direction determined, it is determined that the intersecting point coordinate of chessboard line on chessboard；
Card identification module, for determining current I-go board surface state according to the intersecting point coordinate of chessboard line on the cluster result determined and chessboard。
What the invention solves that Most current note spectrum mode is required for being manually entered is loaded down with trivial details, save a lot of heavy equipment investment, the relative position between photographic head and chess manual need not be controlled, also angle time to shooting, the impact of context reduces requirement, from the tedious work amount largely solving note time spectrum。
For the above and other purpose of the present invention, feature and advantage can be become apparent, preferred embodiment cited below particularly, and coordinate institute's accompanying drawings, it is described in detail below。
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, the accompanying drawing used required in embodiment or description of the prior art will be briefly described below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the premise not paying creative work, it is also possible to obtain other accompanying drawing according to these accompanying drawings。
Fig. 1 is the flow chart of a kind of go card recognition methods disclosed in the embodiment of the present invention；
Fig. 2 is the block diagram of a kind of go card identification device provided by the invention。
Detailed description of the invention
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is only a part of embodiment of the present invention, rather than whole embodiments。Based on the embodiment in the present invention, the every other embodiment that those of ordinary skill in the art obtain under not making creative work premise, broadly fall into the scope of protection of the invention。
The invention discloses a kind of go card recognition methods, as it is shown in figure 1, disclosed in the embodiment of the present invention flow chart of a kind of go card recognition methods, the method includes:
Step S101, carries out the pretreatment pretreated view data of generation to a two field picture of the go card obtained；
Step S102, adopts the method for cluster that the clustering processing of described view data is generated cluster result, and described cluster result, for class value to be divided three classes, represents the white point in chessboard, black color dots and blank spot respectively；
Step S103, carries out image recognition to described pretreated view data, it is determined that in this two field picture, all discernible lines generate sample line；
In the embodiment of the present invention, utilize the Hough transformation in image recognition that view data is carried out image recognition。
Step S104, is fitted described sample line processing the chessboard line determined in chessboard both direction；
Step S105, it is to be determined to both direction on chessboard line intersect, it is determined that the intersecting point coordinate of chessboard line on chessboard；
Step S106, determines current I-go board surface state according to the intersecting point coordinate of chessboard line on the cluster result determined and chessboard。
In the embodiment of the present invention, a two field picture of the go card of described acquisition is carried out pretreatment and includes:
One two field picture of the go card obtained is carried out color enhancement, ashing process, Laplace transform and standardization processing, and the pretreated view data of described generation includes: generate the picture element matrix of this two field picture。
Meanwhile, the present invention also provides for a kind of go card identification device, as in figure 2 it is shown, this device includes:
Pretreatment module 201, generates pretreated view data for a two field picture of the go card obtained is carried out pretreatment；
Clustering processing module 202, for adopting the method for cluster that the clustering processing of described view data is generated cluster result, described cluster result, for class value to be divided three classes, represents the white point in chessboard, black color dots and blank spot respectively；
Sample line generation module 203, for carrying out image recognition to described pretreated view data, it is determined that in this two field picture, all discernible lines generate sample line；
Fitting module 204, processes the chessboard line determined in chessboard both direction for described sample line is fitted；
Module 205 determined by chessboard, for being intersected by the chessboard line in the both direction determined, it is determined that the intersecting point coordinate of chessboard line on chessboard；
Card identification module 206, for determining current I-go board surface state according to the intersecting point coordinate of chessboard line on the cluster result determined and chessboard。
Below in conjunction with concrete enforcement step, the embodiment of the present invention is described in further details:
1, from photographic head, read a two field picture of current chessboard, image is carried out a series of pretreatment (including color enhancement, ashing process, Laplace transform and standardization processing), the process requirement in step after making image reach by the method utilizing image recognition；
Namely to after the pretreatment of image, the picture element matrix of image。
2, use the method for cluster that the pixel class value in picture element matrix is divided into 3 classes (namely representing the white point of chessboard state, black color dots and blank spot), in the present embodiment, picture element matrix can be carried out repeatedly clustering processing, take result several times afterwards preferably standby；
3, use the method for Hough transformation in image recognition to find out all discernible lines in image, and these lines are stored；In the present embodiment, available opencv instrument carries out image recognition。
4, because the line on chessboard has the character of only both direction, so the line obtained is regarded sample line, these sample line are sequentially carried out direct matching, weighted fitting and ransac matching to find the suitable model line to portray in all sample points on a direction, again by these lines as test set, the good line that model is adapted to is stored another place, and all of line in the direction is deleted；
5, repeat the 4th step, from remaining sample point, find the line on another direction that model portrays, and these lines are stored other place；
6, being merged by the line in the both direction of gained, several lines identified near the same line are fused into most suitable line, thus obtain on chessboard 38 lines；
7, being intersected between two by the line in the both direction determined, calculate the coordinate of each intersection point, the coordinate being on chessboard all intersection points, by all coordinates by total big to little sequence, thus obtaining this coordinate relative position on chessboard；
8, it is applied on each point by clustering the result obtained before, it is judged which kind of current point belongs to, thus judging the chess piece state of every bit, thus obtaining the state matrix of a 19*19, is the state of current chessboard。
What the invention solves that Most current note spectrum mode is required for being manually entered is loaded down with trivial details, eliminate the equipment investment of heaviness, time and place that note spectrum occurs are liberated, the embodiment of the present invention utilizes up-to-date OPENCV image recognition technology simultaneously, the relative position between photographic head and chess manual need not be controlled, also angle time to shooting, the impact of context reduces requirement, from the tedious work amount largely solving note time spectrum。
The present invention adopts state-of-the-art opencv instrument at present to carry out image recognition, and image recognition is combined with cluster matching, is the greatest feature of this programme；Wherein comparatively crucial place has 1, arranges suitable parameter, utilizes the method for Hough transformation in opencv to find the satisfactory line in image；2, adjust the computational methods in matching, carry out direct matching, weighted fitting and ransac matching to institute is wired, obtained the adaptive model that the line in chessboard is best；3, utilize the method for cluster that intersection point all of on chessboard is judged, thus obtaining the state of each point。
Applying specific embodiment in the present invention principles of the invention and embodiment are set forth, the explanation of above example is only intended to help to understand method and the core concept thereof of the present invention；Simultaneously for one of ordinary skill in the art, according to the thought of the present invention, all will change in specific embodiments and applications, in sum, this specification content should not be construed as limitation of the present invention。
Claims (10)
1. a go card recognition methods, it is characterised in that described method includes:
One two field picture of the go card obtained is carried out the pretreatment pretreated view data of generation；
The clustering processing of described view data is generated cluster result by the method adopting cluster, and described cluster result, for class value to be divided three classes, represents the white point in chessboard, black color dots and blank spot respectively；
Described pretreated view data is carried out image recognition, it is determined that in this two field picture, all discernible lines generate sample line；
It is fitted described sample line processing the chessboard line determined in chessboard both direction；
Chessboard line in the both direction determined is intersected, it is determined that the intersecting point coordinate of chessboard line on chessboard；
Current I-go board surface state is determined according to the intersecting point coordinate of chessboard line on the cluster result determined and chessboard。
2. go card recognition methods as claimed in claim 1 a, it is characterised in that two field picture of the go card of described acquisition carries out pretreatment and includes:
One two field picture of the go card obtained is carried out color enhancement, ashing process, Laplace transform and standardization processing。
3. go card recognition methods as claimed in claim 2, it is characterised in that the pretreated view data of described generation includes: generate the picture element matrix of this two field picture。
4. go card recognition methods as claimed in claim 1, it is characterised in that described carries out image recognition to described pretreated view data, it is determined that in this two field picture, all discernible lines generation sample line includes:
Utilize the Hough transformation in image recognition that view data is carried out image recognition。
5. go card recognition methods as claimed in claim 1, it is characterised in that described described sample line is fitted processes the chessboard line determined in chessboard both direction includes:
Described sample line is sequentially carried out direct matching, weighted fitting and ransac matching and determines the chessboard line in a direction；
From described sample line, delete the chessboard line determined, remaining model line is carried out again according to the order of sequence direct matching, weighted fitting and ransac matching and determines the chessboard line of other direction。
6. a go card identification device, it is characterised in that described device includes:
Pretreatment module, generates pretreated view data for a two field picture of the go card obtained is carried out pretreatment；
Clustering processing module, for adopting the method for cluster that the clustering processing of described view data is generated cluster result, described cluster result, for class value to be divided three classes, represents the white point in chessboard, black color dots and blank spot respectively；
Sample line generation module, for carrying out image recognition to described pretreated view data, it is determined that in this two field picture, all discernible lines generate sample line；
Fitting module, processes the chessboard line determined in chessboard both direction for described sample line is fitted；
Module determined by chessboard, for being intersected by the chessboard line in the both direction determined, it is determined that the intersecting point coordinate of chessboard line on chessboard；
Card identification module, for determining current I-go board surface state according to the intersecting point coordinate of chessboard line on the cluster result determined and chessboard。
7. go card identification device as claimed in claim 6 a, it is characterised in that two field picture of the described pretreatment module go card to obtaining carries out pretreatment and includes:
One two field picture of the go card obtained is carried out color enhancement, ashing process, Laplace transform and standardization processing。
8. go card identification device as claimed in claim 7, it is characterised in that the pretreated view data of described generation includes: generate the picture element matrix of this two field picture。
9. go card identification device as claimed in claim 6, it is characterised in that described pretreated view data is carried out image recognition and includes by described sample line generation module:
Utilize the Hough transformation in image recognition that view data is carried out image recognition。
10. go card identification device as claimed in claim 6, it is characterised in that described sample line is fitted processing the chessboard line determined in chessboard both direction and includes by described fitting module:
Described sample line is sequentially carried out direct matching, weighted fitting and ransac matching and determines the chessboard line in a direction；
From described sample line, delete the chessboard line determined, remaining model line is carried out again according to the order of sequence direct matching, weighted fitting and ransac matching and determines the chessboard line of other direction。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610016730.XA CN105701496B (en) | 2016-01-12 | 2016-01-12 | A kind of go disk recognition methods based on artificial intelligence technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610016730.XA CN105701496B (en) | 2016-01-12 | 2016-01-12 | A kind of go disk recognition methods based on artificial intelligence technology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105701496A true CN105701496A (en) | 2016-06-22 |
| CN105701496B CN105701496B (en) | 2019-07-05 |
Family
ID=56226176
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610016730.XA Active CN105701496B (en) | 2016-01-12 | 2016-01-12 | A kind of go disk recognition methods based on artificial intelligence technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105701496B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107358221A (en) * | 2017-08-08 | 2017-11-17 | 大连万和海拓文化体育产业有限公司 | The chessboard localization method of spectrum is remembered in a kind of go based on video identification technology automatically |
| CN107480678A (en) * | 2017-09-29 | 2017-12-15 | 北京深度奇点科技有限公司 | A kind of chessboard recognition methods and identifying system |
| CN108491804A (en) * | 2018-03-27 | 2018-09-04 | 腾讯科技（深圳）有限公司 | A kind of method, relevant apparatus and the system of chess game displaying |
| CN109559325A (en) * | 2018-12-03 | 2019-04-02 | 中南大学 | Weiqi chess manual recognition methods based on chess manual RGB image |
| CN110555418A (en) * | 2019-09-08 | 2019-12-10 | 无锡高德环境科技有限公司 | AI target object identification method and system for water environment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20070008319A (en) * | 2005-07-13 | 2007-01-17 | (주)몬도미오 | Korean chess board |
| CN101419705A (en) * | 2007-10-24 | 2009-04-29 | 深圳华为通信技术有限公司 | Video camera demarcating method and device |
| CN102682292A (en) * | 2012-05-10 | 2012-09-19 | 清华大学 | Method based on monocular vision for detecting and roughly positioning edge of road |
| CN102750538A (en) * | 2012-05-16 | 2012-10-24 | 南京大学 | Go competition result analysis method based on image processing technique |
| CN105205447A (en) * | 2015-08-22 | 2015-12-30 | 周立人 | Go identification method based on go image and go board |
-
2016
- 2016-01-12 CN CN201610016730.XA patent/CN105701496B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20070008319A (en) * | 2005-07-13 | 2007-01-17 | (주)몬도미오 | Korean chess board |
| CN101419705A (en) * | 2007-10-24 | 2009-04-29 | 深圳华为通信技术有限公司 | Video camera demarcating method and device |
| CN102682292A (en) * | 2012-05-10 | 2012-09-19 | 清华大学 | Method based on monocular vision for detecting and roughly positioning edge of road |
| CN102750538A (en) * | 2012-05-16 | 2012-10-24 | 南京大学 | Go competition result analysis method based on image processing technique |
| CN105205447A (en) * | 2015-08-22 | 2015-12-30 | 周立人 | Go identification method based on go image and go board |
Non-Patent Citations (1)
| Title |
|---|
| 曾接贤: ""霍夫变换与最小二乘法相结合的直线拟合"", 《南昌航空工业学院学报》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107358221A (en) * | 2017-08-08 | 2017-11-17 | 大连万和海拓文化体育产业有限公司 | The chessboard localization method of spectrum is remembered in a kind of go based on video identification technology automatically |
| CN107480678A (en) * | 2017-09-29 | 2017-12-15 | 北京深度奇点科技有限公司 | A kind of chessboard recognition methods and identifying system |
| CN108491804A (en) * | 2018-03-27 | 2018-09-04 | 腾讯科技（深圳）有限公司 | A kind of method, relevant apparatus and the system of chess game displaying |
| CN108491804B (en) * | 2018-03-27 | 2019-12-27 | 腾讯科技（深圳）有限公司 | Chess game display method, related device and system |
| CN109559325A (en) * | 2018-12-03 | 2019-04-02 | 中南大学 | Weiqi chess manual recognition methods based on chess manual RGB image |
| CN110555418A (en) * | 2019-09-08 | 2019-12-10 | 无锡高德环境科技有限公司 | AI target object identification method and system for water environment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105701496B (en) | 2019-07-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105701496A (en) | Go board surface identification method based on artificial intelligence technology | |
| CN106056064B (en) | A kind of face identification method and face identification device | |
| CN111160269A (en) | Face key point detection method and device | |
| CN107392218A (en) | A kind of car damage identification method based on image, device and electronic equipment | |
| CN110517202A (en) | A kind of vehicle body camera calibration method and its caliberating device | |
| CN107403424A (en) | A kind of car damage identification method based on image, device and electronic equipment | |
| CN112308095A (en) | Picture preprocessing and model training method and device, server and storage medium | |
| WO2021139494A1 (en) | Animal body online claim settlement method and apparatus based on monocular camera, and storage medium | |
| JP2011103561A (en) | System, device and method for processing image, and program | |
| CN109409428A (en) | Training method, device and the electronic equipment of plank identification and plank identification model | |
| CN108093174A (en) | Patterning process, device and the photographing device of photographing device | |
| CN103402058B (en) | A kind of processing method and processing device shooting image | |
| JP2021531571A (en) | Certificate image extraction method and terminal equipment | |
| CN109784415A (en) | The method and device of image-recognizing method and device, training convolutional neural networks | |
| CN114782286A (en) | Defect repairing method, optical repairing device, electronic device and storage medium | |
| CN110321892A (en) | A kind of picture screening technique, device and electronic equipment | |
| CN111222433A (en) | Automatic face auditing method, system, equipment and readable storage medium | |
| CN109635639B (en) | Method, device, equipment and storage medium for detecting position of traffic sign | |
| CN110414522A (en) | A kind of character identifying method and device | |
| CN112101314B (en) | Human body posture recognition method and device based on mobile terminal | |
| CN106778777A (en) | A kind of vehicle match method and system | |
| CN105893578A (en) | Method and device for selecting photos | |
| CN116188720A (en) | Digital person generation method, device, electronic equipment and storage medium | |
| CN113326819A (en) | Robot sketch drawing method and device and robot workbench | |
| JPH03154979A (en) | Graphic shaping device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |