CN102750538B - A kind of weiqi game interpretation of result method based on image processing techniques - Google Patents
A kind of weiqi game interpretation of result method based on image processing techniques Download PDFInfo
- Publication number
- CN102750538B CN102750538B CN201210150815.9A CN201210150815A CN102750538B CN 102750538 B CN102750538 B CN 102750538B CN 201210150815 A CN201210150815 A CN 201210150815A CN 102750538 B CN102750538 B CN 102750538B
- Authority
- CN
- China
- Prior art keywords
- image
- black
- white
- chess piece
- value
- 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
Abstract
The invention discloses a kind of weiqi game interpretation of result method based on image processing techniques, comprise the following steps: step one, the pre-service of checkerboard image: according to the image gathered, by spatial alternation and grey scale interpolation, removes other object gathered in image except chessboard; Step 2, the segmentation of black-and-white piece in checkerboard image: by the image after correction, suitable threshold value is set, is partitioned into black-and-white piece image respectively; Step 3, the detection of black-and-white piece and result judge: filled up respectively through the attribute of the chessboard of opening operation, projection, setting-out, detection line and line point of intersection, blank space chess piece by the black-and-white piece image be partitioned into, result judgement etc. operates the judgement having carried out result of the match.The present invention, when not increasing extra means, carries out image procossing by software approach, judges result of the match quickly and accurately.
Description
Technical field
The present invention relates to image real time transfer and identification field, particularly a kind of weiqi game interpretation of result method based on image processing techniques.
Background technology
Enclosing in board game, the statistics for the chess piece result of the match adopting common chessboard and entity still adopts the method for traditional people's number chess piece substantially, and not only expend time in longer, accuracy many times can not ensure.In today that computer vision technique is popularized day by day, machine auxiliary interpretation is adopted to be a kind of more excellent method.First the art of this patent uses digital camera (or camera) to carry out image acquisition to the chessboard after weiqi game, then carries out machine interpretable by digital image processing techniques.Do like this and not only can greatly reduce the interpretation time, the accuracy of result can also be ensured.Utilize computer technology and digital image processing techniques auto Segmentation black piece and white side in the present invention, the position of location black-and-white piece, obtains the result of playing chess fast.
Summary of the invention
Goal of the invention: technical matters to be solved by this invention is for the deficiencies in the prior art, a kind of weiqi game interpretation of result method based on image processing techniques is provided, thus makes state and the result of the match that can count the chess piece of chessboard when people play Weiqi fast and accurately.
In order to solve the problems of the technologies described above, the invention discloses a kind of weiqi game interpretation of result method based on image processing techniques, comprising the following steps:
Step one, the pre-service of checkerboard image: according to the image gathered, by spatial alternation and grey scale interpolation, removes other object gathered in image except chessboard;
Step 2, the segmentation of black-and-white piece in checkerboard image: by the image after correction, suitable threshold value is set, is partitioned into black-and-white piece image respectively;
Step 3, the detection of black-and-white piece and result judge: filled up respectively through the attribute of the chessboard of opening operation, projection, setting-out, detection line and line point of intersection, blank space chess piece by the black-and-white piece image be partitioned into, result judgement etc. operates the judgement having carried out result of the match.The present invention, when not increasing extra means, carries out image procossing by software approach, judges result of the match quickly and accurately.
In the present invention, preferably, described step one comprises the following steps:
Step (11), in the image gathered, marks the coordinate points at four angles of chessboard, because need in geometric transformation below to use these four coordinate points to carry out affined transformation;
Step (12), image carries out geometric transformation, because checkerboard image is when gathering, have distortion unavoidably, so use geometric transformation that whole checkerboard image is carried out spatial alternation and grey scale interpolation, remove the remainder gathered except chessboard part in checkerboard image, obtain the rectangular image of rule.
In the present invention, preferably, described step 2 comprises the following steps:
Step (21), in the chessboard of go image gathered, after geometric transformation, in image, only there will be three kinds of objects, black chess piece, white chess piece and background color are yellow chessboard, can be partitioned into black chess piece by the direct threshold segmentation method of histogram.Grey level histogram is calculated to the image after correcting, then filtering is carried out by one dimension neighbor smoothing wave filter (hereafter will introduce), filtered grey level histogram there will be peak-to-valley value, the probability that black chess piece occurs before first peak and second peak-to-peak valley point, after first valley point is chessboard and white chess piece, so only need to find the valley in gray histogram curve between first peak value and second peak value to be the segmentation threshold of black chess piece, using the threshold value of this value as segmentation black chess piece, and binaryzation, be 255 by the black chess piece assignment in gray level image, remaining image assignment is 0.
Step (22), because in the image of actual acquisition, after gray processing, chessboard gray-scale value has with the gray-scale value of white chess piece and intersects, and just white chess piece can not be separated with chessboard like this by gray level threshold segmentation.Through finding RGB three-component value of research black, white and yellow chess piece, the three-component value of RGB of black is all smaller, the three-component value of RGB of white is all larger, and in yellow RGB three-component, the value of red green component is larger, and the value of blue component is smaller.In black, white, yellow chess piece, their blue component difference is larger, the blue component of adularescent chess piece is larger, the blue component of black and yellow chess piece is all smaller, so only need the blue component extracted in coloured image, then suitable threshold value is set, namely white chess piece can be separated with chessboard.When selecting threshold value, the same automatic threshold that adopts extracts.Draw the grey level histogram of image blue component after correcting, then filtering is carried out by one dimension neighbor smoothing wave filter, filtered grey level histogram there will be peak-to-valley value, first peak is the probability that in blue component, black chess piece occurs, second peak is the probability that chessboard occurs, the 3rd peak is the probability occurred with white chess piece.So namely the valley asked between second peak value and the 3rd peak value is the threshold value splitting white chess piece.Using the threshold value of this value as the white chess piece of segmentation, and binaryzation, be 255 by the white chess piece assignment in gray level image, remaining image assignment is 0.
In the present invention, preferably, described step 3 comprises the following steps:
Step (31), makes opening operation respectively to the black-and-white piece image be partitioned into, the tiny noise that can occur because Threshold selection is inaccurate when splitting in removal of images like this, can also the border of smooth object;
Step (32), is added the image after opening operation, then does an etching operation, and object boundary point is eliminated in the effect of etching operation, is convenient to like this find out extreme point easily when projecting;
Step (33), uses the method for image projection, projects in the horizontal and vertical directions respectively, and record its value by the image after corrosion;
Step (34), according to the value of the horizontal and vertical direction projection of record, draw drop shadow curve, extreme value is obtained after one dimension neighbor smoothing wave filter, these extreme values just correspond to the position of the chessboard line in chessboard, because the size of chessboard is 19 × 19, so the extreme value obtained also is 19 respectively;
Step (35), according to the chessboard line detected, detect black-and-white piece position, use the window of M × M to go to detect chess piece when detecting chess piece position, M represents window size, M gets 3, after obtaining black-and-white piece position, go chess piece is used to fill up rule: empty to black or white, sky is filled up as black or white, black or white to empty, sky is filled up to be arrived in vain for black or white, white to empty, empty being filled up is arrived black for white, black to sky, and sky is filled up as black, uses above rule to fill up corresponding blank space into black or white side.
Step (36), statistics fills up the number of rear black-and-white piece, then judges result of the match according to laws of the game.
Principle of the present invention is the image according to gathering, first black chess piece and white pawn image is partitioned into respectively, secondly segmentation image is carried out opening operation operation respectively, try again the image addition after opening operation corrosion, then the image after corrosion is projected, detect the chessboard line of 19 × 19 of chessboard, online point of intersection detects black-and-white piece, fill up rule according to the black-and-white piece detected according to go to carry out filling up chess piece, fill up rear statistics black-and-white piece number, use go decision rule to judge result of the match.
Beneficial effect: the present invention carries out image procossing by software approach, after gathering chessboard, can detect every trellis state in chessboard fast and accurately.The present invention has wide practical use in the analysis of statistics weiqi game result and go strategy.
Accompanying drawing explanation
Below in conjunction with the drawings and specific embodiments the present invention done and further illustrates, of the present invention above-mentioned and or otherwise advantage will become apparent.
Fig. 1 is geometry correction schematic diagram.
Fig. 2 is the procedure chart that black chess piece is split in the present invention.
Fig. 3 is the procedure chart that white chess piece is split in the present invention.
Fig. 4 is that the present invention is black in the process flow diagram that chess piece detects and result judges
Fig. 5 is the inventive method simplified flow chart.
Embodiment:
The present invention, core thinking utilizes the image gathered to carry out black-and-white piece segmentation, be partitioned into black-and-white piece image respectively, secondly segmentation image is carried out opening operation operation respectively, try again the image addition after opening operation corrosion, then the projection in horizontal and vertical direction is carried out according to the image be partitioned into, draw the line in chessboard, then online point of intersection finds chess piece, then records the state of chess piece, then according to algorithm, statistics weiqi game result.
As shown in Figure 5, the invention discloses a kind of weiqi game interpretation of result method based on image processing techniques, comprise the following steps:
Step one, the pre-service of checkerboard image: by spatial alternation and grey scale interpolation, removes other object gathered in image except chessboard;
Described step one comprises the following steps: step 11, in the image gathered, marks the coordinate points at four angles of chessboard; Step 12, according to the coordinate points of mark, is removed other object in the image of collection except chessboard, and is transformed to a rectangular image by geometric transformation.
Step 12, to checkerboard image pre-service;
Carry out pre-service to the image gathered, carry out pre-service to image and make it requirement of adaptive algorithm can seem particularly necessary, the Image semantic classification related in the present invention comprises spatial alternation and grey scale interpolation.
Geometric transformation is the spatial mappings relation of image mid point and point, can a bit transform to optional position in image by image, and the continuity and the interlinking that keep converting local feature on the similar i.e. source images of local feature between the two width images of front and back are constant.
Geometric transformation based on bilinear transformation:
Order
G(x，y)＝F(x′，y′)＝F(ax′+by′+cx′y′+d，ex′+fy′+gx′y′+h)(1)
In formula (1), G (x, y) represents the image after converting, and F (x ', y ') represent original image
Namely
x＝ax′+by′+cx′y′+d(2)
y＝ex′+fy′+gx′y′+h(3)
In formula (2), (3), x, y represent the coordinate of the rear image of conversion, and the coordinate of x ', y ' expression original image, a, b, c, d, e, f, g, h represent conversion coefficient
4 points of original image (x '
1, y '
1), (x '
2, y '
2), (x '
3, y '
3), (x '
4, y '
4)
4 point (x of image after correcting
1, y
1), (x
2, y
2), (x
3, y
3), (x
4, y
4)
According to this 4 to point, 8 system of equations can be listed, in equation, have 8 coefficients, only need know 8 equations, just can solve 8 unknown numbers, be by matrix representation:
Order
AX=B can be obtained
According to AX=B, X=A can be obtained
-1b, so just can obtain transformation matrix B.
The image of original image after just can obtaining with up conversion correcting.Schematic diagram before conversion and after conversion as shown in Figure 1.
Step 2, the segmentation of black-and-white piece in checkerboard image: the image after correcting is carried out gray processing process, uses histogram thresholding dividing method, be partitioned into black pawn image; Take out the blue component in the RGB three-component of coloured image after correcting, according to blue component, the same histogram thresholding automatic Segmentation that uses goes out white pawn image;
Described step 2 comprises the following steps: step 21, in the chessboard of go image gathered, after geometric transformation, in image, only there will be three kinds of objects, black chess piece, white chess piece and background color are yellow chessboard, can be partitioned into black chess piece by the direct threshold segmentation method of histogram.When selecting threshold value, adopt automatic threshold extracting method, grey level histogram is calculated to the image after correcting, then filtering is carried out by one dimension neighbor smoothing wave filter, filtered grey level histogram there will be peak-to-valley value, the probability that black chess piece occurs before first peak and second peak-to-peak valley point, after first valley point is chessboard and white chess piece, so only need to find the valley in gray histogram curve between first peak value and second peak value to be the segmentation threshold of black chess piece, using the threshold value of this value as segmentation black chess piece, and binaryzation, be 255 by the black chess piece assignment in gray level image, remaining image assignment is 0.The cutting procedure figure of black chess piece is as Fig. 2;
Step 22, because in the image of actual acquisition, after gray processing, chessboard gray-scale value has with the gray-scale value of white chess piece and intersects, and just white chess piece can not be separated with chessboard like this by gray level threshold segmentation.Through finding RGB three-component value of research black, white and yellow chess piece, the three-component value of RGB of black is all smaller, the three-component value of RGB of white is all larger, and in yellow RGB three-component, the value of red green component is larger, and the value of blue component is smaller.In black, white, yellow chess piece, their blue component difference is larger, the blue component of adularescent chess piece is larger, the blue component of black and yellow chess piece is all smaller, so only need the blue component extracted in coloured image, then suitable threshold value is set, both white chess piece can be separated with chessboard.When selecting threshold value, the same automatic threshold that adopts extracts.Draw the grey level histogram of the blue component of image after correcting, then filtering is carried out by one dimension neighbor smoothing wave filter, filtered grey level histogram there will be peak-to-valley value, first peak is the probability that in blue component, black chess piece occurs, second peak is the probability that chessboard occurs, the 3rd peak is the probability occurred with white chess piece.So namely the valley asked between second peak value and the 3rd peak value is the threshold value splitting white chess piece.Using the threshold value of this value as the white chess piece of segmentation, and binaryzation, be 255 by the white chess piece assignment in gray level image, remaining image assignment is 0.White pawn image cutting procedure figure as shown in Figure 3.
Gray level [0,1 ..., L-1] and the histogram of digital picture of scope is discrete function h (r
k)=n
k, wherein r
kkth level gray scale, n
kthat in image, gray level is r
kpixel occur number.Often remove its each value with the sum of pixel in image (representing with n), to obtain normalized histogram.Therefore, a normalized histogram is by P (r
k)=n
k/ n provides, wherein k=0, and 1 ..., L-1.Briefly, P (r
k) to give gray level be r
kthe probabilistic estimated value occurred.Note, normalized histogrammic all part sum should equal 1.
In formula, n is the summation of pixel in image, n
kbe gray level be r
knumber of pixels, L is possible in image gray level sum.
Histogram thresholding segmentation determines that segmentation threshold carries out Iamge Segmentation according to feature histogrammic in image, the method is fairly simple, first the histogram of computed image, according to the histogram calculated, selection can separate the threshold value of object and background, namely can be partitioned into target or background exactly.
Step 21 and step 22 are all the histogram thresholding dividing methods used, and the segmentation image just used when splitting is different, when splitting black chess piece, use the gray level image of image after correcting to split; When splitting white chess piece, the blue component of image after correcting is used to split.
One dimension neighbor smoothing wave filter:
The simple average value of signal in filtering mask neighborhood is included in during the output of one dimension neighbor smoothing linear filter, in the neighborhood that this smoothing filter filtering mask is determined, signal averaging replaces the value of each pixel of original signal, and this process reduces the sharp-pointed change of one-dimensional signal.If one-dimensional signal h (n), the size of filtering mask is that M, M get odd number, and output h ' (n) of smooth linear wave filter can use following formula to represent:
In formula (6), T is the threshold value of one dimension smothing filtering.This threshold value is set and can removes blurring effect because neighbor smoothing produces: when the difference of the mean value of gray scale put in some points and its neighborhood is no more than the threshold value T of regulation, just retain its former gray-scale value constant; If when being greater than threshold value T, just replace the gray-scale value of this point with their mean value.Make to carry out level and smooth one-dimensional signal in this way, output signal both can have been made relatively more level and smooth, do not changed again the original feature of signal.
Step 3, the detection of black-and-white piece and result judge: filled up respectively through the attribute of the chessboard of opening operation, projection, setting-out, detection line and line point of intersection, blank space chess piece by the black-and-white piece image be partitioned into, result judgement etc. operates the judgement having carried out result of the match.
As shown in Figure 4, described step 3 comprises the following steps: step 31, and the black-and-white piece image be partitioned into is done opening operation respectively;
The process of rear expansion is first corroded, noise that can be tiny on removal of images, and smooth object border during opening operation.Suppose A, B is Z
2in set, A, by B opening operation, represents can be expressed as with mathematic(al) representation:
Opening operation available expression is expressed as:
The expression formula of set is used to be expressed as:
Step 32, is added the image after opening operation, then does an etching operation; Step 33, projects the image after corrosion in the horizontal and vertical directions respectively, and records its value;
The projection of image:
Image projection in one direction can be defined as this image pixel gray level value cumulative rear sum in the direction in which.Image projection and, its feature space, by two boil down to one dimensions, thus by image feature information a small amount of for a large amount of image information boil down tos, not only calculates simple, and has the feature of uniqueness and distinguishability, thus effectively can carry out image recognition.The projection properties of image can mathematically be expressed as follows:
Be provided with a digital picture F (x, y) (wherein x=1,2,3 ..., w-1, w; Y=1,2,3 ..., h-1, h), so this digital picture can represent at the projection properties of transverse axis x-axis and Z-axis y-axis and is called following two formulas:
Horizontal direction projects:
Vertical direction projects:
Wherein: w and h is respectively width and the height of image, and F (x, y) is the gray-scale value of image.
Step 34, according to the value of the horizontal and vertical direction projection of record, draws drop shadow curve, and through one dimension neighbor smoothing wave filter, and obtain extreme value;
Step 35, according to chessboard line, detect black-and-white piece position, size due to chessboard of go is 19 × 19, so the maximum value number in horizontal and vertical direction retains 19 respectively, then draws the chessboard of 19 × 19, in the point of intersection of chessboard, detect the state of chess piece, then record position corresponding to chess piece and then fill up rule according to chess piece, blank space is filled up chess piece;
Step 36, statistics fills up the number of rear black-and-white piece, then judges result of the match according to laws of the game.
The invention provides a kind of weiqi game interpretation of result method based on image processing techniques; the method and access of this technical scheme of specific implementation is a lot; the above is only the preferred embodiment of the present invention; should be understood that; for those skilled in the art; under the premise without departing from the principles of the invention, can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.The all available prior art of each ingredient not clear and definite in the present embodiment is realized.
Claims (1)
1., based on a weiqi game interpretation of result method for image processing techniques, it is characterized in that, comprise the following steps:
Step one, the pre-service of checkerboard image: according to the image gathered, by spatial alternation and grey scale interpolation, removes other object gathered in image except chessboard;
Step 2, the segmentation of black-and-white piece in checkerboard image: by the image after correction, suitable threshold value is set, is partitioned into black-and-white piece image respectively;
Step 3, the detection of black-and-white piece and result judge: filled up respectively through the attribute of the chessboard of opening operation, projection, setting-out, detection line and line point of intersection, blank space chess piece by the black-and-white piece image be partitioned into, result decision to be to complete the judgement of result of the match;
Described step one comprises the following steps:
Step (11), in the image gathered, marks the coordinate points at four angles of chessboard;
Step (12), obtains the image of chessboard by geometric transformation according to the coordinate points marked;
Described step 2 comprises the following steps:
Step (21), gray level image is used in image after calibration, suitable Threshold segmentation is set and black pawn image, specifically comprise: in the chessboard of go image gathered, after geometric transformation, only there will be three kinds of objects in image, black chess piece, white chess piece and background color are yellow chessboard, black chess piece is partitioned into the direct threshold segmentation method of histogram, grey level histogram is calculated to the image after correcting, then filtering is carried out by one dimension neighbor smoothing wave filter, filtered grey level histogram there will be peak-to-valley value, the probability that black chess piece occurs before first peak and second peak-to-peak valley point, after first valley point is chessboard and white chess piece, so only need to find the valley in gray histogram curve between first peak value and second peak value to be the segmentation threshold of black chess piece, using the threshold value of this value as segmentation black chess piece, and binaryzation, be 255 by the black chess piece assignment in gray level image, remaining image assignment is 0,
Step (22), the blue component of coloured image is used in image after calibration, suitable threshold value is set to split white pawn image, specifically comprise: in the image of actual acquisition, after gray processing, chessboard gray-scale value has with the gray-scale value of white chess piece and intersects, just white chess piece can not be separated with chessboard by gray level threshold segmentation like this, through research black, the RGB of white and yellow chess piece three-component value finds, the three-component value of RGB of black is all smaller, the three-component value of RGB of white is all larger, in yellow RGB three-component, the value of red green component is larger, the value of blue component is smaller, black, white, in yellow chess piece, their blue component difference is larger, the blue component of adularescent chess piece is larger, the blue component of black and yellow chess piece is all smaller, so only need the blue component extracted in coloured image, then suitable threshold value is set, namely white chess piece can be separated with chessboard, when selecting threshold value, same employing automatic threshold extracts, draw the grey level histogram of image blue component after correcting, then filtering is carried out by one dimension neighbor smoothing wave filter, filtered grey level histogram there will be peak-to-valley value, first peak is the probability that in blue component, black chess piece occurs, second peak is the probability that chessboard occurs, 3rd peak is the probability occurred with white chess piece, so namely the valley asked between second peak value and the 3rd peak value is the threshold value splitting white chess piece, using the threshold value of this value as the white chess piece of segmentation, and binaryzation, be 255 by the white chess piece assignment in gray level image, remaining image assignment is 0,
Described step 3 comprises the following steps:
Step (31), makes opening operation respectively to the black-and-white piece image be partitioned into, and removes because segmentation threshold selects the incorrect image be partitioned into contain isolated noise;
Step (32), is added the image after opening operation, then does an etching operation; Eliminating object boundary point, being convenient to like this find out extreme point easily when projecting;
Step (33), projects the image after corrosion in the horizontal and vertical directions respectively, and records its value;
Step (34), according to the value of the horizontal and vertical direction projection of record, draws drop shadow curve, and obtains extreme value;
Step (35), according to chessboard line, detects black-and-white piece position, then fills up rule according to chess piece, blank space is filled up chess piece;
Step (36), statistics fills up the number of rear black-and-white piece, then judges result of the match according to laws of the game.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210150815.9A CN102750538B (en) | 2012-05-16 | 2012-05-16 | A kind of weiqi game interpretation of result method based on image processing techniques |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210150815.9A CN102750538B (en) | 2012-05-16 | 2012-05-16 | A kind of weiqi game interpretation of result method based on image processing techniques |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102750538A CN102750538A (en) | 2012-10-24 |
| CN102750538B true CN102750538B (en) | 2016-04-27 |
Family
ID=47030703
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210150815.9A Expired - Fee Related CN102750538B (en) | 2012-05-16 | 2012-05-16 | A kind of weiqi game interpretation of result method based on image processing techniques |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102750538B (en) |
Families Citing this family (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105205447B (en) * | 2015-08-22 | 2018-08-17 | 周立人 | Go recognition methods based on go image and chessboard |
| CN105701496B (en) * | 2016-01-12 | 2019-07-05 | 北京万同科技有限公司 | A kind of go disk recognition methods based on artificial intelligence technology |
| CN105664478A (en) * | 2016-01-15 | 2016-06-15 | 上海斐讯数据通信技术有限公司 | Go composition outcome judgment method and mobile terminal |
| CN105956594A (en) * | 2016-05-10 | 2016-09-21 | 浙江理工大学 | Method of identifying chest piece movement of real chess |
| CN107194944B (en) * | 2017-06-21 | 2019-08-09 | 南京林业大学 | Forest fire image dividing method and device |
| CN107730522A (en) * | 2017-10-12 | 2018-02-23 | 中科视拓（北京）科技有限公司 | A kind of weiqi chess manual recognition methods based on image |
| CN107875625A (en) * | 2017-11-23 | 2018-04-06 | 东华大学 | A kind of voice-based Chinese chess is played chess device |
| CN108509956A (en) * | 2018-03-27 | 2018-09-07 | 深圳大学 | Method, system and the electronic equipment of go victory or defeat judgement based on image procossing |
| CN109559325B (en) * | 2018-12-03 | 2021-02-02 | 中南大学 | Go chess manual identification method based on chess manual RGB image |
| CN110399888B (en) * | 2019-07-25 | 2021-08-27 | 西南民族大学 | Weiqi judging system based on MLP neural network and computer vision |
| CN110766713A (en) * | 2019-10-30 | 2020-02-07 | 上海微创医疗器械（集团）有限公司 | Lung image segmentation method and device and lung lesion region identification equipment |
| CN110909727A (en) * | 2019-11-20 | 2020-03-24 | 北京工业大学 | Image recognition-based go recognition method and program interface |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101000654A (en) * | 2006-12-31 | 2007-07-18 | 常熟理工学院 | Automatic recording method for recognising chess manual by image |
| CN101477687A (en) * | 2009-01-22 | 2009-07-08 | 上海交通大学 | Checkerboard angle point detection process under complex background |
| CN102184544A (en) * | 2011-05-18 | 2011-09-14 | 北京联合大学生物化学工程学院 | Method for correcting deformity and identifying image of go notation |
| CN102412627A (en) * | 2011-11-29 | 2012-04-11 | 安徽继远电网技术有限责任公司 | Image identification-based intelligent transformer substation state monitoring system |
-
2012
- 2012-05-16 CN CN201210150815.9A patent/CN102750538B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101000654A (en) * | 2006-12-31 | 2007-07-18 | 常熟理工学院 | Automatic recording method for recognising chess manual by image |
| CN101477687A (en) * | 2009-01-22 | 2009-07-08 | 上海交通大学 | Checkerboard angle point detection process under complex background |
| CN102184544A (en) * | 2011-05-18 | 2011-09-14 | 北京联合大学生物化学工程学院 | Method for correcting deformity and identifying image of go notation |
| CN102412627A (en) * | 2011-11-29 | 2012-04-11 | 安徽继远电网技术有限责任公司 | Image identification-based intelligent transformer substation state monitoring system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102750538A (en) | 2012-10-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102750538B (en) | A kind of weiqi game interpretation of result method based on image processing techniques | |
| CN107463918B (en) | Lane line extraction method based on fusion of laser point cloud and image data | |
| CN109035276B (en) | Image edge extraction method and device and automatic driving system | |
| CN106960449B (en) | Heterogeneous registration method based on multi-feature constraint | |
| CN108510491B (en) | Method for filtering human skeleton key point detection result under virtual background | |
| CN107316326B (en) | Edge-based disparity map calculation method and device applied to binocular stereo vision | |
| CN109859226B (en) | Detection method of checkerboard corner sub-pixels for graph segmentation | |
| CN109034017A (en) | Head pose estimation method and machine readable storage medium | |
| CN103544491A (en) | Optical character recognition method and device facing complex background | |
| CN106815583A (en) | A kind of vehicle at night license plate locating method being combined based on MSER and SWT | |
| CN102982545A (en) | Image depth estimation method | |
| CN110245600B (en) | Unmanned aerial vehicle road detection method for self-adaptive initial quick stroke width | |
| CN111354047B (en) | Computer vision-based camera module positioning method and system | |
| CN110852992A (en) | Ship trail detection method based on top hat transformation and Radon transformation | |
| CN114693760A (en) | Image correction method, device and system and electronic equipment | |
| CN109389165A (en) | Oil level gauge for transformer recognition methods based on crusing robot | |
| CN104200434B (en) | Non-local mean image denoising method based on noise variance estimation | |
| CN106815851B (en) | A kind of grid circle oil level indicator automatic reading method of view-based access control model measurement | |
| CN111089586B (en) | All-day star sensor star point extraction method based on multi-frame accumulation algorithm | |
| CN111062384B (en) | Vehicle window accurate positioning method based on deep learning | |
| CN110135382B (en) | Human body detection method and device | |
| CN113723432B (en) | Intelligent identification and positioning tracking method and system based on deep learning | |
| CN1331097C (en) | Spot eliminating method for digital image | |
| JP3638845B2 (en) | Image processing apparatus and method | |
| CN109448012A (en) | A kind of method for detecting image edge and 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 | ||
| DD01 | Delivery of document by public notice |
Addressee: Nanjing University Document name: Notification to Go Through Formalities of Registration |
|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160427 Termination date: 20180516 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |