CN106846320A - A kind of picture segmentation method and system positioned based on three-point fix and straight line - Google Patents
A kind of picture segmentation method and system positioned based on three-point fix and straight line Download PDFInfo
- Publication number
- CN106846320A CN106846320A CN201510888787.4A CN201510888787A CN106846320A CN 106846320 A CN106846320 A CN 106846320A CN 201510888787 A CN201510888787 A CN 201510888787A CN 106846320 A CN106846320 A CN 106846320A
- Authority
- CN
- China
- Prior art keywords
- punctuate
- picture
- database
- answering card
- answer
- 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
- 238000000034 method Methods 0.000 title claims abstract description 32
- 230000011218 segmentation Effects 0.000 title claims abstract description 15
- 238000005520 cutting process Methods 0.000 claims abstract description 13
- 238000003860 storage Methods 0.000 claims abstract description 8
- 241000180579 Arca Species 0.000 claims description 6
- 241001442234 Cosa Species 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims 1
- 238000010408 sweeping Methods 0.000 claims 1
- 238000004519 manufacturing process Methods 0.000 description 2
- 210000001747 pupil Anatomy 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000011248 coating agent Substances 0.000 description 1
- 238000000576 coating method Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000009966 trimming Methods 0.000 description 1
Abstract
The invention discloses a kind of picture segmentation method and system positioned based on three-point fix and straight line, every answering card being obtained one by one by scan components, answering card being fabricated to corresponding picture, picture is preserved to the position of setting, data-base recording storage location；Each pictures are read by database, the beginning punctuate of whole answering card is recognized and is terminated punctuate, according to punctuate rotating image, it is ensured that level between picture punctuate；Each pictures are read by database, read the position of each punctuate in picture, according to the length and width in the answer region set in database, by calculating whether next punctuate is to terminate punctuate, find end node, using Graphics methods, a small picture is cut into, and corresponding informance be recorded into correspondence database.The method of the present invention can improve the accuracy rate of network scoring answer area's cutting, make teacher when network scoring is carried out, and will not cause the situation that cannot judge achievement because of answer region cutting mistake.
Description
Technical field
The present invention relates to image positioning analytic technique, Image is applied to, and in particular to one kind is fixed based on three-point fix and straight line
The picture segmentation method and system of position.
Background technology
Existing computer marking mode is：Answering card is scanned with high speed scanner first and is uploaded to by network to scanning end
Go over examination papers server, objective item part according to the model answer for setting, system can programming count result, it is main without manual intervention
The answering card subjective item part surface trimming that sight topic part will upload according to the cutting scheme for providing, net is carried out by the specified person of going over examination papers
On go over examination papers, after subjective item part has all been read and made comments, objective item part and subjective item part can be carried out statistical summaries by system, and
The statistics of correlation can be derived.As shown in Figure 6.
Computer network is goed over examination papers because it has significant advantage in quality of going over examination papers, efficiency of going over examination papers and to the aspect such as improve the quality of teaching,
The today for particularly being improved constantly in the requirement of quality of instruction monitoring and evaluation, it will be increasingly subject to pay attention to and all kinds of be examined as future
The main mode of going over examination papers of examination.
Current computer marking when answering card subjective item partial cut-away piece is realized, by the way of three-point fix, answer by foundation
Upper pre-printed three anchor points of topic card, define a square frame for standard, realize the cutting of picture.This mode can expire substantially
The demand of cutting answering card more than sufficient middle school, but when pupil has coating and drawing at random, there are three anchor points can not be accurately positioned
Situation, cutting answer region when, mistake just occurs.
The content of the invention
To solve the deficiency that prior art is present, the invention discloses a kind of picture recognition side positioned based on three-point fix and straight line
Method and system, the cutting picture that is positioned based on three-point fix and straight line realize Image, reach the mesh in cutting answer region
's.
To achieve the above object, concrete scheme of the invention is as follows：
A kind of picture segmentation method positioned based on three-point fix and straight line, is comprised the following steps：
The scanning of answering card：Obtain every answering card one by one by scan components, answering card is fabricated to corresponding picture, will scheme
Piece is preserved to the position of setting, data-base recording storage location；
Answer sheet picture rotates：Each pictures are read by database, the beginning punctuate of whole answering card are recognized and is terminated punctuate,
According to punctuate rotating image, it is ensured that level between picture punctuate；
Answering card is split：Postrotational each pictures are read by database, the position of each punctuate in picture is read, according to
According to the length and width in the answer region set in database, by calculating whether next punctuate is to terminate punctuate, judge to terminate punctuation bit
Put；Start punctuate in reading picture and terminate to whether there is straight line between punctuate, if it is present the answer region is tried for one
Topic, if there is no straight line, selects the end Mistaken punctuations for thinking to choose, and continually looks for next punctuate, judges whether it is correspondence
End node；
End node is found, the answer region is made into Graphics methods, be cut into a small picture, and corresponding informance is recorded
To in correspondence database.Corresponding informance refers to：Correspondence picture name and question number, examination question numbering in systems etc..Every paper
An ID is all corresponded in database, every pictures of cutting have a small ID, are mapped between ID.
Further, the making of answering card was also needed to before answering card is scanned：Subjective item and objective item are separately positioned on same
Corresponding answer frame is opened on answering card and be respectively provided with, punctuate is started in the corresponding answer inframe setting original position of objective item, if
Put initial punctuate corresponding end punctuate anyhow, initial punctuate and the corresponding straight line of end punctuate are set, answer subjective item is corresponding
Topic inframe is respectively provided with to each subjective item and starts over correspondence punctuate and set line between punctuate.
Further, when answering card is made, also needed to obtain answer quantity and each examination question from database before punctuate
Answer space size, calculate objective item answer region length and width, obtain examination question quantity, obtain objective item answer number and label,
Calculate objective item original position in answering card；Each subjective item individually calculates the position in answering card.
Further, when obtaining every answering card by scan components, the recordable picture information in database, including title and sweep
The time is retouched, is scanned and is staggered the time, position occurs in data-base recording.
Further, during rotating image, anglec of rotation calculation：The position for defining rotary shaft is A points, the position under the point of contact
Be B points, the position that contact is upspring is C points, and a triangle is constituted by tri- summits of ABC, the angle where each summit be angle A,
Angle B, angle C, each summit just to triangle side be a, b, c, calculate angle A the number of degrees.
Further, when calculating the number of degrees of angle A, the coordinate (x, y) of A points is got by the position of picture and the length and width of picture, B
The coordinate (x1, y1) of point and C points, (x2, y2) can be obtained from motionEvent, and each side length of side of triangle is:
Double a=Math.sqrt ((x1-x2) * (x1-x2)+(y1-y2) * (y1-y2))；
Double b=Math.sqrt ((x-x2) * (x-x2)+(y-y2) * (y-y2))；
Double c=Math.sqrt ((x1-x) * (x1-x)+(y1-y) * (y1-y))；
The cosine value for obtaining angle A is double cosA=(b*b+c*c-a*a)/(2*b*c)；
Then the radian of angle A is double arcA=Math.acos (cosA)；
This radian is between 0-PI, to change angle into and use double angleA=arcA*180/Math.PI.
Wherein, motionEvent rotational events Double data types, double precision
Math.sqrt extracts square root
Math.acos cosine
double angleA：Statement variable
Math.PI：Pi, π=3.14159265 ...
Further, also obtained including answering card bar code, inquire about the barcode position of database purchase, barcode position is cut
Figure, makes Graphics methods, and relative coordinate is obtained according to the coordinate points correspondence punctuate of database purchase, is stored in acquisition database
Length and width, new picture is made, in pictorial information storage to database.Every paper has unique mark, bar shaped in database
Code and other examination questions belong to a part for paper.
A kind of picture segmentation method, system positioned based on three-point fix and straight line, including：
The scan module of answering card：Obtain every answering card one by one by scan components, answering card be fabricated to corresponding picture,
Picture is preserved to the position of setting, data-base recording storage location；
Answer sheet picture rotary module：Each pictures are read by database, beginning punctuate and the end of whole answering card is recognized
Punctuate, according to punctuate rotating image, it is ensured that level between picture punctuate；
Answering card splits module：Postrotational each pictures are read by database, the position of each punctuate in picture is read,
According to the length and width in the answer region set in database, by calculating whether next punctuate is to terminate punctuate, judge to terminate punctuate
Position；Start punctuate in reading picture and terminate to whether there is straight line between punctuate, if it is present the answer region is one
Examination question, if there is no straight line, selects the end Mistaken punctuations for thinking to choose, and continually looks for next punctuate, judges whether it is right
The end node answered；End node is found, the answer region is made into Graphics methods, be cut into a small picture, and will be right
Information record is answered in correspondence database.
Beneficial effects of the present invention：
The scanning of answering card of the present invention：Paper is scanned into picture, facilitates subsequent processes to carry out, answer sheet picture rotation：Make inclination
Picture level, facilitate follow-up cutting picture, answering card segmentation：Paper is divided into the pictures of per pass exercise question one, is conveniently goed over examination papers,
The method of the present invention can improve the accuracy rate of network scoring answer area's cutting, evade pupil's paper for causing of scrawling and cut figure
It is the inaccurate risk of positioning, makes teacher when network scoring is carried out, will not causing to judge because of answer region cutting mistake
The situation of achievement.
Brief description of the drawings
Fig. 1 is answer card manufacturing method flow chart；
Fig. 2 is answering card scanning process figure；
Fig. 3 is that answering card bar code obtains flow chart；
Fig. 4 is answering card answer area flow chart；
Fig. 5 picture rotation schematic diagrames；
The existing online paper-marking system flow charts of Fig. 6.
Specific embodiment：
The present invention is described in detail below in conjunction with the accompanying drawings：
Image positions analytic technique, is applied to Image.If so-called image segmentation is exactly to be divided an image into by certain rule
Significant region is done, i.e., the union in each region is whole image, the common factor in each region is 0.
Image by the set that image representation is physically significant connected region, that is, according to the priori of target and background
Knowledge, to the target in image, background is marked, positions, and then target is separated from background or other pseudo- targets.
Specific method is：
As shown in figure 1, answer card manufacturing method：Answer quantity and the answer space size of each examination question, meter are obtained from database
Calculate objective item answer region length and width (objective item notifies to set answer frame), obtain examination question topic quantity, obtain objective item answer number and
Label, calculates objective item original position in answering card, and setting original position starts punctuate, sets original position anyhow right
The end punctuate answered.Original position and the corresponding straight line of end position are set.Subjective item is separately provided answer frame, each subjective item
The position in answering card is individually calculated, setting starts over correspondence punctuate, sets line between punctuate, terminates answering card and makes.
As shown in Fig. 2 answering card is scanned：By scanner, every paper is obtained one by one, paper is fabricated to corresponding picture,
Data-base recording pictorial information (title, sweep time) etc., scans and staggers the time, and position, picture occurs in data-base recording
It is saved in the position of setting, data-base recording storage location；
As shown in figure 5, picture rotation：Each pictures are read according to database, according to the beginning punctuate of whole answering card and right side
Terminate punctuate, according to punctuate rotating image, it is ensured that level between picture punctuate.Anglec of rotation calculation：Define rotary shaft
Position is A points, and the position under the point of contact is B points, and the position that contact is upspring is C points, and a triangle is constituted by tri- summits of ABC
Shape, angle where each summit is angle A, angle B, angle C, each summit just to the side of triangle be a, b, c, calculate the degree of angle A
Number.
The coordinate (x, y) of A points is got by the position of picture and the length and width of picture, the coordinate (x1, y1) of B points and C points, and (x2, y2) can be from
Obtained in motionEvent, each side length of side of triangle is:
Double a=Math.sqrt ((x1-x2) * (x1-x2)+(y1-y2) * (y1-y2))；
Double b=Math.sqrt ((x-x2) * (x-x2)+(y-y2) * (y-y2))；
Double c=Math.sqrt ((x1-x) * (x1-x)+(y1-y) * (y1-y))；
The cosine value for obtaining angle A is double cosA=(b*b+c*c-a*a)/(2*b*c)；
Then the radian of angle A is double arcA=Math.acos (cosA)；
This radian is between 0-PI, to change angle into and use double angleA=arcA*180/Math.PI
As shown in figure 3, answering card bar code acquisition methods：The barcode position of database purchase is inquired about, barcode position is cut into figure,
Make Graphics methods, relative coordinate is obtained according to the coordinate points correspondence punctuate of database purchase, obtain the length stored in database
Width, makes new picture, in pictorial information storage to database.
As shown in figure 4, answering card answer area：Each pictures are read according to database, each punctuate in picture is read, read
Fetch bit is put, and according to the length and width in the answer region set in database, by calculating whether next punctuate is to terminate punctuate, is judged
Terminate punctuate position.Start punctuate in reading picture and terminate to whether there is straight line between punctuate, if it does, select thinking that this is answered
Topic region is one examination question, if there is no straight line, selects the end Mistaken punctuations for thinking to choose, and continually looks for next punctuate,
Judge whether it is corresponding end node.After finding, the answer region is made into Graphics methods, be cut into a small figure
Piece, and corresponding informance be recorded into correspondence database.
Although above-mentioned be described with reference to accompanying drawing to specific embodiment of the invention, not to the limit of the scope of the present invention
System, one of ordinary skill in the art should be understood that on the basis of technical scheme those skilled in the art need not pay
The various modifications or deformation made by going out creative work are still within protection scope of the present invention.
Claims (8)
1. a kind of picture segmentation method positioned based on three-point fix and straight line, it is characterized in that, comprise the following steps：
The scanning of answering card：Obtain every answering card one by one by scan components, answering card is fabricated to corresponding picture, will scheme
Piece is preserved to the position of setting, data-base recording storage location；
Answer sheet picture rotates：Each pictures are read by database, the beginning punctuate of whole answering card are recognized and is terminated punctuate,
According to punctuate rotating image, it is ensured that level between picture punctuate；
Answering card is split：Postrotational each pictures are read by database, the position of each punctuate in picture is read, according to
According to the length and width in the answer region set in database, by calculating whether next punctuate is to terminate punctuate, judge to terminate punctuation bit
Put；Start punctuate in reading picture and terminate to whether there is straight line between punctuate, if it is present the answer region is tried for one
Topic, if there is no straight line, selects the end Mistaken punctuations for thinking to choose, and continually looks for next punctuate, judges whether it is correspondence
End node；
End node is found, the answer region is made into Graphics methods, be cut into a small picture, and corresponding informance is recorded
To in correspondence database.Corresponding informance refers to：Correspondence picture name and question number, examination question numbering in systems etc..Every paper
An ID is all corresponded in database, every pictures of cutting have a small ID, are mapped between ID.
2. a kind of picture segmentation method positioned based on three-point fix and straight line as claimed in claim 1, it is characterized in that, sweeping
The making of answering card is also needed to before retouching answering card：Subjective item and objective item are separately positioned on same answering card and are set respectively
Corresponding answer frame is put, punctuate is started in the corresponding answer inframe setting original position of objective item, initial punctuate is set and is corresponded to anyhow
End punctuate, initial punctuate and the corresponding straight line of end punctuate are set, in the corresponding answer inframe of subjective item to each subjective item
It is respectively provided with and starts over correspondence punctuate and line between punctuate is set.
3. a kind of picture segmentation method positioned based on three-point fix and straight line as claimed in claim 2, it is characterized in that, in system
When making answering card, also needed to obtain answer quantity and the answer space size of each examination question from database before punctuate, calculate
Objective item answer region length and width, obtain examination question quantity, obtain objective item answer number and label, calculate objective item in answer
Original position in card；Each subjective item individually calculates the position in answering card.
4. a kind of picture segmentation method positioned based on three-point fix and straight line as claimed in claim 1, it is characterized in that, pass through
When scan components obtain every answering card, the recordable picture information in database, including title and sweep time, scan and stagger the time,
There is position in data-base recording.
5. a kind of picture segmentation method positioned based on three-point fix and straight line as claimed in claim 1, it is characterized in that, rotation
During picture, anglec of rotation calculation：The position for defining rotary shaft is A points, and the position under the point of contact is B points, what contact was upspring
Position is C points, and a triangle is constituted by tri- summits of ABC, and the angle where each summit is angle A, angle B, angle C, each summit
Just to triangle side be a, b, c, calculate angle A the number of degrees.
6. a kind of picture segmentation method positioned based on three-point fix and straight line as claimed in claim 5, it is characterized in that, calculate
During the number of degrees of angle A, the coordinate (x, y) of A points is got by the position of picture and the length and width of picture, the coordinate of B points and C points
(x1, y1), (x2, y2) can be obtained from motionEvent, and each side length of side of triangle is:
Double a=Math.sqrt ((x1-x2) * (x1-x2)+(y1-y2) * (y1-y2))；
Double b=Math.sqrt ((x-x2) * (x-x2)+(y-y2) * (y-y2))；
Double c=Math.sqrt ((x1-x) * (x1-x)+(y1-y) * (y1-y))；
The cosine value for obtaining angle A is double cosA=(b*b+c*c-a*a)/(2*b*c)；
Then the radian of angle A is double arcA=Math.acos (cosA)；
This radian is between 0-PI, to change angle into and use double angleA=arcA*180/Math.PI；
Wherein, motionEvent rotational events, Double data types, double precision, Math.sqrt extraction of square roots, Math.acos
Cosine, double angleA：Statement variable, Math.PI：Pi, π=3.14159265 ....
7. a kind of picture segmentation method positioned based on three-point fix and straight line as claimed in claim 1, it is characterized in that, also wrap
The acquisition of answering card bar code is included, the barcode position of database purchase is inquired about, barcode position is cut into figure, make Graphics methods,
Relative coordinate is obtained according to the coordinate points correspondence punctuate of database purchase, the length and width stored in database are obtained, new picture is made,
Pictorial information is stored in database.Every paper has unique mark in database, and bar code and other examination questions belong to paper
A part.
8. a kind of picture segmentation method, system positioned based on three-point fix and straight line, it is characterized in that, including：
The scan module of answering card：Obtain every answering card one by one by scan components, answering card be fabricated to corresponding picture,
Picture is preserved to the position of setting, data-base recording storage location；
Answer sheet picture rotary module：Each pictures are read by database, beginning punctuate and the end of whole answering card is recognized
Punctuate, according to punctuate rotating image, it is ensured that level between picture punctuate；
Answering card splits module：Postrotational each pictures are read by database, the position of each punctuate in picture is read,
According to the length and width in the answer region set in database, by calculating whether next punctuate is to terminate punctuate, judge to terminate punctuate
Position；Start punctuate in reading picture and terminate to whether there is straight line between punctuate, if it is present the answer region is one
Examination question, if there is no straight line, selects the end Mistaken punctuations for thinking to choose, and continually looks for next punctuate, judges whether it is right
The end node answered；End node is found, the answer region is made into Graphics methods, be cut into a small picture, and will be right
Information record is answered in correspondence database.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510888787.4A CN106846320A (en) | 2015-12-03 | 2015-12-03 | A kind of picture segmentation method and system positioned based on three-point fix and straight line |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510888787.4A CN106846320A (en) | 2015-12-03 | 2015-12-03 | A kind of picture segmentation method and system positioned based on three-point fix and straight line |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106846320A true CN106846320A (en) | 2017-06-13 |
Family
ID=59151955
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510888787.4A Pending CN106846320A (en) | 2015-12-03 | 2015-12-03 | A kind of picture segmentation method and system positioned based on three-point fix and straight line |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106846320A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107451581A (en) * | 2017-07-11 | 2017-12-08 | 济宁九点连线信息技术有限公司 | A kind of method of batch Automatic Read Overmarginalia answering card |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101334814A (en) * | 2008-04-28 | 2008-12-31 | 华北电力大学 | Automatic scanning and reading system and reading method |
| CN103400124A (en) * | 2013-07-03 | 2013-11-20 | 北京鸿合盛视数字媒体技术有限公司 | Method and device for identifying test paper |
| CN103996022A (en) * | 2014-05-12 | 2014-08-20 | 山东山大鸥玛软件有限公司 | Boundary positioning method for scanned answer sheet image |
| JP2014219949A (en) * | 2013-04-30 | 2014-11-20 | 良信 小村 | Test scoring system |
| CN104820835A (en) * | 2015-04-29 | 2015-08-05 | 岭南师范学院 | Automatic examination paper marking method for examination papers |
-
2015
- 2015-12-03 CN CN201510888787.4A patent/CN106846320A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101334814A (en) * | 2008-04-28 | 2008-12-31 | 华北电力大学 | Automatic scanning and reading system and reading method |
| JP2014219949A (en) * | 2013-04-30 | 2014-11-20 | 良信 小村 | Test scoring system |
| CN103400124A (en) * | 2013-07-03 | 2013-11-20 | 北京鸿合盛视数字媒体技术有限公司 | Method and device for identifying test paper |
| CN103996022A (en) * | 2014-05-12 | 2014-08-20 | 山东山大鸥玛软件有限公司 | Boundary positioning method for scanned answer sheet image |
| CN104820835A (en) * | 2015-04-29 | 2015-08-05 | 岭南师范学院 | Automatic examination paper marking method for examination papers |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107451581A (en) * | 2017-07-11 | 2017-12-08 | 济宁九点连线信息技术有限公司 | A kind of method of batch Automatic Read Overmarginalia answering card |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN202939608U (en) | Test paper inspecting system | |
| US10964060B2 (en) | Neural network-based camera calibration | |
| CN101477696B (en) | Human character cartoon image generating method and apparatus | |
| CN104794948A (en) | Automatic scoring system and method for applying same | |
| US9665759B2 (en) | Decoding method for matrix two-dimensional code | |
| CN105426844B (en) | A kind of answer card identification method | |
| CN111507251A (en) | Method and device for positioning answer area in test question image and electronic equipment | |
| CN110287854B (en) | Table extraction method and device, computer equipment and storage medium | |
| CN102360419A (en) | Method and system for computer scanning reading management | |
| CN110689013A (en) | Automatic marking method and system based on feature recognition | |
| CN105868759A (en) | Method and apparatus for segmenting image characters | |
| CN104463066B (en) | Method for rapidly positioning and recognizing one-dimensional barcode of outer commodity package | |
| CN111860527A (en) | Image correction method, image correction device, computer device, and storage medium | |
| CN107679442A (en) | Method, apparatus, computer equipment and the storage medium of document Data Enter | |
| CN115457565A (en) | OCR character recognition method, electronic equipment and storage medium | |
| CN110909743A (en) | Book checking method and book checking system | |
| CN103488965A (en) | Freight note information inputting and colored color-lump encoding-decoding system | |
| CN106846320A (en) | A kind of picture segmentation method and system positioned based on three-point fix and straight line | |
| CN204537507U (en) | Automatic marking papers system | |
| CN108764248B (en) | Image feature point extraction method and device | |
| CN204641116U (en) | A kind of orientation film | |
| CN110334647A (en) | A kind of parameter format method based on image recognition | |
| CN116129126A (en) | Construction method of segmentation prediction model, picture semantic segmentation labeling method and device | |
| CN115625435A (en) | Intelligent recognition cutting method, device and system | |
| CN111563407B (en) | Model training method, picture direction correcting method and device |
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 |
Application publication date: 20170613 |