CN110232377A - A kind of artificial intelligence points-scoring system that copybook practices calligraphy and method - Google Patents
A kind of artificial intelligence points-scoring system that copybook practices calligraphy and method Download PDFInfo
- Publication number
- CN110232377A CN110232377A CN201910427401.8A CN201910427401A CN110232377A CN 110232377 A CN110232377 A CN 110232377A CN 201910427401 A CN201910427401 A CN 201910427401A CN 110232377 A CN110232377 A CN 110232377A
- Authority
- CN
- China
- Prior art keywords
- image
- scoring
- text
- copybook
- row
- 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.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
- G06V20/63—Scene text, e.g. street names
-
- G—PHYSICS
- G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
- G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
- G09B7/00—Electrically-operated teaching apparatus or devices working with questions and answers
- G09B7/02—Electrically-operated teaching apparatus or devices working with questions and answers of the type wherein the student is expected to construct an answer to the question which is presented or wherein the machine gives an answer to the question presented by a student
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
Abstract
The present invention discloses a kind of intelligent scoring system and method that copybook practices calligraphy, including copybook image capture module, Text segmentation module, text grading module, scoring display module;System passes through acquisition copybook image, and image is handled and divides to obtain the subgraph comprising single text, the feature of subgraph is extracted again and is matched with the characteristics of image of grapholect, then in mobile terminal, screen shows final appraisal result, provides the prompt whether grasped or whether need to practice again.Methods of marking includes: copybook Image Acquisition (mobile phone or plate take pictures), Text segmentation (detecting the imaging region of copybook in the picture using Hough transformation), text scoring (extract subgraph character features with CNN and match with grapholect), scoring display.Rationality during writing practising can be judged using this system and method, accurately be scored in the form of numerical value writing words.Accuracy is high, in terms of all have advantage.
Description
Technical field
The present invention relates to a kind of artificial intelligence points-scoring system that copybook practices calligraphy and methods, and in particular to a kind of copybook white silk
The artificial intelligence points-scoring system and method for calligraphy are practised, artificial intelligence technical field of image processing is belonged to.
Background technique
Chinese-character writing is Sinitic tradition, it has a long history, and is both the load of Chinese thousands of years civilization transfers
Body, and be the important tool of people's exchange, the writing font formed in childhood often influences all one's life of people.But
In today of electronic product high development, handwritten Chinese character also receives huge impact, due to logical in daily life and work
Electronic equipment is crossed to carry out the input of text, needs to use the chance of handwritten Chinese character fewer and fewer, so, more and more Chinese
Chinese character beautiful on the other hand can not be write out.
Currently, in drills to improve one's handwriting, it is general by the way of copybook practice, by imitating copybook, step up
Writing level.But this practice mode has following disadvantage: firstly, lacking necessary points-scoring system, writer is to oneself book
The content write only has the understanding on sense organ, without the judge of system, is difficult to form the standard of judge to the writing level of oneself;
Secondly, writing process is dry as dust, lack interactive, driving force, writer is easy to lose interest.
The existing published patent being related in terms of Chinese-character writing exercise system, as application No. is 201710762777.5
" digitlization writing practising method and system ", application No. is 201710032399.5 " a kind of intelligence is write points-scoring system and side
Method ", application No. is 201710404401.7 " one kind write exercise system and method ", application No. is 2,016,109,573 17.3
" a kind of writing practising methods of marking and writing practising device " etc., generally existing following common disadvantage: be required to by writing
The handwriting input device such as letter stencil, Digitizing plate, on devices writing be different from paper on writing, most of writers without
Method adapts to this change, therefore cannot veritably improve the font and writing style of writer.In view of this, it is necessary to carry out into
One step studies a kind of intelligent scoring system and method that more advanced and practicable copybook practices calligraphy.
Summary of the invention
The purpose of the present invention is to overcome the shortcomings of the existing technology and deficiency, provides a kind of artificial intelligence that copybook practices calligraphy
Energy points-scoring system and method can be accurately with numbers which solve rationality judgment criteria is lacked during writing practising
The form of value scores to writing words.The artificial intelligence points-scoring system and method to be practiced calligraphy based on this copybook, can be concise
Artificial intelligence scoring effectively is carried out to writing words, has the characteristics that code of points is clear, accuracy is high, versatile.
In order to achieve the above object, the technical scheme adopted by the invention is that:
A kind of artificial intelligence points-scoring system that copybook practices calligraphy, comprising: at mobile phone or flat-type mobile terminal, digitlization
Manage device;Copybook image capture module and scoring display module, the number are provided on the mobile phone or flat-type mobile terminal
Text segmentation module, text grading module are provided in word processing unit;At mobile phone or flat-type mobile terminal and digitlization
It manages and passes through wireless blue tooth or wire communication between device;The artificial intelligence points-scoring system passes through acquisition copybook image, and to figure
Picture is handled and divides to obtain the subgraph comprising single text, then extracts the feature of subgraph and the image with grapholect
Whether whether feature is matched, and then shows final appraisal result, and provided according to scoring and to grasp or need to practice again
Prompt;It is characterized by:
The copybook image capture module includes: the camera on mobile phone or flat-type mobile terminal, setting camera point
Resolution is b1×b2;Copybook image capture module is for acquiring handwriting image to obtain original copybook image, and by original word
Note image transmitting is to Text segmentation module;
The Text segmentation module includes: image reception and storage unit, image processing unit;Image receives and storage is single
Member is for wirelessly or non-wirelessly receiving and storing original copybook image, and image processing unit is for adopting copybook image capture module
The original copybook image of collection is handled, and finds each text region and by original copybook image segmentation at multiple subgraphs
Picture forms in each subgraph and includes and only include a hand-written text;
The text grading module includes: subgraph receiving unit, subgraph processing unit, appraisal result transmission unit;
Subgraph receiving unit utilizes convolutional neural networks (Convolutional Neural for obtaining all subgraphs
Network, CNN) feature of each subgraph is extracted, subgraph processing unit is used for the feature of every row handwriting and the row
Grapholect feature compare, scored according to the matching degree between feature single handwriting, and calculate
Its average mark is as total score;The scoring of the single handwriting of every row is ranked up from high to low, (L is most to the higher L that will score
Character image number to be shown is needed eventually, does not include grapholect image) a subgraph and its corresponding scoring, appraisal result send
Unit is used to standard picture and total score passing to scoring display module;
The scoring display module includes: the display screen of mobile phone or flat-type mobile terminal;Scoring display module is used for
Receive subgraph and its corresponding scoring, standard picture and total score of the transmitting of text grading module;And it is shown in mobile phone or flat-type
On the screen of mobile terminal, while showing according to the height of total score whether to need repeatedly to practice the corresponding Chinese character or
Grasp the prompt informations such as the Chinese character.
Further, described image receive and storage unit, image processing unit, subgraph receiving unit, at subgraph
Reason unit, appraisal result transmission unit are either all incorporated into mobile phone or flat-type mobile terminal.
It is practiced calligraphy the methods of marking of artificial intelligence points-scoring system using a kind of copybook as described above, which is characterized in that packet
Include following steps:
Step 1, copybook Image Acquisition:
Writer selects interested copybook, and is practiced, and after completing the writing of one page, reuses mobile phone or plate
Class mobile terminal takes pictures to the page, obtains original copybook image and passes to Text segmentation module;
Step 2, Text segmentation:
Text segmentation module receives and processes the original copybook image from image capture module, and original copybook image is turned
It turns to grayscale image and picture is denoised using filter, detect the imaging region of copybook in the picture using Hough transformation;And
Cause original copybook image different degrees of distortion situation occur according to acquisition apparatus factor and human factor, is inserted by bilinearity
Value carries out geometric correction to picture, obtains the image for being easy to subsequent detection and segmentation;Then the image after correction is handled,
It identifies each text region and original image is divided by multiple subgraphs according to text region, in each subgraph
Include and only include a text, finally obtaining in image includes N row text, and every row includes that M handwriting (does not include
Grapholect), the grapholect segmented image of the i-th row in image is denoted as Si(1≤i≤N), the i-th j-th of row handwriting
Segmented image is denoted as WI, j(1≤i≤N,1≤j≤M);
Step 3, the scoring of artificial intelligence text:
The character features in all subgraphs are extracted with CNN, obtain image WI, jEigenmatrix WCI, j, image SiSpy
Levy matrix SCiAnd it stores.By eigenmatrix SCiWith matrix W CI, j(1≤i≤N, 1≤j≤M) carries out point multiplication operation, the knot of operation
Fruit is the scoring x of j-th of handwriting of the i-th rowSIG(i, j), and the average of the row handwriting is calculated as this article
The total score of word.Character image number L (not including grapholect) to be shown is thought in input, comments all handwritings of the i-th row
Divide xSIG(i, j) is ranked up from high to low, by the subgraph for scoring higher preceding L a after sequence and its corresponding scoring, standard drawing
Picture and total score pass to scoring display module;
Step 4, scoring display:
It receives the subgraph of text grading module transmitting and its corresponds to scoring, standard picture and total score and be shown in mobile terminal
Screen on, show that whether needing repeatedly to practice the Chinese character or grasped the Chinese character etc. mentions according to the height of the row total score
Show information.
Further, in step 3, the specific steps for writing artificial intelligence scoring include:
S1, text scoring start;
S2, obtains the total line number N of text in image, and the handwriting quantity M of every row once needs handwriting to be shown
Picture number L (0≤L≤M) resets counter i=0;
S3, counter start counting i=i+1;
S4, so that the grapholect image S of the i-th rowiIt exports to obtain standard feature matrix SC by CNNi, the i-th j-th of row
(1≤j≤M) image WI, jIt exports to obtain feature WC to be measured by CNNI, j;
S5, by image SiStandard feature matrix SCiRespectively with image WI, jThe eigenmatrix WC to be measured of (1≤j≤M)I, jInto
Row point multiplication operation, to obtain image WI, jScore xSIG(i,j);
S6 stores xSIG(i, j) value;
S7, to all data xSIG(i, j), j=1,2 ... M is ranked up from high to low, is scored after being sorted higher
L picture, the corresponding scoring of storage grapholect image, handwriting image and single image;
S8, according to xSIG(i, j) all values seek all handwriting score x of the i-th rowSIGThe mean value of (i, j)Total score obtained by writing as the text, and store total score;
S9 judges whether to have traversed N number of branch to be evaluated, i.e. i=N;It is completed if so, writing scoring;If it is not, being then recycled to
S3 step, until meeting ergodic condition.
Further, in step 4, the specific steps of the scoring display include:
T1, scoring display start;
T2 obtains text line number N, handwriting the number M, all data x of every row in imageSIG(i, j), xAVE
(i), (1≤i≤N, 1≤j≤M) resets counter i=0;
T3, counter start counting i=i+1;
T4, to all data xSIG(i, j), j=1,2 ... M is ranked up from high to low, the preceding L after being sorted figures
Piece;
T5, by preceding L picture and its total score x of corresponding scoring, the normal pictures of the row and the textAVE(i) it is shown in shifting
On the screen of moved end;
T6 shows total score xAVE(i) the correspondence signal language in section belonging to;
T7 judges whether to have traversed N row data, i.e. i=N;If so, scoring display is completed;If it is not, T3 step is then recycled,
Until meeting ergodic condition.
Compared with prior art, a kind of copybook described in the invention practice points-scoring system and method, have beneficial below
Effect:
(1) score value that writing words are calculated by the characteristic matching based on artificial neural network, by the writing on sense organ
Level is rationally judged as an accurate numerical value;
(2) enhance the interactivity between writer by increasing signal language, increase the interest in writing process, mention
The writing interest of high writer;
(3) not by handwriting input device such as board, Digitizing plates, do not changing the habit write on paper before this
Put the font for veritably improving writer.
Detailed description of the invention
Fig. 1 is that a kind of copybook of the present invention practices calligraphy artificial intelligence points-scoring system block diagram;
Fig. 2 is the Text segmentation method flow diagram that the copybook in the present invention practices calligraphy in artificial intelligence methods of marking;
Fig. 3 is the text methods of marking flow chart that the copybook in the present invention practices calligraphy in artificial intelligence methods of marking;
Fig. 4 is the scoring display methods flow chart in the artificial intelligence methods of marking that the copybook in the present invention practices calligraphy;
Fig. 5 is the original image acquired in the embodiment of the present invention 1;
Fig. 6 is the segmented image obtained after being handled in the embodiment of the present invention 1 by Text segmentation;
Fig. 7 is the 1st row writing calligraphy artificial intelligence scoring figure that mobile terminal screen obtained in the embodiment of the present invention 1 is shown
Picture;
Fig. 8 is the 2nd row writing calligraphy artificial intelligence scoring figure that mobile terminal screen obtained in the embodiment of the present invention 1 is shown
Picture;
Fig. 9 is the 3rd row writing calligraphy artificial intelligence scoring figure that mobile terminal screen obtained in the embodiment of the present invention 1 is shown
Picture;
Figure 10 is the 4th row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 1 is shown
Image;
Figure 11 is the 5th row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 1 is shown
Image;
Figure 12 is the 6th row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 1 is shown
Image;
Figure 13 is the 7th row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 1 is shown
Image;
Figure 14 is the eighth row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 1 is shown
Image;
Figure 15 is the 9th row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 1 is shown
Image;
Figure 16 is the 10th row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 1 is shown
Image;
Figure 17 is the original copybook image acquired in the embodiment of the present invention 2;
Figure 18 is the segmented image obtained after being handled in the embodiment of the present invention 2 by Text segmentation;
Figure 19 is the 1st row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 2 is shown
Image;
Figure 20 is the 2nd row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 2 is shown
Image;
Figure 21 is the 3rd row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 2 is shown
Image;
Figure 22 is the 4th row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 2 is shown
Image;
Figure 23 is the 5th row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 2 is shown
Image;
Figure 24 is the 6th row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 2 is shown
Image;
Figure 25 is the 7th row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 2 is shown
Image;
Figure 26 is the eighth row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 2 is shown
Image;
Figure 27 is the 9th row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 2 is shown
Image;
Figure 28 is the 10th row writing calligraphy artificial intelligence scoring that mobile terminal screen obtained in the embodiment of the present invention 2 is shown
Image.
Specific embodiment
Illustrate embodiments of the present invention below by way of specific specific example, those skilled in the art can be by this specification
Other advantages and efficacy of the present invention can be easily understood for disclosed content.The present invention can also pass through in addition different specific realities
The mode of applying is embodied or practiced, the various details in this specification can also based on different viewpoints and application, without departing from
Various modifications or alterations are carried out under spirit of the invention.
The core technology thought of a kind of copybook artificial intelligence points-scoring system to practice calligraphy and method of the invention is to realize
The method that a kind of pair of handwriting carries out the intelligent scoring of calligraphy, and provide practice opinion.To make technical solution of the present invention more
To be clear, below in conjunction with attached drawing, further the present invention is described in detail.
Fig. 1 is a kind of artificial intelligence points-scoring system block diagram that copybook practices calligraphy in the present invention, comprising: mobile phone or flat-type
Mobile terminal, digitizing treater;It is provided with copybook image capture module on the mobile phone or flat-type mobile terminal and comments
Divide display module, is provided with Text segmentation module, text grading module in the digitizing treater;Mobile phone or flat-type
Pass through wireless blue tooth or wire communication between mobile terminal and digitizing treater;The artificial intelligence points-scoring system is by adopting
Collect copybook image, and image is handled and divides to obtain the subgraph comprising single text, then extracts the feature of subgraph
And matched with the characteristics of image of grapholect, then show final appraisal result, and provided according to scoring whether grasp or
The prompt for whether needing to practice again.
The copybook image capture module includes: the camera on mobile phone or flat-type mobile terminal, setting camera point
Resolution is b1×b2;Copybook image capture module is for acquiring handwriting image to obtain original copybook image, and by original word
Note image transmitting is to Text segmentation module.
The Text segmentation module includes: image reception and storage unit, image processing unit;Image receives and storage is single
Member is for wirelessly or non-wirelessly receiving and storing original copybook image, and image processing unit is for adopting copybook image capture module
The original copybook image of collection is handled, and finds each text region and by original copybook image segmentation at multiple subgraphs
Picture forms in each subgraph and includes and only include a hand-written text.
The text grading module includes: subgraph receiving unit, subgraph processing unit, appraisal result transmission unit;
Subgraph receiving unit utilizes convolutional neural networks (Convolutional Neural for obtaining all subgraphs
Network, CNN) feature of each subgraph is extracted, subgraph processing unit is used for the feature of every row handwriting and the row
Grapholect feature compare, scored according to the matching degree between feature single handwriting, and calculate
Its average mark is as total score;The scoring of the single handwriting of every row is ranked up from high to low, (L is most to the higher L that will score
Character image number to be shown is needed eventually, does not include grapholect image) a subgraph and its corresponding scoring, appraisal result send
Unit is used to standard picture and total score passing to scoring display module.
The scoring display module is the display screen of mobile phone or flat-type mobile terminal;Scoring display module is for receiving
The subgraph of text grading module transmitting and its corresponding scoring, standard picture and total score;And it is shown in mobile phone or flat-type movement
On the screen of terminal, while showing whether need that the corresponding Chinese character is repeatedly practiced or grasped according to the height of total score
The prompt informations such as the Chinese character.
Preferably, described image receives and storage unit, image processing unit, subgraph receiving unit, subgraph are handled
Unit, appraisal result transmission unit are either all incorporated into a mobile phone or flat-type mobile terminal, the mobile phone or flat
Configured at least camera of 10,000,000 pixels, four core processors, at least 32G memory in plate class mobile terminal.
Fig. 2 is the Text segmentation method flow diagram in the present invention, realizes the function of Text segmentation module shown in FIG. 1.Its
Grayscale image is converted by original image and picture is denoised using filter, detects copybook page imaging using Hough transformation
Region.Since acquisition apparatus factor and human factor cause original image different degrees of distortion occur, it is therefore desirable to by double
Linear interpolation carries out geometric correction to picture, to obtain being easy to the image of subsequent detection and segmentation.Image after correction is carried out
Processing, identifies each text region and original image is divided into multiple subgraphs, every height according to text region
Include in image and only comprising a text.Finally obtaining image includes N row text, and every row has M handwriting (not include
Grapholect), the grapholect segmented image of the i-th row in image is denoted as Si(1≤i≤N), the i-th rower j-th of hand-written text of standard
The segmented image of word is denoted as WI, j(1≤i≤N,1≤j≤M)。
Fig. 3 is the artificial intelligence methods of marking flow chart of the writing calligraphy in the present invention, realizes writing calligraphy shown in FIG. 1
Artificial intelligence grading module function.The character features in all subgraphs are extracted with CNN, obtain image WI, jFeature square
Battle array WCI, j, image SiEigenmatrix SCiAnd it stores.By eigenmatrix SCiWith matrix W CI, j(1≤i≤N, 1≤j≤M) is carried out
Point multiplication operation, the result of operation are the scoring x of j-th of handwriting of the i-th rowSIG(i, j), and calculate the row handwriting
Total score of the average as the text.Character image number L (not including grapholect) to be shown is thought in input, to i-th
Capable all handwritings scoring xSIG(i, j) is ranked up from high to low, by the higher preceding L subgraphs that score after sequence
And its corresponding scoring, standard picture and total score pass to scoring display module;
S1, text scoring start;
S2, obtains the total line number N of text in image, and the handwriting number M of every row once needs handwritten Chinese character figure to be shown
The piece number L (0≤L≤M) resets counter i=0;
S3, counter start counting i=i+1;
S4, so that the grapholect image S of the i-th rowiIt exports to obtain standard feature matrix SC by CNNi, the i-th j-th of row
(1≤j≤M) image WI, jIt exports to obtain eigenmatrix WC to be measured by CNNI, j;
S5, by image SiCanonical matrix SCiRespectively with image WI, jThe eigenmatrix WC to be measured of (1≤j≤M)I, jIt carries out a little
Multiplication, to obtain image WI, jScoring xSIG(i,j);
S6 stores xSIG(i, j) value;
S7, to all data x of the i-th rowSIG(i, j), j=1,2 ... M is ranked up from high to low, after being sorted
Preceding L picture, the corresponding scoring of storage grapholect image, handwriting image and single image;
S8, according to xSIG(i, j) all values seek all handwriting score x of the i-th rowSIGThe mean value of (i, j)Total score obtained by writing as the text, and store total score;
S9 judges whether to have traversed N number of branch to be evaluated, i.e. i=N;It is completed if so, writing scoring;If it is not, being then recycled to
S3 step, until meeting ergodic condition.
Fig. 4 is the scoring display methods flow chart in the present invention, realizes the function of scoring display module shown in FIG. 1.It connects
The subgraph of message in-coming word grading module transmitting and its corresponding scoring, standard picture and total score are simultaneously shown on the screen of mobile terminal,
Show whether need repeatedly to practice the text or grasped the prompt informations such as the Chinese character according to the height of the row total score.Tool
The step of body, the scoring display methods are as follows:
T1, scoring display start;
T2 obtains text line number N, the handwriting M, all data x of every row in imageSIG(i, j), xAVE(i), 1
≤ i≤N, 1≤j≤M reset counter i=0;
T3, counter start counting i=i+1;
T4, by the total score x of higher preceding L picture and its corresponding scoring, the normal pictures of the row and the text that scoresAVE
(i) it is shown on the screen of mobile terminal;
T5 judges total score xAVE(i) section belonging to simultaneously shows corresponding signal language on the screen;
T6 judges whether to have traversed N row data, i.e. i=N;If so, scoring display is completed;If it is not, T3 step is then recycled,
Until meeting ergodic condition.
Below with two more specifically embodiments, to technical solution of the present invention further description:
Embodiment 1
Writer selects copybook and practices, and after completing the writing of this page, user claps the page using mobile phone
According to obtaining shown in original image as shown in Figure 5 and pass to Text segmentation module.
Text segmentation module receives and processes the original copybook image from image capture module.Original word textures picture is turned
It turns to grayscale image and picture is denoised using filter, detect copybook page region imaged using Hough transformation.Due to adopting
Collection apparatus factor and human factor cause original image different degrees of distortion occur, it is therefore desirable to by bilinear interpolation to figure
Piece carries out geometric correction, to obtain the image for being easy to detect and divide.Image after correction is handled, identifies each text
Original image is simultaneously divided into multiple subgraphs according to text region by region, is included in each subgraph and is only included
One text.Finally obtaining image includes 10 row texts, 12 column handwritings (not including grapholect), by image i-th
Capable grapholect segmented image is denoted as SiThe segmented image of (1≤i≤10), the i-th j-th of row handwriting is denoted as WI, j(1≤i
≤10,1≤j≤12)。
Fig. 6 is segmented image obtained in present example 1, with the text in all subgraphs of CNN VGG19 model extraction
Word feature obtains image WI, jScale be 512 × 1 eigenmatrix WCI, j, image SiScale be 512 × 1 eigenmatrix
SCiAnd it stores.By image SiEigenmatrix SCiWith image WI, jThe eigenmatrix WC of (1≤i≤10,1≤j≤12)I, jRespectively
Point multiplication operation is carried out, the scoring x of all single handwritings is obtainedSIG(i,j)(1≤i≤10,1≤j≤12)。
Calculate the total score of every row handwriting:
xAVE(1)=84.54, xAVE(2)=83.40, xAVE(3)=85.55, xAVE(4)=78.72, xAVE(5)=
82.95,
xAVE(6)=79.02, xAVE(7)=79.98, xAVE(8)=79.35, xAVE(9)=84.59, xAVE(10)=
80.86。
12 handwriting total scores of every row are ranked up from high to low respectively, the subgraph after being sorted will simultaneously be commented
Preceding 9 subgraphs and its corresponding scoring, standard picture and total score is divided to pass to scoring display module.It is stored in display module
Default score and its export between mapping relationship f (x):
Fig. 7 is the artificial intelligence scoring figure for the 1st row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Picture, total score xAVE(1)=84.55, prompt " not grasping completely, but not far apart from grasping ".
Fig. 8 is the artificial intelligence scoring figure for the 2nd row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Picture, total score xAVE(2)=83.40, prompt " not grasping completely, but not far apart from grasping ".
Fig. 9 is the artificial intelligence scoring figure for the 3rd row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Picture, total score xAVE(3)=85.56, prompt " not grasping completely, but not far apart from grasping ".
Figure 10 is the artificial intelligence scoring for the 4th row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(4)=78.72, " practise handwriting new hand, it is also necessary to add to practice " is prompted.
Figure 11 is the artificial intelligence scoring for the 5th row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(5)=82.95, prompt " not grasping completely, but not far apart from grasping ".
Figure 12 is the artificial intelligence scoring for the 6th row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(6)=79.02, " practise handwriting new hand, it is also necessary to add to practice " is prompted.
Figure 13 is the artificial intelligence scoring for the 7th row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(7)=79.98, " practise handwriting new hand, it is also necessary to add to practice " is prompted.
Figure 14 is the artificial intelligence scoring for the eighth row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(8)=79.35, " practise handwriting new hand, it is also necessary to add to practice " is prompted.
Figure 15 is the artificial intelligence scoring for the 9th row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(9)=84.59, prompt " not grasping completely, but not far apart from grasping ".
Figure 16 is the artificial intelligence scoring for the 10th row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(10)=80.86, prompt " not grasping completely, but not far apart from grasping ".
Embodiment two
Writer selects copybook and carries out the interested copybook of practice user's selection, and is practiced.When completing this page
After writing, user takes pictures to the page using mobile phone, obtains shown in original image as shown in figure 17 and passes to text point
Cut module.
Text segmentation module receives and processes the original image from image capture module.Gray scale is converted by original image
Scheme and filter is used to denoise picture, detects copybook page region imaged using Hough transformation.Due to acquisition equipment because
Element and human factor cause original image different degrees of distortion occur, it is therefore desirable to be carried out by bilinear interpolation to picture several
What is corrected, to obtain the image for being easy to detect and divide.Image after correction is handled, identifies each text region
And original image is divided by multiple subgraphs according to text region, include in each subgraph and only comprising a text
Word.Finally obtaining image includes 10 row texts, 12 column handwritings (not including grapholect), by the mark of the i-th row in image
Quasi- Text segmentation image is denoted as SiThe segmented image of (1≤i≤10), the i-th j-th of row handwriting is denoted as WI, j(1≤i≤10,1
≤j≤12)。
Figure 18 is segmented image obtained in present example 2, with the text in all subgraphs of CNN VGG19 model extraction
Word feature obtains image WI, jScale be 512 × 1 eigenmatrix WCI, j, image SiScale be 512 × 1 eigenmatrix
SCiAnd it stores.By image SiEigenmatrix SCiWith image WI, jThe eigenmatrix WC of (1≤i≤10,1≤j≤12)I, jRespectively
Point multiplication operation is carried out, the scoring x of all single handwritings is obtainedSIG(i,j)(1≤i≤10,1≤j≤12)。
Calculate the total score of every row handwriting:
xAVE(1)=81.58, xAVE(2)=73.55, xAVE(3)=74.70, xAVE(4)=73.30, xAVE(5)=
69.23,
xAVE(6)=66.04, xAVE(7)=76.54, xAVE(8)=70.44, xAVE(9)=63.05, xAVE(10)=
70.12。
12 handwriting total scores of every row are ranked up from high to low respectively, the subgraph after being sorted will simultaneously be commented
Preceding 9 subgraphs and its corresponding scoring, standard picture and total score is divided to pass to scoring display module.It is stored in display module
Default score and its export between mapping relationship f (x):
Figure 19 is the artificial intelligence scoring for the 1st row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(1)=81.58, " luxurious platinum " is prompted.
Figure 20 is the artificial intelligence scoring for the 2nd row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(2)=73.55, " unyielding silver " is prompted.
Figure 21 is the artificial intelligence scoring for the 3rd row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(3)=74.70, " unyielding silver " is prompted.
Figure 22 is the artificial intelligence scoring for the 4th row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(4)=73.30, " unyielding silver " is prompted.
Figure 23 is the artificial intelligence scoring for the 5th row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(5)=69.23, prompt " brave bronze ".
Figure 24 is the artificial intelligence scoring for the 6th row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(6)=66.04, prompt " brave bronze ".
Figure 25 is the artificial intelligence scoring for the 7th row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(7)=76.54, " luxurious platinum " is prompted.
Figure 26 is the artificial intelligence scoring for the eighth row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(8)=70.44, " unyielding silver " is prompted.
Figure 27 is the artificial intelligence scoring for the 9th row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(9)=63.05, prompt " brave bronze ".
Figure 28 is the artificial intelligence scoring for the 10th row writing calligraphy that mobile terminal screen obtained in present example 1 is shown
Image, total score xAVE(10)=70.12, " unyielding silver " is prompted.
It can be seen from the above embodiments that, the present invention is the characteristic matching based on artificial neural network, calculates and writes text
Writing level on sense organ is rationally judged as an accurate numerical value by the score value of word;Enhanced by increasing signal language
Interactivity between writer increases the interest in writing process, improves the writing interest of writer;Not by writing
The handwriting input device such as plate, Digitizing plate, do not change the habit write on paper this under the premise of veritably improve writer
Font.
The principle of the present invention and its effect is only illustrated, and is not intended to limit the present invention.It is any to be familiar with this technology
Personage all without departing from the spirit and scope of the present invention, carries out modifications and changes to above-described embodiment.Therefore, belonging to such as
Have all that usually intellectual is completed without departing from the spirit and technical ideas disclosed in the present invention etc. in technical field
Modifications and changes are imitated, should be covered by the claims of the present invention.
Although the embodiments of the present invention has been shown and described above, it is to be understood that above-described embodiment is example
Property, it is not considered as limiting the invention, those skilled in the art are not departing from the principle of the present invention and objective
In the case where can make changes, modifications, alterations, and variations to the above described embodiments within the scope of the invention.The scope of the present invention
By appended claims and its equivalent limit.
Claims (5)
1. a kind of artificial intelligence points-scoring system that copybook practices calligraphy, comprising: mobile phone or flat-type mobile terminal, digitized processing
Device;Copybook image capture module and scoring display module, the number are provided on the mobile phone or flat-type mobile terminal
Change in processing unit and is provided with Text segmentation module, text grading module;Mobile phone or flat-type mobile terminal and digitized processing
Pass through wireless blue tooth or wire communication between device;The artificial intelligence points-scoring system passes through acquisition copybook image, and to image
Handled and divide to obtain the subgraph comprising single text, then extract subgraph feature and with the image of grapholect spy
Whether sign is matched, and then shows final appraisal result, and provided according to scoring and to be grasped or whether need to practice again mentions
Show;It is characterized by:
The copybook image capture module includes: the camera on mobile phone or flat-type mobile terminal, sets resolution ratio of camera head
For b1×b2;Copybook image capture module is for acquiring handwriting image to obtain original copybook image, and by original copybook figure
As being transferred to Text segmentation module;
The Text segmentation module includes: image reception and storage unit, image processing unit;Image receives and storage unit is used
In wirelessly or non-wirelessly receiving and storing original copybook image, image processing unit is for collected to copybook image capture module
Original copybook image is handled, and finds each text region and by original copybook image segmentation at multiple subgraphs, shape
At including in each subgraph and only comprising a hand-written text;
The text grading module includes: subgraph receiving unit, subgraph processing unit, appraisal result transmission unit;Subgraph
As receiving unit is for obtaining all subgraphs, convolutional neural networks (CNN) is utilized to extract the feature of each subgraph, subgraph
Processing unit is for comparing the feature of every row handwriting and the grapholect feature of the row, according between feature
It scores with degree single handwriting, and calculates its average mark as total score;The single handwriting of every row is commented
Divide and be ranked up from high to low, will score higher L subgraph and its corresponding scoring, and appraisal result transmission unit will be for that will mark
Quasi- image and total score pass to scoring display module;
The scoring display module includes: the display screen of mobile phone or flat-type mobile terminal;Scoring display module is for receiving
The subgraph of text grading module transmitting and its corresponding scoring, standard picture and total score;And it is shown in mobile phone or flat-type movement
On the screen of terminal, while showing whether need that the corresponding Chinese character is repeatedly practiced or grasped according to the height of total score
The prompt informations such as the Chinese character.
2. a kind of artificial intelligence points-scoring system that copybook practices calligraphy as described in claim 1, it is characterised in that: described image connects
Receive and storage unit, image processing unit, subgraph receiving unit, subgraph processing unit, appraisal result transmission unit or
It is all to be incorporated into mobile phone or flat-type mobile terminal.
3. being practiced calligraphy the methods of marking of artificial intelligence points-scoring system using a kind of copybook described in claim 1, which is characterized in that
The following steps are included:
Step 1, copybook Image Acquisition:
Writer selects interested copybook, and is practiced, and after completing the writing of one page, reuses mobile phone or flat-type is moved
Dynamic terminal takes pictures to the page, obtains original copybook image and passes to Text segmentation module;
Step 2, Text segmentation:
Text segmentation module receives and processes the original copybook image from image capture module, converts original copybook image to
Grayscale image simultaneously denoises picture using filter, detects the imaging region of copybook in the picture using Hough transformation;And according to
Acquisition apparatus factor and human factor cause original copybook image different degrees of distortion situation occur, pass through bilinear interpolation pair
Picture carries out geometric correction, obtains the image for being easy to subsequent detection and segmentation;Then the image after correction is handled, is identified
Original image is simultaneously divided into multiple subgraphs according to text region by each text region, includes in each subgraph
And only comprising a text, finally obtaining in image includes N row text, and every row includes that M handwriting (does not include standard
Text), the grapholect segmented image of the i-th row in image is denoted as Si(1≤i≤N), the segmentation of the i-th j-th of row handwriting
Image is denoted as WI, j(1≤i≤N,1≤j≤M);
Step 3, the scoring of artificial intelligence text:
The character features in all subgraphs are extracted using convolutional neural networks (CNN), obtain image WI, jEigenmatrix
WCI, j, image SiEigenmatrix SCiAnd it stores.By eigenmatrix SCiWith matrix W CI, j(1≤i≤N, 1≤j≤M) is carried out a little
Multiplication, the result of operation are the scoring x of j-th of handwriting of the i-th rowSIG(i, j), and calculate the row handwriting
Total score of the average as the text.Character image number L (not including grapholect) to be shown is thought in input, to the i-th row
All handwritings score xSIG(i, j) is ranked up from high to low, by the subgraph for the higher preceding L of scoring after sequence and
It corresponds to scoring, standard picture and total score and passes to scoring display module;
Step 4, scoring display:
It receives the subgraph of text grading module transmitting and its corresponds to scoring, standard picture and total score and the screen for being shown in mobile terminal
On curtain, show whether need repeatedly to practice the Chinese character or grasped the prompts such as Chinese character letter according to the height of the row total score
Breath.
The methods of marking of artificial intelligence points-scoring system 4. a kind of copybook as claimed in claim 3 practices calligraphy, which is characterized in that
In step 3, the specific steps for writing artificial intelligence scoring include: S1, and text scoring starts;S2 obtains the text in image
The total line number N of word, the handwriting quantity M of every row, once need handwriting picture number L (0≤L≤M) to be shown, and resetting counts
Device i=0;S3, counter start counting i=i+1;S4, so that the grapholect image S of the i-th rowiIt is exported and is marked by CNN
Quasi- eigenmatrix SCi, the i-th j-th of row (1≤j≤M) image WI, jIt exports to obtain feature WC to be measured by CNNI, j;
S5, by image SiStandard feature matrix SCiRespectively with image WI, jThe eigenmatrix WC to be measured of (1≤j≤M)I, jIt carries out a little
Multiplication, to obtain image WI, jScore xSIG(i,j);
S6 stores xSIG(i, j) value;
S7, to all data xSIG(i, j), j=1,2 ... M is ranked up from high to low, and the higher L that scores after being sorted
Picture, the corresponding scoring of storage grapholect image, handwriting image and single image;
S8, according to xSIG(i, j) all values seek all handwriting score x of the i-th rowSIGThe mean value of (i, j)
Total score obtained by writing as the text, and store total score;
S9 judges whether to have traversed N number of branch to be evaluated, i.e. i=N;It is completed if so, writing scoring;If it is not, being then recycled to S3
Step, until meeting ergodic condition.
The methods of marking of artificial intelligence points-scoring system 5. a kind of copybook as claimed in claim 3 practices calligraphy, which is characterized in that
In step 4, the specific steps of the scoring display include:
T1, scoring display start;
T2 obtains text line number N, handwriting the number M, all data x of every row in imageSIG(i, j), xAVE(i), (1
≤ i≤N, 1≤j≤M), reset counter i=0;
T3, counter start counting i=i+1;
T4, to all data xSIG(i, j), j=1,2 ... M is ranked up from high to low, the preceding L picture after being sorted;
T5, by preceding L picture and its total score x of corresponding scoring, the normal pictures of the row and the textAVE(i) it is shown in mobile terminal
Screen on;
T6 shows total score xAVE(i) the correspondence signal language in section belonging to;
T7 judges whether to have traversed N row data, i.e. i=N;If so, scoring display is completed;If it is not, T3 step is then recycled, until
Until meeting ergodic condition.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2019100110381 | 2019-01-07 | ||
CN201910011038 | 2019-01-07 |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110232377A true CN110232377A (en) | 2019-09-13 |
Family
ID=67860916
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910427401.8A Withdrawn CN110232377A (en) | 2019-01-07 | 2019-05-22 | A kind of artificial intelligence points-scoring system that copybook practices calligraphy and method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110232377A (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110659702A (en) * | 2019-10-17 | 2020-01-07 | 黑龙江德亚文化传媒有限公司 | Calligraphy copybook evaluation system and method based on generative confrontation network model |
CN110796065A (en) * | 2019-10-26 | 2020-02-14 | 深圳市锦上科技有限公司 | Calligraphy practicing scoring method and system based on image recognition and computer readable medium |
CN110991438A (en) * | 2019-11-29 | 2020-04-10 | 广东马上信息科技有限公司 | Font scoring method, system and storage medium thereof |
CN111160133A (en) * | 2019-12-12 | 2020-05-15 | 佛山科学技术学院 | Method and device for evaluating similarity of copybook characters |
CN111554149A (en) * | 2020-05-15 | 2020-08-18 | 黑龙江德亚文化传媒有限公司 | System and method for copybook scoring |
CN112364723A (en) * | 2020-10-26 | 2021-02-12 | 深圳市合众望电子有限公司 | Calligraphy practicing learning method and device |
CN113344014A (en) * | 2021-08-03 | 2021-09-03 | 北京世纪好未来教育科技有限公司 | Text recognition method and device |
CN113610866A (en) * | 2021-07-28 | 2021-11-05 | 上海墨说科教设备有限公司 | Method, device and equipment for cutting calligraphy practice image and storage medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104715256A (en) * | 2015-03-04 | 2015-06-17 | 南昌大学 | Auxiliary calligraphy exercising system and evaluation method based on image method |
WO2016041423A1 (en) * | 2014-09-16 | 2016-03-24 | 科大讯飞股份有限公司 | Intelligent scoring method and system for text objective question |
CN107609575A (en) * | 2017-08-21 | 2018-01-19 | 北京华文众合科技有限公司 | Calligraphy evaluation method, calligraphy evaluating apparatus and electronic equipment |
-
2019
- 2019-05-22 CN CN201910427401.8A patent/CN110232377A/en not_active Withdrawn
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016041423A1 (en) * | 2014-09-16 | 2016-03-24 | 科大讯飞股份有限公司 | Intelligent scoring method and system for text objective question |
CN104715256A (en) * | 2015-03-04 | 2015-06-17 | 南昌大学 | Auxiliary calligraphy exercising system and evaluation method based on image method |
CN107609575A (en) * | 2017-08-21 | 2018-01-19 | 北京华文众合科技有限公司 | Calligraphy evaluation method, calligraphy evaluating apparatus and electronic equipment |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110659702A (en) * | 2019-10-17 | 2020-01-07 | 黑龙江德亚文化传媒有限公司 | Calligraphy copybook evaluation system and method based on generative confrontation network model |
CN110796065A (en) * | 2019-10-26 | 2020-02-14 | 深圳市锦上科技有限公司 | Calligraphy practicing scoring method and system based on image recognition and computer readable medium |
CN110991438A (en) * | 2019-11-29 | 2020-04-10 | 广东马上信息科技有限公司 | Font scoring method, system and storage medium thereof |
CN111160133A (en) * | 2019-12-12 | 2020-05-15 | 佛山科学技术学院 | Method and device for evaluating similarity of copybook characters |
CN111554149A (en) * | 2020-05-15 | 2020-08-18 | 黑龙江德亚文化传媒有限公司 | System and method for copybook scoring |
CN112364723A (en) * | 2020-10-26 | 2021-02-12 | 深圳市合众望电子有限公司 | Calligraphy practicing learning method and device |
CN113610866A (en) * | 2021-07-28 | 2021-11-05 | 上海墨说科教设备有限公司 | Method, device and equipment for cutting calligraphy practice image and storage medium |
CN113610866B (en) * | 2021-07-28 | 2024-04-23 | 上海墨说科教设备有限公司 | Method, device, equipment and storage medium for cutting calligraphy practicing image |
CN113344014A (en) * | 2021-08-03 | 2021-09-03 | 北京世纪好未来教育科技有限公司 | Text recognition method and device |
CN113344014B (en) * | 2021-08-03 | 2022-03-08 | 北京世纪好未来教育科技有限公司 | Text recognition method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110232377A (en) | A kind of artificial intelligence points-scoring system that copybook practices calligraphy and method | |
CN106127108B (en) | A kind of manpower image region detection method based on convolutional neural networks | |
CN107563385A (en) | License plate character recognition method based on depth convolution production confrontation network | |
CN107240330A (en) | System that handwriting display method and system, the intelligence of practising handwriting practise handwriting terminal and intelligence is practised handwriting | |
CN105787522B (en) | Handwriting-based writing attitude evaluation method and system | |
CN107239801A (en) | Video attribute represents that learning method and video text describe automatic generation method | |
CN104915667B (en) | A kind of answering card identifying and analyzing method and system based on mobile terminal | |
CN106980856A (en) | Formula identification method and system and symbolic reasoning computational methods and system | |
CN105869085A (en) | Transcript inputting system and method for processing images | |
CN101615244A (en) | Handwritten plate blank numbers automatic identifying method and recognition device | |
CN108537218A (en) | A kind of identifying processing method and device of answering card | |
CN110516554A (en) | A kind of more scene multi-font Chinese text detection recognition methods | |
CN109872305A (en) | It is a kind of based on Quality Map generate network without reference stereo image quality evaluation method | |
CN110163567A (en) | Classroom roll calling system based on multitask concatenated convolutional neural network | |
CN110659612A (en) | Digital marking method and system based on paper pen improvement | |
CN106778717A (en) | A kind of test and appraisal table recognition methods based on image recognition and k nearest neighbor | |
CN111540253A (en) | Intelligent hard-tipped pen calligraphy practicing system and scoring method | |
CN104679967B (en) | A kind of method for judging psychological test reliability | |
CN103455826B (en) | Efficient matching kernel body detection method based on rapid robustness characteristics | |
CN108052936B (en) | Automatic inclination correction method and system for Braille image | |
CN113191309A (en) | Method and system for recognizing, scoring and correcting handwritten Chinese characters | |
CN206115752U (en) | Ar augmented reality teaching system | |
CN107958261B (en) | Braille point detection method and system | |
CN110796065A (en) | Calligraphy practicing scoring method and system based on image recognition and computer readable medium | |
CN117095414A (en) | Handwriting recognition system and recognition method based on dot matrix paper pen |
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 | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190913 |
|
WW01 | Invention patent application withdrawn after publication |