CN102541347A - Automatic identification system and method of handwriting Chinese character - Google Patents
Automatic identification system and method of handwriting Chinese character Download PDFInfo
- Publication number
- CN102541347A CN102541347A CN2011103604419A CN201110360441A CN102541347A CN 102541347 A CN102541347 A CN 102541347A CN 2011103604419 A CN2011103604419 A CN 2011103604419A CN 201110360441 A CN201110360441 A CN 201110360441A CN 102541347 A CN102541347 A CN 102541347A
- Authority
- CN
- China
- Prior art keywords
- image
- handwriting
- data
- gradient
- screen
- 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
Images
Landscapes
- Character Discrimination (AREA)
Abstract
The invention relates to an automatic identification system and a method of a handwriting Chinese character. The system comprises an EMB-4650 main board, wherein expanded sub-interfaces of the EMB-4650 main board are a low voltage differential signaling (LVDS) interface, a digital video interactive (DVI) interface, a universal serial bus (USB) interface and a computer-output microfilm (COM) interface which are respectively connected with an AUO display screen, a DVI-LVDS signal converting board, a USB controller and a global system for mobile communication (GSM) module. The DVI-LVDS signal converting board is connected with another AUO display screen, and the USB controller is connected with a touch screen. The identification method comprises the steps of (1) inputting handwriting Chinese character data, (2) preprocessing the handwriting Chinese character data, (3) extracting shape features of the handwriting Chinese character data, and (4) identifying the handwriting Chinese character data. By means of the automatic identification system and the method of the handwriting Chinese character, a novel polar coordinate feature recognition method with font centroid as a reference point. Users can record, store and send important information in real time by utilizing a handwriting mode, thereby achieving the matching between handwriting Chinese characters and handwriting Chinese characters. The system integrates high confidentiality and high-speed transmission and is suitable for intelligence agents to collect and screen the important information.
Description
Technical field
The present invention relates to a kind of method that utilization Computer Automatic Extraction takes down notes feature, develop a kind of hand-written handwriting automatic recognition system and method, using image segmentation algorithm and person's handwriting is searched and extracted to the distinguished number based on statistical nature, this handwriting identification method can be widely applied to the fields such as finance, law.
Background technology
Person's handwriting is a kind of behavioural characteristic of quite stable, and different people has different person's handwritings, and handwriting is easily obtained.With developing rapidly for information digitalization, handwritten form writer verification turns into research topic active in computer vision and area of pattern recognition.It is well known that document expert can identify the true and false of person's handwriting, but Computer Automatic Extraction handwriting characteristic, and mirror determines whether it is true or false, and particularly still has certain difficulty to a small amount of person's handwriting.
The research of computer handwriting identification starts from the 60's of last century earliest.In recent years, because computer handwriting identification technology is widely used in fields such as public security, finance, education assessments, it is set just progressively to turn into a very active but higher research topic of difficulty in computer vision and area of pattern recognition, Chinese and overseas scholars conducts extensive research to this, but fails to obtain gratifying solution so far.
Because the ICT starting of developed country is early, development is fast, it is popularized and usage degree has reached higher level, and their computer management and automatic technology also have suitable popularity rate and higher up-to-dateness, thus computer identity authentication techniques are progressively widely noted and are studied.The article that the existing several signatures of in the market differentiate abroad at present occurs, but is due to the particularity of character structure, and the research work differentiated for Chinese character handwriting is made slow progress.Again due to the difference of culture background etc., western countries differentiate that Study on Problems is less for orient characters, therefore the research institution of relevant issues is distributed in Asia mostly, and main research institution includes Busan, Korea university at present, Institute of Automation, Chinese Academy of sociences's pattern-recognition laboratory, the Chinese Academy of Sciences calculates institute.In addition, the country there are many companies being all directed to the exploitation of writer verification system aspects at present, and achieve some achievements, but not having also till now of really coming into operation.
The content of the invention
It is an object of the invention to the defect existed for prior art, a kind of hand-written handwriting automatic recognition system and method are provided, diary in being worked using daily life is design blank, writing style is taken down notes with reference to personal pair of pages, using computer technology and the lookup algorithm of innovation, information searching is realized for personal handwritten handwriting.The characteristics of many design features of the system from Chinese character in itself, comprehensive analysis handwritten form, given up huge standard character library, directly by " person's handwriting to person's handwriting " retrieve it is convenient, both note extracting general character from individual character, the assurance for individual character emphasized again.
To achieve these goals, design of the invention is as follows：The present invention is the hand-written script identification based on image, realizes locating function.This technology belongs to off line character recognition category, and system searching function need to only find same or close character in the storehouse that user is write, and need not judge that the character that user is write corresponds specifically to which of Chinese character word.The system object to be dealt with is two dimensional image, by after institute's writing of Chinese characters and background separation, by a series of pretreatment, finds out the position where picture file identical or similar character with institute writing of Chinese characters while displaying it.Whole search procedure can be divided into following components：Acquisition of information（Hand-written character is inputted）, pretreatment（Including removing background, Character segmentation etc.）, single character feature extracts, compares selection.
Conceived according to foregoing invention, the present invention uses following technical proposals：
A kind of hand-written handwriting automatic recognition system, including one piece by IntelOne piece of EMB-4650 mainboard that the master control of Atom Z510P chips makes, it is characterised in that the sub-interface of the mainboard extension has LVDS interface, DVI interface, USB interface and com interface；These interfaces are connected with AUO display screens, DVI-LVDS signal transducers, USB controller, gsm module respectively；Thus DVI-LVDS signal transducers are connected with other AUO display screens, and USB controller is connected with touch-screen.When user writes data on the touchscreen, these data are sent to EMB-4650 mainboards by USB controller, while EMB-4650 mainboards transfer data to display screen, realize display in real time；Algorithm process is identified to these data in EMB-4650 mainboards simultaneously, and the result after processing, which is then sent to display screen, shows, reaches easy human-computer interaction real-time operation effect.If it is desirable, the character font data editted can be sent to gsm module by com port, the real-time transmission of data is realized.
Font is write by user on the touchscreen, character font data is input to storage device, is pre-processed by processor, then extracted using the polar coordinate method for applying for protection herein after the shape facility of single character, selection determines inputted character.Meanwhile, the character result after the input and processing of user is shown on LCDs.If it is desirable, can be wirelessly transferred using gsm module to input information, it is allowed to more convenient and timely transmits information.
In said system, the power supply of system is powered by the way of external power supply.The 18V AC powers are obtained from civil power 220V50Hz by a power-supply filter and a transformer, accessed the voltage conversion circuit, 18V alternating currents are changed into 12V, 5V, 3.3V and 1.8V power supply by the voltage conversion circuit, and working power is provided for system modules.
In said system, touch-screen is then the USB interface that mainboard is connected to by change-over panel.The change-over panel completes the signal of input is converted into usb signal with system host to be connected, and controls by corresponding software the performance of touch-screen.
In said system, B089AW01 V1 are a colour dot lattice LCDs being made up of TFT LCD flat panels, a set of drive circuit and LED backlight system.Screen display supports WSVGA (1024 (H) x 600 (V)) forms and is equipped with the 262k color data-drivens of LED backlight driving circuit.All input signals are all compatible LVDS interfaces.This one of LVDS interface for being connected to mainboard of two pieces of display screens, another piece of change-over panel by DVI-LVDS is connected to the DVI interface of mainboard.DVI-LVDS completes the LVDS signals that the DVI signals of input are converted into being adapted to LCDs.
In said system, two gsm modules are the GSM/GPRS wireless transceivers of damp Chinese T-1008 series.Module is main by Modem main parts（Including chip, peripheral circuit etc.）, voltage-stabilized power supply, and three parts of antenna are constituted.GPRS moulds are connected to mainboard by chip com port.
A kind of hand-written handwriting automatic identifying method, is identified, it is characterised in that concrete operation step is as follows using said system：Step 1：The input of handwriting data；Step 1：The pretreatment of handwriting data；Step 3：The Shape Feature Extraction of handwriting data；Step 4：Handwriting data are recognized.
The present invention is compared with the prior art compared with following obvious prominent substantive distinguishing features and remarkable advantage：
This hand-written handwriting automatic identifying method is that recognition methods is that person's handwriting is searched, and sets up the polar coordinates feature recognition new method put on the basis of person's handwriting font barycenter, recognition accuracy is high；And easy-to-look-up written information, removes numerous and diverse papery search procedure from, realize that notes graphic form is preserved in real time, the holding time is permanent.
Single screen before the present invention is different from is shown, using shuangping san mode, related setting is carried out in software, it could be arranged to double screen independence display pattern and joint display pattern, basic display unit could be arranged to digital display during double screen independence display pattern, secondary display is set to MID, while the setting of screen rotation can also be carried out.This disposable display pattern can meet the requirement of various occasions, neither lose visual effect again easy to carry.
The font face feature extraction algorithm of the present invention is advanced, it is different from traditional X-Y axial coordinate recognition methods, the polar coordinates recognition methods proposed using the barycenter of font as the center of circle of innovation, solve the Chinese character for tiled configuration or left, center, right structure, the setting of partitioning algorithm threshold value and personal writing style are a pair of contradictions for being relatively difficult to solve, and improve the discrimination of font.
Brief description of the drawings
Fig. 1 is the structured flowchart of hand-written handwriting automatic recognition system of the invention.
Fig. 2 is hand-written handwriting automatic identification flow chart of the invention.
Fig. 3 is the handwriting data prediction flow chart in Fig. 2.
Fig. 4 is the image binaryzation flow chart in Fig. 3.
Fig. 5 is the Character Font Recognition flow chart in Fig. 1.
Embodiment：
Details are as follows for the case combination accompanying drawing that is preferable to carry out of the present invention：
Embodiment one：Referring to Fig. 1, this hand-written handwriting automatic recognition system includes one piece by IntelOne piece of EMB-4650 mainboard that the master control of Atom Z510P chips makes, its sub-interface extended has LVDS interface, DVI interface, USB interface and com interface；These interfaces are connected with AUO display screens, DVI-LVDS signal transducers, USB controller, gsm module respectively；Thus DVI-LVDS signal transducers are connected with other AUO display screens, and USB controller is connected with touch-screen.When user writes data on the touchscreen, these data are sent to EMB-4650 mainboards by USB controller, while EMB-4650 mainboards transfer data to display screen, realize display in real time；Algorithm process is identified to these data in EMB-4650 mainboards simultaneously, and the result after processing, which is then sent to display screen, shows, reaches easy human-computer interaction real-time operation effect.If it is desirable, the character font data editted can be sent to gsm module by com port, the real-time transmission of data is realized.The present invention is described in further details with reference to the preferred embodiments of the present invention and accompanying drawing.
Embodiment two：Referring to Fig. 2, automatic writing pen body automatic recognition system implementation method is identified using said system, and concrete operations comprise the following steps：
Step 1：The input of handwriting data：It is divided into surface acoustic wave touch screen, resistive touch screen, capacitive touch screen, infrared-type touch-screen etc. according to the difference of touch-screen operation principle several.Touch-screen used by the present invention is resistive touch screen.Resistive touch screen is arranged on display surface, is mainly made up of two layers of transparency conducting layer, is insulated between two conductive layers.From electrical characteristic, equivalent to one resistive element of each conductive layer, in conductive layer two ends making alive, it is possible to set up corresponding relation between plan-position and pressure drop.When being touched, two conductive layers are contacted at contact, by means of contact, the voltage of reflection planes position is drawn, contact position is obtained after measurement.Touch-screen is read these number of contacts evidence is stored in the form of image data for main frame.
Step 2：The pretreatment of handwriting data, referring to Fig. 3.Data prediction flow chart of the present invention is as shown in Figure 4.Mainboard（1）From touch-screen（7）The view data deposited is read in, mainboard is input to（1）In register.
Step 2-1：256 color bitmaps turn gray-scale map.
A) grey scale mapping table is calculated
BMap [i]=(BYTE) (0.299 * lpbmi- can be obtained by the relational expression Y=0.299R+0.587G+0.114B between luminance signal Y and three primary colours R, G, B>bmiColors[i].rgbRed
+0.587 * lpbmi->bmiColors[i].rgbGreen
+0.114 * lpbmi->bmiColors[i].rgbBlue + 0.5);
Wherein lpbmi is the pointer for pointing to BITMAPINFO structures
B) DIB palette three primary colours components are updated
lpbmi->bmiColors[i].rgbRed = i;
lpbmi->bmiColors[i].rgbGreen = i;
lpbmi->bmiColors[i].rgbBlue = i;
C) color index of each pixel is changed（Gray value is changed into according to grey scale mapping table）
for(i = 0; i < lHeight; i++)
{ for(j = 0; j < lWidth; j++)
{ lpSrc = (unsigned char*)lpDIBBits + lLineBytes * (lHeight - 1 - i) + j;
*lpSrc = bMap[*lpSrc];
}
}
（LHeight and lWidth is respectively the height and width of entire image in procedure above section）
Step 2-2：Binaryzation：Realize binaryzation in the present invention to gray-scale map by the way of hard -threshold, image binaryzation flowchart program flow chart is as shown in Figure 4.A certain threshold value Threshold is set, the data of gray level image are divided into two parts using Threshold:Pixel group more than or equal to threshold value and the pixel group less than threshold value.For example, input gray level image function is f (x, y), output bianry image function is g (x, y), then
, the black and white distribution of entire image is determined by threshold value Threshold.It is emphasized that as characteristic body and background are separated into this target, we have then write and special go background function so that color background and word to be fundamentally divided into the bitmap of two width " independent mutually ", rather than the two is merely divided into two regions, this is different from our processing modes in place of traditional binaryzation thought.
Step 2-3：Go discrete noise：When a good article especially fermented during writing needing to choose one's words carefully, people often have every now and then by pen toward putting tactile custom on paper, this allows for being mixed into a certain amount of noise in the information that we preserve in real time, accurate lookup person's handwriting after this also just gives brings additional unfavorable factor, so denoising must be carried out first.
Basic thought is scanning entire image, and the pixel for being determined as discrete point is set into white.The nesting of two functions has been used in the operation for completing to go discrete noise, one function is void RemoveScatterNoise (HDIB hDIB), the function is outer layer functions, the function of mainly completing is to judge whether pixel is discrete point, if being then set to white, and DeleteScaterJudge () function is have invoked when carrying out discreteness judgement；Another function is bool DeleteScaterJudge (), the function employs recursive thought, investigate up and down and upper left, upper right, lower-left, the direction of bottom right eight, if black color dots, then call and itself carry out recurrence, due to itself being bool types, so return value is TRUE or FALSE, obtained in void RemoveScatterNoise (HDIB hDIB), if then pixels illustrated point is discrete point to FALSE, just white is set to, it is achieved thereby that removing discrete noise effect.
Step 2-4：Grads Sharp：The purpose of image sharpening processing is to become more fully apparent fuzzy image, and image sharpening typically has two methods：One kind is the differential method, and another is high-pass filter method.Grads Sharp of the present invention is exactly a kind of the more commonly used method in the differential method.
If image is
, definition
Point
The gradient vector at placeFor：
Thus formula can obtain such conclusion：The numerical value of gradient is exactly
The increased amount of unit distance institute on its maximum rate of change direction.
For discrete digital picture, above formula can be rewritten into： 
For convenience of calculation, it would however also be possible to employ approximate formula (1) below：
If directly using Grad
To represent image, even
=
, then have above-mentioned formula (1) visible：In its slow local value very little of image conversion（Corresponding to dark images）；And it is very big in the value that outline etc. converts faster place.Image is being apparent from after gradient algorithm so as to reach the purpose of sharpening.
Cause this unfavorable factor of dark images for the slow local gradient very little of image conversion, the solution that we use is to a threshold value
, if less than the threshold value
, then keep former ash angle value constant；If greater than or equal to threshold value, then it is entered as：
Step 2-5：Refinement：Refine and refer to seek the central bone partial routine of a sub-picture, central bone portion is to describe one of key character of image geometry and topological property, processed image, which refine, to be helped to protrude features of shape and reduce the information content of redundancy.
The characteristic that the thinning algorithm that the present invention is used has：One in piece image
Region, marks title P1, P2 ... P8, wherein P1 to be located at center to each point.If P1=1（That is stain）If, below 4 conditions meet simultaneously, delete P1（P1=0）.①2
NZ(P1)
6；②
=1；3. P2*P4*P8=0 or 
1；4. P2*P4*P6=0 or
1；The step for being repeated for each point in image, untill all point all unsuppressible-suppressions.
Step 3：The Shape Feature Extraction of handwriting data：The shape facility of handwritten Chinese character includes gradient, centre-of gravity shift, character stretching, extension etc., and these feature explicit physical meanings reflect a part of writing style of writing people.Under normal circumstances, this feature can be described with square.
For a width
Character picture
, its p+q rank square is defined as：
Wherein,
, the physical significance of each rank square is described as follows：
Represent the range of extension of image in the horizontal direction；
Represent the range of extension of image in vertical direction；
Represent the gradient of image（>0, which represents that image is upper to the left, tilts,
<0, which represents that image is upper to the right, tilts）；
Represent the centre-of gravity shift degree of image in the horizontal direction（
>0 represents that center of gravity is to the left,<0 represents that center of gravity is to the right）；
Represent the centre-of gravity shift degree of image in vertical direction（>0 represents that center of gravity is on the upper side,<0 represents that center of gravity is on the lower side）；
Represent the balance degree that image level stretches（
>0 represents that the horizontal stretching degree of image bottom is bigger than image top,
<0 represents that the horizontal stretching degree of image bottom is smaller than image top）；
Represent the balance degree of image vertical stretching（
>0 represents that the vertical stretching degree of image right part is bigger than image left part,
<0 represents that the vertical stretching degree of image right part is smaller than image left part）.
The normalized shape facility of some following can be exported by central moment computed above：
A) length-width ratio：
B) font gradient：
；Wherein
C) elongation degree：
；Wherein
Calculating formula it is as follows：
D) range of extension：
；Wherein size represents the size of character pattern, i.e.,
The central moment of three ranks is because its value may be just or be negative, in fact, they are just and by negative two parts to constitute all by basic function.
The normalization characteristic extracted by third central moment is as follows：
1. horizontal-shift degree：
2. vertical shift degree：
3. horizontal stretching equilibrium degree：
4. vertical stretching equilibrium degree：
For corresponding tagged word, their characteristic vector is obtained
, the distance of definition is as follows：
Step 4：Handwriting data are recognized：The half of word height by origin of word barycenter, after normalization is radius, sets up polar coordinate system, fan-shaped average piecemeal is carried out to whole circumference, block number n is variable.Whether the final word for judging the whether similar Chinese character currently write of two width bitmaps with occurring in the notes preserved in the past there is similitude to a certain extent to follow same criterion：Only when in the corresponding n blocks region of two width bitmaps every piece of area pixel point it is total when difference is respectively less than the threshold value pre-set two-by-two, just the two is considered as similar.And it is adjustable that n, which fixes as variable be equivalent to the precision setting of lookup,.Obviously, fuzzy search is considered it is considered that searching during n=12, and accurate lookup is then considered during n=360.Fig. 5 character font data identification process figures.After character font data identification is completed, computer shows the font recognized on screen.After font editor is completed, by software control, character font data can be sent to receiving station by wireless GSM network, realize real-time Transmission.
Claims (2)
1. a kind of hand-written handwriting automatic recognition system, including one piece by IntelThe EMB-4650 mainboards that the master control of Atom Z510P chips makes（1）, it is characterised in that the mainboard（1）The sub-interface of extension has LVDS interface, DVI interface, USB interface and com interface；These interfaces respectively with an AUO display screen（2）, a DVI-LVDS signal transducer（3）, a USB controller（4）And gsm module（5）It is connected；Thus DVI-LVDS signal transducers（3）With another AUO display screen（6）It is connected, USB controller（4）With touch-screen（7）It is connected；When user writes data on the touchscreen, these data are sent to EMB-4650 mainboards by USB controller, while EMB-4650 mainboards transfer data to display screen, realize display in real time；Algorithm process is identified to these data in EMB-4650 mainboards simultaneously, and the result after processing, which is then sent to display screen, shows, reaches easy human-computer interaction real-time operation effect；The character font data editted can be sent to gsm module by com port, realize the real-time transmission of data.
2. a kind of hand-written handwriting automatic identifying method, is identified, it is characterised in that concrete operation step is as follows using the hand-written handwriting automatic recognition system described in claims 1：
Step 1：The input of handwriting data：When the touch screen is touched, touch-screen in the form of image data stores these data, is read for main frame;
Step 2：The pretreatment of handwriting data：Mainboard 1 is from touch-screen（7）The view data deposited is read in, mainboard is input to（1）In register；
Step 2-1:Touch-screen（7）The color bitmap of data 256 of input turns gray-scale map according to grey scale mapping table method；
Step 2-2：Binaryzation：Binaryzation is realized to gray-scale map by the way of hard -threshold；
Step 2-3：Go discrete noise：Denoising is carried out to the noise that touch is mixed into；
Step 2-4：Grads Sharp：Image sharpening processing is carried out using the differential method：
If image is
, definition
Point
The gradient vector at place
For：
Thus formula obtains conclusion：The numerical value of gradient is
The increased amount of unit distance institute on its maximum rate of change direction；
For discrete digital picture, above formula is rewritten into：
For convenience of calculation, it would however also be possible to employ approximate formula (1) below：
If directly using GradTo represent image, even
=
, then have above-mentioned formula (1) visible：In its slow local value very little of image conversion（Corresponding to dark images）；And it is very big in the value that outline etc. converts faster place；Image is being apparent from after gradient algorithm so as to reach the purpose of sharpening；
Cause this unfavorable factor of dark images for the slow local gradient very little of image conversion, solution is to a threshold value
, if less than the threshold value
, then keep former ash angle value constant；If greater than or equal to threshold value, then it is entered as
：
Step 2-5：Refinement：Finger seeks the central bone partial routine of a sub-picture, and central bone portion is to describe one of key character of image geometry and topological property, and refinement is carried out to processed image to be helped to protrude features of shape and reduce amount of redundant information；
The characteristic that the thinning algorithm of use has：One in piece image
Region, marks title P1, P2 ... P8, wherein P1 to be located at center to each point；If P1=1, i.e. stain, if below 4 conditions meet simultaneously, delete P1（P1=0）；①2NZ(P1)
6；②=1；3. P2*P4*P8=0 or
1；4. P2*P4*P6=0 or
1；The step for being repeated for each point in image, untill all point all unsuppressible-suppressions；
Step 3：The Shape Feature Extraction of handwriting data：The shape facility of handwritten Chinese character includes gradient, centre-of gravity shift, character and stretched, and this feature is described with square；
For a width
Character picture, its p+q rank square is defined as：
Wherein,, the physical significance of each rank square is described as follows：
Represent the range of extension of image in the horizontal direction；
Represent the range of extension of image in vertical direction；Represent the gradient of image
>0, which represents that image is upper to the left, tilts,
<0, which represents that image is upper to the right, tilts；
Represent the centre-of gravity shift degree of image in the horizontal direction
>0 represents that center of gravity is to the left,<0 represents that center of gravity is to the right；
Represent the centre-of gravity shift degree of image in vertical direction
>0 represents that center of gravity is on the upper side,
<0 represents that center of gravity is on the lower side；
The balance degree that image level stretches is represented,
>0 represents that the horizontal stretching degree of image bottom is bigger than image top,
<0 represents that the horizontal stretching degree of image bottom is smaller than image top；
The balance degree of image vertical stretching is represented,>0 represents that the vertical stretching degree of image right part is bigger than image left part,
<0 represents that the vertical stretching degree of image right part is smaller than image left part；
By some normalized shape facilities that central moment export computed above is following：
Length-width ratio： 
Font gradient：
；Wherein
Elongation degree：；Wherein
Calculating formula it is as follows：
Range of extension：
；Wherein size represents the size of character pattern, i.e.,
The central moment of three ranks is because its value may be just or be negative, in fact, they are just and by negative two parts to constitute all by basic function；
1. horizontal-shift degree： 
2. vertical shift degree：
3. horizontal stretching equilibrium degree：
4. vertical stretching equilibrium degree：
For corresponding tagged word, their characteristic vector is obtained
, the distance of definition is as follows：
Step 4：Handwriting data are recognized：The half of word height by origin of word barycenter, after normalization is radius, sets up polar coordinate system, fan-shaped average piecemeal is carried out to whole circumference, block number n is variable；
Whether the final word for judging the whether similar Chinese character currently write of two width bitmaps with occurring in the notes preserved in the past there is similitude to a certain extent to follow same criterion：Only when in the corresponding n blocks region of two width bitmaps every piece of area pixel point it is total when difference is respectively less than the threshold value pre-set two-by-two, just the two is considered as similar；And it is adjustable that n, which fixes as variable be equivalent to the precision setting of lookup,；
Obviously, fuzzy search is considered it is considered that searching during n=12, and accurate lookup is then considered during n=360；After character font data identification is completed, computer shows the font recognized on screen.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011103604419A CN102541347A (en) | 2011-11-15 | 2011-11-15 | Automatic identification system and method of handwriting Chinese character |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011103604419A CN102541347A (en) | 2011-11-15 | 2011-11-15 | Automatic identification system and method of handwriting Chinese character |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102541347A true CN102541347A (en) | 2012-07-04 |
Family
ID=46348375
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011103604419A Pending CN102541347A (en) | 2011-11-15 | 2011-11-15 | Automatic identification system and method of handwriting Chinese character |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102541347A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105094383A (en) * | 2014-05-13 | 2015-11-25 | 联想（新加坡）私人有限公司 | Smart pen pairing and connection |
| CN107918935A (en) * | 2017-12-05 | 2018-04-17 | 深圳市北电仪表有限公司 | The images match cutter device and cutting method of embedded system |
| WO2018103226A1 (en) * | 2016-12-07 | 2018-06-14 | 广州视源电子科技股份有限公司 | Method and device for generating handwriting format document on the basis of picture |
| CN110737364A (en) * | 2019-11-06 | 2020-01-31 | 广州炫视智能科技有限公司 | Control method for touch writing acceleration under android systems |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN85100085A (en) * | 1985-04-01 | 1986-07-16 | 清华大学 | Recognition method of printed Chinese character recognition device |
| CN1128423C (en) * | 1999-04-21 | 2003-11-19 | 中国科学院自动化研究所 | Handwriting identifying method based on grain analysis |
-
2011
- 2011-11-15 CN CN2011103604419A patent/CN102541347A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN85100085A (en) * | 1985-04-01 | 1986-07-16 | 清华大学 | Recognition method of printed Chinese character recognition device |
| CN1128423C (en) * | 1999-04-21 | 2003-11-19 | 中国科学院自动化研究所 | Handwriting identifying method based on grain analysis |
Non-Patent Citations (8)
| Title |
|---|
| MILAN SONKA等: "《图像处理、分析与机器视觉》", 31 January 2011, 清华大学出版社 * |
| 何斌等: "《Visual C++ 数字图像处理》", 30 April 2001, 人民邮电出版社 * |
| 唐伟成: "手写英文字符识别***", 《中国优秀硕士学位论文全文数据库（信息科技辑）》 * |
| 李小泉: "自动检测***关键技术研究", 《武汉理工大学硕士学位论文》 * |
| 杨阳等: "基于图像分析的滚动轴承表面缺陷识别技术研究", 《华东交通大学学报》 * |
| 陈维克等: "印鉴识别方法和技术研究", 《中国水运》 * |
| 黄佳森: "实现个人手写笔体检索的全新算法——质心圆周分块算法", 《信息***工程》 * |
| 黄佳森等: "基于Intel Menlow平台EMB－4650开发板的即写即发、个人双屏触控PDA实现", 《信息与电脑》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105094383A (en) * | 2014-05-13 | 2015-11-25 | 联想（新加坡）私人有限公司 | Smart pen pairing and connection |
| CN105094383B (en) * | 2014-05-13 | 2021-05-11 | 联想（新加坡）私人有限公司 | Smart pen pairing and connection |
| WO2018103226A1 (en) * | 2016-12-07 | 2018-06-14 | 广州视源电子科技股份有限公司 | Method and device for generating handwriting format document on the basis of picture |
| CN107918935A (en) * | 2017-12-05 | 2018-04-17 | 深圳市北电仪表有限公司 | The images match cutter device and cutting method of embedded system |
| CN107918935B (en) * | 2017-12-05 | 2022-03-08 | 深圳市北电仪表有限公司 | Image matching cutting device and cutting method of embedded system |
| CN110737364A (en) * | 2019-11-06 | 2020-01-31 | 广州炫视智能科技有限公司 | Control method for touch writing acceleration under android systems |
| CN110737364B (en) * | 2019-11-06 | 2023-12-12 | 广州炫视智能科技有限公司 | Control method for touch writing acceleration under android system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108647681B (en) | A kind of English text detection method with text orientation correction | |
| Zhang et al. | Text extraction from natural scene image: A survey | |
| US5933526A (en) | Handwritten keyboardless entry computer system | |
| US4972496A (en) | Handwritten keyboardless entry computer system | |
| US9052755B2 (en) | Overlapped handwriting input method | |
| CN108958636B (en) | Intelligent copybook device and copybook writing detection method | |
| CN103093196A (en) | Character interactive input and recognition method based on gestures | |
| TWI461962B (en) | Computing device for performing functions of multi-touch finger gesture and method of the same | |
| CN104361313A (en) | Gesture recognition method based on multi-kernel learning heterogeneous feature fusion | |
| CN104408449A (en) | Intelligent mobile terminal scene character processing method | |
| CN105205449A (en) | Sign language recognition method based on deep learning | |
| CN102541347A (en) | Automatic identification system and method of handwriting Chinese character | |
| Holanda et al. | Development of OCR system on android platforms to aid reading with a refreshable braille display in real time | |
| CN107992483A (en) | The method, apparatus and electronic equipment of translation are given directions for gesture | |
| CN109359566A (en) | The gesture identification method of hierarchical classification is carried out using finger characteristic | |
| CN102957786A (en) | Mobile phone system with functions of drawing board and method for applying mobile phone system | |
| CN107992867A (en) | The method, apparatus and electronic equipment of translation are given directions for gesture | |
| CN201654804U (en) | Character recognizer | |
| Li et al. | Interest point detection using imbalance oriented selection | |
| CN105308535A (en) | Hands-free assistance | |
| CN201532625U (en) | Electromagnetic board capable of displaying input trace in real time | |
| CN102033633A (en) | Coordinate positioning device for displaying input locus in real time | |
| CN105138169A (en) | Touch panel control device based on gesture recognition | |
| Li et al. | Research and implementation of dynamic gesture recognition system based on ZYNQ | |
| CN206080520U (en) | Intelligence touch -type vital capacity tester |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20120704 |