CN103699890B - A kind of scanning input device and scan input method - Google Patents
A kind of scanning input device and scan input method Download PDFInfo
- Publication number
- CN103699890B CN103699890B CN201210365732.1A CN201210365732A CN103699890B CN 103699890 B CN103699890 B CN 103699890B CN 201210365732 A CN201210365732 A CN 201210365732A CN 103699890 B CN103699890 B CN 103699890B
- Authority
- CN
- China
- Prior art keywords
- character
- input
- image
- control command
- scanning
- 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.)
- Active
Links
Landscapes
- Character Input (AREA)
- Character Discrimination (AREA)
Abstract
The invention discloses a kind of scanning input device, including：Imaging sensor, for the character focal imaging on scanning medium, generating image, being sent into controller and handled；Controller, judges that scanned image is character or control command for handling described image, and according to scan mode or recognition result, performs the order if being control command if what is inputted, otherwise preserve character.The present invention realizes control command and the accessible input of character, the random inquiry of word, and realize calculator function by the way that the information such as character or symbol are printed on portable simulating keyboard.Control command and the accessible input of character, make equipment break away from keyboard, reduce cost, reduce volume.Meanwhile compared with handwriting input, there is higher input efficiency, save the operating time of user.
Description
Technical field
The present invention relates to field of optical character recognition, and in particular to a kind of scanning input device and scan input method, lead to
Overscanning is controlled order and the input of character.
Background technology
In existing computer, portable equipment, main input mode is button, touch-screen, hand-written etc..Key-press input side
Formula needs hardware supported with touch-screen input mode, while increasing cost, increases the volume of equipment.Handwriting input mode is benefited from
The hand-written rate limitation in family, causes input efficiency relatively low.
The appearance of optical character identification (Optical Character Recognition, OCR) technology, for extensive text
Rapidly inputting for this is provided convenience so that paper document switchs to Electronic saving mode and is possibly realized.
Using software identification technology, the patent document that number of patent application is 200620077723.2 discloses one kind word
Scanning or camera technique directly consult electronic dictionary.The electronic dictionary is inputted built-in scanning using digital translative mode
Word word image information captured by device, is converted into digital information, the information with being stored in character library contrasts, then directly
Connect and find out the word, and translated.
The patent document that number of patent application is 200620123568 discloses a kind of portable image recognition apparatus.The dress
After the imaging sensor put obtains image, identification converter interception first word, by OCR conversions and dictionary lookup, is completed
One process for taking word-identification-translation-display.
Scheme realizes fast inputting words disclosed in above patent document, the papery text for being printed with word
Shelves, image is gathered using image capture device, and is realized image recognition and looked into word technology.
Although however, image recognition used by patent document 200620123568 and patent document 200620077723.2
Technology can realize fast inputting words, but the technology typing and can only search existing word in paper document, for
The word being stored in complex dielectrics, the character or hand-written character being such as imprinted on the image of complex background, discrimination be not high.Separately
Outside, on blackboard, on billboard or character is given an oral account, can not be inputted by scanning, typing can not be completed and look into word work, therefore
The technology can not meet the diversified typing demand of user.
And in above patent document, by performing key, confirm scanning, during result mistake of serving as interpreter, selected with lower key
A correct word is selected to correct.These buttons support increase equipment volume, make device appearance not succinct enough, operational motion is numerous
It is trivial, reduce user friendly.
The content of the invention
The invention discloses a kind of scanning input device, it can be achieved that the scanning of control command and character inputs, and realize
The switching of character and control command, equipment appearance is succinct, and operation is more smooth, enhances user friendly, scanning input dress
Put and carry simulating keyboard, scanning input device can realize the Freely input of character, meet the diversified demand of user.
Scanning input device proposed by the present invention includes：Imaging sensor, for scanning medium on control command or
Character focal imaging, generates image, is sent into controller and is handled, and being made on the scanning medium has control command and character
At least one, the control command scanned on medium or character are formed by hand-written or printing；Controller, for described image
Handled, and judge that scanned image is character or control command according to scan mode or recognition result, if input
It is that control command then performs the order, otherwise preserves character information.
Wherein, which further comprises trigger device, light path device and optical lens, wherein trigger device
For when scanning input device contact scanning medium is scanned, producing working signal, to control light path device to form light path,
Light path produce the scanned dieletric reflection of light after, then through optical lens converge in imaging sensor be used for imaging sensor into
Picture.
Wherein, which further comprises：Image processing unit, for carrying out image preprocessing to described image；Know
Other engine, is identified the image after image processing unit is handled, and recognition result is the image of input；Judging unit, is used
According to multiple image displacement come judge scanning input be control command or character or by push button signalling judge input
Be control command or character or according to scan input picture recognition result judge input be control command or word
Symbol, if judge input is control command, performs the order of this image representative, and identification knot is preserved if being character if what is inputted
Fruit.
The character, which includes numeral, letter or word, the control command, includes symbol or pattern, and control command is not limited to
Terminate order, delete command, phonetic order, look into word translator command, return command, exit command, language switching command, moving to left life
Make, move up order, moving down order and/or shift-right command；The control command made on the scanning medium and at least the one of character
Kind forms keyboard layout.
The present invention also proposes a kind of scan input method, including：Step 1, to the control command or character on scanning medium
Focal imaging, generates image, is sent into controller and is handled, and being made on the scanning medium has control command and character at least
One kind, the control command scanned on medium or character are formed by hand-written or printing；Step 2, described image is handled,
And judge that scanned image is character or control command according to scan mode or recognition result, if input is control command
The order is then performed, otherwise preserves character.
Wherein step 1 further comprises inputting character and control command according to one kind in following three kinds of modes：First
Kind mode：When needing input control order, scanning input device is placed in scanning medium control command, extruding triggering dress
Put, scanning input device transfixion, acquisition control order, when needing to input character, scanning input device is placed in scanning
On medium character, trigger device is extruded, motion scan input unit gathers input character；The second way：When needing to input
During control command, by extruding trigger device, while a certain button of scanning input device is pinned to scan input control order,
And then not by the button when inputting character；The third mode：When needing input control order, collection represents control command
For specific image, it is necessary to when inputting character, then collection is different from the image of the specific image.
Wherein step 2 further comprise according in following three kinds of modes it is a kind of come judge input be character or control
Order：First way：According to multiple image displacement come judge scan input be control command or character；The second way：
What it is by push button signalling judgement input is control command or character；The third mode：According to the identification to scanning input picture
As a result judge input is control command or character, if judging, input is control command, performs the order of this image representative,
Recognition result is preserved if being character if input.
The present invention also proposes another scanning input device, which includes：Scanning element, for scanning medium
Control command or character focal imaging, generate image, be sent into controller handled, it is described scanning medium on make have control
Order and at least one of character, the control command scanned on medium or character are formed by hand-written or printing；Image procossing
Unit, handles described image, and judges that scanned image is character or control according to scan mode or recognition result
System order, performs the order if being control command if what is inputted, otherwise preserves character.
Wherein described scanning element inputs character and control command according to one kind in following three kinds of modes：The first side
Formula：When needing input control order, scanning input device is placed in scanning medium control command, trigger device is extruded, sweeps
Input unit transfixion is retouched, acquisition control order, when needing to input character, scanning medium word is placed in by scanning input device
Fu Shang, extrudes trigger device, and motion scan input unit gathers input character；The second way：Ordered when needing input control
When making, by extruding trigger device, while a button of scanning input device is pinned to scan input control order, and input word
The button is not pressed during symbol then；The third mode：When needing input control order, collection represents the specific pattern of control command
As, it is necessary to when inputting character, then collection is different from the image of the specific image；
Described image processing unit according in following three kinds of modes it is a kind of come judge input be character or control order
Order：First way：According to multiple image displacement come judge scan input be control command or character；The second way：It is logical
Cross push button signalling judgement input is control command or character；The third mode：According to the identification knot to scanning input picture
What fruit judged input is control command or character, if judging, input is control command, performs the order of this image representative, if
Input is that character then preserves recognition result.
The present invention can be carried on scanning medium by the way that the information such as character or symbol are printed on, and mould is formed on the scanning medium
Intend keyboard, or writing needs to gather the content of input on scanning medium, scanning input device needs to input by scanning
Character or control command, realize the accessible input of control command and character.It is thereby achieved that the random inquiry of word,
And realize calculator function.Control command and the accessible input of character, make equipment break away from keyboard, reduce cost, tailored appearance,
Reduce volume.Meanwhile compared with handwriting input, there is higher input efficiency, save the operating time of user.
Brief description of the drawings
Fig. 1 is the structure diagram of scanning input device of the present invention；
Fig. 2 a are the simulating keyboard schematic diagram for including letter and symbol of the present invention；
Fig. 2 b are the another simulating keyboard schematic diagram including letter and symbol of the present invention；
Fig. 2 c are another simulating keyboard schematic diagram for including numbers and symbols of the present invention；
Fig. 3 is scan input method flow chart of the present invention；
Fig. 4 is the flow chart that word and phrase are inputted using the scan input method of the present invention；
Fig. 5 is the flow chart calculated using the scan input method of the present invention.
Embodiment
For the object, technical solutions and advantages of the present invention are more clearly understood, below in conjunction with specific embodiment, and reference
Attached drawing, the present invention is described in more detail.
The present invention provides a kind of scanning input device and scan input method, it can be achieved that control command input and character are defeated
Enter, which can be equipped with a simulating keyboard, and can scan by different way on simulating keyboard or other scannings
Character or symbol on medium, by image recognition technology, realize control command input and the free switching of character input, so that
Realize that the on-keyboard of character is rapidly input, not by keyboard with hand-written, the character in various media is inputted, realizes no barrier
Hinder input.
Fig. 1 is the functional structure chart of scanning input device of the present invention.With reference to Fig. 1, which includes housing 1,
Controller 2, imaging sensor 3, display screen 4, optical lens 5, perforate 6, trigger device 7, light path device 8.
Controller 2 is arranged on inside housing 1, is preferably embeded processor, for controlling textual scan to input, into style of writing
The processing such as word identification.Imaging sensor 3 is arranged on inside housing 1, and is connected with controller 2, for the text on scanning medium
Word focal imaging, generates image, is sent into controller 2 and carries out the image preprocessings such as image enhancement, gray proces or image binaryzation
And identifying processing, the handling result of controller 2 show that on the display screen 4 imaging sensor 3 can obtain simulating keyboard or other
Scan the image of the character or control command on medium.A perforate 6 is opened up with 3 corresponding position of imaging sensor in shell 1,
So that imaging sensor 3 obtains the character on simulating keyboard or scanning data through perforate 6.Trigger device 7 is used for defeated when scanning
When entering device contact scanning medium and being scanned, working signal is produced, to control light path device 8 to form light path, what light path produced
After the scanned dieletric reflection of light, then directly converge in imaging sensor 3 through optical lens 5 and be imaged for imaging sensor 3.Light
Road device 8 is preferably using light path illumination LED.
Above-mentioned simulating keyboard can be any portable recognizable medium, such as the packaging leather sheath or papery material of this equipment
Material etc., have it is small, the characteristics of being convenient for carrying.Numeral, letter, symbol are included but are not limited on simulating keyboard, can also
It is other characters, control command or other figures.In addition simulating keyboard can be any portable recognizable medium, such as originally
The packaging leather sheath or paper material of equipment.Character or symbol on other above-mentioned scanning media can be by users in recognizable medium
Upper writing is formed.
Fig. 2 a are a kind of schematic diagram of simulating keyboard of the present invention.As shown in Figure 2 a, which includes letter and symbol.
Wherein "=" is about set to end mark, and symbol " ← " is about set to delete operation, and blank image is about set to space.
Fig. 2 b are the schematic diagram of another simulating keyboard of the invention.As shown in Figure 2 b, the simulating keyboard include letter and
Symbol, whereinAbout it is set to phonetic order,About it is set to and looks into word translator command,About it is set to delete command,About it is set to return command, " ← ", which is about set to, to move to left, and " ↑ ", which is about set to, to be moved up, and " ↓ " agreement moves down, and " → ", which is about set to, to move to right.
Scanning input device can come input control order and character according to any of following three mode：The first side
Formula：When needing input control order, scanning input device is placed on simulating keyboard or other scanning medium control commands, is led to
Extruding trigger device 7 is crossed, opens LED illumination, optical lens 5 works, and scanning input device transfixion, acquisition control order is simultaneously
Controller 2 is sent into by imaging sensor, scanning input device is placed in simulation by scanning input device when needing to input character
On keyboard or other scanning media, by extruding trigger device 7, LED illumination is opened, optical lens 5 works, motion scan dress
Put, collection input character is simultaneously sent into controller 2 by imaging sensor.The second way：When needing input control order, move
Dynamic scanning input device pins certain button at the same time, motion scan input unit during character input, not button.The third side
Formula：When needing input control order, collection represents the specific image of control command, it is necessary to when inputting character, then collection is different
In the image of the specific image, in this fashion, control command and character are all the sides by motion scan input unit
Formula inputs, or is inputted by way of click.
Controller 2 further comprises image processing unit 201, identification engine 202 and judging unit 203.
Image processing unit 201 carries out image preprocessing, including image enhancement, ash to the character and control command that are scanned
Degree processing and binary conversion treatment etc..Described image enhancing is to add some information or conversion number to original image by certain means
According to selectively protruding in image some unwanted features in feature interested or suppression (cover) image, make image
Match with eye response characteristic.Described image binaryzation is exactly that the gray value of the pixel on image is arranged to 0 or 1,
Whole image is exactly showed into obvious black and white effect.The binaryzation of image is conducive to the further processing highlighted, makes image
Become simple, and data volume reduces, and can highlight the profile of target interested.
The image after the processing of image processing unit 201 is identified in identification engine 202.Identify that engine 202 is further
Including OCR engine and grader identification engine.
If scanning input device is according to above-mentioned the first and the second way input character and control command, judge
Unit 203 can directly according to multiple image displacement and push button signalling come judge input be character or control command, if defeated
What is entered is character, then calls OCR engine identification, if control command, then directly invokes grader identification.
If scanning input device is to input character and control command according to the third above-mentioned mode, engine 202 is identified
Call OCR engine to be identified first, recognition result confidence level is judged according to identification informations such as identification distances, if recognition result is not
It is credible, then judging unit 203 judge input for customized control command, which is User Defined
Control command, such asDeng, there is no these control commands in standard Internal Code storehouse, need to continue grader identification, controlled
The recognition result of order.If OCR recognition results are credible, it is divided into two kinds of situations, one for input is character, and two are for input
The control command that can be found in standard Internal Code storehouse, such as ' ← ', the present invention carry out the implication of such a control command in advance
Agreement, if being identified as such control command, corresponding control command is performed according to the implication of agreement.
Bianry image is identified in OCR engine, mainly includes：Character extracts, and character feature extraction, feature compares, and knows
Other result post processing.The character extraction refers to extract image intermediate character region, filters out existing above and below the character
The blank column of other black pixels, blank line and the character or so.The feature extraction relatively refers to calculate above-mentioned character with feature
Feature, compared with the feature templates in character library, obtain feature most similar character as recognition result.The recognition result
Post processing refers to handle above-mentioned recognition result according to use environment, changes unreasonable result.
Grader identification refers to regard the feature of control command image as classification foundation, the different recognition classifier of correspondence,
The characteristics of image needed is also different, and recognition classifier compares this characteristics of image using the prior art and identified with grader in engine in advance
The distance between feature of each control command standard form first preserved, selects feature with this characteristics of image at a distance of nearest standard
Recognition result of the corresponding control command of template as this image.
Fig. 3 is the flow chart of scan input method of the present invention.With reference to Fig. 3, this method includes following steps：
Step 1, the image of the character or control command on various media or simulating keyboard is obtained using scanning input device.
In the step, character and control command are inputted according to one kind in following three kinds of modes：
First way：When needing input control order, scanning input device is placed in scanning medium control command,
Extrude trigger device, scanning input device transfixion, acquisition control order, when needing to input character, by scanning input dress
It is placed on scanning medium character, extrudes trigger device, motion scan input unit gathers input character.
The second way：When needing input control order, by extruding trigger device, while scanning input device is pinned
A button to scan input control order, and input character when then not by the button.
The third mode：When needing input control order, collection represents the specific image of control command, it is necessary to input word
Fu Shi, then collection is different from the image of the specific image, and in this fashion, control command and character are swept by movement
The mode for retouching input unit inputs, or is inputted by way of click.
Gather the character of input or control command is sent into the processing unit 201 of controller 2 by imaging sensor 3.
Step 2, image is carried out to the character and control command that are scanned using the image processing unit 201 in controller 2
Pretreatment.Described image pretreatment includes image enhancement, gray proces or binaryzation etc..
Described image enhancing is to add some information or conversion data to original image by certain means, is selectively protruded
Some unwanted features in feature interested or suppression (cover) image, make image and eye response characteristic phase in image
Matching.
Described image gray proces are that gray level image is analyzed and handled.Gray level image reflects the bright of entire image
The distribution of grade is spent, the excursion of each pixel is 0-255.Gray proces include generation grey level histogram, gray-level correction,
Extract the processing such as characteristics of image, image sharpening.
Described image binaryzation is exactly that the gray value of the pixel on image is arranged to 0 or 1, that is, by whole image
Show obvious black and white effect.The binaryzation of image is conducive to the further processing highlighted, image is become simple, and number
Reduce according to amount, the profile of target interested can be highlighted.
Step 3, treated image is identified using the identification engine 202 in controller.
Identification engine 202 further comprises OCR engine and grader identification device.If step 1 selection the first and
The second way inputs character and control command, then judging unit 203 can be directly according to multiple image displacement and push button signalling
It is character or control command judge input, if input is character, calls OCR engine identification, if control life
Order, then directly invoke grader identification.
But if the third input mode is selected to carry out input character and control command in step 1, then in the step 3
In, after being identified first with OCR technique, recognition result confidence level is judged according to identification informations such as identification distances, if recognition result is not
Credible, then it is user-defined control command for customized control command, the customized control command to input, such as
Deng, there is no these control commands in standard Internal Code storehouse, need to continue grader identification, obtain the recognition result of control command.
If OCR recognition results are credible, it is divided into two kinds of situations, one for input is character, and two for input be in standard Internal Code storehouse
The control command that can be found, such as ' ← ', the present invention arranges the implication of such a control command in advance, if be identified as
Such control command, then perform corresponding control command according to the implication of agreement.
Bianry image is identified in OCR engine, mainly includes：Character extracts, and character feature extraction, feature compares, and knows
Other result post processing.The character extraction refers to extract image intermediate character region, filters out existing above and below the character
The blank column of other black pixels, blank line and the character or so.The feature extraction relatively refers to calculate above-mentioned character with feature
Feature, compared with the feature templates in character library, obtain feature most similar character as recognition result.The recognition result
Post processing refers to handle above-mentioned recognition result according to use environment, changes unreasonable result.
Grader identification refers to regard the feature of control command image as classification foundation, the different recognition classifier of correspondence,
The characteristics of image needed is also different, and recognition classifier compares this characteristics of image with being pre-saved in identification engine using the prior art
The distance between the feature of each control command standard form, select feature with this characteristics of image at a distance of nearest standard form pair
Recognition result of the control command answered as this image.
Step 4, corresponding action is performed according to the recognition result in step 3, this is performed if being control command if what is inputted
Order, otherwise preserves character information.
The method of the invention for being scanned input using simulating keyboard is further illustrated below by way of example,
In following example, scanning input device is illustrated by taking wand as an example.
1 ＞ of ＜ examples
Input of the first way in above-mentioned steps 1 into line character and control command is used in this example embodiment.Character is defeated
Fashionable motion scan pen；When control command inputs, wand is placed in above control command, and pen is static, gathers image.Controller
Thus judge input is character or control command.I.e. such as the first way of step 1, during wand motion scan, input
For character, during wand stationary acquisition image, the order in order to control of input.If input is control command, controller tune
Identified with grader, if input is character, controller calls OCR engine.Such as want looking up words BALL, then such as Fig. 4
It is shown, perform following steps：
Step 1, word " BALL " in motion scan pen scanning data., will scanning because inputting character " BALL "
Pen is placed in above scanning data, motion scan pen, gathers frame image sequence, picture frame sequence then is spliced into a width panorama sketch
Picture.
Step 2, image procossing and and identification are carried out to panoramic picture.Wherein image procossing includes image enhancement and image two
Value is handled.Image enhancement weakens noise using median filtering method.Median filtering method is a kind of nonlinear smoothing technology, it will
The gray value of each pixel is arranged to the intermediate value of all pixels point gray value in the point neighborhood window.Image binaryzation is adopted
With otsu methods, i.e. maximum variance between clusters.The method presses the gamma characteristic of image, divides the image into background and prospect two
Point.Inter-class variance between background and prospect is bigger, illustrates that the two-part difference for forming image is bigger, divides when part prospect is wrong
Being divided into prospect for background or background mistake all can cause two parts difference to diminish.Therefore, it is that the segmentation of inter-class variance maximum means
Misclassification probability is minimum.
Since wand is to be scanned input by mobile mode, controller can be according to two field picture displacement
Directly judge to be inputted is character, therefore calls OCR engine that the character of input is identified, OCR technique identifying processing
Process is to carry out character cutting first, obtains each character, then each character is identified.Character recognition process,
Each character intermediate character region is extracted first, is filled into character existing background pixel up and down, is then extracted character feature
And compared with template characteristic in character repertoire, find the nearest character of feature in character library makes as recognition result, last basis
Recognition result is adjusted with environment.
Assuming that the recognition result of scanning input device is " BALI ", last character identification mistake, can perform step 3
Modify, step 4 can also be performed and directly look into word, because dictionary can provide most similar word list, by selecting up and down
Obtain correct result.
Step 3, word last character is changed, then looks into word, step is as follows：
Step 3.1, cursor is now placed in character " I " place in display screen, and wand is placed in simulating keyboardTop,
Pen is static, gathers image and carries out image procossing and grader identifying processing.Controller detects that this time input is ordered in order to control,
The order of this input picture representative is performed, that is, deletes character " I ", in the step, since wand remains static collection figure
Picture, therefore it is judged as control command, so being identified for control command come calling classification device.
Step 3.2, wand is placed in above simulating keyboard " L ", motion scan scan image simultaneously carries out image procossing
And OCR identifying processings.Controller detects that this time input is character, preserves recognition result " L ".Word " BALL " spelling at this time is just
Really.
Step 3.3, wand is placed onTop, pen is static, gathers image and carries out image procossing and grader knowledge
Manage in other places.Controller detects that this time input is ordered in order to control, and the identification of calling classification device, this input figure is performed according to recognition result
As the order represented, carry out looking into word, look into word the results show on a display screen.
Step 4, wand is placed onTop, pen is static, gathers image and carries out at image procossing and grader identification
Reason.Controller detects that this time input is ordered in order to control, is performed the order of this input picture representative, is carried out looking into word, word
" BALI " is not found, and dictionary provides most similar word list：BALL, BALT.
Step 5, wand is placed on " ↓ " top, pen is static, gathers image and carries out at image procossing and grader identification
Reason.Controller detects that this time input is ordered in order to control, is performed the order of this input picture representative, is selected word downwards
" BALL ", and provide translation result and show on a display screen.
2 ＞ of ＜ examples
Input of the second way in step 1 into line character and control command is used in this example embodiment.
When being inputted according to the second way in step 1, when wand movement is detected simultaneously by key press signal, represent
Input is control command；Wand is mobile but when not having key press signal, represent input is character.Therefore, control life
The input mode of order pins certain button on a shell while being motion scan pen, gather image；During character input, only movement is swept
Pen, not button.What controller 2 can directly judge input according to push button signalling is control command or character, if control
System order, then calling classification device identifies, if input is character, calls OCR engine.Such as want looking up words " lovely "
And " love ", step is as follows：
Step 1, word " lovely " in motion scan pen scanning data, the function of setting on wand shell be good for not by
During pressing, scanning input character is represented.Scanning process is identical with the scanning process in 1 ＞ of ＜ examples.Set on wand shell
Function key can decide whether it is the multiplexing for having button on the dedicated button of input control order or shell.
Step 2, image procossing and OCR technique identifying processing are carried out to panoramic picture.At image preprocessing and character recognition
Reason is identical with 1 ＞ of ＜ examples, and recognition result " lovely " is preserved after the completion of to be identified.
Step 3, wand is placed onTop, thumb pins certain button to motion scan pen at the same time, gathers image and carries out
Image procossing and grader identifying processing.Controller 2 detects that this time input is ordered in order to control, performs the representative of this input picture
Order, carries out looking into word, query result is shown on a display screen.
Step 5, cursor is now placed in character " y " place in display screen, and wand is placed in simulating keyboardTop, is moved
Thumb pins certain button to dynamic wand at the same time, gathers image and carries out image procossing and grader identifying processing.Controller detects
Ordered in order to control to this input, perform the order of this input picture representative, delete character " y ".
Step 6, same step 5 is operated, deletes character " l ".
Step 7, same step 3 is operated, display screen display " love " looks into word result.
3 ＞ of ＜ examples
Input of the third mode in step 1 into line character and control command is used in this example embodiment.
Wand motion scan character or control command, after controller is identified with OCR technique first, according to identification distance etc.
Identification information judges recognition result confidence level, if recognition result is insincere, input for customized control command, need to continue
Grader identification is carried out, obtains the recognition result of control command.If OCR recognition results are credible, it is divided into two kinds of situations, one is defeated
What is entered is character, and two for input be the control command that can be found in standard Internal Code storehouse, if being identified as control command,
Corresponding control command is performed according to the implication of agreement.
Such as phrase look at are inquired about by wand, comprise the following steps that：
Step 1, wand is placed in simulating keyboard above alphabetical " l ", the image of camera collection alphabetical " l " simultaneously passes through
Imaging sensor is transmitted to processing unit and carries out image preprocessing, and processing result image is sent to OCR identification devices to this image
Handled and identified, and recognition result is preserved.
In the step, simulating keyboard is using the keyboard shown in Fig. 2 a.Image preprocessing includes image enhancement and image two-value
Change is handled, and image binaryzation uses otsu methods, i.e. maximum variance between clusters.The method presses the gamma characteristic of image, by image
It is divided into background and prospect two parts.Inter-class variance between background and prospect is bigger, illustrates the two-part difference for forming image
Bigger, when part, prospect mistake is divided into background or background mistake is divided into prospect and all two parts difference can be caused to diminish.Therefore, it is side between class
Poor maximum segmentation means misclassification probability minimum.
Image enhancement weakens noise using median filtering method.Median filtering method is a kind of nonlinear smoothing technology, it will
The gray value of each pixel is arranged to the intermediate value of all pixels point gray value in the point neighborhood window.
Step 2, the mode according to step 1 is gathered and inputted " o " " o " " k " respectively.
Step 3, space is inputted.It is space during input blank image of making an appointment, wand is placed in simulating keyboard blank
Place, gathers blank image, this character is preserved as space.
Step 4, respective image input " a " " t " is then gathered in the way of step 1.
Step 5, end of input character.Agreement "=" is used as end mark, and wand is placed on simulating keyboard "=" character
Side, takes character and carries out image procossing and optical character recognition process, recognition result is determined as end mark, then end of input.
Step 6, look into word and show translation result on a display screen.
Wherein termination character is not preserved, and existing input character is formed to the word or phrase of a completion, searches word
Allusion quotation is translated.So as to obtain the information such as the explanation of phrase look at, finally by phrase and explanation presentation of information in display screen
On.
If wand, it is necessary to remove, can be placed in simulating keyboard " ← " top by character input mistake, " ← " word is inputted
Symbol, carries out character deletion.
4 ＞ of ＜ examples
The present invention cannot be only used for word enquiring, can also realize calculator function, as illustrated in the flow chart of figure 5.
Fig. 2 c are to be printed on numeral with calculating the simulating keyboard of symbol.Calculate symbol and be not limited to institute's lettering symbol on simulating keyboard,
Also other complex calculation symbols can be included.Such as need to calculate (39+47) * 62.Scanning input device is sequentially placed into analog key
" (" " 3 " " 9 " "+" ... is waited above character, is gathered its image and is handled and identified in disk.As formula, knot is recently entered
Beam marks "=", to determine that input is completed.Finally calculated, as a result 5332, the results are shown on display screen.Wherein
Processing and recognition methods are identical with method therefor in 1 ＞ of ＜ examples.
Particular embodiments described above, has carried out the purpose of the present invention, technical solution and beneficial effect further in detail
Describe in detail it is bright, it should be understood that the foregoing is merely the present invention specific embodiment, be not intended to limit the invention, it is all
Within the spirit and principles in the present invention, any modification, equivalent substitution, improvement and etc. done, should be included in the guarantor of the present invention
Within the scope of shield.
Claims (10)
1. a kind of scanning input device, including：
Imaging sensor, for scanning medium on control command or character focal imaging, generate image, be sent into controller into
Row is handled, and being made on the scanning medium has control command and at least one of character, the control command on the scanning medium
Or character is formed by hand-written or printing；
Controller, for handling described image, judged according to multiple image displacement scanned image be character also
It is control command, performs the order if being control command if what is inputted, otherwise preserve character.
2. device according to claim 1, it is characterised in that the scanning input device further comprises trigger device, light
Road device and optical lens, wherein trigger device are used for when scanning input device contact scanning medium is scanned, and produce work
Make signal, to control light path device to form light path, after the scanned dieletric reflection of light that light path produces, then converge to through optical lens
It is used to imaging sensor in imaging sensor be imaged.
3. the apparatus of claim 2, it is characterised in that controller further comprises：
Image processing unit, for carrying out image preprocessing to described image；
Identify engine, the image after image processing unit is handled is identified, character or control of the recognition result for input
Order；
Judging unit, for according to multiple image displacement come judge scanning input be control command or character, if judging defeated
What is entered is control command, then performs the order of this image representative, and recognition result is preserved if being character if what is inputted.
4. device according to claim 3, it is characterised in that the character includes numeral, letter or word, the control
Order include symbol or pattern, and be not limited to terminate order, delete command, phonetic order, look into word translator command, return order
Make, exit command, shift-left command, move up order, move down order and/or shift-right command；The control made on the scanning medium
Order and at least one of character form keyboard layout.
5. a kind of scan input method, the method comprising the steps of：
Step 1, to the control command or character focal imaging on scanning medium, image is generated, controller is sent into and is handled, institute
State to make on scanning medium and have control command and at least one of character, the control command or character on the scanning medium are by hand
Write or print formation；
Step 2, described image is handled, and judges that scanned image is character or control according to multiple image displacement
System order, performs the order if being control command if what is inputted, otherwise preserves character.
6. according to the method described in claim 5, it is characterized in that, step 1 further comprises according in following three kinds of modes
One kind inputs character and control command：First way：When needing input control order, scanning input device is placed in and is swept
Retouch in medium control command, extrude trigger device, scanning input device transfixion, acquisition control order, when need input word
Fu Shi, scanning input device is placed on scanning medium character, extrudes trigger device, and motion scan input unit inputs to gather
Character；The second way：When needing input control order, by extruding trigger device, while scanning input device is pinned
Then not by the button when one button is to scan input control order, and input character；The third mode：When needing input control
During order, collection represents the specific image of control command, it is necessary to when inputting character, then collection is different from the figure of the specific image
Picture.
7. according to the method described in claim 6, it is characterized in that,
The character, which includes numeral, letter or word, the control command, includes symbol or pattern, and is not limited to terminate life
Make, delete command, phonetic order, look into word translator command, return command, exit command, shift-left command, move up order, move down life
Order and/or shift-right command；The control command and at least one of character made on the scanning medium forms keyboard layout.
8. a kind of scanning input device, which includes：
Scanning element, for the control command or character focal imaging on scanning medium, generating image, being sent into controller and carry out
Processing, making on the scanning medium has a control command and at least one of character, control command on the scanning medium or
Character is formed by hand-written or printing；
Image processing unit, handles described image, and judges that scanned image is word according to multiple image displacement
Symbol or control command, perform the order if being control command if what is inputted, otherwise preserve character information.
9. device according to claim 8, it is characterised in that the scanning element is according to one kind in following three kinds of modes
To input character and control command：First way：When needing input control order, scanning input device is placed in scanning and is situated between
In matter control command, extruding trigger device, scanning input device transfixion, acquisition control order, when needing to input character,
Scanning input device is placed on scanning medium character, extrudes trigger device, motion scan input unit gathers input character；
The second way：When needing input control order, by extruding trigger device, while a button of scanning input device is pinned
Then not by the button during scanning input control order, and input character；The third mode：When needing input control order
When, collection represents the specific image of control command, it is necessary to when inputting character, then collection is different from the image of the specific image.
10. device according to claim 9, it is characterised in that
The character, which includes numeral, letter or word, the control command, includes symbol or pattern, and is not limited to terminate life
Make, delete command, phonetic order, look into word translator command, return command, exit command, language switching command, shift-left command, on
Move order, move down order and/or shift-right command；The control command made on the scanning medium and at least one formation of character
Keyboard layout.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210365732.1A CN103699890B (en) | 2012-09-27 | 2012-09-27 | A kind of scanning input device and scan input method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210365732.1A CN103699890B (en) | 2012-09-27 | 2012-09-27 | A kind of scanning input device and scan input method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103699890A CN103699890A (en) | 2014-04-02 |
| CN103699890B true CN103699890B (en) | 2018-05-04 |
Family
ID=50361413
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210365732.1A Active CN103699890B (en) | 2012-09-27 | 2012-09-27 | A kind of scanning input device and scan input method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103699890B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104598903B (en) * | 2015-02-04 | 2018-11-06 | 浙江科技学院 | The adaptive character recognition device of object temperature and method |
| CN106412008A (en) * | 2016-08-26 | 2017-02-15 | 乐视控股（北京）有限公司 | Identifier correcting method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1998002792A2 (en) * | 1996-07-16 | 1998-01-22 | Wizcom Technologies Ltd. | Data processor input device |
| CN2733456Y (en) * | 2004-02-20 | 2005-10-12 | 杨学勤 | Portable intelligent translation pen |
| CN101444079A (en) * | 2004-12-06 | 2009-05-27 | 威思克技术有限公司 | Handheld text scanner |
| CN102073847A (en) * | 2010-12-24 | 2011-05-25 | 汉王科技股份有限公司 | Scanning device and scanning method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102135838A (en) * | 2011-05-05 | 2011-07-27 | 汉王科技股份有限公司 | Method and system for partitioned input of handwritten character string |
-
2012
- 2012-09-27 CN CN201210365732.1A patent/CN103699890B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1998002792A2 (en) * | 1996-07-16 | 1998-01-22 | Wizcom Technologies Ltd. | Data processor input device |
| CN2733456Y (en) * | 2004-02-20 | 2005-10-12 | 杨学勤 | Portable intelligent translation pen |
| CN101444079A (en) * | 2004-12-06 | 2009-05-27 | 威思克技术有限公司 | Handheld text scanner |
| CN102073847A (en) * | 2010-12-24 | 2011-05-25 | 汉王科技股份有限公司 | Scanning device and scanning method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103699890A (en) | 2014-04-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11182069B2 (en) | Managing real-time handwriting recognition | |
| US9934430B2 (en) | Multi-script handwriting recognition using a universal recognizer | |
| EP2306270B1 (en) | Character input method and system | |
| US9589198B2 (en) | Camera based method for text input and keyword detection | |
| US6040825A (en) | Input/display integrated information processing device | |
| KR100858358B1 (en) | Method and apparatus for user-interface using the hand trace | |
| US20140363082A1 (en) | Integrating stroke-distribution information into spatial feature extraction for automatic handwriting recognition | |
| US20140361983A1 (en) | Real-time stroke-order and stroke-direction independent handwriting recognition | |
| WO2014200736A1 (en) | Managing real - time handwriting recognition | |
| CN107451582A (en) | A kind of graphics context identifying system and its recognition methods | |
| CN114937278A (en) | Text content extraction and identification method based on line text box word segmentation algorithm | |
| CN103699890B (en) | A kind of scanning input device and scan input method | |
| US20160334941A1 (en) | Method and Device for Optical Handwriting Recognition | |
| CN113687724A (en) | Candidate character display method and device and electronic equipment | |
| Gao et al. | A sequence labeling based approach for character segmentation of historical documents | |
| JP3113747B2 (en) | Character recognition device and character recognition method | |
| KR20050003075A (en) | A mobile communication device with optical character recognition and electronic dictionary, an electronic dictionary with optical character recognition | |
| CN112118491A (en) | Bullet screen generation method and device and computer readable storage medium | |
| Dharsini et al. | Devanagri character image recognition and conversion into text using long short term memory | |
| Zhang et al. | A methodological study of document layout analysis | |
| CN113157194B (en) | Text display method, electronic equipment and storage device | |
| CN118034501A (en) | Virtual keyboard input system and method based on eye tracking technology | |
| CN117764088A (en) | Translation method, translation device, electronic equipment and storage medium | |
| JP2016197362A (en) | Range specification program, range specification method, and range specification device | |
| Sabni et al. | Laser projection virtual keyboard: a laser and image processing based human-computer interaction device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |