CN109214379A - Multi-functional point reading based on image recognition tracer technique gives directions part and reading method - Google Patents
Multi-functional point reading based on image recognition tracer technique gives directions part and reading method Download PDFInfo
- Publication number
- CN109214379A CN109214379A CN201811281549.7A CN201811281549A CN109214379A CN 109214379 A CN109214379 A CN 109214379A CN 201811281549 A CN201811281549 A CN 201811281549A CN 109214379 A CN109214379 A CN 109214379A
- Authority
- CN
- China
- Prior art keywords
- identification
- read
- reading
- face
- image recognition
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/142—Image acquisition using hand-held instruments; Constructional details of the instruments
-
- G—PHYSICS
- G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
- G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
- G09B5/00—Electrically-operated educational appliances
- G09B5/06—Electrically-operated educational appliances with both visual and audible presentation of the material to be studied
- G09B5/062—Combinations of audio and printed presentations, e.g. magnetically striped cards, talking books, magnetic tapes with printed texts thereon
Abstract
The invention discloses a kind of, and the Multi-functional point reading based on image recognition tracer technique gives directions part and reading method, which reads to give directions part to include for the hand-held part of user's gripping and for the identification part of image recognition；The side of the identification part is provided at least two identification faces, and the characteristic image that can will be distinguished from each other open is respectively arranged on each identification face, and respectively identification face respectively corresponds different read-on-command functions.Different read-on-command functions is arranged by the front and back sides for identification part for this method, realizes the Function Extension for reading indication part.This method is read to give directions part using aforementioned point, and the image for giving directions part is read by the real-time collection point of photographic device, determines that the identification face for giving directions part to present is read at image midpoint by image recognition, the difference in the identification face in read procedure according to presentation triggers different read-on-command functions.Multiple identification faces are arranged by giving directions on part in a reading in the present invention, realize a variety of read-on-command functions, greatly extend the functionality of the point-of-reading system based on image recognition tracer technique.
Description
Technical field
The present invention relates to a kind of points to read to give directions part, particularly relates to a kind of Multi-functional point reading based on image recognition tracer technique
Give directions part and reading method.
Background technique
Talking pen be the New Generation of Intelligent made of optical encoding identification technology and the exploitation of digital voice technology read and
Learning tool can be realized simultaneously and read, is re-reading, with various functions such as reading, recording, amusements.Its technical principle is first will be on books
Content encoded by OID, and coding is printed on special books with special printing technology, then user is read with point
Pen scanning books coding can be carried out identifying and playing corresponding voice.User is when using talking pen scanning book content, point
The sound-content and be scanned bookish content and combine that pen issues are read, passes through this process and realizes augmented reality.This point
Reading pen must cooperate the books for being printed with specific coding to be just able to achieve read-on-command function, and books cost of manufacture is high, and cannot utilize
The existing picture and text publication being widely present in the market.
The point-of-reading system independent of specific coding is had developed thus, and this point-of-reading system is shot by photographic device
The version object page and point are read to give directions part, read to give directions the horizontal position of part and vertical high by image recognition tracer technique real-time monitoring point
Degree plays point corresponding with predeterminable area when detecting that tracking point reads that part is given directions to carry out click to predeterminable area on the page
Pronunciation frequency.In the point-of-reading system, point is read to give directions part that any object that can be clicked by image recognition, such as hand can be used
Refer to, prefabricated point reads stick, and the homemade point of user reads magic stick etc..But it has a single function, and functionality is abundant not as good as traditional talking pen.
Summary of the invention
The purpose of the present invention is to provide a kind of high Multi-functional point readings based on image recognition tracer technique of positioning accuracy
Give directions part and reading method.
To achieve the above object, the Multi-functional point reading based on image recognition tracer technique designed by the present invention gives directions part,
Including the hand-held part held for user and for the identification part of image recognition；The side of the identification part is provided at least two identifications
Face is respectively arranged with the characteristic image that can will be distinguished from each other open on each identification face, and respectively identification face respectively correspond it is different
Read-on-command function (including system mode, point pronunciation frequency or question and answer audio etc.).
Preferably, the front end of the identification part is provided with for pinpoint click anchoring area, the click anchoring area setting
There is the characteristic image being convenient for image recognition and distinguishing with each identification face.It is horizontal fixed to be carried out using the lesser click anchoring area of area
Position, can navigate to the lesser trigger area of area, improve horizontal positioning accuracy；It is carried out using the biggish identification face of area high
Degree positioning, can obtain higher areal calculation precision, to improve altitude location precision；Biggish identification face additionally aids
Quickly capture click anchoring area.
Preferably, the area for clicking anchoring area is the 1/1000~1/10 of single identification face area.
Preferably, the prompt word of corresponding function is respectively arranged on each identification face.For example, when two identification faces respectively correspond
When Chinese speech and English voice, prompt word is respectively " Chinese ", " English ".
Preferably, the identification part is sheet, and front and back sides are respectively set to an identification face.
Preferably, the identification part is prismatic, cylindrical or elliptical cylinder-shape, is surrounded at least two on side
Identification face；Or: to be spherical, the extended line on face around hand-held part is provided at least two identification faces for the identification part.
Preferably, each identification face is overlapped and shares a part of characteristic image.When the read-on-command function for needing to be arranged is more, point is read
When giving directions the face number of part cannot meet the requirements, adjacent identification face is allowed to be overlapped, such as adjacent the first prismatic surface and the second prism
Face is respectively as an identification face, while two prismatic surface adjoiners respectively take half (not being suitable for sheet) to identify as third
Face.
Invention also provides a kind of reading method based on image recognition tracer technique, this method is using aforementioned any
Kind point is read to give directions part, and the image of indication part is read by the real-time collection point of photographic device, determines that image midpoint is read by image recognition
The different triggerings in the identification face for giving directions part to present, the identification face in a read procedure (including hovering and clicking) according to presentation are different
Read-on-command function.The corresponding audio content in each identification face can according to need setting, to identify that the point in faces is read to give directions comprising two
For part, its corresponding read-on-command function in the first identification face can be set to Chinese speech, the corresponding point in the second identification face reads function
It can be set as English voice；Or first identification face correspond to word pronunciation, second identifies that face corresponds to example sentence voice；Again or
One identification face corresponds to words pronunciation, and the second identification face corresponds to words explanation.
Preferably, when image midpoint read give directions part be rendered as two or more identification faces when, in read procedure triggering with
The corresponding read-on-command function of combination in the identification face of presentation.
Preferably, this method is accurately positioned by the way that the click anchoring area of identification part front end is arranged in.
Preferably, this method judges a little to read to give directions whether part enters trigger area as follows: tracking point is read to give directions
The coordinate that click anchoring area on part is fastened in page coordinates, when the one or more summits coordinate for clicking anchoring area enters trigger area
Coordinate range when, judge to read that a part is given directions to enter the trigger area.
Preferably, this method judges a little to read to give directions whether part has carried out click action as follows: calculating point and reads to refer to
The real-time area ratio z of identification face and the page on the image that photographic device obtains on point part, and height is clicked with just reaching
Critical area ratio z1It is compared, if z≤z1, that is, judge that point reads that part is given directions to carry out click action.The program passes through camera shooting dress
It sets lower point and reads the size relation of indication part image and page-images to judge height of the reading image relative to the page, do not need to know
Other depth information, exploitativeness greatly improve, and cost is greatly reduced.Simultaneously using the area of publication page captured in real-time as ginseng
According to real-time area ratio is calculated, avoids photographic device and publication placement position and publication thickness causes calculated result
Adverse effect.
Optionally, the critical area ratio z1It is set as a reading and gives directions the identification face of part and the physical area ratio z of the page01
~1.8 times.
Preferably, this method further includes the steps that typing point is read to give directions part in point-of-reading system, comprising:
1) it lays flat a reading and gives directions part, keep the first identification face-up；It reads by photographic device that part is given directions to shoot, it is right
The photo of shooting carries out limb recognition or obtains the characteristic image in the first identification face by subscriber frame choosing, is selected and is obtained by subscriber frame
Click the characteristic image of anchoring area；To simplify identification, click anchoring area all directions pattern is identical, only need to obtain a characteristic image i.e.
It can；
2) overturning point is read to give directions part, keeps the second identification face-up；It reads by photographic device that part is given directions to shoot, it is right
The photo of shooting carries out limb recognition or obtains the characteristic image in the second identification face by subscriber frame choosing.The rest may be inferred, can obtain
Obtain other corresponding characteristic images in identification face.
Preferably, this method further includes the process that making point reads voice packet: shooting in advance or the scanning publication page, by it
On need set-point to read voice region be set as trigger area, and record needed for each read-on-command function corresponding with each trigger area
Audio content.
Preferably, this method reads that part is given directions to realize more read-on-command functions using multiple points being distinguished from each other out, and switching is not
Same point is read to give directions part that can switch different read-on-command functions.
Compared with prior art, the beneficial effects of the present invention are: by read give directions part on multiple identification faces are set,
It realizes a variety of read-on-command functions, greatly extends the functionality of the point-of-reading system based on image recognition tracer technique；Meanwhile each point
Reading function can read that part is given directions to realize by the point of rotation, easy to operate, rich in interest.
Detailed description of the invention
Fig. 1 (a), Fig. 1 (b) are respectively the structural representation that Multi-functional point reading provided by embodiment 2 gives directions the front and back sides of part
Figure.
Fig. 2 (a), Fig. 2 (b) are that point provided by embodiment 2,3 is read to give directions the structural schematic diagram of part.
Fig. 3 is the structural schematic diagram of point-of-reading system provided by embodiment 4.
Fig. 4 is the position view of photographic device in embodiment 4, point reading indication part and page.
Fig. 5 is the schematic diagram that trigger area is arranged in embodiment 6 on the page.
Fig. 6 is the schematic diagram that physical area ratio is obtained in embodiment 5.
Wherein: point reads that part 1, hand-held part 1.1, identification part 1.2, first is given directions to identify that face 1.3, second identifies face 1.4, clicks
Anchoring area 1.5, identification face 1.6, photographic device 2, point-of-reading device 3, picture recognition module 3.1, mode switch module 3.2, position chase after
Track module 3.3, event trigger module 3.4, data packet production module 3.5, storage device 4, audio playing apparatus 5, the page 6, touching
Send out region 6.1
Specific embodiment
The following further describes the present invention in detail with reference to the accompanying drawings and specific embodiments.
Embodiment 1
As shown in Figure 1, present embodiments providing a kind of Multi-functional point reading indication part based on image recognition tracer technique, packet
It includes for the hand-held part 1.1 of user's gripping and for the identification part 1.2 of image recognition.
The front and back sides of identification part 1.2 are respectively arranged with an identification face 1.2 (containing the first identification face 1.3 and the second identification face
1.4), two identification faces 1.2 are respectively arranged with the characteristic image that can be distinguished from each other open and be convenient for image recognition, and respectively
Corresponding different read-on-command function can be used for switching different point pronunciation frequency or point reading/question-answering mode.It is each to know to avoid confusion
The prompt word of corresponding function is provided on other face 1.2.The point reads that the different identification faces 1.2 of part is given directions to respectively correspond different functions
Or audio, it carries out overturning switching with can be convenient, improves the functionality and interest of point-of-reading system.If need to realize, more points is read
Function also can be used more points and read to give directions the combination of part in addition to using mode in embodiment 2,3.
The front end of identification part 1.2 is provided with for pinpoint click anchoring area 1.5, and the front and back sides of click anchoring area 1.5 are equal
It is provided with the characteristic image being convenient for image recognition and can distinguishing with the subject image in each identification face 1.2；It is single to click anchoring area 1.5
Face area is the 1/1000~1/10 of single identification 1.2 area of face, and each face image for clicking anchoring area 1.5 can be identical, can also be with
Different (identical image is set as in the present embodiment).
Above-mentioned Multi-functional point reading gives directions part that can sell together with system hardware, can also by user according to aforementioned structure from
Row design, meets individual requirements.It is needed in typing point-of-reading system after the completion of design, steps are as follows for typing:
1) it lays flat a reading and gives directions part, make the first identification face 1.3 upward；By 2 pairs of points of photographic device read that part is given directions to clap
It takes the photograph, limb recognition is carried out to the photo of shooting or obtains the characteristic image in the first identification face 1.3 by subscriber frame choosing, pass through user
Frame choosing obtains the characteristic image for the click anchoring area 1.5 being located on the first identification face 1.3.
2) overturning point is read to give directions part, makes the second identification face 1.4 upward；By 2 pairs of points of photographic device read that part is given directions to clap
It takes the photograph, limb recognition is carried out to the photo of shooting or obtains the characteristic image in the second identification face 1.4 by subscriber frame choosing.
Embodiment 2~3
Fig. 2 (a), 2 (b) are respectively that point provided by embodiment 2,3 is read to give directions part, the two basic structure and 1 phase of embodiment
Together, difference is: 1) anchoring area is clicked in not set click；2) identification part 1.2 is respectively hexagonal prisms and cylinder；3) number in face is identified
Amount is six (angle can only see three in figure)；4) identify that there are certain overlappings in face in embodiment 3；5) other identification faces
Input method referring to typing step 2).
Embodiment 4
As shown in figure 3, present embodiment discloses a kind of point-of-reading systems based on image recognition tracer technique, including such as the following group
At part:
1) point is read to give directions part 1, for clicking the publication page, structure detailed in Example 2.
2) the publication page is arranged in directly above or obliquely above in photographic device 2, reads to give directions part for acquiring the page and point
1 realtime graphic.
3) characteristics of image for identification and is extracted in point-of-reading device 3, and tracking point is read to give directions the position of part 1 and judges to click thing
Part；The point-of-reading device 3 includes point reading mode and question-answering mode both of which, when being set as reading mode, if the event of click hair
Raw then broadcast point pronunciation frequency；When being set as question-answering mode, playback problem audio first, then judge whether user clicks correctly
Trigger area.
The point-of-reading device 3 specifically includes following software modules:
3.1) picture recognition module 3.1, photographic device 2 acquires the page and point reading indication part 1 in image for identification, mentions
Their characteristics of image is taken, and determination is compared by the characteristics of image of current page and the publication page shot in advance and is worked as
The preceding locating page.
In the present embodiment, picture recognition module 3.1 extracts characteristics of image using the image characteristics extraction algorithm of OpenCV.With
Fingerprint is similar, and each page and point are read to give directions part image, all has any different in the unique features of other images, no matter same image is sent out
The male character types in Chinese operas, usu. referring tov the bearded character degree, displacement, light and shade variation, extracted feature is all identical.The image characteristics extraction algorithm of open source projects OpenCV,
In detail visible https: //github.com/MasteringOpenCV/code/tree/master/Chapter3_
The operation logic of PatternDetector module in MarkerlessAR.
3.2) mode switch module 3.2, for system to be switched to a reading mode or question-answering mode, pattern switching can be by
User's manual setting can also automatically switch after meeting switching condition.
3.3) location tracking module 3.3 judges that a reading gives directions whether part 1 is clicked the trigger area on the page,
Return to the trigger area being clicked.
3.4) event trigger module 3.4 for judging whether trigger conditions meet, and executes subsequent action；Event
Trigger condition includes page turning, clicking trigger region, point reading indication part 1 upward read that part 1 is given directions to hover in trigger area by picture, point
Side；Subsequent action includes playing default audio, problem audio, point pronunciation frequency or hovering audio, after partial condition and movement are detailed in
Text.
3.5) data packet makes module 3.5, acquires image data in advance, and recording audio data are arranged trigger condition, and beat
It is bundled into a reading data packet, to distribute；Point reads data packet can both have been edited recording by professional institution, can also be by ordinary user certainly
Edlin is recorded.
4) storage device 4, for storing image data, audio data and trigger condition, wherein the image data stored is
Publication page-images and its characteristics of image, point read to give directions part image and its characteristics of image, and point reads to give directions part and the page
Physical area ratio z0, critical area ratio z1。
5) audio playing apparatus 5, for playing audio, including pronunciation frequency, problem audio, and interactive finger
Order or prompt tone.
The point-of-reading system can be deployed as an independent software and hardware integrated system, such as be integrated with photographic device and audio
The intelligent movable device systems of playing device, system of the independent photographic device in conjunction with intelligent sound box, independent photographic device, independence
System of the audio playing apparatus in conjunction with intelligent movable equipment can also be deployed as system of the aforementioned software and hardware in conjunction with server.
The present embodiment provides a kind of reading method using above system simultaneously, and this method passes through mode switch module
3.2 are set as a reading mode or question-answering mode.
When being set as reading mode, photographic device 2 acquires the page and point is read to give directions the image of part 1, picture recognition module
3.1 pairs of pages and point read that part image is given directions to identify, extract the characteristics of image of the two, read the page in data packet by contrast points
Face characteristics of image determines the page being presently in；3.3 pairs of points of location tracking module are read to give directions position and click of the part 1 on the page
Movement is tracked.
After event trigger module 3.4 detects a certain page, during waiting user action, played pre-recorded
With the associated default audio of the page, broadcasting for default audio is interrupted once detecting that the movement of part 1 (or click) afterwards is given directions in a reading
It puts, after the trigger area that user clicks on the page, transfers corresponding pronunciation frequency file, transfer to audio playing apparatus
5 play out.
When being set as question-answering mode, picture recognition module 3.1 first identifies current page, event trigger module
3.4, which transfer the problem audio corresponding with current page prerecorded, transfers to audio playing apparatus 5 to play out, and prompts user's point
Preset trigger area corresponding with problem answers on current page is hit, and judges whether user passes through a reading and give directions 1 point of part
The trigger area has been hit, judging result voice is returned.
Point reading mode and question-answering mode both of which may be configured as manual switching, can also be by mode switch module 3.2
Automatically switch.Automatic switchover can be switched over according to specific webpage, such as when photographic device 2 detects cover, and point reads dress
It sets 3 and is switched to a reading mode, be then switched to question-answering mode when detecting back cover；It can also be switched over according to specific pattern, example
A reading mode such as is stamped in certain pages (such as unit surveys the page certainly) or question-answering mode figure, system detection automatically switch after
For reading mode or question-answering mode.
Professional institution and ordinary user can make 3.5 making point of module by data packet and read data packet, the steps include:
1) user takes out a books, the electronic pictures (cover of the page needed for camera or scanner acquisition point-of-reading system
For item must be clapped), the page unrelated with point-of-reading system can not acquire.
2) user creates a data packet in " input system " (data packet production module), is named to packet, names
It can be the title that book name or other specific crowds can distinguish, and set the corresponding electronic pictures of cover, back cover." typing
System " can generate unique data packet id for data packet." input system " can be mobile phone or tablet computer APP, be also possible to
Website or software systems.
2) characteristics of image of the page is extracted by picture recognition module 3.1, picture recognition module 3.1 is existing mature skill
Art can be deployed in APP or website.
3) enclosed region is set in the electronic pictures of the page as trigger area, the trigger area stored on each page is sat
Mark.
4) corresponding with trigger area pronunciation frequency and problem audio are recorded, and establishes the pass between audio and trigger area
Connection.
5) it records the page and is opening default audio when not clicking on also, and establish being associated between default audio and the page.
Module 3.5 is made by data packet, ordinary user may be that the printed matters such as books, picture album, desk calendar, photograph album are made by oneself
Adopted voice, such as father is after drawing this recording audio, child, which opens, draws this and when clicking, with the associated audio of click on area
It will play automatically, auxiliary child reads character learning；Teacher can be the every page or a few pages of recordings teaching audios of teaching material, student
When opening those page, teaching audio is just automatic to be played, auxiliary student's study.
In view of common publication currently on the market does not make any label to can put reading field, reader possibly can not be quick
Trigger area is found, the present embodiment is also that trigger area increases hovering audio, i.e., when a reading gives directions part 1 by trigger area
Side, and when not dropping to the critical altitude of triggering click action or less, it plays for showing to read to give directions a part 1 by trigger area
Prompt tone, be quickly found out trigger area present on the page convenient for user, improve user experience.
In the present embodiment, location tracking module 3.3 reads that part is given directions to carry out horizontal tracking and height with the following method
Tracking, to judge whether user clicks trigger area:
1) horizontal tracking
Tracking point reads the coordinate for giving directions part 1 to fasten in page coordinates, when the part point of 1 front end one or multiple click-through are given directions in a reading
When entering the coordinate range of trigger area, judge that a reading gives directions part 1 to enter the trigger area.
2) height is tracked
2.1) as shown in figure 4, in read procedure, 2 captured in real-time point of the photographic device reading being fixed on above publication refers to
Point part 1 and the publication page, the location tracking module 3.3 of point-of-reading system calculate point in real time and read to give directions the identification face 1.6 of part 1 and go out
Area ratio of the version object page in shooting image, is denoted as real-time area ratio z.
Real-time area of the publication page in shooting image can also be using publication cover (or back cover) in shooting figure
Area as in is substituted, and when calculating real-time pictures ratio z, no longer needs to calculate real-time surface of other pages in shooting image
Product.
2.2) when set-point reads to give directions part 1 close to publication to the position for just triggering click, point is read to give directions the identification of part 1
The area ratio of face 1.6 and the publication page in shooting image is critical area ratio z1；Critical area ratio z1It can be with actual test
It obtains, the identification face 1.6 for reading indication part 1 and an actual physics area ratio z for the publication page can also be preset as01~1.8
Times, specific proportional numerical value is advisable in order to user's operation.And actual physics area ratio z0It can be according to publication size (title page one
As it is on the books), point read give directions part 1 1.6 areal calculation of identification face obtain；It can also obtain as follows: read to give directions by
Part 1 and publication are laid flat in the same plane, obtain the identification face 1.6 that point under same picture reads indication part 1 by photographic device 2
The page-images of image and publication, the range of the two is selected by subscriber frame, and system selects the figure for calculating the two according to the frame of user
As area ratio, value and actual physics area ratio z0Equal, later approach is mainly used for publication or point is read to give directions 1 ruler of part
Very little unknown situation.
If 2.3) real-time pictures ratio z > z1, then point-of-reading system determines that current point reading gives directions part 1 to be in and do not click on state；If z
≤z1, then point-of-reading system determines that current point reads that part 1 is given directions to be in click state.
3) when reading that a part 1 is given directions to carry out click action, and the one or more points in its click on area enters trigger region
When in the coordinate range in domain, judge that a reading indicator clicks trigger area.
4) multipage identification and processing
For being the publication of multipage after expansion, when occurring multiple " recognizable " pages within the scope of video camera, system can
To be carried out according to the page, activation respective page where the relative position judgement at present of reading part and multiple pages according still further to single page
Processing.When calculating real-time area ratio, current page area is calculated after can identifying single page by picture recognition module 3.1,
Or take shooting image in entire area divided by page number.
The system and method read to give directions the size relation of part image and page-images to judge a little by putting under photographic device
Height of the reading image relative to the page does not need identification depth information, and exploitativeness greatly improves, while cost is greatly reduced.
A kind of typical case environment of the invention is presented above, can preferably embody this hair using the system and method
Bright Multi-functional point reading gives directions the technical effect of part, but the present invention can also be read using other points based on image recognition tracer technique
System.
Embodiment 5
Present embodiments providing Multi-functional point reading provided by Application Example 1 gives directions part to carry out the method read,
In embodiment 4 on the basis of reading method, the detection of 1.2 front and back sides of identification part is increased, specifically:
Under reading mode, when the location tracking module 3.3 of point-of-reading system detects that clicking anchoring area 1.5 clicks on the page
Trigger area, if the identification face arrived of current shooting is the first identification face 1.3, event trigger module 3.4 is transferred and trigger region
The corresponding first pronunciation frequency in domain transfers to audio playing apparatus 5 to play；If the identification face that current shooting arrives is the second identification face
1.4, then event trigger module 3.4 transfers second point pronunciation frequency corresponding with trigger area and transfers to the broadcasting of audio playing apparatus 5.Two
The corresponding audio in identification face can according to need setting, such as first pronunciation frequency is Chinese speech, and second point pronunciation frequency is English
Literary voice；For another example first pronunciation frequency is word pronunciation, and second point pronunciation frequency is example sentence voice；For another example first pronunciation frequency is word
Word pronunciation, second point pronunciation frequency are that words is explained.First pronunciation frequency, second point pronunciation frequency are being made by professional institution or user
Make point to read to record when data packet, and is associated with trigger area.
The present embodiment carries out horizontal location using the lesser click anchoring area of area, can navigate to the lesser trigger region of area
Domain improves horizontal positioning accuracy；Altitude location is carried out using the biggish identification face of area, higher area ratio can be obtained
Computational accuracy, so that determining that the process of click action is more stable reliable by area ratio；Also using biggish identification face
Help picture recognition module and quickly captures click anchoring area 1.
Embodiment 6
The present embodiment provides 1 midpoint of embodiment and reads to give directions part and 5 midpoint of embodiment so that children draw this " picture Yunnan greatly " as an example
The concrete application of reading method.
1) making point reads data packet
1.1) children are chosen to draw this " greatly picture Yunnan ", user by photographic device 2 or scanner acquire page-images (including
Cover and back), point-of-reading system (hereinafter referred to as system) automatic identification characteristics of image.
1.2) as shown in figure 5, trigger area 6.1, and first pronunciation frequency of typing is arranged in user on the image of the page 6
(Chinese), second point pronunciation frequency (English), problem audio, default audio and audio are closed with the page, the corresponding of trigger area
System；Trigger area 6.1 is the closure being made of a several vertex connecting line segments rule or irregular figure, and user is in page figure
After upper drafting trigger area 6.1, its coordinate of peripheral vertex relative to page coordinates system of system automatic collection, to obtain touching
Send out the coordinate range in region.
2) data configuration
2.1) user is read to give directions each face input system of part 1 with 2 shooting point of photographic device.
2.2) placed side by side in (not being overlapped) on same plane reading to give directions part 1 and drawing this cover, it will with photographic device 2
Point is read to give directions part 1 and draws this picture photographing in same picture.
2.3) according to aforementioned picture, rectangle frame is voluntarily drawn by user in systems, frame choosing, which is surrounded, draws this cover and point reading
1 profile of part is given directions, as shown in Figure 6.System calculates a reading automatically and part 1 is given directions to identify according to two rectangle frame regions of setting
The area in portion 1.2 and this cover area ratio value relationship is drawn, as " physical area ratio z0", such as z in Fig. 60=1/10=0.1.
2.4) set-point reads to give directions the click anchoring area 1.5 of part 1: point is read to give directions 1.2 picture of identification part of part 1 larger, first is that
In order to prompt the effect in the user face, two read that a part 1 is given directions to be easier to be captured by photographic device 2 also for allowing, but click anchor
Area 1.5 can be set to lesser region, realize that more accurate point is read.
2.5) the critical area ratio z that setting click event triggers1=physical area ratio z0× 1.5, when photographic device 2 identifies
To read give directions part 1 identify face area and page area (being calculated with cover area) the i.e. real-time area ratio z of ratio≤
Critical area ratio z1, i.e. when z≤0.15, reach the height condition that triggering is clicked, judge that user is clicked.
3) read operation is put
3.1) user is ready to draw this, opens system, input data packet id or name keys, or scanning cover searches for number
According to packet；If the multiple disclosed data packets of keyword match, system, which can return to multiple packets, allows user to select one；If user sweeps
Describe this cover to search for data packet, system will identify cover image feature, and inquire the corresponding data of the cover in the database
Packet returns to user, if the multiple disclosed data packets of cover matching, system, which can return to multiple packets, allows user to select one.
3.2) point-of-reading system (smart phone is used in the present embodiment) is fixed on and is drawn on this oblique upper desk lamp by user, system
It detects and draws this cover or according to user's operation inlet point reading mode.
3.3) when system identification is to a certain page 6, if the page setup has default audio, transfer automatically default audio into
Row plays, and interrupts automatically when detecting user's click action；Above reading indication part 1 is by trigger area 6.1, and do not decline
When below to the critical altitude of triggering click action, play for showing that hovering audio of the part 1 Jing Guo trigger area is given directions in a reading.
3.4) user by the trigger area 6.1 on the direction page 6 of click anchoring area 1.5 of reading indication part 1 and clicks,
System is read to give directions the corresponding audio of the picture broadcasting upward of part 1 according to, and Chinese sound is played when picture is the first identification face 1.3 upward
Frequently, English audio is played when picture is the second identification face 1.4 upward.
4) question and answer operate
4.1) system detection is to drawing this back cover or enter question-answering mode according to user's operation.
4.2) when system detection is to the page 6 for being provided with question and answer audio, problem audio corresponding to the page is played automatically,
The problem of such as children: " please point out the elephant on the page ".
4.3) user with point read give directions a part 1 be directed toward the page on answer a trigger area 6.1, such as herein be directed toward the page on have
The region of elephant picture.
4.4) when system detection to read give directions part 1 a click anchoring area 1.5 intersect with event trigger region in page-images and
Height is when meeting click conditional, and triggering, which executes, judge that answer to wrong service logic, and passes through mobile phone speaker in the form of sound
Feed back to user.
5) trigger area clicks judgement
In point read procedure, during a reading is given directions part 1 to be moved to target by user, in order not to allow moving process accidentally to touch
Hair, system calculate real-time area ratio z at regular intervals (such as 0.2 second): working as z > critical area ratio z1When, it can be understood as also not
It is clicked；Point reads to give directions the fitting of part 1 or close to the page, z is in physical area ratio z0With critical area ratio z1Between, i.e. z0≤z≤
z1When, then it is judged as that height has met click conditional.
System judges whether click anchoring area 1.5 touches the trigger area on the page, judgement side in the horizontal direction again
Formula are as follows: whether system acquires in real time clicks each vertex of anchoring area 1.5 in the coordinate of page coordinates system, judge the coordinate in trigger area
Coordinates regional in, such as: certain triangle click each vertex facing pages coordinate system of anchoring area 1.5 real-time coordinates be (X1,Y1)、
(X2,Y2)、(X3,Y3), certain rectangle trigger area upper left, upper right, lower-left, each vertex facing pages coordinate system in bottom right coordinate
For (x1,y1)、(x2,y2)、(x3,y3), (x4, y4), when click anchoring area 1.5 any vertex such as (Xi,Yi) (i=1,2,3) triggering
In region, i.e. x1≤Xi≤x2And y3≤Yi≤y1When, judge that click anchoring area 1.5 has touched trigger area in the horizontal direction.
Claims (10)
1. a kind of Multi-functional point reading based on image recognition tracer technique gives directions part, it is characterised in that: including what is held for user
Hand-held part (1.1) and identification part (1.2) for image recognition；The side of the identification part (1.2) is provided at least two identifications
Face (1.6) respectively identifies and is respectively arranged with the characteristic image that can will be distinguished from each other open on face (1.6), and respectively identifies face (1.6)
Respectively correspond different read-on-command functions.
2. the Multi-functional point reading according to claim 1 based on image recognition tracer technique gives directions part, it is characterised in that: institute
The front end for stating identification part (1.2) is provided with for pinpoint click anchoring area (1.5), and the click anchoring area (1.5) is provided with
Convenient for image recognition and the characteristic image that is distinguished with each identification face (1.6).
3. the Multi-functional point reading according to claim 2 based on image recognition tracer technique gives directions part, it is characterised in that: institute
Stating and clicking the area of anchoring area (1.5) is the 1/1000~1/10 of single identification face (1.6) area.
4. the Multi-functional point reading according to claim 1 based on image recognition tracer technique gives directions part, it is characterised in that: each
The prompt word of corresponding function is respectively arranged on identification face (1.6).
5. the Multi-functional point reading according to any one of claims 1 to 4 based on image recognition tracer technique gives directions part,
Be characterized in that: the identification part (1.2) is sheet, and front and back sides are respectively set to an identification face (1.6).
6. the Multi-functional point reading according to any one of claims 1 to 4 based on image recognition tracer technique gives directions part,
Be characterized in that: the identification part (1.2) is prismatic, cylindrical or elliptical cylinder-shape, is surrounded at least two on side
Identification face (1.6)；Or: the identification part (1.2) is spherical shape, the extended line on face around hand-held part (1.1) be provided with to
Few two identification faces (1.6).
7. the Multi-functional point reading according to claim 6 based on image recognition tracer technique gives directions part, it is characterised in that: each
Identification face (1.6) is overlapped and shares a part of characteristic image.
8. a kind of reading method based on image recognition tracer technique, it is characterised in that: this method uses such as claim 1~7
Any one of described in point read give directions part (1), by photographic device (2) in real time collection point read give directions part (1) image, pass through figure
Determine that the identification face (1.6) for giving directions part (1) to present is read at image midpoint as identifying, according to the identification face of presentation in read procedure
(1.6) difference triggers different read-on-command functions.
9. the reading method according to claim 8 based on image recognition tracer technique, it is characterised in that: when image midpoint
When reading that part (1) is given directions to be rendered as two or more identification faces (1.6), the combination with the identification face presented is triggered in read procedure
Corresponding read-on-command function.
10. the reading method according to claim 9 based on image recognition tracer technique, it is characterised in that: this method is logical
The click anchoring area (1.5) being arranged in identification part (1.2) front end is crossed to be accurately positioned.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811281549.7A CN109214379B (en) | 2018-10-23 | 2018-10-23 | Multifunctional point reading pointing piece based on image recognition tracking technology and point reading method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811281549.7A CN109214379B (en) | 2018-10-23 | 2018-10-23 | Multifunctional point reading pointing piece based on image recognition tracking technology and point reading method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109214379A true CN109214379A (en) | 2019-01-15 |
| CN109214379B CN109214379B (en) | 2022-02-15 |
Family
ID=64998209
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811281549.7A Active CN109214379B (en) | 2018-10-23 | 2018-10-23 | Multifunctional point reading pointing piece based on image recognition tracking technology and point reading method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109214379B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110929709A (en) * | 2019-10-25 | 2020-03-27 | 北京光年无限科技有限公司 | Method and device for converting point-reading content into sketch finger-reading content based on OID |
| CN111079495A (en) * | 2019-06-09 | 2020-04-28 | 广东小天才科技有限公司 | Point reading mode starting method and electronic equipment |
| CN111091531A (en) * | 2019-05-29 | 2020-05-01 | 广东小天才科技有限公司 | Click recognition method and electronic equipment |
Citations (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080259355A1 (en) * | 2007-04-18 | 2008-10-23 | Ming-Yen Lin | Method of recognizing and tracking multiple spatial points |
| CN201489559U (en) * | 2009-04-02 | 2010-05-26 | 马洪生 | Optical touch-reading pen with multiple output modes |
| CN201749435U (en) * | 2010-07-14 | 2011-02-16 | 东莞市步步高教育电子产品有限公司 | Optical point reading pen |
| EP2299406A2 (en) * | 2008-07-09 | 2011-03-23 | Gwangju Institute of Science and Technology | Multiple object tracking method, device and storage medium |
| CN201788597U (en) * | 2010-04-02 | 2011-04-06 | 郑海山 | Reading pen |
| CN102033663A (en) * | 2010-09-30 | 2011-04-27 | 广东威创视讯科技股份有限公司 | Camera surface positioning system and pen color identification method |
| CN102393998A (en) * | 2011-12-08 | 2012-03-28 | 余江 | General remote controller based on image coding identification |
| CN202661964U (en) * | 2012-05-05 | 2013-01-09 | 山东长征教育科技有限公司 | Optical recognition wireless talking pen |
| CN202677648U (en) * | 2012-06-06 | 2013-01-16 | 北京外研通教育科技有限公司 | Interactive teaching device |
| CN103035134A (en) * | 2012-12-10 | 2013-04-10 | 张肃 | Image touch and talk playing system and mage touch and talk playing method |
| CN202976538U (en) * | 2012-12-10 | 2013-06-05 | 张肃 | Image touch-read playing system |
| CN203102633U (en) * | 2013-03-12 | 2013-07-31 | 北京阳光接力教育科技有限公司 | Touch reading and touch watching device and touch reading and touch watching system |
| CN203300063U (en) * | 2013-04-08 | 2013-11-20 | 许卫 | Story machine |
| CN105022582A (en) * | 2015-07-20 | 2015-11-04 | 广东小天才科技有限公司 | Function triggering method for talking terminal and talking terminal |
| CN107731020A (en) * | 2017-11-07 | 2018-02-23 | 广东欧珀移动通信有限公司 | Multi-medium play method, device, storage medium and electronic equipment |
| CN107748645A (en) * | 2017-09-27 | 2018-03-02 | 努比亚技术有限公司 | Reading method, mobile terminal and computer-readable recording medium |
| CN107907975A (en) * | 2017-12-29 | 2018-04-13 | 东莞市宇光光电科技有限公司 | Imaging lens module, imaging lens and talking pen |
| CN207425138U (en) * | 2017-09-09 | 2018-05-29 | 惠州市深科信飞科技有限公司 | Talking pen with liquid crystal display |
| CN108399349A (en) * | 2018-03-22 | 2018-08-14 | 腾讯科技（深圳）有限公司 | Image-recognizing method and device |
| CN207818003U (en) * | 2017-11-02 | 2018-09-04 | 河南书网教育科技股份有限公司 | A kind of Multi-functional point reading learning system |
-
2018
- 2018-10-23 CN CN201811281549.7A patent/CN109214379B/en active Active
Patent Citations (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080259355A1 (en) * | 2007-04-18 | 2008-10-23 | Ming-Yen Lin | Method of recognizing and tracking multiple spatial points |
| EP2299406A2 (en) * | 2008-07-09 | 2011-03-23 | Gwangju Institute of Science and Technology | Multiple object tracking method, device and storage medium |
| CN201489559U (en) * | 2009-04-02 | 2010-05-26 | 马洪生 | Optical touch-reading pen with multiple output modes |
| CN201788597U (en) * | 2010-04-02 | 2011-04-06 | 郑海山 | Reading pen |
| CN201749435U (en) * | 2010-07-14 | 2011-02-16 | 东莞市步步高教育电子产品有限公司 | Optical point reading pen |
| CN102033663A (en) * | 2010-09-30 | 2011-04-27 | 广东威创视讯科技股份有限公司 | Camera surface positioning system and pen color identification method |
| CN102393998A (en) * | 2011-12-08 | 2012-03-28 | 余江 | General remote controller based on image coding identification |
| CN202661964U (en) * | 2012-05-05 | 2013-01-09 | 山东长征教育科技有限公司 | Optical recognition wireless talking pen |
| CN202677648U (en) * | 2012-06-06 | 2013-01-16 | 北京外研通教育科技有限公司 | Interactive teaching device |
| CN202976538U (en) * | 2012-12-10 | 2013-06-05 | 张肃 | Image touch-read playing system |
| CN103035134A (en) * | 2012-12-10 | 2013-04-10 | 张肃 | Image touch and talk playing system and mage touch and talk playing method |
| CN203102633U (en) * | 2013-03-12 | 2013-07-31 | 北京阳光接力教育科技有限公司 | Touch reading and touch watching device and touch reading and touch watching system |
| CN203300063U (en) * | 2013-04-08 | 2013-11-20 | 许卫 | Story machine |
| CN105022582A (en) * | 2015-07-20 | 2015-11-04 | 广东小天才科技有限公司 | Function triggering method for talking terminal and talking terminal |
| CN207425138U (en) * | 2017-09-09 | 2018-05-29 | 惠州市深科信飞科技有限公司 | Talking pen with liquid crystal display |
| CN107748645A (en) * | 2017-09-27 | 2018-03-02 | 努比亚技术有限公司 | Reading method, mobile terminal and computer-readable recording medium |
| CN207818003U (en) * | 2017-11-02 | 2018-09-04 | 河南书网教育科技股份有限公司 | A kind of Multi-functional point reading learning system |
| CN107731020A (en) * | 2017-11-07 | 2018-02-23 | 广东欧珀移动通信有限公司 | Multi-medium play method, device, storage medium and electronic equipment |
| CN107907975A (en) * | 2017-12-29 | 2018-04-13 | 东莞市宇光光电科技有限公司 | Imaging lens module, imaging lens and talking pen |
| CN108399349A (en) * | 2018-03-22 | 2018-08-14 | 腾讯科技（深圳）有限公司 | Image-recognizing method and device |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111091531A (en) * | 2019-05-29 | 2020-05-01 | 广东小天才科技有限公司 | Click recognition method and electronic equipment |
| CN111091531B (en) * | 2019-05-29 | 2023-09-22 | 广东小天才科技有限公司 | Click recognition method and electronic equipment |
| CN111079495A (en) * | 2019-06-09 | 2020-04-28 | 广东小天才科技有限公司 | Point reading mode starting method and electronic equipment |
| CN111079495B (en) * | 2019-06-09 | 2024-03-22 | 广东小天才科技有限公司 | Click-to-read mode starting method and electronic equipment |
| CN110929709A (en) * | 2019-10-25 | 2020-03-27 | 北京光年无限科技有限公司 | Method and device for converting point-reading content into sketch finger-reading content based on OID |
| CN110929709B (en) * | 2019-10-25 | 2022-11-22 | 北京光年无限科技有限公司 | Method and device for converting point-reading content into sketch finger-reading content based on OID |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109214379B (en) | 2022-02-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109448453A (en) | Point based on image recognition tracer technique reads answering method and system | |
| CN109445588A (en) | Point based on image recognition tracer technique is read to give directions part click judging method | |
| CN104157171B (en) | A kind of point-of-reading system and method thereof | |
| CN106648146B (en) | Dot pattern, information reproducing method using dot pattern, and input/output method | |
| CN109214379A (en) | Multi-functional point reading based on image recognition tracer technique gives directions part and reading method | |
| CN109637286A (en) | A kind of Oral Training method and private tutor's equipment based on image recognition | |
| CN109376612B (en) | Method and system for assisting positioning learning based on gestures | |
| CN109191940B (en) | Interaction method based on intelligent equipment and intelligent equipment | |
| CN105590486A (en) | Machine vision-based pedestal-type finger reader, related system device and related method | |
| CN110111612A (en) | A kind of photo taking type reading method, system and point read equipment | |
| CN104199834A (en) | Method and system for interactively obtaining and outputting remote resources on surface of information carrier | |
| CN111259863B (en) | Playing hand type detection/display method, medium, piano, terminal and server | |
| WO2017047182A1 (en) | Information processing device, information processing method, and program | |
| WO2008004331A1 (en) | Voice outputting method and device linked to images | |
| WO2022052941A1 (en) | Intelligent identification method and system for giving assistance with piano teaching, and intelligent piano training method and system | |
| JP2008123265A (en) | Idea extraction support system and method | |
| WO2018108177A1 (en) | Method for teaching painting using robot, device and robot therefor | |
| CN111027537A (en) | Question searching method and electronic equipment | |
| CN109739353A (en) | A kind of virtual reality interactive system identified based on gesture, voice, Eye-controlling focus | |
| CN111077992B (en) | Click-to-read method, electronic equipment and storage medium | |
| CN111079501B (en) | Character recognition method and electronic equipment | |
| CN111026786B (en) | Dictation list generation method and home education equipment | |
| CN111539408A (en) | Intelligent point reading scheme based on photographing and object recognizing | |
| US11580868B2 (en) | AR-based supplementary teaching system for guzheng and method thereof | |
| Syeda-Mahmood | Indexing for topics in videos using foils |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20210929 Address after: Room 301, block B, platform 18, Jinding Science Park, No. 690 Xuefu Road, Wuhua District, Kunming, Yunnan 650033 Applicant after: Kunming micro Chi Sen Polytron Technologies Inc. Address before: Room 4501b, building 4, Zone C, No. 12, xidawang Road, Chaoyang District, Beijing 100022 Applicant before: BEIJING KUAILE COGNITION TECHNOLOGY Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |