CN101615253B - System and method for instantly identifying file contents - Google Patents
System and method for instantly identifying file contents Download PDFInfo
- Publication number
- CN101615253B CN101615253B CN 200810130719 CN200810130719A CN101615253B CN 101615253 B CN101615253 B CN 101615253B CN 200810130719 CN200810130719 CN 200810130719 CN 200810130719 A CN200810130719 A CN 200810130719A CN 101615253 B CN101615253 B CN 101615253B
- Authority
- CN
- China
- Prior art keywords
- block
- file
- reads
- read
- identification
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Character Discrimination (AREA)
Abstract
The invention discloses a system for instantly identifying file contents, which can instantly identify a file with a special structure. The system comprises a file structure analyzing device, a reading schedule setting device, a positioning device and an identifying module, wherein the file structure analyzing device is used for marking a file into a plurality of blocks according to at least one structural characteristic in the file; the reading schedule setting device is used for setting a reading schedule to read the blocks; the positioning device is used for positioning a block in the reading process; and the identifying device is used for identifying the block in the reading process to output the content of the block in the reading process. The invention can be applied to the robot field to design a robot with the function of imitating people to read files.
Description
Technical field
The present invention relates to a kind of identification system and method, but particularly relate to a kind of system and method for immediately identifying file content.
Background technology
In the daily life, we need change into editable file with Miscellaneous Documents often.In general, file must be scanned into image file earlier, then utilizes optical character identification (Optical Character Recognition, OCR) software, the character in the identification file.Perhaps, utilize scanning identification pen,, word for word scan word for word identification with manual mode.Yet with the file identification, the former lacks maneuverability, and the latter can't handle a large amount of files automatically.
In the field of robot, development robotic vision function is a trend.Robot with immediately identifying ability more near human behavior, also becomes the robot vision field and uses a target of demanding urgently breaking through.If robot can adopt with seeing with the mode reading file of reading as human, use in each field, for example, the field of service type robot has certain potential business opportunity.
Yet, in the traditional reading file technology, utilize high-resolution digital camera (or scanner) once to take (or scanning) whole part of file, again the image of obtaining is done identification.This discrimination method need jumbo memory, and identification process can spend long time.
Another kind of discrimination method utilizes the low resolution video camera to take whole part of file several times, then the partial image of obtaining is done skew corrected individually, and they are bonded into the big image of opening, and again this is opened image greatly and does identification.This discrimination method on the step of skew corrected and map interlinking, needs a large amount of computing times.In addition, use this discrimination method, wayward image quality.
Therefore, above-mentioned traditional discrimination method is not suitable for the content of immediately identifying file, more unlike the habit of true man's reading file.So, be necessary to develop a kind of new discrimination method, when being used in the field of robot, make robot have the function of emulation people reading file.
Summary of the invention
First purpose of the present invention is to provide a kind of file content immediately identifying system and method, and it is the system or the method for identification file content immediately.
Second purpose of the present invention is to provide a kind of file content immediately identifying system and method, and it can identification have the system or the method for the file of ad hoc structure.
The 3rd purpose of the present invention is to provide a kind of file content immediately identifying system and method, and it has the system or the method for emulation people reading file function.
According to above-mentioned purpose of the present invention, the present invention provides a kind of file content immediately identifying system, includes: a document structure analysis device, be used at least one architectural feature according to a file, and this document is marked as several blocks; One reads the flow process setting device, is used to be provided with one and reads flow process to read these blocks; One location device is used for locating one and reads block; And a device for identifying, be used for identification and read block, with export this read in the content of block.
According to above-mentioned purpose of the present invention, the present invention provides a kind of file content immediately identifying method, includes: at least one architectural feature according in the file is marked as several blocks with this document; Be provided with one and read flow process to read these blocks; Middle block is read in location one; And identification this read middle block, with export this read in the content of block.
Utilization the present invention can the various dissimilar file contents of immediately identifying, and for example, books and newspapers, map, music score, project blue print, pipeline wiring diagram etc. have the file of ad hoc structure property.
Under natural scene, authentic document possibly present torsional deformation, and the present invention can utilize vision detecting and the skill of following the trail of, and confirms the position of file, and considers the problem that ornaments are crooked.In addition, can increase the resolution of block image through amplifying the label pad in the file, thereby improve the sense of block content.
The present invention can be applicable to robot and reads on the dissimilar files; It is to adopt with seeing with the skill of reading; Can reach the effect of immediately identifying, and can be under the situation that aquatic foods are artificially got involved less, let robot accomplish the identification of heap file in regular turn and reach the purpose of reading.In addition, also can transfer the file content after the identification to voice signal, let robot according to file content and read aloud.
In the field of robot, the present invention can be applicable to like intelligence development educational robot, amusement and recreation robot, reaches medical auxiliary robot etc., also might be applied to other field.
Description of drawings
Fig. 1 shows the synoptic diagram of file content immediately identifying of the present invention system.
Fig. 2 shows the process flow diagram of file content immediately identifying method of the present invention.
Fig. 3 shows a kind of example that is applied to the discrimination method of the English file of identification.
Embodiment
Fig. 1 shows the synoptic diagram of file content immediately identifying of the present invention system.File content immediately identifying of the present invention system 10 comprises that mainly a document structure analysis device 121, reads flow process setting device 122, a location device 133 and a device for identifying 136.Have some characteristic in a structural file, for example, the paragraph in the English file or with the word of blank spaces etc.The present invention utilizes the characteristic of this structural file, and document structure analysis device 121 becomes several blocks with file mark, reads flow process setting device 122 and is provided with one and reads flow process to read these blocks of document structure analysis device 121 marks.Locating device 133 receives this that read that flow process setting device 122 is provided with and reads flow process.When this reads flow process when execution, the block that is reading locating device 133 location one.After locating device 133 was accomplished this block that is reading of location, this block that is reading of device for identifying 136 identifications was to export the content of this block that is reading.
Fig. 2 shows the process flow diagram of file content immediately identifying method of the present invention.Please be simultaneously with reference to figure 1 and Fig. 2.Be example with the English file of identification below, as embodiments of the invention.
At first, in step S202, whether vision detecting exists with follow-up mechanism 110 detecting files, if exist then confirm the position (step S204) of file.The position of file possibly change the position because of various factors, and at this moment, the vision detecting is searched file with follow-up mechanism 110 in a scope, if find this part file, then replaces the position of original record with new position.
In step S206, when the vision detecting detected file with follow-up mechanism 110, document structure analysis device 121 became block with each word or sign flag with blank spaces, and these blocks are commonly referred to as the word block at this.
In step S208, read flow process setting device 122 flow process that reads that reads by these word blocks of document structure analysis device 121 marks is set.One is the most simply read mode is that basipetal mode reads these word blocks according to right by a left side.
In step S230, according to step S208 set read flow process, locating device 133 is word for word done the action of location to these word blocks.Locating device 133 control one motor 144, with the camera lens of an image capture unit 145 facing to the next word block that will be read.The word block that the camera lens of image capture unit 145 faces representes that this word block is the word block in reading.Locating device 133 is all carried out same positioning step to each word block.
In step S232, the word block capture during image capture unit 145 reads each, the image of being obtained can be deposited into the file of various image formats, like the BMP image shelves of uncompressed, or the JPEG image shelves through compressing.Perhaps, with the image of the being obtained memory that writes direct.Because the resolution of considering to the image data of being obtained is too low, in this step, can amplify this word block in reading, obtain the image data of high-resolution, can solve the problem that is difficult for identification because of the composition pixel of word very little like this.
In step S236, the image data that image capture unit 145 is obtained is sent to device for identifying 136.Device for identifying 136 is with optical character identification (Optical Character Recognition; OCR) image data of this word block in reading of technological identification is then exported the content of this word block.The content of this word block of output can be the character code like ASCII (American Standard Code for Information Interchange), can directly be editor or convert other signal again at general PC.
In step S238, the content of this word block is converted into a voice signal through voice conversion device 137.
More than, if be provided with among the step S208 read flow process and accomplish the time, whether then step S202 detecting is got back to by system has another part file to exist.Otherwise, get back to step S230, continue to carry out location, capture, the identification of next word block.
Moreover locating device 133 also can be located a local block of the word block in reading, and for example, forms the character of this word.At this moment, image capture unit 145 is respectively to each character capture, each character of device for identifying 136 identifications.Afterwards, again with the synthetic word of the character group after the identification.
Fig. 3 shows a kind of example that is applied to the discrimination method of the English file of identification.The image of the word block of being obtained by step S230, S232 can carry out identification according to the following step.With word " robot " is example, at first confirms the target character position, the initial character " r " of word for example, and capture the image (step S356) of this character " r ".With the normalization of the image of this character " r ", that is, the image of acquisition character is adjusted into fixed size (step S358).Transfer the image of this character " r " to black-and-white image, this moment, the colour of each pixel was 0 or 1, that is binaryzation (step S360).Step S362 then captures the characteristic of the digital date after this binaryzation, is attached to the data bank of the sample character set of before being trained.Step S362, with the characteristic and the training sample character set of the character that is captured " r " compare, identification.If all characters " r ", " o ", " b ", " o ", and " t " all accomplish identification, then finish this word of identification, otherwise continue identification character late (step S368).Step S370 continues to confirm next target character position, like " o ".So, again with the synthetic word of the character group after the identification.
Be noted that, when in step S206, a structural file being done the block mark, can use plural architectural feature to carry out the mark of block.For example, can be divided into paragraph, ranks to English file, reach word, carry out the mark of block according to these three kinds of architectural features.Then, set the flow process that reads of these three kinds of structures, for example, read first word of first section first row earlier.
In addition, according to the present invention, except that above-mentioned be that block is the embodiment of identification with the word, be that the embodiment of block can implement equally with paragraph or ranks.
Among the present invention; More particularly; Image capture unit can use the low resolution PTZ video camera (Pan Tilt Zoom camera) that generally is used for video monitoring; This kind video camera can wide-angle rotates, tilts, focusing automatically, high magnification are amplified, and the demand of looking is carried fixing at one or the platform that moves on, be rich in motor-driven and independence.
Claims (10)
1. file content immediately identifying system, it is characterized in that: this system comprises:
One document structure analysis device is used at least one architectural feature according to a file, and this document is marked as several blocks;
One reads the flow process setting device, is used to be provided with one and reads flow process to read these blocks;
One location device is used for locating one and reads block; And
One device for identifying, this reads block to be used for identification, to export the content that this reads middle block.
2. file content immediately identifying according to claim 1 system, it is characterized in that: this system also comprises a voice conversion device, is used for converting this content that reads block into a voice signal.
3. file content immediately identifying according to claim 1 system, it is characterized in that: this locating device reads middle block through controlling a motor to locate this.
4. file content immediately identifying according to claim 1 system; It is characterized in that: this system also comprises an image capture unit; This reads block and becomes an image data to be used for capture; Wherein this device for identifying identification this read in the image data of block, with export this read in the content of block.
5. file content immediately identifying according to claim 1 system is characterized in that: this locating device location this read in a local block of block, this device for identifying identification should the part block, to export the content of this part block.
6. file content immediately identifying method, it is characterized in that: this method comprises:
At least one architectural feature according in the file is marked as several blocks with this document;
Be provided with one and read flow process to read these blocks;
Middle block is read in location one; And
This reads middle block identification, to export the content that this reads middle block.
7. file content immediately identifying method according to claim 6 is characterized in that: it is a voice signal that this method also comprises this content that reads middle block of conversion.
8. file content immediately identifying method according to claim 6; It is characterized in that: this method also comprise capture this read in block become an image data; Wherein in the step of identification, be identification this read in the image data of block, with export this read in the content of block.
9. file content immediately identifying method according to claim 6 is characterized in that: this method also comprises the location, and this reads a local block of middle block.
10. file content immediately identifying method according to claim 9 is characterized in that: this method also comprises identification should the part block, to export the content of this part block.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 200810130719 CN101615253B (en) | 2008-06-27 | 2008-06-27 | System and method for instantly identifying file contents |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 200810130719 CN101615253B (en) | 2008-06-27 | 2008-06-27 | System and method for instantly identifying file contents |
Publications (2)
Publication Number | Publication Date |
---|---|
CN101615253A CN101615253A (en) | 2009-12-30 |
CN101615253B true CN101615253B (en) | 2012-12-05 |
Family
ID=41494883
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN 200810130719 Expired - Fee Related CN101615253B (en) | 2008-06-27 | 2008-06-27 | System and method for instantly identifying file contents |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101615253B (en) |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1773523A (en) * | 2004-11-08 | 2006-05-17 | 乐金电子(昆山)电脑有限公司 | Character identification and sound outputting apparatus and method for portable infomation terminal machine with photographic head |
CN200997199Y (en) * | 2007-01-24 | 2007-12-26 | 蒋清晓 | Automatic intelligent reader for blind |
CN201153280Y (en) * | 2008-01-28 | 2008-11-19 | 中兴通讯股份有限公司 | Mobile phone having traffic light recognition function |
-
2008
- 2008-06-27 CN CN 200810130719 patent/CN101615253B/en not_active Expired - Fee Related
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1773523A (en) * | 2004-11-08 | 2006-05-17 | 乐金电子(昆山)电脑有限公司 | Character identification and sound outputting apparatus and method for portable infomation terminal machine with photographic head |
CN200997199Y (en) * | 2007-01-24 | 2007-12-26 | 蒋清晓 | Automatic intelligent reader for blind |
CN201153280Y (en) * | 2008-01-28 | 2008-11-19 | 中兴通讯股份有限公司 | Mobile phone having traffic light recognition function |
Also Published As
Publication number | Publication date |
---|---|
CN101615253A (en) | 2009-12-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US7805307B2 (en) | Text to speech conversion system | |
EP0920189A3 (en) | Image processing apparatus | |
CN1619580A (en) | Information identification method of full-filling information card | |
EP1434419A3 (en) | Image processing apparatus and image processing method | |
US20060290999A1 (en) | Image processing apparatus and network system | |
EP1920852A3 (en) | Mail sorting system and method of sorting mails | |
CN101615253B (en) | System and method for instantly identifying file contents | |
US20090324139A1 (en) | Real time document recognition system and method | |
JP2012194837A (en) | Image processing device, method, program, and recording medium | |
JP2011258129A (en) | Handwritten character separation device, handwritten character separation method, and handwritten character separation program | |
CN100511267C (en) | Graph and text image processing equipment and image processing method thereof | |
KR19990006421A (en) | A system for processing and displaying information relating to an image captured by a camera | |
JP4383429B2 (en) | Form image processing method and apparatus | |
JP4894184B2 (en) | Teaching material processing apparatus, teaching material processing method, and teaching material processing program | |
JP2508975B2 (en) | Electronic blackboard | |
JPS5572281A (en) | Collating method for print of seal | |
CN114926850A (en) | Document identification method, device, equipment and medium | |
CN117649670A (en) | Document layout analysis model training method, application method, computer device and computer readable storage medium | |
JP2001291085A (en) | Method and system for registering electronic file | |
JP2823350B2 (en) | Multimedia input device | |
JPS638988A (en) | Character reader | |
JP2000020677A (en) | Image capture/massage display device | |
JPS58207184A (en) | Recording information recognizer | |
JPH01284988A (en) | Method and device for information processing | |
JPS6325772A (en) | Filing system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20121205 Termination date: 20160627 |