CN107392212A - A kind of image information method for quickly identifying - Google Patents
A kind of image information method for quickly identifying Download PDFInfo
- Publication number
- CN107392212A CN107392212A CN201710590654.8A CN201710590654A CN107392212A CN 107392212 A CN107392212 A CN 107392212A CN 201710590654 A CN201710590654 A CN 201710590654A CN 107392212 A CN107392212 A CN 107392212A
- Authority
- CN
- China
- Prior art keywords
- array
- value
- image
- database
- dimension array
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/758—Involving statistics of pixels or of feature values, e.g. histogram matching
Abstract
The invention provides a kind of image information method for quickly identifying, it is characterised in that:Target image information is represented with the numerical matrix or sequence being made up of the first value and second value;The database for including various image informations is established, the database represents reference image information by the referential data matrix or sequence being made up of the first value and second value;The array of target image is divided into more parts, the array of reference picture in database is also divided into equal parts, then by the array of target image and reference picture it is overall and and relevant position each part and contrasted respectively, the successfully then further content of in one's duty first value of detailed comparisons and second value is contrasted, until contrast successfully reading completely.By first rough contrast, the mode of detailed contrast again is contrasted successfully roughly, is avoided to all detailed contrast identification of each reference picture in database, is improved information and search plain matching degree and search plain efficiency, save match time.
Description
Technical field
The present invention relates to a kind of image information method for quickly identifying, available for identification image, figure, Chinese character, letter, number
The target informations such as word, fingerprint, face, belong to information discriminating technology field.
Background technology
At present, the identification of the information such as image, figure, Chinese character, letter, numeral, fingerprint, face is to commonly use in life and extremely close
The technology of key.Now, monitor is assembled with public places such as airport, stadium, supermarkets, crowd is monitored.
Such as monitoring system is installed on airport, to prevent terrorist's boarding.Such as the Automatic Teller Machine of bank, user's card and password
It is stolen, enchashment gold will be emitted by other people.
Therefore, the target information method for quickly identifying such as a kind of image, fingerprint, face how is researched and developed, to avoid the above situation
Generation, save the extensive work time, improve operating efficiency, be those skilled in the art be directed to solve problem.
The content of the invention
The technical problem to be solved in the present invention is to provide one kind can quickly identify image, figure, Chinese character, letter, numeral,
The method of the target informations such as fingerprint, face.
In order to solve the above-mentioned technical problem, the technical scheme is that providing a kind of image information method for quickly identifying,
It is characterized in that:Step is:
Step 1:Target image information is amplified, causes the information on image by pixel unit group until image is amplified to
Into, white pixel is then arranged to the first value, non-white color pixel is arranged to second value, to obtain by the first value and second
It is worth the numerical matrix or sequence of the target image of composition;
Step 2:The method of installation steps 1, establish and include the databases of various image informations, the database passes through by the
The referential data matrix or sequence of one value and second value composition represent reference image information;
Step 3:When the numerical matrix or sequence of the target image are one-dimension array, the one-dimension array of target image is drawn
It is divided into more parts, the one-dimension array of the reference picture in database is also divided into equal parts, then by target image and reference
The one-dimension array of image it is overall and and relevant position each part and contrasted respectively, contrast successfully then further detailed
The content of in one's duty first value and second value is contrasted, is successfully read until contrasting completely;
When the numerical matrix or sequence of the target image are two-dimensional array, by its all row and in a row, become one-dimensional
Array, then identified with the method for one-dimension array;Or regard each row of two-dimensional array as an one-dimension array respectively, then with one-dimensional
The method of array identifies each row of two-dimensional array successively.
Preferably, in the step 1, white pixel is arranged to 0, non-white color pixel and is arranged to 1, or white pixel is set
It is set to 1, non-white color pixel and is arranged to 0, so information on equal to whole image is to be ranked into number by many 0 and 1 is regular
What group was shown.
It is highly preferred that in the step 3, the one-dimension array of target image is divided into { A, B, C } three form, A, B, C divide
Not Biao Shi target image the overall sum of one-dimension array, the sum of the one-dimension array first half of target image, the one of target image
The sum of dimension group latter half;The one-dimension array of reference picture in database is also divided by identical method, by target image
A, B, C and database in reference picture A, B, C contrast, contrast successfully then further 0 and in detailed comparisons A, B, C
1 content, successfully read until contrasting completely.
Further, by first rough contrast, the mode of detailed contrast again is contrasted successfully roughly, is avoided in database
The all detailed contrast identification of each reference picture, improve information and search and plain matching degree and search plain efficiency, save match time.
The present invention represents image information by 0~1 array, and image is divided into more parts, using array it is overall and it is and each
Part and contrasted respectively, contrast successfully then further detailed comparisons;By first rough contrast, contrast roughly successfully detailed again
The mode contrasted to the greatest extent, avoid to all detailed contrast identification of each reference picture in database, improve information and search plain matching degree
With search plain efficiency, save match time, can quickly identify the mesh such as image, figure, Chinese character, letter, numeral, fingerprint, face
Mark information.
Brief description of the drawings
Fig. 1 is the one-dimension array recognition methods flow chart that image information quickly identifies;
Fig. 2 is the two-dimensional array recognition methods flow chart that image information quickly identifies;
Fig. 3 is schematic diagram of the Chinese character " four " as specific identification object.
Embodiment
With reference to specific embodiment, the present invention is expanded on further.It should be understood that these embodiments are merely to illustrate the present invention
Rather than limitation the scope of the present invention.In addition, it is to be understood that after the content of the invention lectured has been read, people in the art
Member can make various changes or modifications to the present invention, and these equivalent form of values equally fall within the application appended claims and limited
Scope.
Fig. 1 is the flow chart of the one-dimension array recognition methods of image information method for quickly identifying, and this method includes following step
Suddenly:
First step S1:Target information picture is amplified, causes the information on picture by pixel list until picture is amplified to
Position composition, white pixel is then arranged to the first value, non-white color pixel is arranged to second value, with obtain by the first value with
White pixel (in embodiments of the present invention, is arranged to 0, non-white picture by the picture numerical matrix or sequence of second value composition
Element is arranged to 1, is to be ranked into what array was shown by many 0 and 1 is regular so equal to the information on whole pictures.
If the one-dimension array (matrix) of pictorial information is [1010010010......10110011010].
Meanwhile establish for identify image, figure, Chinese character, letter, numeral, fingerprint, face information database, it is described
Reference picture, figure, Chinese character, the word that database includes the referential data matrix being made up of the first value and second value or sequence represents
Mother, numeral, fingerprint, face information.
Second step S2:The one-dimension array of pictorial information is divided into { A, B, C } form, A, B, C represent that the array is whole respectively
The sum of body, the array first half and, the array latter half and, only need to by A, B, C and database A, B, C contrast, it is right
Then read than success;
3rd step S3:Or one-dimension array is divided into { A, B1, B2, C1, C2 } form, A, B1, B2, C1, C2 are represented respectively
Overall sum of the array, the first half of the array first half and, the latter half of the array first half with after the array
The first half of half part and, the latter half of the array latter half and, only need to be by A, B1, B2, C1, C2 and database
A, B1, B2, C1, C2 are contrasted, and are contrasted and are successfully then read;
4th step S4:Or by that analogy, the one-dimension array of pictorial information is divided into
[1010010010......10110011010] [A, N1, N2 ... Nk] form, A, Ni (i=1,2 ..., k) represent the number respectively
Overall sum of group, array Ni parts and, first will [A, N1, N2 ... Nk] and database in each one-dimension array A, N1,
N2 ... Nk is contrasted, and contrasts successfully further detailed comparisons, relatively then identifies target after success in detail.
For two-dimensional array (matrix), one-dimension array can be become, by its all row and in a row so as to use one-dimension array
Method identification.Also can be identified with the specific process of two-dimensional array, it is specific such as Fig. 2.
Step 1:If original picture information composition is m rows, the two-dimensional array (matrix) that n is arranged
[1010010010......10110011010]
[1011010011......10110111010]
......
[0011010110......01110111010]
Step 2:Two-dimensional array (matrix) is divided into
[1010010010......10110011010][A1、B1、C1]
[1011010011......10110111010][A2、B2、C2]
[0011010110......01110111010][Ak、Bk、Ck]
[d1、d2、......dn][Ak1、Bk1、Ck1]
[e1、e2、......en][Ak2、Bk2、Ck2]
[f1、f2、......fn][Ak3、Bk3、Ck3]
D1, d2 (j=1,2..., n) ..., dn represents the sum of array jth row respectively;E1, e2 (j=1,2...,
N) ..., en represent respectively the array jth row first half and;F1, f2 (j=1,2..., n) ..., fn difference tables
Show the array jth row latter half and;Its Ai, Bi, Ci (i=1,2..., k, k1, k2, k3) represent that the row of array i-th is whole respectively
The sum of body, first half and, latter half and;First Ai, Bi, Ci of each two-dimensional array in Ai, Bi, Ci and database are contrasted,
Successfully further detailed comparisons are contrasted, relatively then identify target after success in detail.
Step 3:Or two-dimensional array (matrix) is divided into
[1010010010......10110011010][A1、N11、N12、…N1k]
[1011010011......10110111010][A2、N21、N22、…N2k]
……
[0011010110......01110111010][Am、Nm1、Nm2、…Nmk]
[b11、b12、......b1n]
[b21、b22、......b2n]
……
[bx1、bx2、......bxn]
B11, b12 ..., b1n (j=1,2 ..., n) represent the sum of array jth row respectively;Bi1, bi2 (i=2,3 ...,
S, j=1,2 ..., n) ... ..., bsn represent respectively the array jth row i-th section and;Wherein A, Nuv (v=1,2 ..., k < n, u
=1,2 ..., 1 < m) represent respectively the array u rows v part and;First by bij, Auv respectively with each two-dimemsional number in database
The bij of group, Auv contrast, contrast successfully further detailed comparisons, relatively then identify target after success in detail.
To sum up, array (matrix) is divided into { A, B, C } form by the present invention first, and its A, B, C represent that the array is whole respectively
The sum of body, the array first half and, the array latter half and, by A, B, C and database A, B, C contrast, that is, pass through
Matrix contrasts with matrix, unsuccessful then to continue to contrast with the other information in database, until finding corresponding information, contrast
Successful then reading.The Target Photo such as this image, figure, Chinese character, letter, numeral, fingerprint, face information has carried out first
Wheel " and " rapid comparison, a large amount of unnecessary detailed contrast identifications are eliminated, information is improved and searches plain matching degree and search plain effect
Rate, save match time.
In embodiments of the present invention, these 0 and 1 are found out by way of amplification picture one by one, for target
Information, it is only necessary to identify Chinese character and letter, also have numeral, so the present invention will establish special character library, by identifying on picture
Information, contrasted with the Chinese character in character library, letter, numeral, then read out corresponding information, can thus be extracted
The information gone out in picture, i.e., it similarly can recognize that the target informations such as image, fingerprint, face.
Chinese character " four " as shown in Figure 3, form are that the Song typeface is small by four, and the amplification of its sectional drawing is just obtained to many white blocks by 16*16
With black fast composition, white block is set to 0, black patch is set to 1, obtains as follows:
Two-dimensional array (matrix):
0000000000000000
0000000000000000
0111111111111100
0100010001000100
0100010001000100
0100010001000100
0100010001000100
0100010001000100
0100100001000100
0100100000111100
0101000000000100
0110000000000100
0100000000000100
0111111111111100
0100000000000100
0000000000000000
First method, by one-dimension array method of identification:
The information in picture is identified, and reads information therein.It is to amplify picture first, Chinese character therein is become and is imaged
The matrix of element composition, white are 0, non-white 1, then to become one-dimension array as follows:
″00000000000000000000000000000000011111111111110001000100010001000100
01000100010001000100010001000100010001000100010001000100010001001000010001000
10010000011110001010000000001000110000000000100010000000000010001111111111111
0001000000000001000000000000000000″
Exemplified by the form of { A, B, C }, one-dimension array is divided into { A, B, C } form first, its A, B, C represent the number respectively
Overall sum of group, the array first half and, the array latter half with.{ A, B, C } form calculus result corresponding to the array
For { 66,33,33 }, will the array be changed into:
″00000000000000000000000000000000011111111111110001000100010001000100
01000100010001000100010001000100010001000100010001000100010001001000010001000
10010000011110001010000000001000110000000000100010000000000010001111111111111
0001000000000001000000000000000000 66 33 33″., first will be identified during matching is searched for
In A, B, C and character library of image array each array A, B, C contrast, if { A, B, C } is unequal, skip contrast it is next, if phase
Equal to identify successfully Deng the array of detailed comparisons 0~1 again, unequal continuation contrasts with the other information in storehouse, until finding phase
Corresponding information, contrast and successfully then read.The searching method eliminates many detailed 0~1 array matchings, significantly carries
High search recognition efficiency.
Second method, by two-dimensional array (matrix) method of identification:
Illustrated by taking each three sections of segmentations read group total of row and column as an example:First by numerical matrix or the full line of sequence, OK
First half, row latter half summation, adds the columns and rows that new picture element matrix is formed after picture element matrix.Therefore picture element matrix becomes
For:
[0000000000000000 0 0 0
0000000000000000 0 0 0
0111111111111100 1 3 7 6
0100010001000100 4 2 2
0100010001000100 4 2 2
0100010001000100 4 2 2
0100010001000100 4 2 2
0100010001000100 4 2 2
0100100001000100 4 2 2
0100100000111100 6 2 4
0101000000000100 3 2 1
0110000000000100 3 2 1
0100000000000100 2 1 1
0111111111111100 13 7 6
0100000000000100 2 1 1
0000000000000000 0 0 0
0 13 3 3 4 7 2 2 2 8 3 3 3 13 0 0
0 6 1 1 1 6 1 1 1 6 1 1 1 6 0 0
0 7 2 2 3 1 1 1 1 2 2 2 2 7 0 0]
During matching is searched for, first by the row { A, B, C } of identified image array, arrange in { D, E, F } and character library
The row { A, B, C } of each array, row { D, E, F } contrast, if row { A, B, C }, row { D, E, F } it is unequal, skip contrast it is next, if
The equal array of detailed comparisons again 0~1, equal to identify successfully, unequal continuation contrasts with the other information in storehouse, until finding
Corresponding information, contrast and successfully then read.By row { A, B, C }, row { D, E, F } searching method eliminates many detailed
0~1 array matching, search recognition efficiency is greatly improved.
Example is divided into above with three sections to be illustrated.The patented method can be according to the picture element matrix of search identification object
Scale, summation and the segmentation summation search matching of row any section are split using row any section.If row any section segmentation summation or
Row any section segmentation summation it is unequal, skip contrast it is next, if the equal array of detailed comparisons again 0~1.It is equal to be identified as
Work(, unequal continuation contrast with other picture element matrix information in storehouse, until finding corresponding information, contrast and successfully then read
.Summation is split by row any section and row any section segmentation summation searching method eliminates many 0~1 detailed arrays
Match somebody with somebody, search recognition efficiency is greatly improved.
It is described above, only presently preferred embodiments of the present invention, it is not any to the present invention in form and substantial limitation,
It should be pointed out that for those skilled in the art, on the premise of the inventive method is not departed from, can also make
Some improvement and supplement, these are improved and supplement also should be regarded as protection scope of the present invention.All those skilled in the art,
Without departing from the spirit and scope of the present invention, when made using disclosed above technology contents it is a little more
Dynamic, modification and the equivalent variations developed, it is the equivalent embodiment of the present invention;Meanwhile all substantial technologicals pair according to the present invention
The variation, modification and evolution for any equivalent variations that above-described embodiment is made, still fall within the scope of technical scheme
It is interior.
Claims (4)
1. a kind of image information method for quickly identifying, it is characterised in that step is:
Step 1:Target image information is amplified, until image is amplified to so that the information on image is made up of pixel unit
, white pixel is then arranged to the first value, non-white color pixel is arranged to second value, to obtain by the first value and second value
The numerical matrix or sequence of the target image of composition;
Step 2:The method of installation steps 1, establishes the database for including various image informations, and the database passes through by the first value
Reference image information is represented with the referential data matrix or sequence of second value composition;
Step 3:When the numerical matrix or sequence of the target image are one-dimension array, the one-dimension array of target image is divided into
More parts, the one-dimension array of the reference picture in database is also divided into equal parts, then by target image and reference picture
One-dimension array it is overall and and relevant position each part and contrasted respectively, contrast successfully then further detailed comparisons
The content of in one's duty first value and second value, successfully reads until contrasting completely;
When the numerical matrix or sequence of the target image are two-dimensional array, by its all row and in a row, become one-dimension array,
Identified again with the method for one-dimension array;Or regard each row of two-dimensional array as an one-dimension array respectively, then use one-dimension array
Method identify each row of two-dimensional array successively.
A kind of 2. image information method for quickly identifying as claimed in claim 1, it is characterised in that:In the step 1, by white
Pixel is arranged to 0, non-white color pixel and is arranged to 1, or white pixel is arranged into 1, non-white color pixel and is arranged to 0, is so equal to
Information on whole image is to be ranked into what array was shown by many 0 and 1 is regular.
A kind of 3. image information method for quickly identifying as claimed in claim 2, it is characterised in that:In the step 3, by target
The one-dimension array of image is divided into { A, B, C } three form, and A, B, C represent the overall sum of the one-dimension array of target image, mesh respectively
The sum of the one-dimension array first half of logo image, the sum of the one-dimension array latter half of target image;Reference chart in database
The one-dimension array of picture is also divided by identical method, by A, B, C couple of the reference picture in A, B, C and database of target image
Than then further 0 and 1 content in detailed comparisons A, B, C being contrasted successfully, until contrast successfully reading completely.
A kind of 4. image information method for quickly identifying as claimed in claim 3, it is characterised in that:Contrasted by first rough, slightly
The successfully mode of detailed contrast again is slightly contrasted, avoids to all detailed contrast identification of each reference picture in database, improves
Information searches plain matching degree and searches plain efficiency, saves match time.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710590654.8A CN107392212A (en) | 2017-07-19 | 2017-07-19 | A kind of image information method for quickly identifying |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710590654.8A CN107392212A (en) | 2017-07-19 | 2017-07-19 | A kind of image information method for quickly identifying |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107392212A true CN107392212A (en) | 2017-11-24 |
Family
ID=60337095
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710590654.8A Pending CN107392212A (en) | 2017-07-19 | 2017-07-19 | A kind of image information method for quickly identifying |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107392212A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109325138A (en) * | 2018-07-12 | 2019-02-12 | 上海电机学院 | A kind of image rapid identification method based on extension in conjunction with sub-pix matrix |
CN110598020A (en) * | 2019-09-23 | 2019-12-20 | 北京永亚普信科技有限责任公司 | Binary image retrieval method |
Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020135018A1 (en) * | 2001-03-23 | 2002-09-26 | Mitsubishi Denki Kabushiki Kaisha | Thin film magnetic memory device writing data of a plurality of bits in parallel |
US20070271013A1 (en) * | 2006-05-18 | 2007-11-22 | Applied Perception Inc. | Vision guidance system and method for identifying the position of crop rows in a field |
CN101122995A (en) * | 2007-09-11 | 2008-02-13 | 北京大学 | Binary image digit water mark embedding, extraction method and device |
CN101183430A (en) * | 2007-12-13 | 2008-05-21 | 中国科学院合肥物质科学研究院 | Handwriting digital automatic identification method based on module neural network SN9701 rectangular array |
CN101207405A (en) * | 2007-12-14 | 2008-06-25 | 西安华迅微电子有限公司 | Method for capturing pseudo-code sequence |
CN101206259A (en) * | 2007-12-13 | 2008-06-25 | 西安华迅微电子有限公司 | Method for capturing multi-constellation navigation weak signal based on digital matched filter |
CN101630365A (en) * | 2009-06-05 | 2010-01-20 | 天津大学 | Method for extracting and describing DAISY-based feature with mirror face turning invariance |
CN101695135A (en) * | 2009-10-20 | 2010-04-14 | 大连捷成实业发展有限公司 | SDI numerical error code detection based on FPGA and correction algorithm thereof |
CN101739565A (en) * | 2009-12-04 | 2010-06-16 | 华东交通大学 | Large-capacity pattern recognition method |
CN101877065A (en) * | 2009-11-26 | 2010-11-03 | 南京信息工程大学 | Extraction and identification method of non-linear authentication characteristic of facial image under small sample condition |
CN102622608A (en) * | 2012-02-23 | 2012-08-01 | 北京航空航天大学 | High-spectrum image classification method based on nonlinear time series analysis |
CN103118413A (en) * | 2013-02-04 | 2013-05-22 | 重庆邮电大学 | IPv6 (Internet protocol version 6) industrial wireless sensor network routing method based on definiteness scheduling |
CN103336983A (en) * | 2013-06-08 | 2013-10-02 | 上海电机学院 | Bar-code-based bill generation system and recognition method thereof |
CN103543989A (en) * | 2013-11-11 | 2014-01-29 | 镇江中安通信科技有限公司 | Adaptive parallel processing method aiming at variable length characteristic extraction for big data |
CN104468090A (en) * | 2014-11-12 | 2015-03-25 | 辽宁大学 | Chinese character password encoding method based on image pixel coordinates |
CN105913218A (en) * | 2016-04-01 | 2016-08-31 | 上海电机学院 | Electronic invoice invoicing reimbursement method and electronic invoice reimbursement information extraction method |
-
2017
- 2017-07-19 CN CN201710590654.8A patent/CN107392212A/en active Pending
Patent Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020135018A1 (en) * | 2001-03-23 | 2002-09-26 | Mitsubishi Denki Kabushiki Kaisha | Thin film magnetic memory device writing data of a plurality of bits in parallel |
US20070271013A1 (en) * | 2006-05-18 | 2007-11-22 | Applied Perception Inc. | Vision guidance system and method for identifying the position of crop rows in a field |
CN101122995A (en) * | 2007-09-11 | 2008-02-13 | 北京大学 | Binary image digit water mark embedding, extraction method and device |
CN101183430A (en) * | 2007-12-13 | 2008-05-21 | 中国科学院合肥物质科学研究院 | Handwriting digital automatic identification method based on module neural network SN9701 rectangular array |
CN101206259A (en) * | 2007-12-13 | 2008-06-25 | 西安华迅微电子有限公司 | Method for capturing multi-constellation navigation weak signal based on digital matched filter |
CN101207405A (en) * | 2007-12-14 | 2008-06-25 | 西安华迅微电子有限公司 | Method for capturing pseudo-code sequence |
CN101630365A (en) * | 2009-06-05 | 2010-01-20 | 天津大学 | Method for extracting and describing DAISY-based feature with mirror face turning invariance |
CN101695135A (en) * | 2009-10-20 | 2010-04-14 | 大连捷成实业发展有限公司 | SDI numerical error code detection based on FPGA and correction algorithm thereof |
CN101877065A (en) * | 2009-11-26 | 2010-11-03 | 南京信息工程大学 | Extraction and identification method of non-linear authentication characteristic of facial image under small sample condition |
CN101739565A (en) * | 2009-12-04 | 2010-06-16 | 华东交通大学 | Large-capacity pattern recognition method |
CN102622608A (en) * | 2012-02-23 | 2012-08-01 | 北京航空航天大学 | High-spectrum image classification method based on nonlinear time series analysis |
CN103118413A (en) * | 2013-02-04 | 2013-05-22 | 重庆邮电大学 | IPv6 (Internet protocol version 6) industrial wireless sensor network routing method based on definiteness scheduling |
CN103336983A (en) * | 2013-06-08 | 2013-10-02 | 上海电机学院 | Bar-code-based bill generation system and recognition method thereof |
CN103543989A (en) * | 2013-11-11 | 2014-01-29 | 镇江中安通信科技有限公司 | Adaptive parallel processing method aiming at variable length characteristic extraction for big data |
CN104468090A (en) * | 2014-11-12 | 2015-03-25 | 辽宁大学 | Chinese character password encoding method based on image pixel coordinates |
CN105913218A (en) * | 2016-04-01 | 2016-08-31 | 上海电机学院 | Electronic invoice invoicing reimbursement method and electronic invoice reimbursement information extraction method |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109325138A (en) * | 2018-07-12 | 2019-02-12 | 上海电机学院 | A kind of image rapid identification method based on extension in conjunction with sub-pix matrix |
CN109325138B (en) * | 2018-07-12 | 2022-07-15 | 上海电机学院 | Image rapid identification method based on combination of expansion and sub-pixel matrix |
CN110598020A (en) * | 2019-09-23 | 2019-12-20 | 北京永亚普信科技有限责任公司 | Binary image retrieval method |
CN110598020B (en) * | 2019-09-23 | 2022-05-03 | 北京永亚普信科技有限责任公司 | Binary image retrieval method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Paliwal et al. | Tablenet: Deep learning model for end-to-end table detection and tabular data extraction from scanned document images | |
Kwon | Threshold selection based on cluster analysis | |
Li et al. | A novel approach to hyperspectral band selection based on spectral shape similarity analysis and fast branch and bound search | |
Yang et al. | A novel copy-move forgery detection algorithm via two-stage filtering | |
Shuai et al. | Fingerprint indexing based on composite set of reduced SIFT features | |
Alvarez et al. | Texton theory revisited: A bag-of-words approach to combine textons | |
Chen et al. | Using binarization and hashing for efficient SIFT matching | |
Roy et al. | Document seal detection using GHT and character proximity graphs | |
Ding et al. | Single sample per person face recognition with KPCANet and a weighted voting scheme | |
Singh et al. | Content-based image retrieval based on supervised learning and statistical-based moments | |
Zhou et al. | Enhanced locality-sensitive hashing for fingerprint forensics over large multi-sensor databases | |
Song et al. | Robust and parallel Uyghur text localization in complex background images | |
CN107392212A (en) | A kind of image information method for quickly identifying | |
Wan et al. | Data driven multi-index hashing | |
Pengcheng et al. | Fast Chinese calligraphic character recognition with large-scale data | |
CN109325138B (en) | Image rapid identification method based on combination of expansion and sub-pixel matrix | |
Li et al. | A combined feature representation of deep feature and hand-crafted features for person re-identification | |
Tankala et al. | A content-based image retrieval scheme with object detection and quantised colour histogram | |
Saravanan | Information retrieval using image attribute possessions | |
Foo et al. | Discovery of image versions in large collections | |
Goswami et al. | High level shape representation in printed Gujarati character | |
Sun et al. | Product Re-identification System in Fully Automated Defect Detection | |
CN108520079B (en) | Migo search engine | |
Cen et al. | An improved algorithm for locality-sensitive hashing | |
Singh et al. | An algorithm and hardware design for very fast similarity search in high dimensional space |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171124 |