CN108615058A - A kind of method, apparatus of character recognition, equipment and readable storage medium storing program for executing - Google Patents
A kind of method, apparatus of character recognition, equipment and readable storage medium storing program for executing Download PDFInfo
- Publication number
- CN108615058A CN108615058A CN201810442986.6A CN201810442986A CN108615058A CN 108615058 A CN108615058 A CN 108615058A CN 201810442986 A CN201810442986 A CN 201810442986A CN 108615058 A CN108615058 A CN 108615058A
- Authority
- CN
- China
- Prior art keywords
- character
- images
- recognized
- matching
- fuzzy
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- 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
Abstract
The invention discloses a kind of method, apparatus of character recognition, equipment and computer readable storage mediums, including：Collected images to be recognized is pre-processed, in order to obtain each character in the images to be recognized；Each character in the images to be recognized is subjected to template matches with the fuzzy matching template library pre-established respectively, obtains the preliminary matching result and matching rate of each character in the images to be recognized；To there is the character of multiple matching results and/or low matching rate to carry out structure characteristic analysis in the images to be recognized, the final matching result for having the character of multiple matching results and/or low matching rate in the images to be recognized is obtained.Method, apparatus, equipment and computer readable storage medium provided by the present invention improve the accuracy rate and efficiency of character recognition.
Description
Technical field
The present invention relates to area of pattern recognition, more particularly to a kind of method, apparatus of character recognition, equipment and calculating
Machine readable storage medium storing program for executing.
Background technology
Pattern-recognition be current image processing field an important branch, by with machine replace human eye to things into
Row judges, can greatly improve the speed and efficiency of identification, has higher application value.Character recognition technologies can help people
The multiple-task in life is completed, application is very extensive, thus has obtained quick development.
Using existing character recognition technologies identification complex background (influence of the factors such as color, illumination), image inclination, deformation,
When damaged, fuzzy character to be identified, identification error is larger.The method of character recognition generally comprises template matches in the prior art
Method and structure recognition algorithm.It is straight to be difficult to find that the template form for being generally applicable to all characters is realized in traditional template matching method
The matching primitives connect, and be difficult to distinguish the similar character of shape, therefore traditional template matching method is wanted in character recognition accuracy rate
It is difficult to meet the requirements in the case of asking higher.Although and structure recognition algorithm can identify that complicated pattern, needs treat knowledge
The stroke feature of malapropism symbol extracts, and when character stroke is more, calculation amount is very big, has seriously affected the effect of character recognition
Rate；And character to be identified quality of input image it is bad when, structure recognition algorithm tends not to extract the stroke of character to be identified
Feature.
In summary as can be seen that how to improve the accuracy rate of character recognition and efficiency is to have problem to be solved at present.
Invention content
The object of the present invention is to provide a kind of method, apparatus of character recognition, equipment and computer readable storage medium,
Character recognition accuracy rate in the prior art and less efficient is solved the problems, such as.
In order to solve the above technical problems, the present invention provides a kind of method of character recognition, including：To collected to be identified
Image is pre-processed, in order to obtain each character in the images to be recognized；By each word in the images to be recognized
Symbol carries out template matches with the fuzzy matching template library that pre-establishes respectively, obtain each character in the images to be recognized just
The matching result and matching rate of step；To there is the progress of the character of multiple matching results and/or low matching rate in the images to be recognized
Structure characteristic analysis obtains final for having the character of multiple matching results and/or low matching rate in the images to be recognized
With result.
Preferably, the fuzzy matching template library includes：Arabic numerals, English alphabet and Chinese character；Wherein it is described I
Uncle's number includes 0 to 9, and the English alphabet includes 26 capitalizations：A to Z, the Chinese character include the Chinese Character of predetermined number
Symbol.
Preferably, the fuzzy matching template library that pre-establishes includes：
Several reference pictures for meeting preset condition are chosen for each character of the fuzzy matching template library, using described
Halcon algorithms read every width reference picture corresponding to each character respectively；Using the halcon algorithms to each word
The corresponding every width reference picture of symbol carries out denoising, gray processing, binaryzation, removal noise spot, becomes a full member, standardized processing, from
And obtain several corresponding standard bianry images that foreground character pixels are 1 and background pixel is 0 described in each character；It will be described
It is normalized after being superimposed after whole standard bianry images alignment corresponding to each character, to obtain each word
The Fuzzy Template of symbol；Fuzzy Template corresponding to English alphabet character and numerical character is a kind of Fuzzy Template library；Chinese character
Corresponding Fuzzy Template is two class Fuzzy Template libraries.
Preferably, described that collected images to be recognized is pre-processed, in order to obtain in the images to be recognized
Each character includes：
Images to be recognized is acquired, denoising, gray processing, feature extraction are carried out to the images to be recognized using halcon algorithms
And remove the operation of noise spot other than character；Using halcon segmentation operators to the figure to be identified of noise spot other than removal character
As being split, to be partitioned into each character in the images to be recognized；It extracts successively each in the images to be recognized
A character, and become a full member to each character after extraction and standardization, in order to obtain and the Fuzzy Template library
The binary image for each character that graphics standard matches.
Preferably, each character by the images to be recognized respectively with the fuzzy matching template library that pre-establishes
Template matches are carried out, the preliminary matching result for obtaining each character in the images to be recognized includes with matching rate：
Judge each character generic in pretreated images to be recognized；It respectively will be every in the images to be recognized
A character is compared one by one with all template characters in the Fuzzy Template library corresponding to each character；Corresponding to each character
Comparison result in meet the template character of preset condition, the matching result as the character exports.
Preferably, described to thering are multiple matching results and/or the character of low matching rate to tie in the images to be recognized
Structure signature analysis obtains the final matching for having the character of multiple matching results and/or low matching rate in the images to be recognized
As a result include：
Using structure feature recognizer, analyzing has multiple matching results and/or low matching rate in the images to be recognized
Character structure feature；To there are multiple matching results and/or low matching rate character to be similar to character with it and carry out default aspect ratio
Right, the comparison result by presetting feature determines the word for having multiple matching results and/or low matching rate in the images to be recognized
The final matching results of symbol.
The present invention also provides a kind of devices of character recognition, including：
Preprocessing module, for being pre-processed to collected images to be recognized, in order to obtain the figure to be identified
Each character as in；
Fuzzy matching module, for by each character in the images to be recognized respectively with the fuzzy matching that pre-establishes
Template library carries out template matches, obtains the preliminary matching result and matching rate of each character in the images to be recognized；
Characteristics analysis module, for the character to having multiple matching results and/or low matching rate in the images to be recognized
Structure characteristic analysis is carried out, obtaining has the final of the character of multiple matching results and/or low matching rate in the images to be recognized
Matching result.
Preferably, the preprocessing module is specifically used for：Images to be recognized is acquired, waits knowing to described using halcon algorithms
Other image carries out denoising, gray processing, feature extraction and the operation for removing noise spot other than character；Utilize halcon segmentation operators
The images to be recognized of noise spot other than removal character is split, to be partitioned into each word in the images to be recognized
Symbol；Each character in the images to be recognized is extracted successively, and each character after extraction is become a full member and standardized
Processing, in order to obtain the binary image of each character to match with Fuzzy Template library graphics standard.
The present invention also provides a kind of equipment of character recognition, including：
Memory, for storing computer program；Processor realizes above-mentioned one kind when for executing the computer program
The step of method of character recognition.
The present invention also provides a kind of computer readable storage medium, meter is stored on the computer readable storage medium
The step of calculation machine program, the computer program realizes a kind of method of above-mentioned character recognition when being executed by processor.
The method of character recognition provided by the present invention, after being pre-processed to images to be recognized, extraction is described to be identified
Each character in image；By each character in the images to be recognized respectively in the fuzzy matching template library that pre-establishes
Matching template is matched, and each character preliminary matching result and matching rate are obtained；It is multiple to having in the images to be recognized
The character of matching result and/or low matching rate carries out structure characteristic analysis, compares the knot of character to be matched and multiple matching results
Structure feature determines the final matching results of character to be matched；The structure feature of the character of low matching rate is analyzed, determines final matching
As a result.The method of character recognition provided by the present invention, in conjunction with fuzzy template matching method and charcter topology method for feature analysis, first
Fuzzy matching is carried out to images to be recognized, structure characteristic analysis is carried out to the character that matching result is not known, determines word to be identified
The matching result of symbol improves the accuracy of character recognition；And due to need not be carried out to all characters in image to be detected
Structure characteristic analysis, therefore improve the speed of character recognition.
Description of the drawings
It, below will be to embodiment or existing for the clearer technical solution for illustrating the embodiment of the present invention or the prior art
Attached drawing is briefly described needed in technology description, it should be apparent that, the accompanying drawings in the following description is only this hair
Some bright embodiments for those of ordinary skill in the art without creative efforts, can be with root
Other attached drawings are obtained according to these attached drawings.
Fig. 1 is the flow chart of the first specific embodiment of the method for character recognition provided by the present invention；
Fig. 2 is the flow chart of second of specific embodiment of the method for character recognition provided by the present invention；
Fig. 3 is the flow chart of the third specific embodiment of the method for character recognition provided by the present invention；
Fig. 4 is a kind of structure diagram of the device of character recognition provided in an embodiment of the present invention.
Specific implementation mode
Core of the invention is to provide a kind of method, apparatus of character recognition, equipment and computer readable storage medium,
Improve the accuracy rate and efficiency of character recognition.
In order to enable those skilled in the art to better understand the solution of the present invention, with reference to the accompanying drawings and detailed description
The present invention is described in further detail.Obviously, described embodiments are only a part of the embodiments of the present invention, rather than
Whole embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art are not making creative work premise
Lower obtained every other embodiment, shall fall within the protection scope of the present invention.
Referring to FIG. 1, Fig. 1 is the flow of the first specific embodiment of the method for character recognition provided by the present invention
Figure；Concrete operation step is as follows：
Step S101：Collected images to be recognized is pre-processed, it is each in the images to be recognized in order to obtain
A character；
The images to be recognized can be captured by camera, can also pass through local load；When being captured by camera, it is desirable that phase
Machine pixel reaches 3,000,000 pixels or more.
After collecting the images to be recognized, using halcon algorithms to the images to be recognized carry out denoising, gray processing,
The operation of noise spot other than feature extraction and removal character；
Due to that may include multiple characters in the images to be recognized, using halcon segmentation operators to removing other than character
The images to be recognized of noise spot is split, to be partitioned into each character in the images to be recognized；
Each character in the images to be recognized is extracted successively, and each character after extraction is become a full member and marked
Quasi-ization processing, in order to obtain the binary image of each character to match with Fuzzy Template library graphics standard.
Step S102：By each character in the images to be recognized respectively with the fuzzy matching template library that pre-establishes into
Row template matches obtain the preliminary matching result and matching rate of each character in the images to be recognized；
The fuzzy matching template library includes Arabic numerals, English alphabet and Chinese character；The wherein described Arabic numerals packet
0 to 9 is included, the English alphabet includes 26 capitalizations：A to Z, the Chinese character include the chinese character of predetermined number.
When establishing the fuzzy matching template library, it is full that each character first for the fuzzy matching template library chooses several
The reference picture of sufficient preset condition reads every width reference picture corresponding to each character using the halcon algorithms respectively；
After carrying out denoising to every width reference picture corresponding to each character using halcon operators, to each character
Corresponding every width reference picture carries out gray processing processing, obtains the gray scale of every width reference picture corresponding to each character
Image.
It should be noted that can be the pixel of each character selection no less than 20 in the fuzzy template matching library
Higher reference picture.In the present embodiment, can be each word in order to reduce workload when fuzzy template matching library is established
Symbol chooses 20 reference pictures；In order to improve the accurate of fuzzy template matching storehouse matching result while ensureing computational efficiency
Property, can be reference picture of each character selection more than 20.
After every width gray level image corresponding to each character carries out binary conversion treatment, the spy in the halcon is utilized
Algorithms of Selecting is levied, the noise spot other than character in every width bianry image corresponding to each character is removed；It seeks described every
The angle of inclination of character zone in every width bianry image corresponding to a character, using affine transformation by every width bianry image
Become a full member, and every width bianry image is normalized to default size, to obtain several corresponding foreground characters described in each character
The standard reference image that pixel is 1 and background pixel is 0.
It is normalized after being superimposed after whole standard bianry images alignment corresponding to each character, to
Obtain the Fuzzy Template of each character.Each pixel on the Fuzzy Template of character actually all corresponds to a probability
Value, the probability value represent the possibility that target occurs in the point；If a certain point value is 1 such as in Fuzzy Template, show in institute
It all includes the point to have on the two-value figure for participating in statistics, and the possibility for being object pixel is 1, is the possibility of black background pixel
Property is 0；Vice versa.
Wherein, the Fuzzy Template corresponding to English alphabet character and numerical character is a kind of Fuzzy Template library；Chinese character
Corresponding Fuzzy Template is two class Fuzzy Template libraries.
Step S103：To there is the character of multiple matching results and/or low matching rate to carry out structure in the images to be recognized
Signature analysis obtains the final matching knot for having the character of multiple matching results and/or low matching rate in the images to be recognized
Fruit.
In the present embodiment, first by the powerful image processing function of halcon algorithms, effectively eliminate due to figure to be identified
Character can not divide the problem of extraction caused by image quality amount is bad, compare traditional template matches, carried with fuzzy template matching
The recognition capability that height is incomplete for character, deforms, and for the character that template matches can not accurately identify, further use structure
Characteristics algorithm is identified, and to increase the accuracy rate of character recognition, while also improving the efficiency of character recognition.
On the basis of the above embodiments, the present embodiment is passing through fuzzy mould to each character in the images to be recognized
It is special to there is the character of multiple matching results and/or low matching rate to carry out horizontal line feature, vertical line in the identification image after plate matching
The structure characteristic analysis such as sign, closed area, position relationship and symmetrical feature, so that it is determined that having in the images to be recognized multiple
The final matching results of the character of matching result and/or low matching rate.
Referring to FIG. 2, Fig. 2 is the flow of second of specific embodiment of the method for character recognition provided by the present invention
Figure；Concrete operation step is as follows：
Step S201：Images to be recognized is acquired, the images to be recognized is pre-processed, institute is extracted in order to divide
State each character in images to be recognized；
Step S202：Judge each character generic in the images to be recognized；
Will be upper and lower in the bianry image after each character standard in the images to be recognized obtained after pretreatment, left,
It is a block of pixels that all target pixel points being interconnected on right four direction, which form a connected region,.All English
The pixel block number of alphabetic character and numerical character is 1, and pixel block number is chinese character more than 1, is thus divided into character
Two major classes.
Fuzzy Template corresponding to English alphabet character and numerical character is a kind of Fuzzy Template library；Chinese character is corresponding
Fuzzy Template is two class Fuzzy Template libraries.
Step S203：Respectively by each character in the images to be recognized and the Fuzzy Template library corresponding to each character
In all template characters be compared；
When the character in the images to be recognized be English alphabet character or numerical character when, by character to be matched with it is described
Character in a kind of Fuzzy Template library is compared；It, will be to be matched when the character in the images to be recognized is digital alphabet
Character is compared with the character in the two classes Fuzzy Template library.
Step S204：The template character for meeting preset condition in comparison result corresponding to each character, as the character
Matching result output；
The character match result output that result is unique and matching rate is more than 90% in comparison result, will be met, and think this
Character match success, the character match result of output is the character.It should be noted that the accuracy rate requirement to character recognition
Can be to meet that result is unique and matching rate is more than 95% by condition setting that character match result exports when higher.It can also
According to the actual demand condition appropriate for reducing the output of character match result.
Step S205：It is compared to there is the character of multiple matching results and/or low matching rate to choose in the images to be recognized
After feature, will there are multiple matching results and/or low matching rate character to be similar to character with it according to the comparison feature chosen and carry out spy
Sign compare, to export the multiple matching result and/or low matching rate character matching result.
In template matches, it is the similar character of some shapes to fail to accurately identify, and is that many experiments show by analysis
Because selected chinese character appearance difference is big, unique value can be all obtained in template matches, and there is good matching effect.Therefore template
It is similar number and English character that matching, which fails identification, such as B and 8,2 and Z etc., and here by taking digital and English alphabet as an example,
Structure feature recognizer is illustrated, if having shape similar character in selected Chinese character, using same method.
In charcter topology feature recognition algorithms, the horizontal line feature of character, vertical line feature, closed area, position can be passed through
It sets relationship and symmetrical feature is identified.Wherein, horizontal line is defined as 0.6 times of length more than standard drawing image width, vertical line definition
For 0.6 times more than standard drawing image height.Horizontal line feature extraction takes the method scanned to image progressive, vertical line feature extraction to take
The method that image is scanned by column.
In the present embodiment, according to the structure of a large amount of template matches, the character identified completely grouping will be failed, with once six
Illustrate the process of charcter topology analysis for group likeness in form character：
First group of likeness in form character is 2 and Z, carries out horizontal line signature analysis to character, scanning is Z to there is two horizontal lines, is swept
It is 2 to retouch to a horizontal line；
Second group of likeness in form character is E and F, carries out horizontal line signature analysis to character, scanning is E to there is three horizontal lines, is swept
It is F to retouch to two horizontal lines；
It is 6,8 and B that third group, which is similar to character, is first judged according to the closed area of character, and there are one closed areas
It is 6；Vertical line signature analysis is carried out for 8 and B, scanning is B to there is vertical line, and no vertical line is 8；
4th group of likeness in form character is M, N and H, carries out horizontal line signature analysis to character picture, scanning to horizontal line is H；It is right
In M and N, it is that its own is matched after carrying out left and right mirror transformation, since M is symmetrical figure, therefore matching value is big
For M, matching value is small for N；
5th group of likeness in form character is 1,7 and T, and first by character or so mirror image and after matching, matching value minimum is 7；For 1
Image is split as two parts up and down, scans and arrive in upper part by the position that can be occurred according to the horizontal line scanned with T
Horizontal line is T；
6th group of similar character is O, D and G, and closed area analysis is carried out to character, and no closed area is G, for O and
D, scanning are D to vertical line, and vertical line is not O.
During character recognition, the accurate of character recognition is directly influenced to the handling result of the image collected
Rate.For the knowledge of Arabic numerals, 26 capitalization English letters and specific Chinese character this three categories characters in the present embodiment
Not, the method for being handled collected images to be recognized using halcon algorithms, and proposing fuzzy template matching, is waiting knowing
When going out to have a large amount of incompleteness, fuzzy character in other image, the speed of character recognition can be accelerated；Fuzzy template matching method passes through
Pixel distinguishes Chinese character and other characters, and classification carries out template matches, reduces unnecessary calculating process, further to improve
The efficiency of character recognition.After carrying out fuzzy template matching to the character in images to be recognized, to having to a matching result and/or
The character of low matching rate carries out structure characteristic analysis, and small part likeness in form character is carried out structure feature identification, is greatly avoided
The complexity for directly carrying out structure feature identification, accelerates the efficiency of character feature extraction.In conclusion the present embodiment will obscure
Template matching method and charcter topology method for feature analysis combine, and improve the speed and accuracy of character recognition.
Based on the above embodiments, the present embodiment is by taking Recognition of License Plate Characters as an example, to character recognition provided herein
Method further illustrates, to keep technical solution provided herein clearer, complete.It should be noted that the application
Embodiment can be applied not only in Recognition of License Plate Characters, can also be applied in the field that other need character recognition technologies.
General domestic car plate all includes English alphabet, number and Chinese character three categories, and the Chinese character in car plate includes all
Referred to as, number includes 0 to 9 for provinces and cities, and English alphabet includes whole 26 capitalizations.In Recognition of License Plate Characters, can apply
Fuzzy template matching library including whole Chinese characters in common use carries out fuzzy template matching；In order to reduce character recognition it is required when
Between, characters on license plate fuzzy template matching library can be re-established, in characters on license plate fuzzy template matching library, the matching of two class templates
The chinese character in library only needs to include all provinces and cities' abbreviations.A kind of template matches library in characters on license plate fuzzy template matching library
It is identical as a kind of template matches library in above-described embodiment.
Referring to FIG. 3, Fig. 3 is the flow of the third specific embodiment of the method for character recognition provided by the present invention
Figure；Concrete operation step is as follows：
Step S301：Obtain the image of car plate to be identified；
The image of the car plate to be identified can be captured by pixel in 3,000,000 or more camera, and local can also be passed through
Load.
Step S302：The image of the car plate to be identified is pre-processed using halcon algorithms, it is described to obtain
The standard binary image of all characters in car plate to be identified；
Denoising is carried out to the image of the car plate to be identified using halcon algorithms, binaryzation is divided, and extraction is become a full member, and is marked
The operation of standardization, detailed process is identical as the operating procedure in above-described embodiment, and details are not described herein.
Step S303：By the character in the car plate to be identified and obscuring in characters on license plate fuzzy template matching library
Template matches, and obtains the preliminary matches result of character in the car plate to be identified；
When carrying out fuzzy template matching, the classification of character can be first judged, by the mould of character to be matched corresponding thereto
Paste template matches library is compared, and accelerates the rate of character recognition.
Step S304：To thering are multiple matching results and/or the low character of matching rate to carry out structure in the car plate to be identified
Signature analysis, so that it is determined that there is the matching result of multiple matching results and/or the low character of matching rate.
It in the present embodiment, can be according to the horizontal line feature of character, vertical line feature, closed area, position relationship and right
Claim feature that likeness in form character is identified, specific steps are same as the previously described embodiments, and details are not described herein.
The present embodiment combines fuzzy template matching method and charcter topology method for feature analysis, improves Recognition of License Plate Characters
Speed and accuracy.
Referring to FIG. 4, Fig. 4 is a kind of structure diagram of the device of character recognition provided in an embodiment of the present invention；Specific dress
It sets and may include：
Preprocessing module 100, it is described to be identified in order to obtain for being pre-processed to collected images to be recognized
Each character in image；
Fuzzy matching module 200, for obscuring each character in the images to be recognized with what is pre-established respectively
Matching template library carries out template matches, obtains the preliminary matching result and matching rate of each character in the images to be recognized；
Characteristics analysis module 300, for the word to having multiple matching results and/or low matching rate in the images to be recognized
Symbol carries out structure characteristic analysis, and obtaining has the character of multiple matching results and/or low matching rate most in the images to be recognized
Whole matching result.
The device of the character recognition of the present embodiment is for realizing the method for character recognition above-mentioned, therefore the dress of character recognition
The embodiment part of the method for the visible character recognition hereinbefore of specific implementation mode in setting, for example, preprocessing module 100,
Fuzzy matching module 200, characteristics analysis module 300 are respectively used to step S101, S102 in the method for realizing above-mentioned character recognition
And S103, so, specific implementation mode is referred to the description of corresponding various pieces embodiment, and details are not described herein.
The specific embodiment of the invention additionally provides a kind of equipment of character recognition, including：Memory, for storing computer
Program；Processor, the step of a kind of method of above-mentioned character recognition is realized when for executing the computer program.
The specific embodiment of the invention additionally provides a kind of computer readable storage medium, the computer readable storage medium
On be stored with computer program, the computer program realizes a kind of step of the method for above-mentioned character recognition when being executed by processor
Suddenly.
Each embodiment is described by the way of progressive in this specification, the highlights of each of the examples are with it is other
The difference of embodiment, just to refer each other for same or similar part between each embodiment.For being filled disclosed in embodiment
For setting, since it is corresponded to the methods disclosed in the examples, so description is fairly simple, related place is referring to method part
Explanation.
Professional further appreciates that, unit described in conjunction with the examples disclosed in the embodiments of the present disclosure
And algorithm steps, can be realized with electronic hardware, computer software, or a combination of the two, in order to clearly demonstrate hardware and
The interchangeability of software generally describes each exemplary composition and step according to function in the above description.These
Function is implemented in hardware or software actually, depends on the specific application and design constraint of technical solution.Profession
Technical staff can use different methods to achieve the described function each specific application, but this realization is not answered
Think beyond the scope of this invention.
The step of method described in conjunction with the examples disclosed in this document or algorithm, can directly be held with hardware, processor
The combination of capable software module or the two is implemented.Software module can be placed in random access memory (RAM), memory, read-only deposit
Reservoir (ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technology
In any other form of storage medium well known in field.
Above to the method, apparatus, equipment and computer readable storage medium of character recognition provided by the present invention into
It has gone and has been discussed in detail.Principle and implementation of the present invention are described for specific case used herein, the above implementation
The explanation of example is merely used to help understand the method and its core concept of the present invention.It should be pointed out that for the general of the art
, without departing from the principle of the present invention, can be with several improvements and modifications are made to the present invention for logical technical staff, this
A little improvement and modification are also fallen within the protection scope of the claims of the present invention.
Claims (10)
1. a kind of method of character recognition, which is characterized in that including：
Collected images to be recognized is pre-processed, in order to obtain each character in the images to be recognized；
Each character in the images to be recognized is subjected to template matches with the fuzzy matching template library pre-established respectively, is obtained
To the preliminary matching result and matching rate of each character in the images to be recognized；
To there is the character of multiple matching results and/or low matching rate to carry out structure characteristic analysis in the images to be recognized, obtain
There is the final matching result of the character of multiple matching results and/or low matching rate in the images to be recognized.
2. the method as described in claim 1, which is characterized in that the fuzzy matching template library includes：
Arabic numerals, English alphabet and Chinese character；The wherein described Arabic numerals include 0 to 9, and the English alphabet includes 26
Capitalization：A to Z, the Chinese character include the chinese character of predetermined number.
3. method as claimed in claim 2, which is characterized in that the fuzzy matching template library that pre-establishes includes：
Several reference pictures for meeting preset condition are chosen for each character of the fuzzy matching template library, using described
Halcon algorithms read every width reference picture corresponding to each character respectively；
Denoising, gray processing, two-value are carried out to every width reference picture corresponding to each character using the halcon algorithms
Change, removal noise spot, become a full member, standardized processing, is 1 to obtain several corresponding foreground character pixels described in each character
And the standard bianry image that background pixel is 0；
It is normalized after being superimposed after whole standard bianry images alignment corresponding to each character, to obtain
The Fuzzy Template of each character；
Fuzzy Template corresponding to English alphabet character and numerical character is a kind of Fuzzy Template library；Chinese character is corresponding fuzzy
Template is two class Fuzzy Template libraries.
4. method as claimed in claim 3, which is characterized in that it is described that collected images to be recognized is pre-processed, with
Include convenient for obtaining each character in the images to be recognized：
Acquire images to be recognized, using halcon algorithms to the images to be recognized carry out denoising, gray processing, feature extraction and
Remove the operation of noise spot other than character；
The images to be recognized of noise spot other than removal character is split using halcon segmentation operators, it is described to be partitioned into
Each character in images to be recognized；
Each character in the images to be recognized is extracted successively, and each character after extraction is become a full member and standardized
Processing, in order to obtain the binary image of each character to match with Fuzzy Template library graphics standard.
5. method as claimed in claim 4, which is characterized in that each character by the images to be recognized respectively with
The fuzzy matching template library pre-established carries out template matches, obtains the preliminary matching of each character in the images to be recognized
As a result include with matching rate：
Judge each character generic in pretreated images to be recognized；
Respectively by each character in the images to be recognized and all templates in the Fuzzy Template library corresponding to each character
Character is compared one by one；The template character for meeting preset condition in comparison result corresponding to each character, as the character
Matching result output.
6. method as claimed in claim 5, which is characterized in that described to having multiple matching results in the images to be recognized
And/or the character of low matching rate carries out structure characteristic analysis, obtaining has multiple matching results and/or low in the images to be recognized
The final matching result of the character of matching rate includes：
Using structure feature recognizer, the word for having multiple matching results and/or low matching rate in the images to be recognized is analyzed
The structure feature of symbol；
To there are multiple matching results and/or low matching rate character to be similar to character with it and carry out default aspect ratio pair, and pass through default spy
The comparison result of sign determines that the final matching for having the character of multiple matching results and/or low matching rate in the images to be recognized is tied
Fruit.
7. a kind of device of character recognition, which is characterized in that including：
Preprocessing module, for being pre-processed to collected images to be recognized, in order to obtain in the images to be recognized
Each character；
Fuzzy matching module, for by each character in the images to be recognized respectively with the fuzzy matching template that pre-establishes
Library carries out template matches, obtains the preliminary matching result and matching rate of each character in the images to be recognized；
Characteristics analysis module, for there is the progress of the character of multiple matching results and/or low matching rate in the images to be recognized
Structure characteristic analysis obtains final for having the character of multiple matching results and/or low matching rate in the images to be recognized
With result.
8. device as claimed in claim 7, which is characterized in that the preprocessing module is specifically used for：
Acquire images to be recognized, using halcon algorithms to the images to be recognized carry out denoising, gray processing, feature extraction and
Remove the operation of noise spot other than character；
The images to be recognized of noise spot other than removal character is split using halcon segmentation operators, it is described to be partitioned into
Each character in images to be recognized；
Each character in the images to be recognized is extracted successively, and each character after extraction is become a full member and standardized
Processing, in order to obtain the binary image of each character to match with Fuzzy Template library graphics standard.
9. a kind of equipment of character recognition, which is characterized in that including：
Memory, for storing computer program；
Processor, realizing a kind of character recognition as described in any one of claim 1 to 6 when for executing the computer program
The step of method.
10. a kind of computer readable storage medium, which is characterized in that be stored with computer on the computer readable storage medium
Program realizes a kind of side of character recognition as described in any one of claim 1 to 6 when the computer program is executed by processor
The step of method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810442986.6A CN108615058A (en) | 2018-05-10 | 2018-05-10 | A kind of method, apparatus of character recognition, equipment and readable storage medium storing program for executing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810442986.6A CN108615058A (en) | 2018-05-10 | 2018-05-10 | A kind of method, apparatus of character recognition, equipment and readable storage medium storing program for executing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108615058A true CN108615058A (en) | 2018-10-02 |
Family
ID=63662717
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810442986.6A Pending CN108615058A (en) | 2018-05-10 | 2018-05-10 | A kind of method, apparatus of character recognition, equipment and readable storage medium storing program for executing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108615058A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109558875A (en) * | 2018-11-14 | 2019-04-02 | 广州同略信息科技有限公司 | Method, apparatus, terminal and storage medium based on image automatic identification |
| CN109800762A (en) * | 2019-01-28 | 2019-05-24 | 安阳工学院 | A kind of fuzzy license plate recognizer based on the Dynamic Matching factor |
| CN110363195A (en) * | 2019-06-18 | 2019-10-22 | 深圳壹账通智能科技有限公司 | Graphical verification code recognition methods, device, readable storage medium storing program for executing and terminal device |
| CN110555441A (en) * | 2019-09-10 | 2019-12-10 | 杭州橙鹰数据技术有限公司 | character recognition method and device |
| CN110991522A (en) * | 2019-11-29 | 2020-04-10 | 山东万腾电子科技有限公司 | Character recognition method and system and industrial intelligent gateway |
| CN111079735A (en) * | 2018-10-22 | 2020-04-28 | 莱芜钢铁集团电子有限公司 | Method and device for recognizing end face characters of casting blank |
| CN111401381A (en) * | 2020-03-27 | 2020-07-10 | 白志青 | Digital monitoring system and monitoring method for slow door imaging recognition instrument |
| CN111539383A (en) * | 2020-05-22 | 2020-08-14 | 浙江蓝鸽科技有限公司 | Formula knowledge point identification method and device |
| CN111783766A (en) * | 2020-07-10 | 2020-10-16 | 上海淇毓信息科技有限公司 | Method and device for recognizing image characters step by step and electronic equipment |
| CN114580429A (en) * | 2022-01-26 | 2022-06-03 | 云捷计算机软件(江苏)有限责任公司 | Artificial intelligence-based language and image understanding integrated service system |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101398894A (en) * | 2008-06-17 | 2009-04-01 | 浙江师范大学 | Automobile license plate automatic recognition method and implementing device thereof |
| CN102314615A (en) * | 2011-07-30 | 2012-01-11 | 山东电力研究院 | Substation inspection robot-based circuit breaker state template-matching identification method |
-
2018
- 2018-05-10 CN CN201810442986.6A patent/CN108615058A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101398894A (en) * | 2008-06-17 | 2009-04-01 | 浙江师范大学 | Automobile license plate automatic recognition method and implementing device thereof |
| CN102314615A (en) * | 2011-07-30 | 2012-01-11 | 山东电力研究院 | Substation inspection robot-based circuit breaker state template-matching identification method |
Non-Patent Citations (1)
| Title |
|---|
| 谷秋頔: "基于模板匹配的车牌字符识别及其判别函数的研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111079735B (en) * | 2018-10-22 | 2023-12-22 | 莱芜钢铁集团电子有限公司 | Casting blank end face character recognition method and device |
| CN111079735A (en) * | 2018-10-22 | 2020-04-28 | 莱芜钢铁集团电子有限公司 | Method and device for recognizing end face characters of casting blank |
| CN109558875A (en) * | 2018-11-14 | 2019-04-02 | 广州同略信息科技有限公司 | Method, apparatus, terminal and storage medium based on image automatic identification |
| CN109800762A (en) * | 2019-01-28 | 2019-05-24 | 安阳工学院 | A kind of fuzzy license plate recognizer based on the Dynamic Matching factor |
| WO2020253119A1 (en) * | 2019-06-18 | 2020-12-24 | 深圳壹账通智能科技有限公司 | Graphic verification code recognizing method and apparatus, readable storage medium, and terminal device |
| CN110363195A (en) * | 2019-06-18 | 2019-10-22 | 深圳壹账通智能科技有限公司 | Graphical verification code recognition methods, device, readable storage medium storing program for executing and terminal device |
| CN110555441A (en) * | 2019-09-10 | 2019-12-10 | 杭州橙鹰数据技术有限公司 | character recognition method and device |
| CN110991522A (en) * | 2019-11-29 | 2020-04-10 | 山东万腾电子科技有限公司 | Character recognition method and system and industrial intelligent gateway |
| CN111401381A (en) * | 2020-03-27 | 2020-07-10 | 白志青 | Digital monitoring system and monitoring method for slow door imaging recognition instrument |
| CN111539383A (en) * | 2020-05-22 | 2020-08-14 | 浙江蓝鸽科技有限公司 | Formula knowledge point identification method and device |
| CN111539383B (en) * | 2020-05-22 | 2023-05-05 | 浙江蓝鸽科技有限公司 | Formula knowledge point identification method and device |
| CN111783766A (en) * | 2020-07-10 | 2020-10-16 | 上海淇毓信息科技有限公司 | Method and device for recognizing image characters step by step and electronic equipment |
| CN111783766B (en) * | 2020-07-10 | 2023-02-14 | 上海淇毓信息科技有限公司 | Method and device for recognizing image characters step by step and electronic equipment |
| CN114580429A (en) * | 2022-01-26 | 2022-06-03 | 云捷计算机软件(江苏)有限责任公司 | Artificial intelligence-based language and image understanding integrated service system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108615058A (en) | A kind of method, apparatus of character recognition, equipment and readable storage medium storing program for executing | |
| CN111401372B (en) | Method for extracting and identifying image-text information of scanned document | |
| Gatos et al. | Automatic table detection in document images | |
| CN104751142B (en) | A kind of natural scene Method for text detection based on stroke feature | |
| CN107784301A (en) | Method and apparatus for identifying character area in image | |
| WO1996024114A2 (en) | Method and apparatus for separating foreground from background in images containing text | |
| CN110119741A (en) | A kind of card card image information recognition methods having powerful connections | |
| CN110866529A (en) | Character recognition method, character recognition device, electronic equipment and storage medium | |
| CN111680690A (en) | Character recognition method and device | |
| CN109086772A (en) | A kind of recognition methods and system distorting adhesion character picture validation code | |
| Kölsch et al. | Recognizing challenging handwritten annotations with fully convolutional networks | |
| CN111161281A (en) | Face region identification method and device and storage medium | |
| CN103455816B (en) | Stroke width extraction method and device and character recognition method and system | |
| CN109741273A (en) | A kind of mobile phone photograph low-quality images automatically process and methods of marking | |
| CN110414517B (en) | Rapid high-precision identity card text recognition algorithm used for being matched with photographing scene | |
| CN109508716B (en) | Image character positioning method and device | |
| CN110795955A (en) | System and method for rapidly identifying multiple bar codes | |
| CN111027538A (en) | Container detection method based on instance segmentation model | |
| CN107862314B (en) | Code spraying identification method and device | |
| CN113065404A (en) | Method and system for detecting train ticket content based on equal-width character segments | |
| Siddique et al. | Development of an automatic vehicle license plate detection and recognition system for Bangladesh | |
| CN108133205B (en) | Method and device for copying text content in image | |
| CN111445402A (en) | Image denoising method and device | |
| CN115965987A (en) | Table character structured recognition method based on heterogeneous architecture | |
| CN115546796A (en) | Non-contact data acquisition method and system based on visual computation |
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 |
Application publication date: 20181002 |
|
| RJ01 | Rejection of invention patent application after publication |