CN106709484A - Number identification method of digital instrument - Google Patents
Number identification method of digital instrument Download PDFInfo
- Publication number
- CN106709484A CN106709484A CN201510776461.2A CN201510776461A CN106709484A CN 106709484 A CN106709484 A CN 106709484A CN 201510776461 A CN201510776461 A CN 201510776461A CN 106709484 A CN106709484 A CN 106709484A
- Authority
- CN
- China
- Prior art keywords
- character
- barycenter
- monocase
- midline
- longitudinal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
- G06V10/225—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition based on a marking or identifier characterising the area
-
- 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 provides a number identification method of a digital instrument. The method comprises a character image is preprocessed, and a single character image is obtained by edge cutting; the width and height of a character are calculated and extracted, and the character 1 is identified according to the width to height ratio of the single-character image; and the completeness of the single-character image is determined, if the character is complete, vertical braiding is carried out in a central line position of the width of the single-character image, communication areas in the single-character image is calculated by scanning, and the characters 0 as well as 2 to 9 are identified in a classified manner according to the number of communication areas. According to the method, structural features of the same character in different fonts are extracted to establish a character identification model, the method is suitable for different types of digital instruments, the identification accuracy is higher, and the robustness is higher.
Description
Technical field
The present invention relates to image identification technical field, more particularly to a kind of numeral of digital instrument
Recognition methods.
Background technology
At present, manual inspection, transformer station's track are replaced using transformer station track machine people extensively
Robot can be used to complete the daily tour of converting equipment, infrared measurement of temperature, equipment state inspection etc.
Work.The digital instrument shape of transformer station varies, and the numeral of various digital instruments shows
Show that part mainly uses two kinds of display formats:LED seven segment codes display format and LCD block letter
Display format.Due to being influenceed by light irradiation and instrument and equipment oneself factor, using number
Word image processing technique carries out automatic identification to the reading of digital instrument, still faces discrimination poor
And the low problem of recognition efficiency.
In the art, the Chinese patent of Application No. 201410624087.X proposes one kind
Printing digit recognizing method based on template matches, by by the digital masterplate figure of multiple fonts
As import database in, by normalization after the digital picture of segmentation carry out binary conversion treatment after,
Digital picture to be identified pixel corresponding with masterplate is overlapped calculating again, it is folded by statistics
The result of calculation being added to carries out digital identification.Using the side of the stencil matching described in the patented technology
Method needs to be normalized the character picture being divided into, not only to the segmentation of numeric area
It is required that it is high, and increased amount of calculation.
The Chinese patent of Application No. CN201310209889.X proposes one kind and uses both sides
The digit recognition method of three quant's signs, suitable dimension is normalized to by by single character picture
Afterwards, entered by the left envelope of calculating character, right envelope, the head breadth, the waist parameter such as wide with pin wide
Line number word is recognized.Numeral identification is carried out using the quant's sign of both sides three and is related to calculating parameter, to list
The size of individual digital Character segmentation there are certain requirements, but be typically different the digital instrument institute of type
The printing type face size of use is simultaneously differed, therefore there is a problem of poor for applicability.
In sum, deposited in the practical application of the digital identification techniques of digital instrument at present
Digit recognition methods in a lot, but the method robustness that numeral is recognized mostly is poor, Zhi Nengshi
For the reading of the digital instrument of single type;Or, there is the big recognition efficiency of operand low
Problem.
The content of the invention
Present invention seek to address that problem as described above.It is an object of the present invention to provide one kind
Solve the digit recognition method of any one a kind of digital instrument in problem above.Specifically
Ground, the present invention is provided being capable of strong robustness, it is adaptable to which the reading of polytype digital instrument is known
Other digit recognition method.
According to the first aspect of the invention, there is provided a kind of digit recognition method of digital instrument,
Comprise the following steps:
Character picture is pre-processed, monocase image is obtained after carrying out edge cuts;
The width and height of character are calculated and extract, the ratio of width to height identification word according to monocase image
Symbol 1;
The integrality for performing monocase image judges, if character is complete, in monocase image
Longitudinal threading is carried out at the midline position of width, scanning calculates the connected region in monocase image
Domain, the number Classification and Identification character 0,2,3,4,5,6,7,8,9 according to connected region.
Further, when the number of the connected region is 4, then Direct Recognition monocase figure
As being character 8.
Further, when the number of the connected region is 1, then Direct Recognition monocase figure
As being printed character 7.
Further, when the number of the connected region is 0, in monocase picture altitude
Horizontal threading is carried out at line position, if longitudinal midline first half left pixel summation is significantly less than the right side
Side pixel summation, then recognize the character for character 7, if longitudinal midline first half left pixel is total
Significantly greater than right pixel summation, then recognize that the character is character 4.
Further, when the number of the connected region is 3, in monocase picture altitude
Horizontal threading is carried out at line position, the word if the barycenter of character connected domain is located at more than halfline
Symbol is identified as character 9, the character recognition if the barycenter of character connected domain is located at below halfline
It is character 6.
Further, when the number of the connected region is 2, if two barycenter pair of connected domain
Title is distributed in longitudinal midline both sides and ordinate is located at halfline, then recognize that the character is character
0;If two barycenter asymmetric distributions of connected domain are recognizing the character in longitudinal midline both sides
It is character 4;
Horizontal threading is carried out at the midline position of monocase picture altitude, if two connected domains
On the right side of longitudinal midline and respectively positioned at the both sides up and down of transversal centerline, then can recognize that should for barycenter
Character is character 3;If two barycenter of connected domain are in the left and right sides of longitudinal midline and two matter
The bottom left section and upper right portion that the heart is located in four regions that transverse and longitudinal center line is divided respectively are then known
The character is not character 2;If two barycenter of connected domain are in the left and right sides of longitudinal midline and two
Individual barycenter is located at upper left and lower right-most portion in four regions that transverse and longitudinal center line is divided respectively
Then recognize that the character is character 5.
Further, when the number of the connected region is 4, then Direct Recognition monocase figure
As being character 8;When the number of the connected region is 1, then Direct Recognition monocase image
It is printed character 7;When the number of the connected region is 0, in monocase picture altitude
Horizontal threading is carried out at midline position, if longitudinal midline first half left pixel summation is significantly less than
Right pixel summation, then recognize the character for character 7, if longitudinal midline first half left pixel
Summation is significantly greater than right pixel summation, then recognize that the character is character 4;The connected region
Number be 3 when, horizontal threading is carried out at the midline position of monocase picture altitude, if word
It is character 9 that the barycenter of symbol connected domain is located at more than halfline the then character recognition, if character is connected
It is character 6 that the barycenter in domain is located at below halfline the then character recognition;The connected region
When number is 2, if the barycenter of two connected domains is symmetrically distributed in longitudinal midline both sides and ordinate
Positioned at halfline, then recognize that the character is character 0;If asymmetric point of the barycenter of two connected domains
Longitudinal midline both sides are distributed in, are recognizing that the character is character 4;In monocase picture altitude
Horizontal threading is carried out at line position, if two barycenter of connected domain are on the right side of longitudinal midline and divide
Not Wei Yu transversal centerline both sides up and down, then can recognize that the character be character 3;If two connections
The barycenter in domain is divided positioned at transverse and longitudinal center line respectively in the left and right sides of longitudinal midline and two barycenter
Four regions in bottom left section and upper right portion then recognize the character be character 2;If two
The barycenter of connected domain is in the left and right sides of longitudinal midline and two barycenter are located at transverse and longitudinal center line respectively
Upper left and lower right-most portion in four regions for dividing then recognize that the character is character 5.
Further, the step of pretreatment includes:Suppress noise processed, and/or image
White balance processes, and/or carries out the breakpoint joint treatment of mathematical morphology.
Further, methods described is further included:The integrality for performing monocase image
In the step of judgement, if character is imperfect, character integrity filling is further carried out:
The end points number of monocase image top is calculated, by two if top end points are for 2
End points is attached treatment, is spread round about from breakpoint if top end points are for 1
Until reaching boundary to complete character integrity filling.
Further, methods described is further included:The integrality for performing monocase image
In the step of judgement, if character is imperfect, character integrity filling is further carried out:
The end points number of monocase image base is calculated, by two if bottom number of endpoint is 2
End points is attached treatment, is spread round about from breakpoint if bottom number of endpoint is 1
Until reaching boundary to complete character integrity filling.
Compared with prior art, the method pair based on character feature identification numeral proposed by the present invention
Digital instrument is all suitable for, and is set up by extracting the architectural feature of same character of different fonts
The model of character recognition, recognition accuracy is higher, and robustness is stronger.
The following description for exemplary embodiment, other spies of the invention is read with reference to the drawings
Property feature and advantage will become apparent.
Brief description of the drawings
The accompanying drawing for being incorporated into specification and constituting a part for specification shows the present invention
Embodiment, and together with the description be used for explain principle of the invention.In the drawings,
Similar reference is used to represent similar key element.Drawings in the following description are of the invention
Some embodiments, rather than whole embodiments.For those of ordinary skill in the art, exist
On the premise of not paying creative work, other accompanying drawings can be obtained according to these accompanying drawings.
Fig. 1 schematically illustrates the digital identification side of digital instrument in the embodiment of the present invention
Method flow chart;
Fig. 2 schematically illustrates the digital identification side of the digital instrument of the embodiment of the present invention
The specific implementation flow chart of method;
Fig. 3 schematically illustrates digital " 3 " and numeral " 5 " in the image of collection in worksite
Form schematic diagram;
Fig. 4 schematically illustrates digital " 4 " and " 2 " the latter half and there is incomplete signal
Figure.
Specific embodiment
To make the purpose, technical scheme and advantage of the embodiment of the present invention clearer, below will knot
Close the embodiment of the present invention in accompanying drawing, the technical scheme in the embodiment of the present invention is carried out it is clear,
It is fully described by, it is clear that described embodiment is a part of embodiment of the invention, rather than
Whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not having
The every other embodiment obtained under the premise of creative work is made, protection of the present invention is belonged to
Scope.It should be noted that in the case where not conflicting, embodiment and reality in the application
Applying the feature in example can mutually be combined.
The present invention provides a kind of digit recognition method of digital instrument, for Contemporary Digital formula instrument
The low problem of recognition efficiency poor for applicability present in the digital identification techniques of table, is worn using longitudinal direction
The mode of line and horizontal threading is identified to the digital reading of instrument, big to difference to realize
Small, different fonts identifications, can be applicable various different types of digital instruments.Further
Ground, can also realize there is the incomplete numeral reading in a little part for what is occurred in rotary digital instrument
Several automatic identifications.
Because LED formula data display instrument will not occur the incomplete situation of character, therefore numeral
The incomplete scene of the reading of formula instrument is:Numeral in rotary digital instrument only have top or
Architectural feature is imperfect caused by bottom small part is blocked.
The present invention provides a kind of digit recognition method of digital instrument, as shown in figure 1, including
Following steps:Character picture is pre-processed, monocase image is obtained after carrying out edge cuts;
The width and height of character are calculated and extract, the ratio of width to height identification character according to monocase image
1;The integrality for performing monocase image judges, if character is complete, in monocase image
Longitudinal threading is carried out at the midline position of width, scanning calculates the connected region in monocase image
Domain, the number Classification and Identification character 0,2,3,4,5,6,7,8,9 according to connected region.
After carrying out longitudinal threading and scanning the connected region in calculating numeral, according to connected region
Number character can be divided into 5 groups:(LED numerals 4 and 7 and printed portions body number
Word 7), (printing digital 7), (0,2,3, printing digital 4,5), (6,
9) with (8).This 5 groups connected domain number is respectively 0,1,2,3,4.
Further, when implementing, the digit recognition method of described digital instrument, bag
Include following steps:
Step S101, is pre-processed to character picture, and individual character is obtained after carrying out edge cuts
Symbol image;
Step S102, calculates and extracts width W, the height H of character, according to monocase figure
The ratio of width to height of picture recognizes character " 1 ";
Step S103, the integrality for performing monocase image judges, if character is complete,
Longitudinal threading is carried out at the midline position of monocase picture traverse, scanning calculates monocase image
Interior connected region, if connected region is 4, the Direct Recognition monocase image is character 8;
If connected region is 1, the Direct Recognition monocase image is printed character 7;
Step S104, if connected domain number is 0, at the midline position of monocase picture altitude
Horizontal threading is carried out, if to be significantly less than right pixel total for longitudinal midline first half left pixel summation
With the character is then recognized for character 7, if longitudinal midline first half left pixel summation is substantially big
In right pixel summation, then recognize that the character is character 4;
Step S105, if connected domain number is 3, at the midline position of monocase picture altitude
Horizontal threading is carried out, the character recognition is if the barycenter of character connected domain is located at more than halfline
Character 9, the character recognition is character 6 if the barycenter of character connected domain is located at below halfline.
Step S106, when connected domain number is 2, if the barycenter of two connected domains symmetrically divides
It is distributed in longitudinal midline both sides and ordinate is located at halfline, then recognizes that the character is character 0;If
Two barycenter asymmetric distributions of connected domain are recognizing that the character is character in longitudinal midline both sides
4;
Horizontal threading is carried out at the midline position of monocase picture altitude, if two connected domains
On the right side of longitudinal midline and respectively positioned at the both sides up and down of transversal centerline, then can recognize that should for barycenter
Character is character 3;If two barycenter of connected domain are in the left and right sides of longitudinal midline and two matter
The bottom left section and upper right portion that the heart is located in four regions that transverse and longitudinal center line is divided respectively are then known
The character is not character 2;If two barycenter of connected domain are in the left and right sides of longitudinal midline and two
Individual barycenter is located at upper left and lower right-most portion in four regions that transverse and longitudinal center line is divided respectively
Then recognize that the character is character 5.
Further, in step s 103, if character is imperfect, word is further carried out
The step of symbol integrality is filled:
The end points number of monocase image top and/or bottom is calculated, if top end points are 2
Two end points are then attached treatment, if top end points for if 1 from breakpoint to opposite
Spread until reaching boundary to complete character integrity filling in direction;If bottom number of endpoint is
Two end points are attached treatment by 2, from breakpoint to phase if bottom number of endpoint is 1
Opposite direction is spread until reaching boundary to complete character integrity filling.
Application example:
Fig. 2 is the solution stream using the digit recognition method of digital instrument of the invention
Cheng Tu, the monocase to collecting performs the following steps:
Step S201:Character picture is pre-processed to be influenceed with suppressing noise and reducing light,
If breakpoint, morphology processing connection breakpoint is carried out to numeral, and character is cut
Cut with so that the surrounding of character does not have excess margins.
Step S202:Calculate and extract width W, the height H of character.
Step S203:The ratio of width to height based on character recognizes character " 1 ", no matter because whether 1 is residual
Lack, its ratio of width to height will not change.
Step S204:Integrality according to character judges whether to need the integrality into line character
Filling.If character is complete, step S205 is directly carried out, otherwise enter line character complete
Whole property filling.
The incomplete incomplete situation is:Numeral in rotary digital instrument only have top or
Person bottom small part is blocked and shows imperfect, cause architectural feature gather it is imperfect and to know
The phenomenon of difficulty is not caused.Integrality is carried out to character with bottom incompleteness for top is incomplete respectively
The method of filling is:
The numeral incomplete for top, it judges the end points number at the top of numeral:If 2
Two end points are then carried out into treatment connection, is spread round about from breakpoint if 1 straight
Filled with completing character integrity to boundary is reached.
The numeral incomplete for bottom, it judges the end points number at the top of numeral:If 2
Two end points are then carried out into treatment connection, is spread round about from breakpoint if 1 straight
Filled with completing character integrity to boundary is reached.
Step S205:Longitudinal threading is carried out at the midline position of character duration, and scans meter
The connected region counted in word.Statistical computation is carried out by connected region, can be character point
It is 5 groups:(LED numerals 4 and 7 and printed portions body numeral 7), (printing digital
7), (0,2,3, printing digital 4,5), (6,9) and (8).This 5 groups
Connected domain number is respectively 0,1,2,3,4.
First group of connected domain number of character is 0, can allow digital longitudinal midline for line of demarcation,
The number of pixels of halfline H/2 above line of demarcation arranged on left and right sides is only counted, if left side summation
Right pixel is significantly less than, is judged as 7, be otherwise 4.
Second group of connected domain number of character is 1, because the connected domain number of block letter 7 is 1,
Can individually judge.
3rd group of connected domain number of character is 2, can by with longitudinal midline as line of demarcation,
Judge the centroid position of connected domain.If two barycenter of connected domain are with line of demarcation in symmetrical point
Cloth, and ordinate is near H/2, and then may determine that to be character 0;If two connections
The barycenter in domain is located at the both sides in longitudinal line of demarcation, but without obvious symmetry distribution, then may be used
To be judged as printed character 4;It is again horizontal line of demarcation with H/2, if two connected domains
Barycenter is located at the upper and lower of horizontal line of demarcation respectively on the right side in longitudinal line of demarcation, then can determine whether
It is character 3;If two barycenter of connected domain are in the left and right sides in longitudinal line of demarcation, longitudinal direction
Character zone is divided into four regions by line of demarcation and horizontal line of demarcation, if barycenter is located at lower left quarter
Divide and upper right portion is then judged as character 2, conversely, if located in upper left and lower right-most portion
It is then 5.
4th group of connected domain number is 3, can be by carrying out the barycenter of connected domain to former character
Calculate and obtain.May determine that to be 9 if in more than H/2, be otherwise judged as 6.
5th group of connected domain number is 4, then can directly be judged as 8.
Substation field is differed due to digital table equipment source, causes digital morphological different, application
Technical scheme, is adapted to the number of the variform of polytype digital instrument
Word image, as shown in figure 3, digital " 3 " and numeral " 5 " in listing the image of collection in worksite
Form, the digital picture of these digital instruments can be using numeral identification side of the invention
Method is accurately quickly recognized.
In addition, for rotary digital instrument, due to the not exclusively caused character of numerical monitor
Part is incomplete, the integrality by recognizing monocase image of the invention, and incomplete numeral is carried out
Image completion treatment ensures the digital integrity of monocase image, then carry out numeral identification can also
Obtain good recognition effect.Fig. 4 is being collected from substation field of passing back of crusing robot
Picture, there is incompleteness in numeral " 4 " and " 2 " in Fig. 4 the latter half, by using this
Method in invention, still can obtain preferable recognition effect.For rotary digital instrument
In the part that occurs of numeral " 4 " it is incomplete, be identified using digit recognition method of the invention
When, the collection on architectural feature does not influence, therefore can according to complete character the step of carry out
Treatment.
Compared with prior art, the method pair based on character feature identification numeral proposed by the present invention
Digital instrument is all suitable for, and is set up by extracting the architectural feature of same character of different fonts
The model of character recognition, recognition accuracy is higher, and robustness is stronger.
Descriptions above can combine implementation individually or in a variety of ways, and this
A little variants are all within protection scope of the present invention.
Finally it should be noted that:The above embodiments are merely illustrative of the technical solutions of the present invention,
Rather than its limitations.Although being described in detail to the present invention with reference to the foregoing embodiments,
It will be understood by those within the art that:It can still be remembered to foregoing embodiments
The technical scheme of load is modified, or carries out equivalent to which part technical characteristic;
And these modifications or replacement, the essence of appropriate technical solution is departed from each reality of the invention
Apply the spirit and scope of a technical scheme.
Claims (10)
1. a kind of digit recognition method of digital instrument, it is characterised in that comprise the following steps:
Character picture is pre-processed, monocase image is obtained after carrying out edge cuts;
The width and height of character are calculated and extract, the ratio of width to height identification character 1 according to monocase image;
The integrality for performing monocase image judges, if character is complete, in monocase picture traverse
Longitudinal threading is carried out at midline position, scanning calculates the connected region in monocase image, according to connected region
The number Classification and Identification character 0,2,3,4,5,6,7,8,9 in domain.
2. the digit recognition method of digital instrument as claimed in claim 1, it is characterised in that
When the number of the connected region is 4, then the Direct Recognition monocase image is character 8.
3. the digit recognition method of digital instrument as claimed in claim 1, it is characterised in that
When the number of the connected region is 1, then the Direct Recognition monocase image is printed character 7.
4. the digit recognition method of digital instrument as claimed in claim 1, it is characterised in that
When the number of the connected region is 0, transverse direction is carried out at the midline position of monocase picture altitude
Threading, if longitudinal midline first half left pixel summation is significantly less than right pixel summation, recognizes the word
It is character 7 to accord with, if longitudinal midline first half left pixel summation is significantly greater than right pixel summation, is known
The character is not character 4.
5. the digit recognition method of digital instrument as claimed in claim 1, it is characterised in that
When the number of the connected region is 3, transverse direction is carried out at the midline position of monocase picture altitude
Threading, the character recognition is character 9 if the barycenter of character connected domain is located at more than halfline, if character
It is character 6 that the barycenter of connected domain is located at below halfline the then character recognition.
6. the digit recognition method of digital instrument as claimed in claim 1, it is characterised in that
When the number of the connected region is 2, if the barycenter of two connected domains is symmetrically distributed in longitudinal midline
Both sides and ordinate are located at halfline, then recognize that the character is character 0;If the barycenter of two connected domains is non-
Longitudinal midline both sides are symmetrically distributed in, are recognizing that the character is character 4;
Horizontal threading is carried out at the midline position of monocase picture altitude, if the barycenter of two connected domains exists
The right side of longitudinal midline and respectively positioned at transversal centerline both sides up and down, then can recognize that the character be character 3;
If two barycenter of connected domain are in the left and right sides of longitudinal midline and two barycenter are located at transverse and longitudinal center line respectively
Bottom left section and upper right portion in four regions for dividing then recognize that the character is character 2;If two companies
The barycenter in logical domain is in the left and right sides of longitudinal midline and two barycenter are located at transverse and longitudinal center line is divided four respectively
Upper left and lower right-most portion in individual region then recognize that the character is character 5.
7. the digit recognition method of digital instrument as claimed in claim 1, it is characterised in that
When the number of the connected region is 4, then the Direct Recognition monocase image is character 8;
When the number of the connected region is 1, then the Direct Recognition monocase image is printed character 7;
When the number of the connected region is 0, transverse direction is carried out at the midline position of monocase picture altitude
Threading, if longitudinal midline first half left pixel summation is significantly less than right pixel summation, recognizes the word
It is character 7 to accord with, if longitudinal midline first half left pixel summation is significantly greater than right pixel summation, is known
The character is not character 4;
When the number of the connected region is 3, transverse direction is carried out at the midline position of monocase picture altitude
Threading, the character recognition is character 9 if the barycenter of character connected domain is located at more than halfline, if character
It is character 6 that the barycenter of connected domain is located at below halfline the then character recognition;
When the number of the connected region is 2, if the barycenter of two connected domains is symmetrically distributed in longitudinal midline
Both sides and ordinate are located at halfline, then recognize that the character is character 0;If the barycenter of two connected domains is non-
Longitudinal midline both sides are symmetrically distributed in, are recognizing that the character is character 4;In monocase picture altitude
Horizontal threading is carried out at line position, if two barycenter of connected domain are on the right side of longitudinal midline and are located at respectively
The both sides up and down of transversal centerline, then can recognize that the character is character 3;If two barycenter of connected domain are vertical
The left and right sides to the midline and two barycenter are located at the lower-left in four regions that transverse and longitudinal center line is divided respectively
Part and upper right portion then recognize that the character is character 2;If two barycenter of connected domain are in longitudinal midline
The left and right sides and two barycenter are located at upper left and the right side in four regions that transverse and longitudinal center line is divided respectively
Lower part then recognizes that the character is character 5.
8. the digit recognition method of digital instrument as claimed in claim 1, it is characterised in that institute
The step of stating pretreatment includes:Suppress noise processed, and/or the treatment of image white balance, and/or enter line number
Learn morphologic breakpoint joint treatment.
9. the digit recognition method of the digital instrument as any one of claim 1 to 8, its
It is characterised by, methods described is further included:The step of integrality for performing monocase image judges
In, if character is imperfect, further carry out character integrity filling:
The end points number of monocase image top is calculated, enters two end points if top end points are for 2
Row connection is processed, and is spread round about from breakpoint up to reaching border if top end points are for 1
Sentence completion character integrity filling.
10. the digit recognition method of the digital instrument as any one of claim 1 to 8, its
It is characterised by, methods described is further included:The step of integrality for performing monocase image judges
In, if character is imperfect, further carry out character integrity filling:
The end points number of monocase image base is calculated, enters two end points if bottom number of endpoint is 2
Row connection is processed, and is spread round about from breakpoint up to reaching border if bottom number of endpoint is 1
Sentence completion character integrity filling.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510776461.2A CN106709484B (en) | 2015-11-13 | 2015-11-13 | Digital identification method of digital instrument |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510776461.2A CN106709484B (en) | 2015-11-13 | 2015-11-13 | Digital identification method of digital instrument |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106709484A true CN106709484A (en) | 2017-05-24 |
CN106709484B CN106709484B (en) | 2022-02-22 |
Family
ID=58930181
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510776461.2A Active CN106709484B (en) | 2015-11-13 | 2015-11-13 | Digital identification method of digital instrument |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106709484B (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106326934A (en) * | 2016-08-27 | 2017-01-11 | 厦门市朗星节能照明股份有限公司 | Dice throw-out point number judging method and device |
CN107240185A (en) * | 2017-06-23 | 2017-10-10 | 深圳怡化电脑股份有限公司 | A kind of crown word number identification method, device, equipment and storage medium |
CN109858480A (en) * | 2019-01-08 | 2019-06-07 | 北京全路通信信号研究设计院集团有限公司 | A kind of digital instrument recognition methods |
CN110533003A (en) * | 2019-09-06 | 2019-12-03 | 兰州大学 | A kind of threading method license plate number recognizer and equipment |
CN110659645A (en) * | 2019-08-05 | 2020-01-07 | 沈阳工业大学 | Character recognition method for digital instrument |
CN112703538A (en) * | 2018-12-12 | 2021-04-23 | 株式会社东芝 | Reading support system, mobile object, reading support method, program, and storage medium |
CN112766271A (en) * | 2021-01-12 | 2021-05-07 | 齐鲁工业大学 | Method and system for identifying digital display panel |
CN114926841A (en) * | 2022-07-22 | 2022-08-19 | 广州能源检测研究院 | Electronic water meter reading image identification method and device based on improved threading method |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030198386A1 (en) * | 2002-04-19 | 2003-10-23 | Huitao Luo | System and method for identifying and extracting character strings from captured image data |
CN1584921A (en) * | 2004-06-10 | 2005-02-23 | 上海交通大学 | Identifying method for fragmentary printed digits |
US20090274392A1 (en) * | 2008-05-01 | 2009-11-05 | Zhigang Fan | Page orientation detection based on selective character recognition |
CN101916364A (en) * | 2010-07-06 | 2010-12-15 | 浙江大学 | Adaptive dial digital identification device and method |
CN102254159A (en) * | 2011-07-07 | 2011-11-23 | 清华大学深圳研究生院 | Interpretation method for digital readout instrument |
CN103886294A (en) * | 2014-03-24 | 2014-06-25 | 北京国网富达科技发展有限责任公司 | Method and system for recognizing number of movement times of arrester for convertor station based on images |
CN103984930A (en) * | 2014-05-21 | 2014-08-13 | 南京航空航天大学 | Digital meter recognition system and method based on vision |
CN104392205A (en) * | 2014-10-24 | 2015-03-04 | 浙江力石科技股份有限公司 | Abnormal vehicle license plate recognition method and system |
CN104463157A (en) * | 2014-11-07 | 2015-03-25 | 李宇 | Electronic identification method for handwritten characters |
CN104700092A (en) * | 2015-03-26 | 2015-06-10 | 南京理工大学 | Small-character number identification method based on template and feature matching |
-
2015
- 2015-11-13 CN CN201510776461.2A patent/CN106709484B/en active Active
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030198386A1 (en) * | 2002-04-19 | 2003-10-23 | Huitao Luo | System and method for identifying and extracting character strings from captured image data |
CN1584921A (en) * | 2004-06-10 | 2005-02-23 | 上海交通大学 | Identifying method for fragmentary printed digits |
US20090274392A1 (en) * | 2008-05-01 | 2009-11-05 | Zhigang Fan | Page orientation detection based on selective character recognition |
CN101916364A (en) * | 2010-07-06 | 2010-12-15 | 浙江大学 | Adaptive dial digital identification device and method |
CN102254159A (en) * | 2011-07-07 | 2011-11-23 | 清华大学深圳研究生院 | Interpretation method for digital readout instrument |
CN103886294A (en) * | 2014-03-24 | 2014-06-25 | 北京国网富达科技发展有限责任公司 | Method and system for recognizing number of movement times of arrester for convertor station based on images |
CN103984930A (en) * | 2014-05-21 | 2014-08-13 | 南京航空航天大学 | Digital meter recognition system and method based on vision |
CN104392205A (en) * | 2014-10-24 | 2015-03-04 | 浙江力石科技股份有限公司 | Abnormal vehicle license plate recognition method and system |
CN104463157A (en) * | 2014-11-07 | 2015-03-25 | 李宇 | Electronic identification method for handwritten characters |
CN104700092A (en) * | 2015-03-26 | 2015-06-10 | 南京理工大学 | Small-character number identification method based on template and feature matching |
Non-Patent Citations (2)
Title |
---|
崔行臣等: "数显仪表数字实时识别系统的设计与实现", 《计算机工程与设计》 * |
李艺: "仪表表盘的数字字符自动识别技术的研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106326934A (en) * | 2016-08-27 | 2017-01-11 | 厦门市朗星节能照明股份有限公司 | Dice throw-out point number judging method and device |
CN107240185A (en) * | 2017-06-23 | 2017-10-10 | 深圳怡化电脑股份有限公司 | A kind of crown word number identification method, device, equipment and storage medium |
CN107240185B (en) * | 2017-06-23 | 2019-09-20 | 深圳怡化电脑股份有限公司 | A kind of crown word number identification method, device, equipment and storage medium |
CN112703538A (en) * | 2018-12-12 | 2021-04-23 | 株式会社东芝 | Reading support system, mobile object, reading support method, program, and storage medium |
CN112703538B (en) * | 2018-12-12 | 2023-06-30 | 株式会社东芝 | Reading support system, moving object, reading support method, program, and storage medium |
CN109858480A (en) * | 2019-01-08 | 2019-06-07 | 北京全路通信信号研究设计院集团有限公司 | A kind of digital instrument recognition methods |
CN110659645A (en) * | 2019-08-05 | 2020-01-07 | 沈阳工业大学 | Character recognition method for digital instrument |
CN110659645B (en) * | 2019-08-05 | 2023-01-31 | 沈阳工业大学 | Character recognition method for digital instrument |
CN110533003A (en) * | 2019-09-06 | 2019-12-03 | 兰州大学 | A kind of threading method license plate number recognizer and equipment |
CN110533003B (en) * | 2019-09-06 | 2022-09-20 | 兰州大学 | Threading method license plate number recognition method and equipment |
CN112766271A (en) * | 2021-01-12 | 2021-05-07 | 齐鲁工业大学 | Method and system for identifying digital display panel |
CN114926841A (en) * | 2022-07-22 | 2022-08-19 | 广州能源检测研究院 | Electronic water meter reading image identification method and device based on improved threading method |
Also Published As
Publication number | Publication date |
---|---|
CN106709484B (en) | 2022-02-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106709484A (en) | Number identification method of digital instrument | |
CN102254159B (en) | Interpretation method for digital readout instrument | |
CN104182750B (en) | A kind of Chinese detection method based on extreme value connected domain in natural scene image | |
CN104680161A (en) | Digit recognition method for identification cards | |
US10846549B2 (en) | Marker for occluding foreign matter in acquired image, method for recognizing foreign matter marker in image and book scanning method | |
CN106529537A (en) | Digital meter reading image recognition method | |
CN105205488B (en) | Word area detection method based on Harris angle points and stroke width | |
CN104778470B (en) | Text detection based on component tree and Hough forest and recognition methods | |
CN103413145B (en) | Intra-articular irrigation method based on depth image | |
CN104990926A (en) | TR element locating and defect detecting method based on vision | |
CN104680130A (en) | Chinese character recognition method for identification cards | |
CN107610128A (en) | The method for inspecting and device of a kind of oil level indicator | |
CN101777124A (en) | Method for extracting video text message and device thereof | |
CN106446952A (en) | Method and apparatus for recognizing score image | |
CN101576958B (en) | Extraction method of rectangle seal graphics with border based on form feature | |
CN104077577A (en) | Trademark detection method based on convolutional neural network | |
CN106295655A (en) | A kind of transmission line part extraction method patrolling and examining image for unmanned plane | |
CN107506762A (en) | A kind of achievement method for automatically inputting based on graphical analysis | |
CN108509988B (en) | Test paper score automatic statistical method and device, electronic equipment and storage medium | |
CN104463138A (en) | Text positioning method and system based on visual structure attribute | |
CN103886294A (en) | Method and system for recognizing number of movement times of arrester for convertor station based on images | |
CN106355592A (en) | Educational toy suite and its circuit elements and electric wires identifying method thereof | |
CN107818321A (en) | A kind of watermark date recognition method for vehicle annual test | |
CN106485710A (en) | A kind of element mistake part detection method and device | |
CN109409211B (en) | Processing method, processing device and storage medium for Chinese character skeleton stroke segments |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |