CN110414525A - A kind of character identifying method and device - Google Patents
A kind of character identifying method and device Download PDFInfo
- Publication number
- CN110414525A CN110414525A CN201910701844.1A CN201910701844A CN110414525A CN 110414525 A CN110414525 A CN 110414525A CN 201910701844 A CN201910701844 A CN 201910701844A CN 110414525 A CN110414525 A CN 110414525A
- Authority
- CN
- China
- Prior art keywords
- circle
- center
- radius
- character
- feature
- 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
- 238000000034 method Methods 0.000 title claims abstract description 56
- 239000000463 material Substances 0.000 claims abstract description 78
- 239000000284 extract Substances 0.000 claims abstract description 3
- 238000000605 extraction Methods 0.000 claims description 7
- 238000009751 slip forming Methods 0.000 claims description 2
- 238000012545 processing Methods 0.000 abstract description 10
- 238000010586 diagram Methods 0.000 description 9
- 238000001514 detection method Methods 0.000 description 5
- 238000012163 sequencing technique Methods 0.000 description 3
- 235000007926 Craterellus fallax Nutrition 0.000 description 1
- 240000007175 Datura inoxia Species 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 235000020281 long black Nutrition 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/60—Rotation of whole images or parts thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/42—Global feature extraction by analysis of the whole pattern, e.g. using frequency domain transformations or autocorrelation
- G06V10/422—Global feature extraction by analysis of the whole pattern, e.g. using frequency domain transformations or autocorrelation for representing the structure of the pattern or shape of an object therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/148—Segmentation of character regions
- G06V30/153—Segmentation of character regions using recognition of characters or words
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Geometry (AREA)
- Character Input (AREA)
- Character Discrimination (AREA)
Abstract
The embodiment of the present application provides a kind of character identifying method and device, is related to technical field of image processing, comprising: determines first center of circle comprising the border circular areas around ocra font ocr in images to be recognized;Obtain the first radius of the inscribed circle around ocra font ocr and the second radius of circumscribed circle;According to the first radius and the second radius, determine passed through by first straight line, the adjacent two edges rectangular area parallel with Shui Ping, vertical direction respectively;It is rotation angle step and using first center of circle as rotation axis points using preset angle, along preset direction of rotation continuous rotation at least once around ocra font ocr region, extracts the feature for being located at the picture material of rectangular area after rotation every time;Indicatrix is formed according to extracted feature, obtains the feature for belonging to maximum on indicatrix;The picture material characterized to acquired feature carries out character recognition.Character recognition is carried out using scheme provided by the embodiments of the present application, can be realized to the identification around ocra font ocr.
Description
Technical field
This application involves technical field of image processing, more particularly to a kind of character identifying method and device.
Background technique
With the development of artificial intelligence technology, more and more application scenarios can be related to character recognition.
When carrying out character recognition in the prior art, usually horizontal positioned character is identified.
However, needing to identify to around ocra font ocr in some application scenarios, for example, it is desired to in round seal image
Circular ocra font ocr identify etc..
Given this, it is desirable to provide one kind can identify the character recognition scheme of circulating type character.
Summary of the invention
The embodiment of the present application is designed to provide a kind of character identifying method and device, to realize to around ocra font ocr
Identification.Specific technical solution is as follows:
In a first aspect, the embodiment of the present application provides a kind of character identifying method, which comprises
The center of circle comprising the border circular areas around ocra font ocr in images to be recognized is determined, as first center of circle;
Obtain the first radius of the inscribed circle around ocra font ocr and the second radius of circumscribed circle, wherein described inscribed
The center of circle in the round center of circle and circumscribed circle is first center of circle;
According to first radius and the second radius, determine by first straight line pass through and adjacent two edges respectively with level
The parallel rectangular area of direction, vertical direction, wherein the first straight line are as follows: by the vertical straight line in first center of circle;
It is rotation angle step and using first center of circle as rotation axis points using preset angle, along preset rotation side
To the continuous rotation circular ocra font ocr region at least once, and extract the figure for being located at the rectangular area after rotation every time
As the feature of content;
According to extracted feature, indicatrix is formed, and obtains the feature for belonging to maximum on the indicatrix;
The picture material characterized to acquired feature carries out character recognition, obtains recognition result.
Second aspect, the embodiment of the present application provide a kind of character recognition device, and described device includes:
Center of circle determining module, for determining the center of circle comprising the border circular areas around ocra font ocr in images to be recognized, as
First center of circle;
Radius obtains module, for obtain the inscribed circle around ocra font ocr the first radius and circumscribed circle the second half
Diameter, wherein the center of circle of the inscribed circle and the center of circle of circumscribed circle are first center of circle;
Area determination module is passed through and adjacent for determining according to first radius and the second radius by first straight line
Two sides rectangular area parallel with horizontal direction, vertical direction respectively, wherein the first straight line are as follows: pass through described first
The vertical straight line in the center of circle;
Characteristic extracting module, for being to rotate angle step and using first center of circle as rotary shaft using preset angle
Point, along the preset direction of rotation continuous rotation circular ocra font ocr region, and position after each rotation of extraction at least once
In the feature of the picture material of the rectangular area;
Feature obtains module, for forming indicatrix according to extracted feature, and obtains and belongs on the indicatrix
In the feature of maximum;
Character recognition module, the picture material for being characterized to acquired feature carry out character recognition, are identified
As a result.
The third aspect, the embodiment of the present application provide a kind of electronic equipment, including processor and machine readable storage medium,
The machine readable storage medium is stored with the machine-executable instruction that can be executed by the processor, and the processor is by institute
It states machine-executable instruction to promote: realizing character identifying method step described in the embodiment of the present application.
Fourth aspect, the embodiment of the present application provide a kind of machine readable storage medium, are stored with machine-executable instruction,
When being called and being executed by processor, the machine-executable instruction promotes the processor: realizing described in the embodiment of the present application
Character identifying method step.
It as seen from the above, is rotation with preset angle when carrying out character recognition using scheme provided by the embodiments of the present application
Gyration step-length and using the center of circle of the border circular areas where the ocra font ocr as rotation axis points, continuously revolves along preset direction of rotation
Turn at least once around ocra font ocr region, and is located at the feature of the picture material of rectangular area after each rotation of extraction, according to
According to extracted feature, indicatrix is formed, and obtains the feature for belonging to maximum on indicatrix, to acquired feature institute
The picture material of characterization carries out character recognition, obtains recognition result.Since above-mentioned rectangular area is to be passed through by first straight line and phase
Adjacent two sides rectangular area parallel with horizontal direction, vertical direction respectively.Again since above-mentioned first straight line is by circulating type
The vertical straight line in the center of circle of border circular areas where character.So can be by after rotating around ocra font ocr to above-mentioned rectangular area
It is considered horizontal positioned character, and then the character being located in above-mentioned rectangular area can be identified, to realizes to ring
Around the identification of ocra font ocr.
Detailed description of the invention
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of application for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is a kind of flow diagram of character identifying method provided by the embodiments of the present application;
Fig. 2 is the flow diagram of another character identifying method provided by the embodiments of the present application;
Fig. 3 a is inscribed circle provided by the embodiments of the present application and circumscribed circle schematic diagram;
Fig. 3 b is the first edge image provided by the embodiments of the present application;
Fig. 3 c is second of edge image provided by the embodiments of the present application;
Fig. 3 d is the third edge image provided by the embodiments of the present application;
Fig. 3 e is the 4th kind of edge image provided by the embodiments of the present application;
Fig. 3 f is the 5th kind of edge image provided by the embodiments of the present application;
Fig. 4 is a kind of structural schematic diagram of character recognition device provided by the embodiments of the present application;
Fig. 5 is the structural schematic diagram of a kind of electronic equipment provided by the embodiments of the present application.
Specific embodiment
Below in conjunction with the attached drawing in the embodiment of the present application, technical solutions in the embodiments of the present application carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of embodiments of the present application, instead of all the embodiments.It is based on
Embodiment in the application, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall in the protection scope of this application.
Due to needing to identify that the embodiment of the present application provides one kind thus to around ocra font ocr in concrete application scene
Character identifying method and device.
In one embodiment of the application, a kind of character identifying method is provided, this method can be applied to electronic equipment.
This method comprises:
The center of circle comprising the border circular areas around ocra font ocr in images to be recognized is determined, as first center of circle;
Obtain the first radius of the inscribed circle around ocra font ocr and the second radius of circumscribed circle, wherein above-mentioned inscribed circle
The center of circle and the center of circle of circumscribed circle are first center of circle;
According to the first radius and the second radius, determine by first straight line pass through and adjacent two edges respectively with horizontal direction,
The parallel rectangular area of vertical direction, wherein above-mentioned first straight line are as follows: by the vertical straight line in first center of circle;
It is rotation angle step and using first center of circle as rotation axis points using preset angle, connects along preset direction of rotation
Continuous rotation is located at the spy of the picture material of rectangular area around ocra font ocr region, and after each rotation of extraction at least once
Sign;
According to extracted feature, indicatrix is formed, and obtains the feature for belonging to maximum on features described above curve;
The picture material characterized to acquired feature carries out character recognition, obtains recognition result.
Due to above-mentioned rectangular area be passed through by first straight line and adjacent two edges respectively with horizontal direction, vertical direction
Parallel rectangular area.Again since above-mentioned first straight line is by around the vertical straight of the center of circle of border circular areas where ocra font ocr
Line.So being considered horizontal positioned character after rotating around ocra font ocr to above-mentioned rectangular area, and then can be right
Character in above-mentioned rectangular area is identified, to realize to the identification around ocra font ocr.
Character identifying method provided by the embodiments of the present application is described in detail below by specific embodiment.
Referring to Fig. 1, Fig. 1 is a kind of flow diagram of character identifying method provided by the embodiments of the present application, the above method
Including following S101-S106.
S101: the center of circle comprising the border circular areas around ocra font ocr in images to be recognized is determined, as first center of circle.
Around the character that ocra font ocr is along circular arc arrangement.For example, with reference to Fig. 3 a, the character " UVWX that circumferentially arranges in figure
Seal scholar trade Co., Ltd " is around ocra font ocr.
Wherein, the character for including in above-mentioned circular ocra font ocr can be the squares such as Chinese character, Japanese character, Korea character
Shape character can also be the non-box-shaped character such as English character, method Chinese character, German character.
It, can be according to round recognition methods (for example, the round knowledge based on Hough transform in one embodiment of the application
Other method) identification images to be recognized in circle, then detect in the circle whether circumferentially there is foreground point again.If it is detected that
Circle in circumferentially there is foreground point, it may be considered that the circle area defined is comprising around the circle of ocra font ocr
Domain.
It will be appreciated by persons skilled in the art that in the picture, the content of image can be divided into foreground and background.
Wherein, prospect refers to the object that subjective role is played in image, and background refers to the content in image in addition to prospect.Before in image
Scape and background can be according to different application scenarios and determination.For example, prospect can be image in video monitoring scene
In the objects such as people, automobile, building, trees etc. in image in addition to people, automobile are then background, and in character recognition scene
In, prospect is character, and the content in image in addition to the characters is then background.
Based on this, it is properly termed as foreground point for describing the pixel of prospect in image, for describing the pixel of background
It is properly termed as background dot.
It, can be based on being in Identifying Technique of Object detection image specifically, when whether there is foreground point in detection image
The no object existed as prospect, and if it exists, it may be considered that there are foreground points in image.
It can be realized with connected applications scene using any Identifying Technique of Object in the prior art during specific implementation
Object detection, I will not elaborate.
Specifically, the circle in images to be recognized can be detected by the way of carrying out Hough transform to images to be recognized.
In addition, can also determine detected circle other than it can detecte out the circle in images to be recognized using Hough transform
The center of circle, that is, above-mentioned first center of circle.
S102: the first radius of the above-mentioned inscribed circle around ocra font ocr and the second radius of circumscribed circle are obtained.
Wherein, the center of circle of above-mentioned inscribed circle and the center of circle of circumscribed circle are first center of circle.
Since character has certain height, and due to arranging around ocra font ocr along circular arc, it is therefore contemplated that circulating type
Character has inscribed circle and circumscribed circle.Wherein, inscribed circle can be understood as above-mentioned circular ocra font ocr close to above-mentioned first center of circle
The corresponding circle in side.Circumscribed circle can be understood as side corresponding circle of the above-mentioned circular ocra font ocr far from above-mentioned first center of circle
Shape.
For example, with reference to Fig. 3 a, the inscribed circle and circumscribed circle around ocra font ocr are shown.Wherein, " UVWX seal scholar in Fig. 3 a
Trade Co., Ltd " is to surround ocra font ocr, and circle shown in the thick dashed line in figure on the outside of circulating type character is around ocra font ocr
Circumscribed circle, circle shown in the fine dotted line in figure on the inside of circulating type character are the inscribed circle around ocra font ocr.
It is noted that above-mentioned inscribed circle and circumscribed circle are not the circle actually existed in images to be recognized, but
Fictionalize the circle come.
In view of the foregoing, in one embodiment of the application, obtain around ocra font ocr inscribed circle the first radius and
When the second radius of circumscribed circle, the graphics field and non-circulating type character location in above-mentioned border circular areas can be first determined
Domain, the pixel that will include in the zone boundary of above-mentioned border circular areas, above-mentioned graphics field and non-circulating type character region
Point is set as background dot, can make in above-mentioned border circular areas in this way except the pixel for describing to surround ocra font ocr is foreground point
Outside, other pixels are arranged to background dot.
On the basis of the above, the minimum range in first center of circle and above-mentioned border circular areas between foreground point can be obtained, is made
For the first radius of the above-mentioned inscribed circle around ocra font ocr, and obtain in first center of circle and above-mentioned border circular areas between foreground point
Maximum distance, the second radius as the above-mentioned circumscribed circle around ocra font ocr.
Specifically, above-mentioned graphics field can be the region including star graphic, it is also possible to the area of star graphic composition
Domain etc..
In one embodiment, SWT (Stationary Wavelet can be carried out using to above-mentioned border circular areas
Transform, Stationary Wavelet Transform) mode, determine the graphics field in above-mentioned border circular areas.Above-mentioned border circular areas is image
Region carries out SWT processing to above-mentioned border circular areas, that is, carries out SWT processing to image-region, then carries out to processing result
Graphics field detection, to determine the graphics field of the border circular areas.Specifically, can be examined according to graphics field in the prior art
Survey mode carries out graphics field detection to processing result, and I will not elaborate.
Furthermore it is also possible to be determined in above-mentioned border circular areas by the way of carrying out edge extracting to above-mentioned border circular areas
Graphics field.Wherein, it can be realized using any Boundary extracting algorithm in the prior art when carrying out edge extracting, here no longer in detail
It states.Then the graphics field in above-mentioned border circular areas can be determined according to the information such as shape of connected domain in edge extracting result.
In one embodiment of the application, the non-circulating type character region in above-mentioned border circular areas can be by following
Mode determines:
Horizontally arranged character is known otherwise using existing, or the character being vertically arranged is identified
Mode, non-circulating type character recognition is carried out to above-mentioned border circular areas, in this case, it is believed that can successfully be carried out word
The region of symbol identification is non-circulating type character region.
Above-described embodiment is illustrated below with reference to specific example.
Referring to Fig. 3 b- Fig. 3 d, wherein circular edge line is the above-mentioned border circular areas comprising around ocra font ocr in Fig. 3 b
Boundary, pentalpha region are above-mentioned graphics field, and " YZ branch company " region is non-circulating type character region.
Image after setting the pixel for including in the boundary of above-mentioned border circular areas and pentalpha to background dot is as shown in Figure 3c.It will
It is as shown in Figure 3d that the pixel for including in non-circulating type character is set as the image after background dot.
In addition, black line short in Fig. 3 e indicates that above-mentioned first radius, long black line indicate above-mentioned second radius.
S103: according to the first radius and the second radius, determine by first straight line pass through and adjacent two edges respectively with level
The parallel rectangular area of direction, vertical direction.
Wherein, above-mentioned first straight line are as follows: by the vertical straight line in first center of circle.
Specifically, above-mentioned rectangular area can be located at the upside in above-mentioned first center of circle, above-mentioned first center of circle can also be located at
Downside.
In one embodiment of the application, the region in above-mentioned rectangular area positioned at first straight line two sides can be equal
, it is, above-mentioned rectangular area is the axial symmetry image-region using first straight line as symmetry axis.Certainly, the embodiment of the present application
In do not limit yet in above-mentioned rectangular area positioned at first straight line two sides region it is strictly equal.
Based on foregoing description, in one embodiment of the application, above-mentioned rectangle region can be determined by following steps A-C
Domain.
Step A: the side length L of rectangular area is calculated according to following formula:
L=R2-R1+2P
Wherein, R1 indicates that the first radius, R2 indicate that the second radius, P indicate default value.
Above-mentioned default value P can be determined according to subsequent rotary course bring pixel-shift.A kind of realization of the application
In mode, subsequent rotary course bring pixel-shift is bigger, and the value of above-mentioned P can be set bigger, for example, above-mentioned P
Value can be set to 8,9,10 etc., and the unit of above-mentioned P can be pixel.Conversely, subsequent rotary course bring pixel is inclined
Move smaller, the value of above-mentioned P can be set smaller, for example, the value of above-mentioned P can be set to 2,3,4 etc..It is calculating in this way
When the side length of rectangular area, above-mentioned P is introduced, can effectively mitigate subsequent rotary course bring pixel-shift.
Step B: the coordinate of the lower-left angular vertex of above-mentioned rectangular area is calculated according to following formula:
CL_X=CO_X-L/2
CL_Y=CO_Y+R1-P
Wherein, CL_X、CL_YIndicate cross, the ordinate of above-mentioned lower-left angular vertex, CO_X、CO_YIndicate the horizontal, vertical of first center of circle
Coordinate.
Step C: according to the coordinate of above-mentioned lower-left angular vertex, determine side length be L and adjacent two edges respectively with level side
The rectangular area parallel to, vertical direction.
The rectangular area determined in this way according to above-mentioned steps A-C, from horizontal direction for, be located around ocra font ocr institute
In the centre of border circular areas horizontal direction, from vertical direction for, correspond to around ocra font ocr highest area in the vertical direction
Domain.And it surround the character that horizontal center and vertical highest character are generally horizontally arranged in ocra font ocr.
For example, with reference to Fig. 3 f, figure center is above-mentioned rectangle in the rectangle frame area defined outside " scholar " this character
Region
S104: being rotation angle step and using first center of circle as rotation axis points using preset angle, along preset rotation side
It is located at the picture material of rectangular area around ocra font ocr region, and after each rotation of extraction at least once to continuous rotation
Feature.
Herein, " continuous rotation is at least once " refers to using first center of circle as rotation axis points, with initial position (if not carrying out
Cross rotation)/last time postrotational position (if carrying out rotation) is as starting rotation position, and rotation is at least once.
By taking circulating type character region in Fig. 3 f along clockwise direction continuous rotation 3 times as an example.By circulating type in Fig. 3 f
Character region is rotated in a clockwise direction once, it is assumed that " chapter " this character is located in rectangular area, this is rotation 1
It is secondary.Then circulating type character region in image that " chapter " this character is located in rectangular area is rotated in a clockwise direction
Once, it is assumed that " print " this character is located in rectangular area, this is rotation 2 times." print " this character is finally located at square
Circulating type character region is rotated in a clockwise direction once in image in shape region, it is assumed that " X " this character bit
In in rectangular area, this is rotation 3 times.
Specifically, above-mentioned preset angle can be set according to the requirement in concrete application scene to precision.For example, to essence
It spends more demanding, then can set above-mentioned preset angle to lesser value, but calculating speed can be relatively slow in this way, if otherwise
It is lower to required precision, then biggish value can be set by above-mentioned preset angle, required precision can met in this way
Meanwhile accelerating calculating speed.
For example, above-mentioned preset angle can be set to 3 degree, 5 degree etc..
It above-mentioned preset direction of rotation can be for counterclockwise, or clockwise.It specifically can be according to answering
It is set with the habit that character in scene is written.For example, the habit that character is written in application scenarios is book from left to right
It writes, then above-mentioned preset direction of rotation can be for counterclockwise.
Above-mentioned circular ocra font ocr region can be entire above-mentioned border circular areas, can also be only comprising around type-word
The region of symbol, it is, a part of above-mentioned border circular areas, the application are defined not to this.
In addition, can rotate a circle, that is, after 360 degree when being rotated to above-mentioned circular ocra font ocr region,
Terminate rotation.
Furthermore since the thought that scheme provided by the embodiments of the present application carries out character recognition is: when in ocra font ocr
After each character is rotated into above-mentioned rectangular area, it is converted into the character of level of approximation placement, so as to realize that character is known
Not, given this situation, it is respectively positioned on the left side of above-mentioned first straight line in above-mentioned circular ocra font ocr, or is respectively positioned on above-mentioned first
When the right of straight line, rotation can also be terminated after rotating 180 degree.
The feature mode for the picture material for being located at above-mentioned rectangular area after rotating is extracted in following embodiments shown in Fig. 2
It is described, wouldn't be described in detail here.
S105: according to extracted feature, indicatrix is formed, and obtains the feature for belonging to maximum on indicatrix.
The above-mentioned available multiple features of S104, these are characterized in the presentation of specific value, that is, above-mentioned S104 can
, in this case can be by the way of curve matching to obtain the value of multiple features, each spy that above-mentioned S104 is obtained
The value of sign carries out curve fitting, to obtain an indicatrix.Wherein, the every bit on this indicatrix is rotation
It is located at the value of the feature of the picture material of rectangular area in the process.Maximum point on this indicatrix is characterized on curve
Point, so the value of above-mentioned maximum point is also the value of the feature of the picture material positioned at rectangular area, it is, above-mentioned
The value of maximum point corresponds to the value of the feature of the picture material of rectangular area, and the value of maximum point also characterizes in image
Feature possessed by holding.Again since the maximum point on indicatrix is the maximum point of local value on indicatrix, so, pole
The value of the feature of the picture material of the corresponding rectangular area of value of big value point is also in circulating type character zone rotary course
Positioned at the local maxima value of the feature of the picture material of rectangular area.Furthermore the value of feature is bigger, illustrate this feature institute table
The picture material of sign more has feature abundant, and therefore, the picture material that the value of these maximum points is characterized has part
The most abundant feature.
In addition, when annular character includes the whole region of a character in rotary course, in above-mentioned rectangular area, rectangle
The feature of picture material is able to reflect the feature of entire character in region, and feature is compared with horn of plenty, and in above-mentioned rectangular area only
When partial region comprising a character, the feature of picture material is then merely capable of the spy of reflection partial character in rectangular area
Sign, therefore, feature will not be very abundant.In consideration of it, after annular character is rotated to above-mentioned rectangular area, in rectangular area in image
It is most abundant that the feature of appearance can reach part.
From the discussion above it is known that the feature for belonging to maximum on features described above curve is considered complete character
The feature of picture material in rectangular area behind above-mentioned rectangular area.
S106: the picture material characterized to acquired feature carries out character recognition, obtains recognition result.
In this step, acquired feature is to belong to the feature of maximum on the indicatrix obtained in above-mentioned S105,
It is exactly, it is considered to be complete character is rotated into the feature of picture material in rectangular area behind above-mentioned rectangular area.Again due to upper
Stating rectangular area is to be passed through by first straight line and the adjacent two edges rectangular area parallel with horizontal direction, vertical direction respectively,
And above-mentioned first straight line is by the vertical straight line around the center of circle of border circular areas where ocra font ocr.So acquired in above-mentioned
It is not only may be considered that in the picture material that feature is characterized comprising complete character, included complete character can also be considered
It is horizontal positioned.In this case, character recognition can be carried out using existing character recognition mode.
As seen from the above, when carrying out character recognition using the scheme that above-mentioned each embodiment provides, it is with preset angle
Rotate angle step and using the center of circle of the border circular areas where the ocra font ocr as rotation axis points, it is continuous along preset direction of rotation
Rotation is located at the feature of the picture material of rectangular area around ocra font ocr region, and after each rotation of extraction at least once,
According to extracted feature, indicatrix is formed, and obtains the feature for belonging to maximum on indicatrix, to acquired feature
The picture material characterized carries out character recognition, obtains recognition result.Due to above-mentioned rectangular area be passed through by first straight line and
The adjacent two edges rectangular area parallel with horizontal direction, vertical direction respectively.Again since above-mentioned first straight line is by surrounding
The vertical straight line in the center of circle of border circular areas where ocra font ocr.So can be with after rotating around ocra font ocr to above-mentioned rectangular area
It is considered as horizontal positioned character, and then the character being located in above-mentioned rectangular area can be identified, thus realization pair
Around the identification of ocra font ocr.
In one embodiment of the application, referring to fig. 2, Fig. 2 provides the flow diagram of another character identifying method,
Compared with above-mentioned embodiment illustrated in fig. 1, above-mentioned S104 can be realized by following S104A-S104D.
S104A: being rotation angle step and using first center of circle as rotation axis points using preset angle, along preset rotation
Direction continuous rotation above-mentioned circular ocra font ocr region at least once.
S104B: counting the quantity of foreground point in picture material to be processed, as the first quantity.
Wherein, above-mentioned picture material to be processed are as follows: be located at the picture material of above-mentioned rectangular area after rotation every time.
S104C: counting the quantity for meeting the foreground point of preset condition in picture material to be processed, as the second quantity.
Wherein, above-mentioned preset condition are as follows: adjacent pixel is foreground point or along the vertical direction in the horizontal direction
Neighbor pixel is foreground point.
It is assumed that pixel is Pic (x, y), then adjacent pixel in the horizontal direction are as follows: Pic (x-1, y) and Pic
(x+1, y), along the vertical direction adjacent pixel are as follows: Pic (x, y-1) and Pic (x, y+1).
On this basis, the foreground point for meeting above-mentioned preset condition refers to:
In the case where Pic (x y) is foreground point, Pic (x-1, y) and Pic (x+1, y) are foreground point or Pic
(x, y-1) and Pic (x, y+1) are foreground point.
In other examples, above-mentioned preset condition are as follows: adjacent pixel is foreground point and along perpendicular in the horizontal direction
Histogram to neighbor pixel be foreground point.
S104D: it is in varying numbers in zero in the case where first, ratio of second quantity relative to the first quantity is calculated, is made
For the feature of picture material to be processed.
In one embodiment of the application, when above-mentioned first quantity is equal to zero, above-mentioned image to be processed can be directly set
The feature of content is zero.
It should be noted that can first pass through above-mentioned S104A during executing above-mentioned S104A-S104D and carry out one
Then secondary rotation executes above-mentioned S104B-S104D for the picture material for being located at above-mentioned rectangular area after rotation, on having executed
It after stating S104B-S104D, returns to above-mentioned S104A and continues rotation process, so circulation is until meet the knot for terminating rotation
Beam condition.
For example, above-mentioned termination condition, which can be, has rotated through 360 degree, or has rotated through preset other angles,
Such as, 180 degree etc..
Since position of the above-mentioned circular ocra font ocr in images to be recognized is arbitrary, so the first character identified
Symbol is not necessarily the first character in ocra font ocr, so after completing character recognition, it can also be each to what is identified
The sequence of character is adjusted.
Again due to along preset direction of rotation, although the first character around ocra font ocr is deposited between a last character
In character, but the opposite direction along above-mentioned direction of rotation, it can between the first character and last character of ocra font ocr
It can be that character is not present.For example, with reference to Fig. 3 f, along clockwise direction, around ocra font ocr " the limited public affairs of UVWX seal scholar's trade
Department " in first character " U " between a last character " department " there are character " V ", " W ", " X " " print ", " chapter ", " scholar ",
" trade ", " easy ", " having ", " limit ", " public affairs ", " department ".And opposite direction along clockwise direction is namely counterclockwise, above-mentioned first
There is no characters between a last character " department " for a character " U ".When the region there is no character is rotated to above-mentioned rectangle region
Behind domain, since picture material is simple in rectangular area, so the value of the feature of picture material is smaller in rectangular area.
In consideration of it, after obtaining recognition result, can also determine and connect on indicatrix in one embodiment of the application
The curve section that the continuous value no more than preset value is formed, and according to preset direction of rotation, acquired feature and above-mentioned curve
Relative positional relationship of the section on indicatrix adjusts the sequence of character in obtained recognition result.
Wherein, acquired feature and relative position of the above-mentioned curve section on indicatrix may be that acquired feature exists
Before above-mentioned curve section, it is also possible to which acquired feature is after above-mentioned curve section.
Since the value in above-mentioned curve section is no more than preset value, and the every bit on indicatrix is to rotate through
It is located at the feature of the picture material of rectangular area in journey, it is, it is rotation that each value in curve section, which is every bit,
It is located at the feature of the picture material of rectangular area during turning, then illustrates the figure that each value in above-mentioned curve section is characterized
As content is relatively simple, may not include character.So as to think that above-mentioned curve section corresponds in ocra font ocr, along upper
The opposite direction of direction of rotation is stated from first character to the image-region a last character.
Specifically, can be the curve area that determination section length is not less than preset length when determining above-mentioned curve section
Between.Due to some characters are relatively simple or adjacent character between gap it is big, the rotation of image-region in the case of these is supreme
After stating rectangular area, the feature of picture material can be smaller in obtained rectangular area, so having on features described above curve can
Shorter curve section can occur.To exclude such situation, a preset length can be set, when determining curve section is less than this
When preset length, it is believed that the curve section be due to character is relatively simple or adjacent character between gap it is big and obtain.Although
Value is all not more than above-mentioned preset value in these shorter curve sections, but these shorter curve sections are not correspond to edge
The opposite direction of above-mentioned direction of rotation is from first character to the region for not including character a last character, therefore, application
Aforesaid way can effectively exclude this kind of continuum.
When the siding-to-siding block length in a certain curve section is not less than preset length, then curve section then corresponds to rectangular area
Picture material in do not include character region.
Specifically, above-mentioned preset value can be zero.
Due to above-mentioned preset direction of rotation can be clockwise, be also possible to counterclockwise, below assuming that
Character is according in tactic situation from left to right, and a point situation is illustrated.
The first situation, above-mentioned preset direction of rotation are counterclockwise
In this case, after determining above-mentioned curve section, above-mentioned curve area can will be located in acquired feature
Between after first characterized picture material of feature recognition result, as the first character in the ocra font ocr.Also
It is the feature being located at after above-mentioned curve section according to determined feature putting in order from left to right, in acquired feature
The recognition result of the picture material characterized is the last period character in ocra font ocr, and is located at before above-mentioned curve section
The recognition result for the picture material that feature is characterized is the latter section of character in the ocra font ocr, then splice the last period character and
Latter section of character, obtains final recognition result.
For example, according to rotating counterclockwise, then " scholar's trade is limited when identifying to the circular ocra font ocr in Fig. 3 f
The feature of picture material is located at before curve section in these character regions of company ", and " UVWX seal " these words
The feature of picture material is located at after curve section in symbol region, and " U " is first feature after above-mentioned curve section
The recognition result of characterized picture material, this six characters of U, V, W, X, print, chapter be it is successively identified, scholar, trade, easily, have,
Limit, public, to take charge of this seven characters be successively identified, but is identified before U, V, W, X, print, chapter this six characters, on
State the identified sequence of each character with it is above-mentioned it is preset put in order it is consistent.By the knowledge of U, V, W, X, print, chapter this six characters
Other result and scholar, trade, easily, have, limit, recognition result that is public, taking charge of this seven characters splices, the knowledge after sequencing can be obtained
Other result.
Second situation, above-mentioned preset direction of rotation are clockwise
In this case, after determining above-mentioned curve section, above-mentioned curve area can will be located in acquired feature
Between before the characterized picture material of the last one feature recognition result, as the first character in the ocra font ocr.
It is exactly, according to determined feature putting in order from right to left, to be located at before above-mentioned curve section in the feature determined
The recognition result of picture material that is characterized of feature be the last period character around ocra font ocr, and be located at above-mentioned curve section it
The recognition result for the picture material that feature afterwards is characterized is the latter section of character in ocra font ocr.
In one example, in such a way that latter section of character is posterior in preceding, the last period character, splicing the last period character with it is latter
Section character, obtains the reversed character of character after splicing, obtains final recognition result.
For example, when being identified to the circular ocra font ocr in Fig. 3 f, according to being rotated clockwise, then " scholar's chapter prints XWVU "
The feature of picture material is located at before curve section in these character regions, and " the public limit of department has easy trade " these characters
The feature of picture material is located at after curve section in region, " U " be above-mentioned curve section before first feature institute
Characterize the recognition result of picture material, scholar, chapter, print, X, W, V, U this seven characters be it is successively identified, department, it is public, limit, have,
Easily, trade, this six characters are successively identified, but are identified after scholar, chapter, print, X, W, V, U this seven characters,
The identified sequence of above-mentioned each character and it is above-mentioned it is preset put in order it is opposite.With latter section of character in preceding, the last period character
Posterior mode, splicing scholar, chapter, print, the recognition result of X, W, V, U this seven characters and department, public affairs, limit, have, easily, trade this six
The recognition result of character, character after being spliced " department, it is public, limit, have, easily, trade, scholar, chapter, print, X, W, V, U ".Then according to from
The right side is reversed to left sequence, obtains the reversed character of character after splicing, that is, obtain " U, V, W, X, print, chapter, scholar, trade, easily,
Have, limit, is public, department ", the recognition result after sequencing can be obtained.
In another example, the reversed character of the last period character and the reversed word of latter section of character can be obtained respectively
Symbol, then splices the reversed character of the last period character and the reversed character of latter section of character again, obtains final recognition result.
For example, when being identified to the circular ocra font ocr in Fig. 3 f, according to being rotated clockwise, then " scholar's chapter prints XWVU "
The feature of picture material is located at before curve section in these character regions, and " the public limit of department has easy trade " these characters
The feature of picture material is located at after curve section in region, " U " be above-mentioned curve section before first feature institute
Characterize the recognition result of picture material, scholar, chapter, print, X, W, V, U this seven characters be it is successively identified, department, it is public, limit, have,
Easily, this six characters of trade are successively identified, but are identified after scholar, chapter, print, X, W, V, U this seven characters, on
State the identified sequence of each character and it is above-mentioned it is preset put in order it is opposite.Obtain offices and men, chapter, print, X, W, V, U this seven words
The reversed character of symbol: U, V, W, X, print, chapter, scholar and acquisition department, public affairs, limit, have, easily, the reversed character of trade this six characters:
Trade, easily, have, limit, is public, department, the reversed character of the reversed character and latter section of character that splice the last period character obtain " U, V, W, X,
Print, chapter, scholar, trade, easily, have, limit, is public, department ", the recognition result after sequencing can be obtained.
Corresponding with above-mentioned character identifying method, the embodiment of the present application also provides a kind of character recognition devices.
Referring to fig. 4, Fig. 4 provides a kind of structural schematic diagram of character recognition device, which includes:
Center of circle determining module 401 is made for determining the center of circle comprising the border circular areas around ocra font ocr in images to be recognized
For first center of circle;
Radius obtains module 402, for obtain the inscribed circle around ocra font ocr the first radius and circumscribed circle the
Two radiuses, wherein the center of circle of the inscribed circle and the center of circle of circumscribed circle are first center of circle;
Area determination module 403, passed through by first straight line for determining according to first radius and the second radius and
The adjacent two edges rectangular area parallel with horizontal direction, vertical direction respectively, wherein the first straight line are as follows: described in process
The vertical straight line in first center of circle;
Characteristic extracting module 404, for being to rotate angle step and using first center of circle as rotary shaft using preset angle
Point along preset direction of rotation continuous rotation at least once around ocra font ocr region, and is located at square after extracting rotation every time
The feature of the picture material in shape region;
Feature obtains module 405, for forming indicatrix according to extracted feature, and obtains the indicatrix
On belong to the feature of maximum;
Character recognition module 406, the picture material for being characterized to acquired feature carry out character recognition, are known
Other result.
In one embodiment of the application, the characteristic extracting module 404 is specifically used for:
Using preset angle as rotary step and using first center of circle as rotation axis points, along preset direction of rotation pair
The circular ocra font ocr region is gradually rotated;
The quantity for counting foreground point in picture material to be processed, as the first quantity, wherein the picture material to be processed
Are as follows: it is located at the picture material of the rectangular area after rotation every time;
The quantity for meeting the foreground point of preset condition in the picture material to be processed is counted, as the second quantity, wherein
The preset condition are as follows: adjacent pixel is foreground point in the horizontal direction or neighbor pixel along the vertical direction is
Foreground point;
It is in varying numbers in zero in the case where described first, calculate ratio of second quantity relative to first quantity
Value, the feature as the picture material to be processed.
In one embodiment of the application, described device further include:
Section determining module, for determining the curve area continuously formed no more than the value of preset value on the indicatrix
Between;
Sort adjustment module is used for according to the preset direction of rotation, acquired feature and the curve section in institute
The relative positional relationship on indicatrix is stated, the sequence of character in obtained recognition result is adjusted.
In one embodiment of the application, the area determination module 403 is specifically used for:
The side length L of rectangular area is calculated according to following formula:
L=R2-R1+2P
Wherein, R1 indicates that first radius, R2 indicate that second radius, P indicate default value;
The coordinate of the lower-left angular vertex of the rectangular area is calculated according to following formula:
CL_X=CO_X-L/2
CL_Y=CO_Y+R1-P
Wherein, CL_X、CL_YIndicate cross, the ordinate of the lower-left angular vertex, CO_X、CO_YIndicate first center of circle
Horizontal, ordinate;
According to the coordinate of the lower-left angular vertex, determine side length be L and adjacent two edges respectively with horizontal direction, perpendicular
Histogram is to parallel rectangular area.
In one embodiment of the application, the radius obtains module 402, is specifically used for:
Determine the graphics field and non-circulating type character region in the border circular areas;
To include in the zone boundary of the border circular areas, the graphics field and non-circulating type character region
Pixel is set as background dot;
The minimum range in first center of circle and the border circular areas between foreground point is obtained, as the circular type-word
First radius of the inscribed circle of symbol;
The maximum distance in first center of circle and the border circular areas between foreground point is obtained, as the circular type-word
Second radius of the circumscribed circle of symbol.
As seen from the above, when carrying out character recognition using the scheme that above-mentioned each embodiment provides, it is with preset angle
Rotate angle step and using the center of circle of the border circular areas where the ocra font ocr as rotation axis points, it is continuous along preset direction of rotation
Rotation is located at the feature of the picture material of rectangular area around ocra font ocr region, and after each rotation of extraction at least once,
According to extracted feature, indicatrix is formed, and obtains the feature for belonging to maximum on indicatrix, to acquired feature
The picture material characterized carries out character recognition, obtains recognition result.Due to above-mentioned rectangular area be passed through by first straight line and
The adjacent two edges rectangular area parallel with horizontal direction, vertical direction respectively.Again since above-mentioned first straight line is by surrounding
The vertical straight line in the center of circle of border circular areas where ocra font ocr.So can be with after rotating around ocra font ocr to above-mentioned rectangular area
It is considered as horizontal positioned character, and then the character being located in above-mentioned rectangular area can be identified, thus realization pair
Around the identification of ocra font ocr.
Corresponding with above-mentioned character identifying method, the embodiment of the present application also provides a kind of electronic equipment.
The structural schematic diagram of a kind of electronic equipment is provided referring to Fig. 5, Fig. 5, above-mentioned electronic equipment includes: processor 501
With machine readable storage medium 502, the machine readable storage medium 502 is stored with and can be executed by the processor 501
Machine-executable instruction, the processor 501 are promoted by the machine-executable instruction: realizing word provided by the embodiments of the present application
The step of according with recognition methods.
In one embodiment of the application, a kind of character identifying method is provided, this method comprises:
The center of circle comprising the border circular areas around ocra font ocr in images to be recognized is determined, as first center of circle;
Obtain the first radius of the inscribed circle around ocra font ocr and the second radius of circumscribed circle, wherein described inscribed
The center of circle in the round center of circle and circumscribed circle is first center of circle;
According to first radius and the second radius, determine by first straight line pass through and adjacent two edges respectively with level
The parallel rectangular area of direction, vertical direction, wherein the first straight line are as follows: by the vertical straight line in first center of circle;
It is rotation angle step and using first center of circle as rotation axis points using preset angle, along preset rotation side
To continuous rotation at least once around ocra font ocr region, and it is located in the image of the rectangular area after extracting rotation every time
The feature of appearance;
According to extracted feature, indicatrix is formed, and obtains the feature for belonging to maximum on the indicatrix;
The picture material characterized to acquired feature carries out character recognition, obtains recognition result.
It should be noted that its for the character identifying method that above-mentioned processor 501 is promoted to realize by machine-executable instruction
His embodiment, identical as embodiment mentioned by preceding method embodiment part, which is not described herein again.
As seen from the above, when carrying out character recognition using electronic equipment provided in this embodiment, due to above-mentioned rectangular area
It is to be passed through by first straight line and the adjacent two edges rectangular area parallel with horizontal direction, vertical direction respectively.Again due to above-mentioned
First straight line is by the vertical straight line around the center of circle of border circular areas where ocra font ocr.So when around ocra font ocr rotate to
Be considered horizontal positioned character behind above-mentioned rectangular area, so can to the character being located in above-mentioned rectangular area into
Row identification, to realize to the identification around ocra font ocr.
Corresponding with above-mentioned character identifying method, the embodiment of the present application also provides a kind of machine readable storage mediums, should
Machine readable storage medium is stored with machine-executable instruction, and when being called and being executed by processor, the machine is executable to be referred to
Order promotes the processor: the step of realizing character identifying method provided by the embodiments of the present application.
In one embodiment of the application, a kind of character identifying method is provided, this method comprises:
The center of circle comprising the border circular areas around ocra font ocr in images to be recognized is determined, as first center of circle;
Obtain the first radius of the inscribed circle around ocra font ocr and the second radius of circumscribed circle, wherein described inscribed
The center of circle in the round center of circle and circumscribed circle is first center of circle;
According to first radius and the second radius, determine by first straight line pass through and adjacent two edges respectively with level
The parallel rectangular area of direction, vertical direction, wherein the first straight line are as follows: by the vertical straight line in first center of circle;
It is rotation angle step and using first center of circle as rotation axis points using preset angle, along preset rotation side
To continuous rotation at least once around ocra font ocr region, and it is located in the image of the rectangular area after extracting rotation every time
The feature of appearance;
According to extracted feature, indicatrix is formed, and obtains the feature for belonging to maximum on the indicatrix;
The picture material characterized to acquired feature carries out character recognition, obtains recognition result.
It should be noted that other for the character identifying method that above-mentioned machine-executable instruction promotes above-mentioned processor to realize
Embodiment, identical as embodiment mentioned by preceding method embodiment part, which is not described herein again.
As seen from the above, execute the machine-executable instruction that is stored in machine readable storage medium provided in this embodiment into
When line character identifies, due to above-mentioned rectangular area be passed through by first straight line and adjacent two edges respectively with horizontal direction, vertical
The parallel rectangular area in direction.Again since above-mentioned first straight line is by around the vertical of the center of circle of border circular areas where ocra font ocr
Straight line.So being considered horizontal positioned character after rotating around ocra font ocr to above-mentioned rectangular area, and then can
The character being located in above-mentioned rectangular area is identified, to realize to the identification around ocra font ocr.
It should be noted that above-mentioned machine readable storage medium may include random access memory (Random Access
Memory, RAM), it also may include nonvolatile memory (Non-Volatile Memory, NVM), for example, at least a magnetic
Disk storage.Optionally, above-mentioned machine readable storage medium can also be that at least one is located remotely from the storage of aforementioned processor
Device.
Above-mentioned processor can be general processor, including central processing unit (Central Processing Unit,
CPU), network processing unit (Network Processor, NP) etc.;It can also be digital signal processor (Digital Signal
Processing, DSP), it is specific integrated circuit (Application Specific Integrated Circuit, ASIC), existing
It is field programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic device, discrete
Door or transistor logic, discrete hardware components.
Herein, relational terms such as first and second and the like be used merely to by an entity or operation with it is another
One entity or operation distinguish, and without necessarily requiring or implying between these entities or operation, there are any this reality
Relationship or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to the packet of nonexcludability
Contain, so that the process, method, article or equipment for including a series of elements not only includes those elements, but also including
Other elements that are not explicitly listed, or further include for elements inherent to such a process, method, article, or device.
In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including the element
Process, method, article or equipment in there is also other identical elements.
Each embodiment in this specification is all made of relevant mode and describes, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for device,
For electronic equipment and machine readable storage medium embodiment, since it is substantially similar to the method embodiment, so the ratio of description
Relatively simple, the relevent part can refer to the partial explaination of embodiments of method.
The foregoing is merely the preferred embodiments of the application, are not intended to limit the protection scope of the application.It is all
Any modification, equivalent replacement, improvement and so within spirit herein and principle are all contained in the protection scope of the application
It is interior.
Claims (12)
1. a kind of character identifying method, which is characterized in that the described method includes:
The center of circle comprising the border circular areas around ocra font ocr in images to be recognized is determined, as first center of circle;
Obtain the first radius of the inscribed circle around ocra font ocr and the second radius of circumscribed circle, wherein the inscribed circle
The center of circle and the center of circle of circumscribed circle are first center of circle;
According to first radius and the second radius, determine by first straight line pass through and adjacent two edges respectively with horizontal direction,
The parallel rectangular area of vertical direction, wherein the first straight line are as follows: by the vertical straight line in first center of circle;
It is rotation angle step and using first center of circle as rotation axis points using preset angle, connects along preset direction of rotation
It is continuous to rotate the circular ocra font ocr region at least once, and be located in the image of the rectangular area after extracting rotation every time
The feature of appearance;
According to extracted feature, indicatrix is formed, and obtains the feature for belonging to maximum on the indicatrix;
The picture material characterized to acquired feature carries out character recognition, obtains recognition result.
2. the method according to claim 1, wherein the extraction is located at the rectangular area after rotating every time
The feature of picture material, comprising:
The quantity for counting foreground point in picture material to be processed, as the first quantity, wherein the picture material to be processed are as follows:
It is located at the picture material of the rectangular area after rotation every time;
The quantity for meeting the foreground point of preset condition in the picture material to be processed is counted, as the second quantity, wherein described
Preset condition are as follows: adjacent pixel is foreground point in the horizontal direction or neighbor pixel along the vertical direction is prospect
Point;
It is in varying numbers in zero in the case where described first, ratio of second quantity relative to first quantity is calculated,
Feature as the picture material to be processed.
3. according to the method described in claim 2, it is characterized in that, after the acquisition recognition result, further includes:
Determine the curve section continuously formed no more than the value of preset value on the indicatrix;
According to the preset direction of rotation, acquired feature and relative position of the curve section on the indicatrix
Relationship adjusts the sequence of character in obtained recognition result.
4. the method according to claim 1, wherein described according to first radius and the second radius, determination
It is passed through and the adjacent two edges rectangular area parallel with horizontal direction, vertical direction respectively by first straight line, comprising:
The side length L of rectangular area is calculated according to following formula:
L=R2-R1+2P
Wherein, R1 indicates that first radius, R2 indicate that second radius, P indicate default value;
The coordinate of the lower-left angular vertex of the rectangular area is calculated according to following formula:
CL_X=CO_X-L/2
CL_Y=CO_Y+R1-P
Wherein, CL_X、CL_YIndicate cross, the ordinate of the lower-left angular vertex, CO_X、CO_YIndicate the horizontal, vertical of first center of circle
Coordinate;
According to the coordinate of the lower-left angular vertex, determine side length be L and adjacent two edges respectively with horizontal direction, vertical side
To parallel rectangular area.
5. method according to any of claims 1-4, which is characterized in that described to obtain the interior of the circular ocra font ocr
Connect the second radius of round the first radius and circumscribed circle, comprising:
Determine the graphics field and non-circulating type character region in the border circular areas;
The pixel that will include in the zone boundary of the border circular areas, the graphics field and non-circulating type character region
Point is set as background dot;
The minimum range in first center of circle and the border circular areas between foreground point is obtained, as the circular ocra font ocr
First radius of inscribed circle;
The maximum distance in first center of circle and the border circular areas between foreground point is obtained, as the circular ocra font ocr
Second radius of circumscribed circle.
6. a kind of character recognition device, which is characterized in that described device includes:
Center of circle determining module, for determining the center of circle comprising the border circular areas around ocra font ocr in images to be recognized, as first
The center of circle;
Radius obtains module, for obtaining the first radius of the inscribed circle around ocra font ocr and the second radius of circumscribed circle,
Wherein, the center of circle of the inscribed circle and the center of circle of circumscribed circle are first center of circle;
Area determination module is passed through and two adjacent for determining according to first radius and the second radius by first straight line
The side rectangular area parallel with horizontal direction, vertical direction respectively, wherein the first straight line are as follows: pass through first center of circle
Vertical straight line;
Characteristic extracting module, for being rotation angle step and using first center of circle as rotation axis points, edge using preset angle
The preset direction of rotation continuous rotation circular ocra font ocr region at least once, and extract be located at after rotation every time it is described
The feature of the picture material of rectangular area;
Feature obtains module, for forming indicatrix according to extracted feature, and obtains and belongs to pole on the indicatrix
The feature being worth greatly;
Character recognition module, the picture material for being characterized to acquired feature carry out character recognition, obtain recognition result.
7. device according to claim 6, which is characterized in that the characteristic extracting module is specifically used for:
Using preset angle as rotary step and using first center of circle as rotation axis points, along preset direction of rotation to described
It is gradually rotated around ocra font ocr region;
The quantity for counting foreground point in picture material to be processed, as the first quantity, wherein the picture material to be processed are as follows:
It is located at the picture material of the rectangular area after rotation every time;
The quantity for meeting the foreground point of preset condition in the picture material to be processed is counted, as the second quantity, wherein described
Preset condition are as follows: adjacent pixel is foreground point in the horizontal direction or neighbor pixel along the vertical direction is prospect
Point;
It is in varying numbers in zero in the case where described first, ratio of second quantity relative to first quantity is calculated,
Feature as the picture material to be processed.
8. device according to claim 7, which is characterized in that described device further include:
Section determining module, it is continuously bent no more than the curve section that the value of preset value is formed on the indicatrix for determining
Line;
Sort adjustment module is used for according to the preset direction of rotation, acquired feature and the curve section in the spy
The relative positional relationship on curve is levied, the sequence of character in obtained recognition result is adjusted.
9. device according to claim 6, which is characterized in that the area determination module is specifically used for:
The side length L of rectangular area is calculated according to following formula:
L=R2-R1+2P
Wherein, R1 indicates that first radius, R2 indicate that second radius, P indicate default value;
The coordinate of the lower-left angular vertex of the rectangular area is calculated according to following formula:
CL_X=CO_X-L/2
CL_Y=CO_Y+R1-P
Wherein, CL_X、CL_YIndicate cross, the ordinate of the lower-left angular vertex, CO_X、CO_YIndicate the horizontal, vertical of first center of circle
Coordinate;
According to the coordinate of the lower-left angular vertex, determine side length be L and adjacent two edges respectively with horizontal direction, vertical side
To parallel rectangular area.
10. the device according to any one of claim 6-9, which is characterized in that the radius obtains module, specific to use
In:
Determine the graphics field and non-circulating type character region in the border circular areas;
The pixel that will include in the zone boundary of the border circular areas, the graphics field and non-circulating type character region
Point is set as background dot;
The minimum range in first center of circle and the border circular areas between foreground point is obtained, as the circular ocra font ocr
First radius of inscribed circle;
The maximum distance in first center of circle and the border circular areas between foreground point is obtained, as the circular ocra font ocr
Second radius of circumscribed circle.
11. a kind of electronic equipment, which is characterized in that including processor and machine readable storage medium, the machine readable storage
Media storage has the machine-executable instruction that can be executed by the processor, and the processor is by the machine-executable instruction
Promote: realizing any method and step of claim 1-5.
12. a kind of machine readable storage medium, which is characterized in that be stored with machine-executable instruction, by processor call and
When execution, the machine-executable instruction promotes the processor: realizing any method and step of claim 1-5.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910701844.1A CN110414525B (en) | 2019-07-31 | 2019-07-31 | Character recognition method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910701844.1A CN110414525B (en) | 2019-07-31 | 2019-07-31 | Character recognition method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110414525A true CN110414525A (en) | 2019-11-05 |
CN110414525B CN110414525B (en) | 2022-02-25 |
Family
ID=68364686
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910701844.1A Active CN110414525B (en) | 2019-07-31 | 2019-07-31 | Character recognition method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110414525B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110991449A (en) * | 2019-11-28 | 2020-04-10 | 南京工程学院 | Character wheel character-oriented full-character and half-character recognition method |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090022426A1 (en) * | 2007-07-20 | 2009-01-22 | Noboru Yamazaki | Method And Device For Generating Character Data, Method And Control Device For Displaying Character Data, And Navigation Apparatus |
CN102663386A (en) * | 2012-03-27 | 2012-09-12 | 江南大学 | Online detection method of mechanical-bearing annular-distributed pressed character |
CN104239888A (en) * | 2014-09-10 | 2014-12-24 | 河海大学 | Method for positioning steel seal characters annularly distributed on water meter plate |
CN105046257A (en) * | 2015-07-31 | 2015-11-11 | 武汉理工大学 | Character recognition method and system for special marking form of main gear of rear axle |
CN105067638A (en) * | 2015-07-22 | 2015-11-18 | 广东工业大学 | Tire fetal-membrane surface character defect detection method based on machine vision |
US20160224863A1 (en) * | 2015-01-29 | 2016-08-04 | Dain Co., Ltd. | Tire recognition method and apparatus |
CN109657666A (en) * | 2018-11-23 | 2019-04-19 | 汉王科技股份有限公司 | A kind of circular stamp recognition methods and device |
CN109993739A (en) * | 2019-03-29 | 2019-07-09 | 深圳市创业***实业有限公司 | A kind of seal distinguishing method between true and false and device |
-
2019
- 2019-07-31 CN CN201910701844.1A patent/CN110414525B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090022426A1 (en) * | 2007-07-20 | 2009-01-22 | Noboru Yamazaki | Method And Device For Generating Character Data, Method And Control Device For Displaying Character Data, And Navigation Apparatus |
CN102663386A (en) * | 2012-03-27 | 2012-09-12 | 江南大学 | Online detection method of mechanical-bearing annular-distributed pressed character |
CN104239888A (en) * | 2014-09-10 | 2014-12-24 | 河海大学 | Method for positioning steel seal characters annularly distributed on water meter plate |
US20160224863A1 (en) * | 2015-01-29 | 2016-08-04 | Dain Co., Ltd. | Tire recognition method and apparatus |
CN105067638A (en) * | 2015-07-22 | 2015-11-18 | 广东工业大学 | Tire fetal-membrane surface character defect detection method based on machine vision |
CN105046257A (en) * | 2015-07-31 | 2015-11-11 | 武汉理工大学 | Character recognition method and system for special marking form of main gear of rear axle |
CN109657666A (en) * | 2018-11-23 | 2019-04-19 | 汉王科技股份有限公司 | A kind of circular stamp recognition methods and device |
CN109993739A (en) * | 2019-03-29 | 2019-07-09 | 深圳市创业***实业有限公司 | A kind of seal distinguishing method between true and false and device |
Non-Patent Citations (2)
Title |
---|
曾锐: ""***特征提取算法研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
黄剑航: ""基于HALCON的圆环区域字符识别实现"", 《图形图像》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110991449A (en) * | 2019-11-28 | 2020-04-10 | 南京工程学院 | Character wheel character-oriented full-character and half-character recognition method |
Also Published As
Publication number | Publication date |
---|---|
CN110414525B (en) | 2022-02-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111178337B (en) | Human face key point data enhancement method, device and system and model training method | |
TW202011733A (en) | Method and device for performing target sampling on pictures | |
JP4755202B2 (en) | Face feature detection method | |
US8515178B2 (en) | Method and system for image feature extraction | |
CN105550631B (en) | A kind of iris image acquiring method and device | |
CN107580175A (en) | A kind of method of single-lens panoramic mosaic | |
CN107945111B (en) | Image stitching method based on SURF (speeded up robust features) feature extraction and CS-LBP (local binary Pattern) descriptor | |
CN105678806B (en) | A kind of live pig action trail automatic tracking method differentiated based on Fisher | |
CN110333240B (en) | Method and device for detecting appearance of capacitor, storage medium and processor | |
CN111259878A (en) | Method and equipment for detecting text | |
CN108197642A (en) | A kind of seal discrimination method and device | |
Huang et al. | Automatic detection and localization of natural scene text in video | |
CN108537237A (en) | Characteristic area extracting method, device, equipment and the storage medium of bill | |
CN111126205A (en) | Optical remote sensing image airplane target detection method based on rotary positioning network | |
KR20200036098A (en) | Apparatus, method and system for detecting character | |
CN115063566B (en) | AR-based creative product display method and display equipment | |
CN112749696B (en) | Text detection method and device | |
CN110414525A (en) | A kind of character identifying method and device | |
CN116051439A (en) | Method, equipment and storage medium for removing rainbow-like glare of under-screen RGB image by utilizing infrared image | |
CN113570540A (en) | Image tampering blind evidence obtaining method based on detection-segmentation architecture | |
CN109299718A (en) | A kind of character identifying method and device | |
JP2004038980A (en) | Methods for analyzing image and generating threshold, program executing the method, computer-readable medium storing the program, and apparatus executing the method | |
WO2022183907A1 (en) | Image processing method and apparatus, intelligent invoice recognition device, and storage medium | |
US11734790B2 (en) | Method and apparatus for recognizing landmark in panoramic image and non-transitory computer-readable medium | |
Bilan et al. | Technology of Selection and Recognition of Information Objects on Images of the Earth's Surface Based on Multi-Projection Analysis. |
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 |