CN114299275A - Hough transform-based license plate inclination correction method - Google Patents
Hough transform-based license plate inclination correction method Download PDFInfo
- Publication number
- CN114299275A CN114299275A CN202111469904.5A CN202111469904A CN114299275A CN 114299275 A CN114299275 A CN 114299275A CN 202111469904 A CN202111469904 A CN 202111469904A CN 114299275 A CN114299275 A CN 114299275A
- Authority
- CN
- China
- Prior art keywords
- license plate
- image
- hough transform
- inclination correction
- correction method
- 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 32
- 238000012937 correction Methods 0.000 title claims abstract description 25
- 230000009466 transformation Effects 0.000 claims abstract description 14
- 238000001914 filtration Methods 0.000 claims abstract description 13
- 230000011218 segmentation Effects 0.000 claims abstract description 7
- 238000013519 translation Methods 0.000 claims description 11
- 238000001514 detection method Methods 0.000 claims description 10
- 238000013507 mapping Methods 0.000 claims description 8
- 238000003708 edge detection Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000001629 suppression Effects 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 description 14
- 230000009286 beneficial effect Effects 0.000 description 4
- 238000003384 imaging method Methods 0.000 description 4
- 230000004075 alteration Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- 230000002452 interceptive effect Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 238000003702 image correction Methods 0.000 description 1
- 230000002401 inhibitory effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- 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
- G06V10/24—Aligning, centring, orientation detection or correction of the image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
-
- 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
- G06V30/146—Aligning or centring of the image pick-up or image-field
-
- 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
-
- 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
- G06V30/18—Extraction of features or characteristics of the image
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention belongs to the field of intelligent transportation, and discloses a Hough transform-based license plate inclination correction method, wherein a picture to be detected is used for detecting the position of a license plate through a cascade classifier; adjusting the obtained license plate position by using a frame regression method; filtering the color of the license plate image; using a horizontal expansion algorithm for the license plate image; carrying out single identification on the license plate image by using a Canny operator; detecting a discontinuous point boundary by using Hough transform on the license plate image, and confirming a rotation angle; rotating the license plate image by using a rotation algorithm, and stretching according to the inclination angles of the upper frame and the lower frame; and performing character segmentation and recognition on the license plate after the inclination correction is completed. The license plate coordinate obtained by the invention is closer to a real license plate; the method uses horizontal expansion operation on the image, so that the character width of the image is reduced, and the number of interference straight line segments after Hough transformation and the time consumption of Hough transformation are reduced. The method and the device reduce the distortion of the license plate image and improve the accuracy of license plate recognition.
Description
Technical Field
The invention belongs to the technical field of intelligent traffic, and particularly relates to a Hough transform-based license plate inclination correction method.
Background
The inclined license plate correction belongs to one type of image correction processing, and aims to correct the inclined license plate in the image and facilitate license plate recognition by a machine. License plate slope is rectified and is widely used in the intelligent transportation field.
The key of the inclined license plate correction technology lies in how to efficiently and accurately find a theoretical inclined angle, a common method is to find an optimal inclined straight line, and in an actual scene of license plate recognition, the license plate recognition requirement of a side angle is possibly generated, and besides the inclination of the license plate, the expansion and distortion are also caused, so that certain difficulty is caused to a traditional inclined correction algorithm. The existing correction algorithm is mostly aimed at the rotation inclination, does not have the capability of side angle imaging recovery, and cannot meet the actual use requirement. .
Disclosure of Invention
In order to solve the problems of low license plate inclination correction efficiency and low final recognition accuracy caused by side angle imaging in the prior art, the invention provides a license plate inclination correction method based on Hough transform, which comprises the following steps:
detecting the position of a license plate of the picture to be detected by a cascade classifier;
adjusting the obtained license plate position by using a frame regression method, and obtaining a first license plate image from the adjusted license plate position;
converting the first license plate image from an RGB color space to an HSV color space, and performing color filtering to obtain a second license plate image;
removing the influence of character information on differential operation by using a horizontal expansion algorithm on the second license plate image to obtain a third license plate image;
carrying out single identification on the third license plate image by using a Canny operator, and excluding a set with noise;
detecting a discontinuous point boundary of the third license plate image by using Hough transform, and fitting a license plate frame in the image by using a straight line to confirm a rotation angle;
rotating the first license plate image by using a rotation algorithm, and stretching according to the inclination angles of an upper frame and a lower frame;
and performing character segmentation and recognition on the license plate after the inclination correction is completed.
Further, the step of detecting the license plate position of the picture to be detected by the cascade classifier comprises:
and converting the picture to be detected into a gray-scale image to accelerate the detection speed, and detecting the license plate position by using a trained XML classifier file through an Opencv cascade classifier.
Further, the frame regression method comprises the following steps:
translation transformation: Δ x ═ Pwdx(P),Δy=Phdy(P)
Scaling: sw=exp(dw(P)),Sh=exp(dh(P))
Where Δ x, Δ y are the translation, dx(P),dy(P) is the coordinates of the input window P, dw(P) and dh(P) is the width and height of the input window P, PwAnd PhIs a proportionality coefficient;
calculating a loss function: obtaining a predicted value according to the translation transformation and the scale scaling, and obtaining a loss function by minimizing the difference between the predicted value and the true value:
the function optimization objective is:
wherein phi5(Pi) is the feature vector of the input window P, w*Is the parameter to be learned and is,representing the translation amount and scale scaling of a real frame, mu being a hyper-parameter;
using gradient descentBy the method or least squares to obtain w*。
Further, the HSV color filtering step is as follows:
converting the image from the RGB color space to the HSV color space:
V=max(R,G,B)
If H<0 then H=H+360
wherein, R, G, B are color components in RGB color space, and V, S, H are components in HSV color space;
determining HSV color threshold values needing to be reserved according to camera exposure parameters, removing regions with too low S components and V components, and screening out regions with H components within the threshold values.
Further, the Canny edge detection step is as follows:
s051 Gaussian filtering: carrying out weighted average according to the gray values of a certain pixel point (x, y) in the image and the pixel points in the field of the pixel point according to a parameter rule generated by a two-dimensional Gaussian formula, and removing high-frequency noise superposed in the image; the two-dimensional gaussian formula is:
wherein σ is a constant;
s052 calculates gradient image and angle image: the filter obtained by gradient calculation by using a Gaussian filter is used as a gradient detection operator to calculate a gradient image, the closer the pixel point to the central point is, the greater the weight is, the angle image provides guidance for the direction of non-maximum suppression, and for a binary function z, f (x, y), a gradient calculation formula of the point (x, y) is as follows:
s053, carrying out non-maximum suppression on the gradient image to search for the local maximum of the pixel point, and eliminating the problems of edge thickness, width and weak edge interference;
s054, performing edge connection by using double thresholds, regarding a point smaller than a threshold as a false edge to be set to be 0, regarding a point larger than a high threshold as a strong edge to be set to be 1, and further checking pixel points between the low threshold and the high threshold;
and linking the edges in the high-threshold image into the contour to complete edge detection.
Further, the straight line detection step is as follows:
s061 spatial mapping: each pixel point in the binary image is mapped to a curve in a Hough space, and the mapping formula is as follows:
x cosθ+y sinθ=ρ
wherein theta is an included angle between the vertical line and the x axis, and rho is the distance of the polar diameter;
s062, taking a local maximum value, setting a threshold value and filtering an interference straight line;
s063 draws a straight line: two points (x) are arbitrarily selected from the straight line segment obtained in step S0621,y1) And (x)2,y2) The angle of the straight line segment is determined and converted to (-45 DEG, 45 DEG)]The following is calculated:
and (3) calculating an optimal angle set by using a clustering algorithm to obtain the average inclination angle of straight-line segments of the upper and lower frames of the license plate.
Further, the character segmentation method comprises the following steps:
calling a contour extraction function to obtain a character contour;
searching a circumscribed rectangle to solve the problem of Chinese character fracture;
and calling an opencv interface to load a GRU-based sequence model for character recognition.
The invention has the following beneficial effects:
according to the invention, a frame regression algorithm is used for carrying out secondary regression fitting on the license plate frame, so that the obtained license plate coordinate is closer to a real license plate;
the method uses horizontal expansion operation on the binary image to be processed, so that the character width of the binary image is reduced, and the number of interfering straight line segments after Hough transformation and the time consumption of Hough transformation are reduced.
The correction of the upper and lower double inclination angles provided by the invention can restore the distortion of the license plate image generated under the side angle imaging, and is more beneficial to improving the accuracy of license plate recognition.
Drawings
FIG. 1 is a flow chart of the Hough transform-based license plate inclination correction method;
FIG. 2 is a block diagram illustrating a regression process of the frame of the present invention;
FIG. 3 is a flow chart of Canny edge detection in the present invention;
FIG. 4 is a flow chart of line detection according to the present invention.
Detailed Description
The invention is further described with reference to the accompanying drawings, but the invention is not limited in any way, and any alterations or substitutions based on the teaching of the invention are within the scope of the invention.
S01 license plate positioning: converting the picture to be detected into a gray-scale image to accelerate the detection speed, and detecting the license plate position by using a trained XML classifier file through an opencv cascade classifier (Cascade classifier);
and (3) regression of S02 frames: in order to obtain a regression window closer to the real license plate coordinates, the invention uses frame regression to adjust the license plate position obtained in the step 1 by using a frame regression algorithm. The algorithm has the function of searching a mapping function to enable regression point positions obtained by transforming original coordinate point positions to be closer to real coordinate point positions. Fig. 2 is a frame regression flow chart, which includes the following steps:
s021 translational transformation: Δ x ═ Pwdx(P),Δy=Phdy(P)
S022 scale scaling: sw=exp(dw(P)),Sh=exp(dh(P))
Where Δ x, Δ y are the translation, dx(P),dy(P) is the coordinates of the input window P, dw(P) and dh(P) is the width and height of the input window P, PwAnd PhIs a proportionality coefficient;
s023 calculates a loss function: obtaining a predicted value according to the translation transformation and the scale scaling, and obtaining a loss function by minimizing the difference between the predicted value and the true value:
the function optimization objective is:
wherein phi5(Pi) is the feature vector of the input window P, w*Is the parameter to be learned and is,representing the translation amount and scale scaling of a real frame, mu being a hyper-parameter;
w can be obtained by gradient descent method or least square method*。
S03 HSV color filtration: the image obtained in step S02 is converted from the RGB color space to the HSV color space. HSV is a color model facing users, and can better complete color filtering operation. The specific operation steps are as follows:
s031 color space conversion: the image is converted from the RGB color space to the HSV color space according to equations 3, 4, 5, and 6. Wherein, R, G, B are color components in RGB color space, and V, S, H are components in HSV color space.
V=max(R,G,B) (3)
If H<0 then H=H+360 (6)
S032: determining HSV color threshold values needing to be reserved according to camera exposure parameters, removing regions with too low S components and V components, and screening out regions with H components within the threshold values.
S04 horizontal expansion: in the invention, a horizontal expansion algorithm is used before Hough transformation, so that the influence of character information on differential operation is removed as much as possible, and the operation amount of Hough transformation is reduced.
S05 Canny edge detection: and (5) carrying out single identification on the binary image by using a Canny operator, and simultaneously excluding the set with noise. The implementation flow is shown in fig. 3, and the specific implementation steps are as follows:
s051 Gaussian filtering: and carrying out weighted average according to the gray values of a certain pixel point in the image and the pixel points in the field of the pixel point according to the parameter rule generated by the Gaussian company, and removing the high-frequency noise superposed in the image.
Wherein, the two-dimensional Gaussian formula is:
where σ is a constant.
The Gaussian filter used is
Wherein the value of the filter corresponds to the weight of the pixel.
S052 calculates gradient image and angle image: and taking a filter obtained by gradient calculation by using a Gaussian filter as a gradient detection operator to calculate a gradient image, wherein the closer the pixel point is to the central point, the higher the weight is. The angle image provides guidance for the direction of non-maximum suppression, and the calculation formula is as in formula 8:
s053, carrying out non-maximum suppression on the gradient image: and a non-maximum value is used for inhibiting and searching the local maximum value of the pixel point, and the problems of thick and wide edges and weak edge interference are solved.
S054 edge join using dual thresholds: and (3) eliminating false edge interference by using double thresholds, regarding the point smaller than the offset threshold as a false edge to be set to be 0, regarding the point larger than the high threshold as a strong edge to be set to be 1, and further checking pixel points between the low threshold and the high threshold.
Linking the edges into the contour according to the high-threshold image, searching a point meeting a low threshold value in 8 neighborhood points of a breakpoint by the algorithm when reaching the end point of the contour, and collecting a new edge according to the point until the whole image is closed to finish edge detection.
And S06 straight line detection: and (4) detecting the boundary of the discontinuous points by using Hough transform on the picture obtained in the step (4), and fitting a license plate frame in the picture by using a straight line to confirm the rotation angle. The implementation flow is shown in fig. 4, and the specific description of each step is as follows:
s061 spatial mapping: and mapping each pixel point in the binary image to a curve in the Hough space. One coordinate point in the cartesian coordinate plane is transformed into the polar coordinate system, corresponding to one curve in the hough space. The mapping formula is as formula 9
x cosθ+y sinθ=ρ (9)
Wherein theta is an included angle between the vertical line and the x axis, and rho is the distance of the pole diameter.
S062 takes local maxima: and setting a threshold value according to the local maximum value, and filtering the interference straight line.
S063 draws a straight line: any two points on the straight line segment obtained in step S062 are taken to find the angle of the straight line segment according to equation 10 and converted to between (-45 °,45 °).
And (3) calculating an optimal angle set by using a clustering algorithm to obtain the average inclination angle of straight-line segments of the upper and lower frames of the license plate.
S07 license plate rotation correction: and (3) rotating the original license plate image obtained in the step (2) by using a rotation algorithm, and stretching according to the inclination angles of the upper frame and the lower frame.
S08 license plate recognition: and performing character segmentation and recognition on the license plate after the inclination correction is completed. The character segmentation method comprises the following steps: calling a contour extraction function to obtain a character contour; and then searching a circumscribed rectangle to solve the Chinese character fracture problem. And calling an opencv interface to load a GRU-based sequence model for character recognition.
The invention has the following beneficial effects:
according to the invention, a frame regression algorithm is used for carrying out secondary regression fitting on the license plate frame, so that the obtained license plate coordinate is closer to a real license plate;
the method uses horizontal expansion operation on the binary image to be processed, so that the character width of the binary image is reduced, and the number of interfering straight line segments after Hough transformation and the time consumption of Hough transformation are reduced.
The correction of the upper and lower double inclination angles provided by the invention can restore the distortion of the license plate image generated under the side angle imaging, and is more beneficial to improving the accuracy of license plate recognition.
The word "preferred" is used herein to mean serving as an example, instance, or illustration. Any aspect or design described herein as "preferred" is not necessarily to be construed as advantageous over other aspects or designs. Rather, use of the word "preferred" is intended to present concepts in a concrete fashion. The term "or" as used in this application is intended to mean an inclusive "or" rather than an exclusive "or". That is, unless specified otherwise or clear from context, "X employs A or B" is intended to include either of the permutations as a matter of course. That is, if X employs A; b is used as X; or X employs both A and B, then "X employs A or B" is satisfied in any of the foregoing examples.
Also, although the disclosure has been shown and described with respect to one or an implementation, equivalent alterations and modifications will occur to others skilled in the art based upon a reading and understanding of this specification and the annexed drawings. The present disclosure includes all such modifications and alterations, and is limited only by the scope of the appended claims. In particular regard to the various functions performed by the above described components (e.g., elements, etc.), the terms used to describe such components are intended to correspond, unless otherwise indicated, to any component which performs the specified function of the described component (e.g., that is functionally equivalent), even though not structurally equivalent to the disclosed structure which performs the function in the herein illustrated exemplary implementations of the disclosure. In addition, while a particular feature of the disclosure may have been disclosed with respect to only one of several implementations, such feature may be combined with one or other features of the other implementations as may be desired and advantageous for a given or particular application. Furthermore, to the extent that the terms "includes," has, "" contains, "or variants thereof are used in either the detailed description or the claims, such terms are intended to be inclusive in a manner similar to the term" comprising.
Each functional unit in the embodiments of the present invention may be integrated into one processing module, or each unit may exist alone physically, or a plurality of or more than one unit are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium. The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc. Each apparatus or system described above may execute the storage method in the corresponding method embodiment.
In summary, the above-mentioned embodiment is an implementation manner of the present invention, but the implementation manner of the present invention is not limited by the above-mentioned embodiment, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be regarded as equivalent replacements within the protection scope of the present invention.
Claims (7)
1. A license plate inclination correction method based on Hough transform is characterized by comprising the following steps:
detecting the position of a license plate of the picture to be detected by a cascade classifier;
adjusting the obtained license plate position by using a frame regression method, and obtaining a first license plate image from the adjusted license plate position;
converting the first license plate image from an RGB color space to an HSV color space, and performing color filtering to obtain a second license plate image;
removing the influence of character information on differential operation by using a horizontal expansion algorithm on the second license plate image to obtain a third license plate image;
carrying out single identification on the third license plate image by using a Canny operator, and excluding a set with noise;
detecting a discontinuous point boundary of the third license plate image by using Hough transform, and fitting a license plate frame in the image by using a straight line to confirm a rotation angle;
rotating the first license plate image by using a rotation algorithm, and stretching according to the inclination angles of an upper frame and a lower frame;
and performing character segmentation and recognition on the license plate after the inclination correction is completed.
2. The Hough transform-based license plate inclination correction method according to claim 1, wherein the detecting the license plate position of the picture to be detected by a cascade classifier comprises:
and converting the picture to be detected into a gray-scale image to accelerate the detection speed, and detecting the license plate position by using a trained XML classifier file through an Opencv cascade classifier.
3. The Hough transform-based license plate inclination correction method according to claim 1, wherein the border regression method comprises the following steps:
translation transformation: Δ x ═ Pwdx(P),Δy=Phdy(P)
Scaling: sw=exp(dw(P)),Sh=exp(dh(P))
Where Δ x, Δ y are the translation, dx(P),dy(P) is the coordinates of the input window P, dw(P) and dh(P) is the width and height of the input window P, PwAnd PhIs a proportionality coefficient;
calculating a loss function: obtaining a predicted value according to the translation transformation and the scale scaling, and obtaining a loss function by minimizing the difference between the predicted value and the true value:
the function optimization objective is:
wherein phi5(Pi) is the feature vector of the input window P, w*Is the parameter to be learned and is,representing the translation amount and scale scaling of a real frame, mu being a hyper-parameter;
obtaining w by gradient descent or least square method*。
4. The Hough transform-based license plate inclination correction method according to claim 1, wherein the HSV color filtering step is as follows:
converting the image from the RGB color space to the HSV color space:
V=max(R,G,B)
If H<0 then H=H+360
wherein, R, G, B are color components in RGB color space, and V, S, H are components in HSV color space;
determining HSV color threshold values needing to be reserved according to camera exposure parameters, removing regions with too low S components and V components, and screening out regions with H components within the threshold values.
5. The Hough transform-based license plate tilt correction method according to claim 1, wherein the Canny edge detection step is as follows:
s051 Gaussian filtering: carrying out weighted average according to the gray values of a certain pixel point (x, y) in the image and the pixel points in the field of the pixel point according to a parameter rule generated by a two-dimensional Gaussian formula, and removing high-frequency noise superposed in the image; the two-dimensional gaussian formula is:
wherein σ is a constant;
s052 calculates gradient image and angle image: the filter obtained by gradient calculation by using a Gaussian filter is used as a gradient detection operator to calculate a gradient image, the closer the pixel point to the central point is, the greater the weight is, the angle image provides guidance for the direction of non-maximum suppression, and for a binary function z, f (x, y), a gradient calculation formula of the point (x, y) is as follows:
s053, carrying out non-maximum suppression on the gradient image to search for the local maximum of the pixel point, and eliminating the problems of edge thickness, width and weak edge interference;
s054, performing edge connection by using double thresholds, regarding a point smaller than a threshold as a false edge to be set to be 0, regarding a point larger than a high threshold as a strong edge to be set to be 1, and further checking pixel points between the low threshold and the high threshold;
and linking the edges in the high-threshold image into the contour to complete edge detection.
6. The Hough transform-based license plate inclination correction method according to claim 1, wherein the straight line detection step is as follows:
s061 spatial mapping: each pixel point in the binary image is mapped to a curve in a Hough space, and the mapping formula is as follows:
x cosθ+y sinθ=ρ
wherein theta is an included angle between the vertical line and the x axis, and rho is the distance of the polar diameter;
s062, taking a local maximum value, setting a threshold value and filtering an interference straight line;
s063 draws a straight line: two points (x) are arbitrarily selected from the straight line segment obtained in step S0621,y1) And (x)2,y2) The angle of the straight line segment is determined and converted to (-45 DEG, 45 DEG)]The following is calculated:
and (3) calculating an optimal angle set by using a clustering algorithm to obtain the average inclination angle of straight-line segments of the upper and lower frames of the license plate.
7. The Hough transform-based license plate inclination correction method according to claim 1, wherein the character segmentation method comprises:
calling a contour extraction function to obtain a character contour;
searching a circumscribed rectangle to solve the problem of Chinese character fracture;
and calling an opencv interface to load a GRU-based sequence model for character recognition.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111469904.5A CN114299275B (en) | 2021-12-03 | 2021-12-03 | License plate inclination correction method based on Hough transformation |
PCT/CN2022/081179 WO2023097931A1 (en) | 2021-12-03 | 2022-03-16 | Hough transform-based license plate tilt correction method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111469904.5A CN114299275B (en) | 2021-12-03 | 2021-12-03 | License plate inclination correction method based on Hough transformation |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114299275A true CN114299275A (en) | 2022-04-08 |
CN114299275B CN114299275B (en) | 2023-07-18 |
Family
ID=80965515
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111469904.5A Active CN114299275B (en) | 2021-12-03 | 2021-12-03 | License plate inclination correction method based on Hough transformation |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN114299275B (en) |
WO (1) | WO2023097931A1 (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN118314191A (en) * | 2024-06-06 | 2024-07-09 | 无锡市方顺型钢科技有限公司 | Profile steel outline dimension accurate measurement method based on data analysis |
Families Citing this family (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116468640B (en) * | 2023-06-20 | 2023-08-29 | 山东正禾大教育科技有限公司 | Video image enhancement method for Internet teaching |
CN116740180A (en) * | 2023-06-30 | 2023-09-12 | 西北工业大学 | Two-machine near vision relative positioning method based on feature points |
CN117079284B (en) * | 2023-07-26 | 2024-03-05 | 苏州城市学院 | Character and image oriented inclination detection and correction method and system |
CN116912126A (en) * | 2023-07-26 | 2023-10-20 | 新疆大学 | A high-brightness license plate image enhancement method |
CN117078661B (en) * | 2023-10-12 | 2024-01-05 | 济宁久邦工程机械设备有限公司 | Intelligent detection method for production quality of structural member of overhead working truck |
CN117649559B (en) * | 2023-12-12 | 2024-06-14 | 兰州交通大学 | Intelligent parking lot parking space detection system based on deep learning |
CN117522863B (en) * | 2023-12-29 | 2024-03-29 | 临沂天耀箱包有限公司 | Integrated box body quality detection method based on image features |
CN117689677B (en) * | 2024-02-01 | 2024-04-16 | 山东大学日照研究院 | Grinding wheel abrasion state identification method, system, equipment and medium |
CN117974605B (en) * | 2024-02-02 | 2024-08-02 | 山东福茂装饰材料有限公司 | Method for detecting edge sealing defect of plate based on image |
CN117994347B (en) * | 2024-04-07 | 2024-06-11 | 宝鸡市鹏盛鑫有色金属有限责任公司 | High-precision positioning method for flange machining drilling |
CN118097581B (en) * | 2024-04-28 | 2024-06-25 | 山东领军智能交通科技有限公司 | Road edge recognition control method and device |
CN118711192B (en) * | 2024-08-29 | 2024-11-19 | 陕西风华时代环境工程有限公司 | A license plate image processing method |
CN118735952B (en) * | 2024-09-04 | 2025-01-07 | 航天赛博产业技术研究院 | A high-speed image cutting method and device |
CN119229455A (en) * | 2024-09-06 | 2024-12-31 | 深圳市海蓝智能科技有限公司 | A method for visual detection of magnetic ring line sequence characters |
CN118799314B (en) * | 2024-09-12 | 2024-11-15 | 大连环旭智能科技有限公司 | Detection information processing method for wind power and photovoltaic power generation equipment |
CN118781119B (en) * | 2024-09-12 | 2025-02-25 | 国网山东省电力公司聊城供电公司 | Surface damage detection system for power grid equipment based on image analysis |
CN119832555B (en) * | 2025-03-14 | 2025-05-23 | 湖南至简复印机再制造有限公司 | A method and system for automatically correcting printed documents based on deep learning |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6553131B1 (en) * | 1999-09-15 | 2003-04-22 | Siemens Corporate Research, Inc. | License plate recognition with an intelligent camera |
US20170300786A1 (en) * | 2015-10-01 | 2017-10-19 | Intelli-Vision | Methods and systems for accurately recognizing vehicle license plates |
CN109145915A (en) * | 2018-07-27 | 2019-01-04 | 武汉科技大学 | License plate rapid distortion antidote under a kind of complex scene |
CN109886896A (en) * | 2019-02-28 | 2019-06-14 | 闽江学院 | A blue license plate segmentation and correction method |
CN110619335A (en) * | 2019-09-20 | 2019-12-27 | 河南工业大学 | License plate positioning and character segmentation method |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109284757A (en) * | 2018-08-31 | 2019-01-29 | 湖南星汉数智科技有限公司 | A kind of licence plate recognition method, device, computer installation and computer readable storage medium |
CN112598066A (en) * | 2020-12-25 | 2021-04-02 | 中用科技有限公司 | Lightweight road pavement detection method and system based on machine vision |
CN112766273A (en) * | 2021-01-20 | 2021-05-07 | 湘潭大学 | License plate recognition method |
CN112862769A (en) * | 2021-01-28 | 2021-05-28 | 中南大学 | On-line intelligent monitoring method and system for blast furnace slag-iron ratio |
-
2021
- 2021-12-03 CN CN202111469904.5A patent/CN114299275B/en active Active
-
2022
- 2022-03-16 WO PCT/CN2022/081179 patent/WO2023097931A1/en active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6553131B1 (en) * | 1999-09-15 | 2003-04-22 | Siemens Corporate Research, Inc. | License plate recognition with an intelligent camera |
US20170300786A1 (en) * | 2015-10-01 | 2017-10-19 | Intelli-Vision | Methods and systems for accurately recognizing vehicle license plates |
CN109145915A (en) * | 2018-07-27 | 2019-01-04 | 武汉科技大学 | License plate rapid distortion antidote under a kind of complex scene |
CN109886896A (en) * | 2019-02-28 | 2019-06-14 | 闽江学院 | A blue license plate segmentation and correction method |
CN110619335A (en) * | 2019-09-20 | 2019-12-27 | 河南工业大学 | License plate positioning and character segmentation method |
Non-Patent Citations (9)
Title |
---|
于明等: "融合Hough与Radon变换的车牌倾斜校正算法", 《控制工程》 * |
于明等: "融合Hough与Radon变换的车牌倾斜校正算法", 《控制工程》, vol. 20, no. 06, 20 November 2013 (2013-11-20) * |
双锴: "边框回归算法", 《计算机视觉》 * |
双锴: "边框回归算法", 《计算机视觉》, 31 January 2020 (2020-01-31), pages 114 - 117 * |
王志喜 等 * |
王志喜 等, article 霍冠英 等: "边缘检测算法与Canny算子", pages: 199 - 203 * |
许珊珊: "车牌图像倾斜校正算法的研究", 《内江科技》 * |
许珊珊: "车牌图像倾斜校正算法的研究", 《内江科技》, vol. 39, no. 5, 25 May 2018 (2018-05-25), pages 2 * |
霍冠英 等: "边缘检测算法与Canny算子", 《侧扫声呐图像目标分割》, 30 April 2017 (2017-04-30), pages 57 - 62 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN118314191A (en) * | 2024-06-06 | 2024-07-09 | 无锡市方顺型钢科技有限公司 | Profile steel outline dimension accurate measurement method based on data analysis |
Also Published As
Publication number | Publication date |
---|---|
CN114299275B (en) | 2023-07-18 |
WO2023097931A1 (en) | 2023-06-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN114299275A (en) | Hough transform-based license plate inclination correction method | |
CN109086714B (en) | Form recognition method, recognition system and computer device | |
CN109145915B (en) | A Fast Distortion Correction Method for License Plates in Complex Scenes | |
CN109886896B (en) | A blue license plate segmentation and correction method | |
CN103971128B (en) | A kind of traffic sign recognition method towards automatic driving car | |
CN104715252B (en) | A kind of registration number character dividing method of dynamic template combination pixel | |
CN109784344A (en) | An image non-target filtering method for ground plane identification recognition | |
CN106815583B (en) | A night vehicle license plate location method based on the combination of MSER and SWT | |
CN103903018A (en) | Method and system for positioning license plate in complex scene | |
CN109409356B (en) | Multi-direction Chinese print font character detection method based on SWT | |
CN111027544B (en) | MSER license plate positioning method and system based on visual saliency detection | |
CN113903024A (en) | Handwritten bill numerical value information identification method, system, medium and device | |
CN109460722B (en) | Intelligent license plate recognition method | |
CN109583435A (en) | A kind of adaptive ballot method for detecting lane lines | |
CN110060259A (en) | A kind of fish eye lens effective coverage extracting method based on Hough transformation | |
CN111353961A (en) | Document curved surface correction method and device | |
CN106709500A (en) | A Method of Image Feature Matching | |
CN107862319B (en) | Heterogeneous high-light optical image matching error eliminating method based on neighborhood voting | |
CN112767358A (en) | Railway electric locomotive fault detection method based on image feature registration | |
CN101315664A (en) | Text Image Preprocessing Method for Text Recognition | |
CN106778736A (en) | The licence plate recognition method and its system of a kind of robust | |
CN110689003A (en) | Low-illumination imaging license plate recognition method and system, computer equipment and storage medium | |
CN112734816B (en) | Heterogeneous image registration method based on CSS-Delaunay | |
CN116883868A (en) | UAV intelligent cruise detection method based on adaptive image defogging | |
CN116052152A (en) | License plate recognition system based on contour detection and deep neural network |
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 |