CN109214380B - License plate inclination correction method - Google Patents

License plate inclination correction method Download PDF

Info

Publication number
CN109214380B
CN109214380B CN201811062278.6A CN201811062278A CN109214380B CN 109214380 B CN109214380 B CN 109214380B CN 201811062278 A CN201811062278 A CN 201811062278A CN 109214380 B CN109214380 B CN 109214380B
Authority
CN
China
Prior art keywords
corner
point
pixel
value
image
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.)
Active
Application number
CN201811062278.6A
Other languages
Chinese (zh)
Other versions
CN109214380A (en
Inventor
陈世强
杨鼎鼎
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hubei Coland Technology Co ltd
Hubei University for Nationalities
Original Assignee
Hubei Coland Technology Co ltd
Hubei University for Nationalities
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Hubei Coland Technology Co ltd, Hubei University for Nationalities filed Critical Hubei Coland Technology Co ltd
Priority to CN201811062278.6A priority Critical patent/CN109214380B/en
Publication of CN109214380A publication Critical patent/CN109214380A/en
Application granted granted Critical
Publication of CN109214380B publication Critical patent/CN109214380B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/24—Aligning, centring, orientation detection or correction of the image
    • G06V10/243—Aligning, centring, orientation detection or correction of the image by compensating for image skew or non-uniform image deformations

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Image Analysis (AREA)

Abstract

The invention provides a license plate inclination correction method, which comprises the following steps: s1, acquiring a license plate RGB format image, converting the RGB format image into an HSI format image, and extracting an S component diagram in the HSI format image; s2, detecting the corner point coordinates of the S component image; s3, storing the detected corner coordinates into a two-dimensional matrix L; s4, searching a left lower corner, a left upper corner, a right lower corner and a right upper corner; s5, respectively solving two horizontal slopes and two vertical slopes according to the coordinates of the upper left corner point, the upper right corner point, the lower left corner point and the lower right corner point, and solving the average value of the two horizontal slopes and the average value of the two vertical slopes; s6, obtains the inclination angle α from the formula tan α ═ k between the slope and the inclination angle: and S7, finishing the license plate inclination correction according to the inclination angle alpha. The method is simple in calculation, the image is converted into the HSI format, the S component image in the format is calculated, the influence of brightness on product inclination correction is effectively eliminated, and the correction of the license plate inclination can be quickly realized.

Description

License plate inclination correction method
Technical Field
The invention relates to the field of computers, in particular to a license plate inclination correction method.
Background
With the development of national economy, automobiles become a necessary transportation tool for most families, and the number of the automobiles puts a great pressure on traffic management. The intelligent traffic system can well solve the traffic management pressure, and the license plate recognition is taken as a core module of the intelligent traffic system, so that the intelligent traffic system has important significance. The license plate inclination correction is a key step of a license plate recognition system, and directly influences the result of license plate recognition.
Ideally, the acquired license plate image is a rectangle parallel to the horizontal direction. Because the acquired images have certain inclination due to factors such as the shooting angle of the camera, the driving direction and speed of the vehicle, the distance between the lens and the license plate and the like, the license plate images need to be subjected to inclination correction, and a good basis is provided for subsequent license plate segmentation and license plate identification. At present, methods for license plate image inclination correction mainly include methods such as line detection, projection maximum value, corner point detection, principal component analysis and the like. (1) The inclination correction method based on the linear detection mainly comprises a least square fitting method, Hough transformation and Radon transformation methods, the methods finish correction by detecting the linear of the license plate frame, and the algorithm is simple. (2) The inclination correction based on the projection maximum value is completed through a projection method, and the anti-interference capability is strong. (3) The inclination correction based on the corner detection utilizes the characteristics of the corner, which can represent the main characteristics of things with minimum information, and can effectively finish the inclination correction of the license plate. (4) The inclination correction based on the principal component analysis completes the inclination correction by analyzing the main characteristics of the license plate, so that the calculated amount can be simplified, and the correction real-time performance is better. The 4 methods can finish the inclination correction of the license plate, but are sensitive to the light and shade.
Disclosure of Invention
In order to overcome the defects in the prior art, the invention aims to provide a license plate inclination correction method and system.
In order to achieve the above object, the present invention provides a license plate inclination correction method, comprising the following steps:
s1, acquiring a license plate RGB format image, converting the RGB format image into an HSI format image, and extracting an S component diagram in the HSI format image;
s2, detecting the corner point coordinates of the S component image;
s3, storing the detected corner coordinates into a two-dimensional matrix L;
s4, searching a left lower corner, a left upper corner, a right lower corner and a right upper corner;
wherein the abscissa of the lower left corner point is the smallest and the ordinate is the smallest; the abscissa of the upper left corner point is minimum, and the ordinate is maximum; the abscissa of the right lower corner point is maximum, and the ordinate is minimum; the corner point with the maximum abscissa and the maximum ordinate of the upper right corner point;
s5, respectively solving two horizontal slopes and two vertical slopes according to the coordinates of the upper left corner point, the upper right corner point, the lower left corner point and the lower right corner point, and solving the average value of the two horizontal slopes and the average value of the two vertical slopes;
s6, obtains the inclination angle α from the formula tan α ═ k between the slope and the inclination angle:
and S7, finishing the license plate inclination correction according to the inclination angle alpha.
The method is simple in calculation, converts the picture into the HSI format, calculates the S component picture in the format, effectively eliminates the influence of brightness on product inclination correction, and quickly realizes the correction of the license plate inclination.
Further, the step S2 includes the following steps:
s2-1, the gray value change of the pixel in any direction is described by the gray value change in the horizontal direction and the vertical direction, and the description formula is
Figure BDA0001797402620000031
Wherein I (x + u, y + v) ═ I (x, y) + Ixu+Iyv+O(u2,v2) (x, y) is the coordinates of the target pixel point; (u, v) represents the offset of other pixel points relative to the target pixel point; omegau,vAs a weighted window function, I is an image matrix, IxAnd IyRespectively, the first partial derivatives of the image matrix I in the horizontal and vertical directions.
S2-2, obtaining an autocorrelation matrix M of a pixel point (x, y): ignoring the high order infinitesimal minimums in the above description
Figure BDA0001797402620000032
To obtain
Figure BDA0001797402620000033
Wherein the content of the first and second substances,
Figure BDA0001797402620000034
m is an autocorrelation matrix of the pixel point (x, y);
s2-3, making the autocorrelation matrix M equal to 0, and obtaining two non-negative eigenvalues λ of the matrix M1And λ2Let λ be1≥λ2；
S2-4, when lambda2When the first threshold value is larger than the set first threshold value, the target pixel point is an angular point;
when lambda is2Is equal to 0 and lambda1If the value is larger than the first threshold value, indicating that the target pixel point is located at the edge, and executing the step S2-5;
when lambda is1If the value is equal to 0, indicating that the target pixel point is located in the flat area, executing step S2-5;
s2-5, calculating the corner response value R of the pixel point (x, y):
s2-6, traversing all pixel points of the whole S component image, executing the steps S2-1 to S2-6, and calculating the corner response values R of all the pixel points;
s2-7, if the response value of some pixel points is less than Th1If the pixel point is not an angular point, assigning the pixel value as B; if the response value of some pixel points is greater than Th1And is less than Th2The pixel point is a candidate angular point, and the pixel value is assigned as C; if the response value of some pixel points is greater than Th2The pixel point is an angular point, and the pixel value is assigned as D, wherein Th1<Th2And Th1、Th2B, C, D is a non-negative number;
and S2-8, performing non-maximum suppression on all candidate corner points in the image, suppressing pseudo corner points around the real corner points, and obtaining the corner points in the image.
The corner point coordinate confirmation method is simple and high in calculation speed, and missing detection of corner points and false corner points of the bar is reduced.
Further, in the step S2-1,
Figure BDA0001797402620000041
σ is the variance of (x + u) and (y + v), which can effectively reduce the amount of computation.
Further, the corner response value R ═ det (m) -kxtrace2(M), wherein det (M) ═ λ1λ2Is a determinant of the autocorrelation matrix M, trace (M) λ1+λ2K is a constant which is a trace of the autocorrelation matrix, which can effectively reduce the amount of computation.
Further, the Th1(maximum pixel value-minimum pixel value) 1/3, Th2(maximum pixel value-minimum pixel value) × 2/3, where maximum pixel value and minimum pixel value refer to the maximum pixel value and minimum pixel value in the S-component map.
The invention has the beneficial effects that:
1. the method mainly solves the influence of light brightness on license plate inclination correction, and adopts a color model and angular point detection to finish inclination correction.
2. The color model part can respectively extract each color channel and is applied in a targeted manner according to the characteristics of different channels.
3. The corner detection part uses an improved Harris corner detection algorithm, so that the calculation amount can be effectively reduced, and the missing detection corner and the false corner can be reduced.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The above and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
FIG. 1 is a flow chart of the present method;
fig. 2 is a flow chart of corner coordinate detection.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention, and are not to be construed as limiting the present invention.
In the description of the present invention, unless otherwise specified and limited, it is to be noted that the terms "mounted," "connected," and "connected" are to be interpreted broadly, and may be, for example, a mechanical connection or an electrical connection, a communication between two elements, a direct connection, or an indirect connection via an intermediate medium, and specific meanings of the terms may be understood by those skilled in the art according to specific situations.
As shown in fig. 1, the present invention provides a license plate inclination correction method, which comprises the following steps:
and S1, acquiring the license plate RGB format image, converting the RGB format image into an HSI format image, and extracting an S component map in the HSI format image.
And S2, detecting the coordinates of the corner points of the S component image.
The determination of the angular point coordinates can be performed by using a conventional Harris angular point detection method, and the present invention can also perform detection by using an improved Harris angular point detection method, and the detection is performed by the following steps, as shown in fig. 2:
s2-1, the gray value change of the pixel in any direction is described by the gray value change in the horizontal direction and the vertical direction, and the description formula is
Figure BDA0001797402620000061
Wherein I (x + u, y + v) ═ I (x, y) + Ixu+Iyv+O(u2,v2) (x, y) is the coordinates of the target pixel point; (u, v) represents the offset of other pixel points relative to the target pixel point; omegau,vIn order to weight the window function,
Figure BDA0001797402620000062
σ is the variance of (x + u) and (y + v). I is an image matrix, IxAnd IyRespectively representing the first partial derivatives of the image matrix I in the horizontal direction and the vertical direction; since the images are discrete sequences, the partial derivative IxAnd IyCan be obtained by a differential method, i.e. IxCan be obtained by subtracting pixel values of two adjacent lines, IyCan be obtained by subtracting the pixel values of two adjacent columns.
S2-2, obtaining an autocorrelation matrix M of a pixel point (x, y): ignoring the high order infinitesimal minimums in the above description
Figure BDA0001797402620000063
To obtain
Figure BDA0001797402620000064
Wherein the content of the first and second substances,
Figure BDA0001797402620000065
m is an autocorrelation matrix of the pixel point (x, y);
s2-3, knowing that M has two non-negative eigenvalues, marked as lambda, according to the symmetry of the matrix1And λ2Let the autocorrelation matrix M be 0, and obtain two non-negative eigenvalues λ of the matrix M1And λ2Let λ be1≥λ2。
S2-4, when lambda2And when the first threshold value is larger than the set first threshold value, the target pixel point is the angular point.
When lambda is2Is equal to 0 and lambda1If the value is greater than the first threshold, it indicates that the target pixel point is located at the edge, and step S2-5 is performed.
When lambda is1Equal to 0, indicating that the target pixel point is located in the flat area, and go to step S2-5.
Where lambda is2Equal to 0 includes λ2Case of approximately 0, with λ1Equal to 0 also includes λ1Approximately 0. The first threshold is set according to specific conditions, and the first threshold is not less than 150.
And S2-5, calculating the corner response value R of the pixel point (x, y).
Corner response value R ═ det (m) -kxtrace2(M), wherein det (M) ═ λ1λ2Is a determinant of the autocorrelation matrix M, trace (M) λ1+λ2K is a trace of the autocorrelation matrix and is a constant, and for convenience of operation, the value range of k is usually more than or equal to 0.04 and less than or equal to 0.06.
S2-6, traversing all pixel points of the whole S component image, executing the steps S2-1 to S2-6, and calculating the corner response value R of all the pixel points.
S2-7, if the response value of some pixel points is less than Th1If the pixel point is not an angular point, assigning the pixel value as B; if the response value of some pixel points is greater than Th1And is less than Th2The pixel point is a candidate angular point, and the pixel value is assigned as C; if the response value of some pixel points is greater than Th2The pixel point is an angular point, and the pixel value is assigned as D, wherein Th1<Th2And Th1、Th2B, C, D is a non-negative number. Where B is typically 0, C is typically 128, and D is typically 255.
Preferably, the Th is1(maximum pixel value-minimum pixel value) 1/3, Th2(maximum pixel value-minimum pixel value) × 2/3, where maximum pixel value and minimum pixel value refer to the maximum pixel value and minimum pixel value in the S-component map.
And S2-7, performing non-maximum suppression on all candidate corner points in the image, suppressing pseudo corner points around the real corner points, and obtaining all corner points in the image. The non-maximum suppression is to suppress the non-maximum, and find a candidate corner with the largest local pixel value, which is mainly used to eliminate a pseudo corner, suppress the pseudo corners around the real corner, and obtain the corners in the image.
And S3, obtaining coordinates of the corner point when the corner point is obtained, and storing the detected coordinates of the corner point into a two-dimensional matrix L.
And S4, finding a left lower corner point, a left upper corner point, a right lower corner point and a right upper corner point.
Wherein the abscissa of the lower left corner point is the smallest and the ordinate is the smallest; the abscissa of the upper left corner point is minimum, and the ordinate is maximum; the abscissa of the right lower corner point is maximum, and the ordinate is minimum; the corner point with the maximum abscissa and the maximum ordinate of the upper right corner point.
And S5, respectively solving two horizontal slopes and two vertical slopes according to the coordinates of the upper left corner point, the upper right corner point, the lower left corner point and the lower right corner point, and solving the average value of the two horizontal slopes and the average value of the two vertical slopes.
S6, the inclination angle α is obtained from the equation tan α ═ k between the slope and the inclination angle.
And S7, finishing the license plate inclination correction according to the inclination angle alpha.
And after the license plate inclination correction is finished, outputting a binary image of the corrected license plate image for subsequent image processing.
Because the license plate is influenced by the environment, the shot image has more salt and pepper noises, and after the license plate RGB format image is obtained, the image is subjected to median filtering to remove the noises in the image. Median filtering belongs to a non-linear filtering method, i.e. the pixel value at that position is replaced by the median value of the entire window. The median filtering is very effective in smoothing impulse noise, and meanwhile, sharp edges of images can be protected, a median point is selected to replace a value of a pollution point, the processing effect is good, and the impulse filtering is good in salt and pepper noise performance.
And (4) obviously reducing noise in the denoised image, and carrying out edge detection on the denoised image. Since the license plate characters are longitudinal textures, the background is a transverse texture. The Canny edge detection has more detailed effect, and can realize the non-differential detection of horizontal and vertical edges. In order to highlight the character area of the license plate, a Canny edge detection algorithm is improved, so that longitudinal textures can be better detected, and transverse textures are reduced. The improved model is as follows:
Figure BDA0001797402620000081
Figure BDA0001797402620000082
the improved Canny edge detection can better detect vertical edges, reduce interference of transverse edges, highlight characters and reduce the algorithm amount.
Due to the fact that the partial license plate area is broken or adhered after edge detection, morphological processing needs to be conducted on the filtered image, and breakage and adhesion are reduced. If the filtered license plate is well reserved, morphological processing can not be carried out on the license plate, otherwise, the license plate is processed.
The morphological image processing is a neighborhood operation form, a neighborhood structural element method is adopted to carry out specific logic operation on the neighborhood structural element and a binary image corresponding domain at each pixel position, and the result of the logic operation is the corresponding pixel of an output image. The basic operations of image morphology processing comprise erosion and expansion, and opening and closing operations. The most basic operations are erosion and dilation, and other operations are defined on the basis of the two.
The dilation process for image a with structuring element B is defined as:
X＝A⊕B＝{x:B(x)∩A≠Φ}；
the erosion process for image a with structuring element B is defined as:
Figure BDA0001797402620000091
the image A is an RGB format image, X represents each pixel point in the image A, B (X) represents a structural element, phi is a null set, and X is a result of expansion or corrosion of the image A. The result of etching A with B (x) is a set of all points that contain B in A after translating the structuring element B. The result of expanding a with B (x) is a set of points that shift the structural element B such that the intersection of B and a is not empty.
The formula for performing the open operation is as follows,
Figure BDA0001797402620000092
representing that set A is operated on by structural element B;
the formula for performing the close operation is as follows,
Figure BDA0001797402620000093
representing that set a is closed by structuring element B.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
While embodiments of the invention have been shown and described, it will be understood by those of ordinary skill in the art that: various changes, modifications, substitutions and alterations can be made to the embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims (4)

1. A license plate inclination correction method is characterized by comprising the following steps:
s1, acquiring a license plate RGB format image, converting the RGB format image into an HSI format image, and extracting an S component diagram in the HSI format image;
s2, detecting the corner point coordinates of the S component image;
specifically, the step S2 includes the following steps:
s2-1, the gray value change of the pixel in the S component diagram in any direction is described by the gray value change in the horizontal direction and the vertical direction, and the description formula is
Figure FDA0003099227210000011
Wherein I (x + u, y + v) ═ I (x, y) + Ixu+Iyv+O(u2,v2) (x, y) is the coordinate of the target pixel point, and (u, v) represents the offset of other pixel points relative to the target pixel point; omegau,vAs a weighted window function, I is an image matrix, IxAnd IyRespectively representing the first partial derivatives of the image matrix I in the horizontal direction and the vertical direction;
s2-2, obtaining an autocorrelation matrix M of a pixel point (x, y): ignoring the high order infinitesimal minimums in the above description
Figure FDA0003099227210000012
To obtain
Figure FDA0003099227210000013
Wherein the content of the first and second substances,
Figure FDA0003099227210000014
m is an autocorrelation matrix of the pixel point (x, y);
s2-3, making the autocorrelation matrix M equal to 0, and obtaining two non-negative eigenvalues λ of the matrix M1And λ2Let λ be1≥λ2；
S2-4, when lambda2If the value is larger than the first threshold value, the target image isThe pixel points are angular points;
when lambda is2Is equal to 0 and lambda1If the value is larger than the first threshold value, indicating that the target pixel point is located at the edge, and executing the step S2-5;
when lambda is1If the value is equal to 0, indicating that the target pixel point is located in the flat area, executing step S2-5;
s2-5, calculating the corner response value R of the pixel point (x, y):
s2-6, traversing all pixel points of the whole S component image, executing the steps S2-1 to S2-6, and calculating the corner response values R of all the pixel points;
s2-7, if the response value of some pixel points is less than Th1If the pixel point is not an angular point, assigning the pixel value as B; if the response value of some pixel points is greater than Th1And is less than Th2The pixel point is a candidate angular point, and the pixel value is assigned as C; if the response value of some pixel points is greater than Th2The pixel point is an angular point, and the pixel value is assigned as D, wherein Th1<Th2And Th1、Th2B, C, D is a non-negative number;
s2-8, performing non-maximum suppression on all candidate corner points in the image, suppressing pseudo corner points around the real corner points, and obtaining corner points in the image;
s3, storing the detected corner coordinates into a two-dimensional matrix L;
s4, searching a left lower corner, a left upper corner, a right lower corner and a right upper corner;
wherein the abscissa of the lower left corner point is the smallest and the ordinate is the smallest; the abscissa of the upper left corner point is minimum, and the ordinate is maximum; the abscissa of the right lower corner point is maximum, and the ordinate is minimum; the corner point with the maximum abscissa and the maximum ordinate of the upper right corner point;
s5, respectively solving two horizontal slopes and two vertical slopes according to the coordinates of the upper left corner point, the upper right corner point, the lower left corner point and the lower right corner point, and solving the average value of the two horizontal slopes and the average value of the two vertical slopes;
s6, obtains the inclination angle α from the formula tan α ═ k between the slope and the inclination angle:
and S7, finishing the license plate inclination correction according to the inclination angle alpha.
2. The method for correcting the inclination of a license plate of claim 1, wherein in step S2-1,
Figure FDA0003099227210000031
σ is the variance of (x + u) and (y + v).
3. The license plate inclination correction method of claim 1, wherein the corner response value R ═ det (M) -kxtrace2(M), wherein det (M) ═ λ1λ2Is a determinant of the autocorrelation matrix M, trace (M) λ1+λ2K is a constant, which is the trace of the autocorrelation matrix.
4. The method of claim 1, wherein the Th is a value of1(maximum pixel value-minimum pixel value) 1/3, Th2(maximum pixel value-minimum pixel value) × 2/3, where maximum pixel value and minimum pixel value refer to the maximum pixel value and minimum pixel value in the S-component map.
CN201811062278.6A 2018-09-12 2018-09-12 License plate inclination correction method Active CN109214380B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811062278.6A CN109214380B (en) 2018-09-12 2018-09-12 License plate inclination correction method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811062278.6A CN109214380B (en) 2018-09-12 2018-09-12 License plate inclination correction method

Publications (2)

Publication Number Publication Date
CN109214380A CN109214380A (en) 2019-01-15
CN109214380B true CN109214380B (en) 2021-10-01

Family

ID=64984054

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811062278.6A Active CN109214380B (en) 2018-09-12 2018-09-12 License plate inclination correction method

Country Status (1)

Country Link
CN (1) CN109214380B (en)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110500951B (en) * 2019-06-04 2021-03-09 湘潭大学 Car light glass shell size detection method based on machine vision
CN110634222B (en) * 2019-08-27 2021-07-09 河海大学 Bank bill information identification method
CN111207659A (en) * 2020-01-07 2020-05-29 河北科技大学 Chip parameter detection method and device based on capacitor array
CN112749735B (en) * 2020-12-30 2023-04-07 中冶赛迪信息技术(重庆)有限公司 Converter tapping steel flow identification method, system, medium and terminal based on deep learning
CN113506279B (en) * 2021-07-22 2024-07-02 浙江大华技术股份有限公司 Method and device for determining inclination angle of object, storage medium and electronic device
CN113657371B (en) * 2021-10-20 2021-12-21 成都宜泊信息科技有限公司 Camera angle adjusting method and system, storage medium and electronic equipment
CN114882489B (en) * 2022-07-07 2022-12-16 浙江智慧视频安防创新中心有限公司 Method, device, equipment and medium for horizontally correcting rotating license plate
CN117315664B (en) * 2023-09-18 2024-04-02 山东博昂信息科技有限公司 Scrap steel bucket number identification method based on image sequence

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104298966A (en) * 2014-09-17 2015-01-21 电子科技大学 License plate positioning method
CN104598905A (en) * 2015-02-05 2015-05-06 广州中国科学院软件应用技术研究所 License plate positioning method and device
CN106203433A (en) * 2016-07-13 2016-12-07 西安电子科技大学 In a kind of vehicle monitoring image, car plate position automatically extracts and the method for perspective correction
CN106503709A (en) * 2016-10-20 2017-03-15 江苏商贸职业学院 A kind of slag-soil truck characters on license plate intelligent identification Method
CN107909085A (en) * 2017-12-01 2018-04-13 中国科学院长春光学精密机械与物理研究所 A kind of characteristics of image Angular Point Extracting Method based on Harris operators
CN108241859A (en) * 2016-12-26 2018-07-03 浙江宇视科技有限公司 The bearing calibration of car plate and device

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104298966A (en) * 2014-09-17 2015-01-21 电子科技大学 License plate positioning method
CN104598905A (en) * 2015-02-05 2015-05-06 广州中国科学院软件应用技术研究所 License plate positioning method and device
CN106203433A (en) * 2016-07-13 2016-12-07 西安电子科技大学 In a kind of vehicle monitoring image, car plate position automatically extracts and the method for perspective correction
CN106503709A (en) * 2016-10-20 2017-03-15 江苏商贸职业学院 A kind of slag-soil truck characters on license plate intelligent identification Method
CN108241859A (en) * 2016-12-26 2018-07-03 浙江宇视科技有限公司 The bearing calibration of car plate and device
CN107909085A (en) * 2017-12-01 2018-04-13 中国科学院长春光学精密机械与物理研究所 A kind of characteristics of image Angular Point Extracting Method based on Harris operators

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
Harris角点检测与AP聚类结合的车牌定位方法;吴艳;《广西科技大学学报》;20140430;第25卷(第2期);全文 *

Also Published As

Publication number Publication date
CN109214380A (en) 2019-01-15

Similar Documents

Publication Publication Date Title
CN109214380B (en) License plate inclination correction method
CN112950508A (en) Drainage pipeline video data restoration method based on computer vision
CN106846359A (en) Moving target method for quick based on video sequence
CN112819772B (en) High-precision rapid pattern detection and recognition method
CN102831591B (en) Gaussian filter-based real-time defogging method for single image
CN110782477A (en) Moving target rapid detection method based on sequence image and computer vision system
CN107169972B (en) Non-cooperative target rapid contour tracking method
CN111062293B (en) Unmanned aerial vehicle forest flame identification method based on deep learning
CN110580709A (en) Target detection method based on ViBe and three-frame differential fusion
CN109035287B (en) Foreground image extraction method and device and moving vehicle identification method and device
CN107392095A (en) A kind of small IR targets detection algorithm based on mask image
CN107832674B (en) Lane line detection method
CN111709888B (en) Aerial image defogging method based on improved generation countermeasure network
CN109708658B (en) Visual odometer method based on convolutional neural network
Lian et al. A novel method on moving-objects detection based on background subtraction and three frames differencing
CN111353496B (en) Real-time detection method for infrared dim targets
CN112991369A (en) Method for detecting overall dimension of running vehicle based on binocular vision
CN108846844A (en) A kind of sea-surface target detection method based on sea horizon
CN111985314B (en) Smoke detection method based on ViBe and improved LBP
CN107818547A (en) The minimizing technology of the spiced salt and Gaussian mixed noise in a kind of sequence towards twilight image
CN109784145B (en) Target detection method based on depth map and storage medium
CN113205494B (en) Infrared small target detection method and system based on adaptive scale image block weighting difference measurement
CN112163606B (en) Infrared small target detection method based on block contrast weighting
Wu et al. Research on crack detection algorithm of asphalt pavement
CN112365516A (en) Virtual and real occlusion processing method in augmented reality

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