CN111310754A - Method for segmenting license plate characters - Google Patents
Method for segmenting license plate characters Download PDFInfo
- Publication number
- CN111310754A CN111310754A CN201911411732.9A CN201911411732A CN111310754A CN 111310754 A CN111310754 A CN 111310754A CN 201911411732 A CN201911411732 A CN 201911411732A CN 111310754 A CN111310754 A CN 111310754A
- Authority
- CN
- China
- Prior art keywords
- license plate
- chinese character
- characters
- boundary
- segmenting
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000012545 processing Methods 0.000 claims abstract description 20
- 238000001514 detection method Methods 0.000 claims abstract description 17
- 238000010606 normalization Methods 0.000 claims abstract description 9
- 238000006243 chemical reaction Methods 0.000 claims abstract description 5
- 238000012937 correction Methods 0.000 claims abstract description 5
- 238000001914 filtration Methods 0.000 claims description 6
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 238000011176 pooling Methods 0.000 claims 1
- 238000013527 convolutional neural network Methods 0.000 abstract description 5
- 230000011218 segmentation Effects 0.000 abstract description 5
- 238000004364 calculation method Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- 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
- G06V20/63—Scene text, e.g. street names
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of 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
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- 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
- 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
- G06V20/625—License plates
Abstract
The invention discloses a method for segmenting license plate characters, which belongs to the field of license plate recognition and comprises the steps of object detection, convolutional neural network regression prediction, normalization processing, color conversion processing, binarization processing, contour detection, boundary positioning and cutting. According to the method for separating the license plate characters, the positions of the characters are determined according to the set thresholds of the width and the height of the license plate characters through the steps of license plate positioning, angle correction, image normalization and the like, so that the precision of license plate character segmentation is greatly improved, and the subsequent license plate character recognition is guaranteed.
Description
Technical Field
The invention belongs to the technical field of license plate recognition, and particularly relates to a method for segmenting license plate characters.
Background
At present, most of the traditional methods for recognizing the license plate based on deep learning adopt character segmentation and then recognize the segmented characters through a neural network. Due to the fact that a certain angle exists in license plate shooting, character boundary positioning is not accurate, segmentation accuracy is low, interference is caused to a subsequent recognition process, overall recognition accuracy is reduced, and license plate Chinese character recognition accuracy is affected.
Disclosure of Invention
The present invention is directed to a method for segmenting license plate characters, so as to solve the problems mentioned in the background art.
In order to achieve the purpose, the invention provides the following technical scheme:
a method for segmenting license plate characters, the method comprising the steps of:
s1, roughly positioning the license plate by using the trained license plate detection model to obtain a license plate picture;
s2, carrying out regression calculation on the obtained license plate picture through a Convolutional Neural Network (CNN), and carrying out forward propagation once to obtain coordinates of each vertex of the license plate;
s3, performing angle correction through affine transformation by using coordinates of four vertexes of the license plate to obtain a front view of the license plate;
s4, normalization processing, namely adjusting the positioned license plate pictures to be uniform in size;
s5, color conversion processing, namely, converting the color image subjected to the normalization processing in the S4 into a single-channel gray image;
s6, binarization processing, namely filtering the gray level image processed by the S5 by Gaussian filtering and carrying out binarization processing to obtain a binary image;
s7, contour detection, namely, carrying out outer contour detection on the binary image obtained in the S6 to obtain the outer contours of all characters and upper and lower boundary values of the outer contours;
s8, sorting the contours obtained in the S7 according to the X-axis direction of the contour position coordinates;
s9, cutting the Chinese character part according to the boundary values of the upper, lower, left and right sides of the Chinese characters obtained by sequencing in the S8 to obtain a picture of the Chinese character part;
s10, dividing the numbers and the alphabetic characters to obtain the final images of all the license plate characters
The invention also has the following technical characteristics:
the size of the uniform dimension in S4 is 720pix × 224 pix.
The contour detection in S7 is outer contour detection.
The method for locating the boundary in S7 includes the following steps:
s701, determining the width and height of the character cut by the uniform size of the license plate, and determining the width and height values of the license plate characters:
defining the width of the Chinese character as W and the height as H;
setting W <90pix, H <165 pix;
carrying out contour screening according to the set conditions of W and H to obtain each part of the Chinese character part;
s702, obtaining coordinates (x, y) of the upper left corner of each outline of the Chinese character part in the binary image in S701; setting the width of each outline of the Chinese character part as w, the height as h, the leftmost boundary of the Chinese character as the minimum value of x, the rightmost boundary of the Chinese character as the maximum value of x + w, the upper boundary of the Chinese character as the minimum value of y, and the lower boundary of the Chinese character as the maximum value of y + h, so as to obtain the upper, lower, left and right boundary values of the Chinese character part.
Compared with the prior art, the invention has the beneficial effects that: the method for segmenting the license plate characters sets the uniform size of the license plate picture through angle correction and normalization operation by utilizing a license plate front view, and performs preliminary estimation on the position of the Chinese character according to the set threshold values of the width and the height of the Chinese character, so that the outline of a non-Chinese character part can be eliminated, and the segmentation precision is improved; by adopting single-channel picture processing, the influence of factors such as license plate color and light is small, and the precision of license plate recognition is greatly improved.
Drawings
FIG. 1 is a flow chart of the steps of a method for segmenting license plate characters;
FIG. 2 is a front view of a license plate adjusted to a uniform size in a method of segmenting characters of the license plate;
FIG. 3 is a license plate gray scale image after color conversion in the method for segmenting license plate characters;
FIG. 4 is a license plate image after binarization processing and contour detection in the method for segmenting license plate characters;
FIG. 5 is a diagram of portions of license plate characters cut according to contours in a method for segmenting license plate characters;
FIG. 6 is a license plate Chinese character diagram in a method of segmenting license plate characters;
fig. 7 shows all the characters obtained in the method for segmenting the license plate characters (arranged in the order of the license plate characters).
Detailed Description
The present invention will be further described with reference to the following examples.
The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention. The conditions in the embodiments can be further adjusted according to specific conditions, and simple modifications of the method of the present invention based on the concept of the present invention are within the scope of the claimed invention.
Referring to fig. 1 to 6, the present invention provides a method for segmenting license plate characters, including the steps of:
s1, detecting the image, roughly positioning the license plate by adopting an object detection algorithm to obtain a license plate image, and training a license plate positioning model in advance before detection;
s2, performing primary forward calculation on the obtained license plate picture through a trained Convolutional Neural Network (CNN) regression model to obtain coordinates of each vertex of the license plate;
s3, performing angle correction through affine transformation by using coordinates of four vertexes of the license plate to obtain a front view of the license plate;
s4, normalization processing, namely, adjusting the license plate front-view picture to be uniform in size of 720pix × 224pix shown in figure 2;
s5, color conversion processing, converting the color map subjected to the normalization processing in S4 into a single-channel gray map as shown in fig. 3;
s6, binarization processing, namely, filtering the gray scale map processed in S5 by gaussian filtering and performing binarization processing to obtain a binary map as shown in fig. 4;
s7, contour detection, namely, carrying out outer contour detection on the binary image obtained in the step S6 to obtain a character image which is obtained by dividing according to all character contours as shown in the figure 5;
s701, determining the width and height of the character cut by the uniform size of the license plate, and determining the width and height values of the license plate characters:
defining the width of the Chinese character as W and the height as H;
setting W <90pix, H <165 pix;
carrying out contour screening according to the set conditions of W and H to obtain each part of the Chinese character part shown in figure 5;
s702, obtaining coordinates (x, y) of the upper left corner of each outline of the Chinese character part in the binary image in S701; setting the width of each outline of the Chinese character part as w, the height as h, the leftmost boundary of the Chinese character as the minimum value of x, the rightmost boundary of the Chinese character as the maximum value of x + w, the upper boundary of the Chinese character as the minimum value of y, and the lower boundary of the Chinese character as the maximum value of y + h, so as to obtain the upper, lower, left and right boundary values of the Chinese character part.
S8, sequencing the contours obtained in the S7 according to the X-axis direction of the contour position coordinates to obtain partial contours of all Chinese characters and contours of numbers and letters;
and S9, cutting the Chinese character part according to the upper, lower, left and right boundary values of the Chinese character obtained in the S8 to obtain a Chinese character part picture as shown in the figure 6.
And S10, dividing the numbers and the alphabetic characters to obtain final images of all license plate characters and obtain the image shown in the figure 7.
The method for segmenting the license plate characters sets the uniform size of the license plate picture through normalization operation, and performs preliminary estimation on the positions of the Chinese characters according to the set thresholds of the width and the height of the Chinese characters, so that the outlines of non-Chinese character parts can be eliminated, and the segmentation precision is improved; by adopting single-channel picture processing, the influence of factors such as license plate color and light is small, and the precision of license plate recognition is greatly improved.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (4)
1. A method for segmenting license plate characters is characterized by comprising the following steps:
and S1, roughly positioning the license plate by using the trained license plate detection model to obtain a license plate picture.
And S2, inputting the license plate picture into another trained CNN regression model, performing primary forward propagation of CNN, and outputting coordinates of four vertexes of the license plate after continuous convolution and pooling operation.
And S3, performing angle correction through affine transformation by using the coordinates of the four vertexes of the license plate to obtain the front view of the license plate.
S4, further normalizing the license plate front view: adjusting the size of the product to be uniform;
s5, color conversion processing, namely, converting the color image subjected to the normalization processing in the S4 into a single-channel gray image;
s6, binarization processing, namely filtering the gray level image processed by the S5 by Gaussian filtering and carrying out binarization processing to obtain a binary image;
s7, contour detection, namely, carrying out outer contour detection on the binary image obtained in the S6 to obtain the outer contours of all characters and upper and lower boundary values of the outer contours;
s8, sorting the contours obtained in the S7 according to the X-axis direction of the contour position coordinates;
s9, cutting the Chinese character part according to the boundary values of the upper, lower, left and right sides of the Chinese characters obtained by sequencing in the S8 to obtain a picture of the Chinese character part;
and S10, dividing the numbers and the alphabetic characters to obtain final all license plate character pictures.
2. The method of claim 1, wherein the size of the unified space in S4 is 720pix × 224 pix.
3. The method for segmenting license plate characters according to claim 1, wherein the contour detection in the step S7 is outer contour detection.
4. The method for segmenting license plate characters according to claim 1, wherein the method for locating the boundary in S7 comprises the following steps:
s701, determining the width and height of the character cut by the uniform size of the license plate, and determining the width and height values of the license plate characters:
defining the width of the Chinese character as W and the height as H;
setting W <90pix, H <165 pix;
carrying out contour screening according to the set conditions of W and H to obtain each part of the Chinese character part;
s702, obtaining coordinates (x, y) of the upper left corner of each outline of the Chinese character part in the binary image in S701; setting the width of each outline of the Chinese character part as w, the height as h, the leftmost boundary of the Chinese character as the minimum value of x, the rightmost boundary of the Chinese character as the maximum value of x + w, the upper boundary of the Chinese character as the minimum value of y, and the lower boundary of the Chinese character as the maximum value of y + h, so as to obtain the upper, lower, left and right boundary values of the Chinese character part.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911411732.9A CN111310754A (en) | 2019-12-31 | 2019-12-31 | Method for segmenting license plate characters |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911411732.9A CN111310754A (en) | 2019-12-31 | 2019-12-31 | Method for segmenting license plate characters |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111310754A true CN111310754A (en) | 2020-06-19 |
Family
ID=71148212
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911411732.9A Pending CN111310754A (en) | 2019-12-31 | 2019-12-31 | Method for segmenting license plate characters |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111310754A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112017210A (en) * | 2020-07-14 | 2020-12-01 | 创泽智能机器人集团股份有限公司 | Target object tracking method and device |
CN113095320A (en) * | 2021-04-01 | 2021-07-09 | 湖南大学 | License plate recognition method and system and computing device |
CN114166132A (en) * | 2021-11-11 | 2022-03-11 | 中铁大桥科学研究院有限公司 | Vehicle height snapshot measuring method and device |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102402686A (en) * | 2011-12-07 | 2012-04-04 | 北京云星宇交通工程有限公司 | Method for dividing license plate characters based on connected domain analysis |
CN104298976A (en) * | 2014-10-16 | 2015-01-21 | 电子科技大学 | License plate detection method based on convolutional neural network |
CN105335743A (en) * | 2015-10-28 | 2016-02-17 | 重庆邮电大学 | Vehicle license plate recognition method |
CN109145900A (en) * | 2018-07-30 | 2019-01-04 | 中国科学技术大学苏州研究院 | A kind of licence plate recognition method based on deep learning |
CN109657676A (en) * | 2018-12-06 | 2019-04-19 | 河池学院 | Licence plate recognition method and system based on convolutional neural networks |
CN110619327A (en) * | 2018-06-20 | 2019-12-27 | 湖南省瞬渺通信技术有限公司 | Real-time license plate recognition method based on deep learning in complex scene |
-
2019
- 2019-12-31 CN CN201911411732.9A patent/CN111310754A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102402686A (en) * | 2011-12-07 | 2012-04-04 | 北京云星宇交通工程有限公司 | Method for dividing license plate characters based on connected domain analysis |
CN104298976A (en) * | 2014-10-16 | 2015-01-21 | 电子科技大学 | License plate detection method based on convolutional neural network |
CN105335743A (en) * | 2015-10-28 | 2016-02-17 | 重庆邮电大学 | Vehicle license plate recognition method |
CN110619327A (en) * | 2018-06-20 | 2019-12-27 | 湖南省瞬渺通信技术有限公司 | Real-time license plate recognition method based on deep learning in complex scene |
CN109145900A (en) * | 2018-07-30 | 2019-01-04 | 中国科学技术大学苏州研究院 | A kind of licence plate recognition method based on deep learning |
CN109657676A (en) * | 2018-12-06 | 2019-04-19 | 河池学院 | Licence plate recognition method and system based on convolutional neural networks |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112017210A (en) * | 2020-07-14 | 2020-12-01 | 创泽智能机器人集团股份有限公司 | Target object tracking method and device |
CN113095320A (en) * | 2021-04-01 | 2021-07-09 | 湖南大学 | License plate recognition method and system and computing device |
CN114166132A (en) * | 2021-11-11 | 2022-03-11 | 中铁大桥科学研究院有限公司 | Vehicle height snapshot measuring method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110866924B (en) | Line structured light center line extraction method and storage medium | |
CN108171104B (en) | Character detection method and device | |
CN111310760B (en) | Method for detecting alpha bone inscription characters by combining local priori features and depth convolution features | |
CN110647795B (en) | Form identification method | |
CN111310754A (en) | Method for segmenting license plate characters | |
CN112017200B (en) | Fish posture judgment method based on visual recognition | |
CN108133216B (en) | Nixie tube reading identification method capable of realizing decimal point reading based on machine vision | |
CN108846844B (en) | Sea surface target detection method based on sea antenna | |
CN111754538B (en) | Threshold segmentation method for USB surface defect detection | |
CN110674812B (en) | Civil license plate positioning and character segmentation method facing complex background | |
CN115331245B (en) | Table structure identification method based on image instance segmentation | |
CN115063421A (en) | Pole piece region detection method, system and device, medium and defect detection method | |
CN110738139A (en) | NIN license plate recognition method fusing Min-Max targets | |
CN108256518B (en) | Character area detection method and device | |
CN112686265A (en) | Hierarchic contour extraction-based pictograph segmentation method | |
CN113781413B (en) | Electrolytic capacitor positioning method based on Hough gradient method | |
CN110363196B (en) | Method for accurately recognizing characters of inclined text | |
CN112102189B (en) | Line structure light bar center line extraction method | |
CN109671084B (en) | Method for measuring shape of workpiece | |
CN107944393B (en) | Human face nose tip positioning method | |
CN113643290B (en) | Straw counting method and device based on image processing and storage medium | |
CN110276260B (en) | Commodity detection method based on depth camera | |
CN110348452B (en) | Image binarization processing method and system | |
CN111079756B (en) | Form extraction and reconstruction method and equipment in receipt image | |
CN112017201A (en) | Method for judging head and tail postures of fish body in processing and conveying |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200619 |
|
RJ01 | Rejection of invention patent application after publication |