License plate character segmentation method

A character segmentation, character technology, applied in character and pattern recognition, instruments, computer parts and other directions, can solve the problem of reducing the accuracy of license plate positioning, cutting and other problems

Inactive Publication Date: 2012-06-27
SICHUAN JIUZHOU ELECTRIC GROUP
View PDF0 Cites 11 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The technical problem to be solved by the present invention is: in order to overcome the problem that the horizontal projection method also has a local minimum value and the radicals are cut off as the background, a new method for analyzing and segmenting the characters of the license plate is provided. The principle of a connected domain, combined with the fixed height and fixed ratio relationship of the license plate characters, better solves the character segmentation problem of the license plate under complex background conditions, and reduces the requirements for the accuracy of license plate positioning

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • License plate character segmentation method
  • License plate character segmentation method
  • License plate character segmentation method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0034] Embodiment 1, if the process is as figure 1 Described, a kind of license plate character segmentation method, comprises steps:

[0035] Step S1: Determine the upper, lower, left, and right frame areas of the license plate, and filter out the frame area image of the license plate image, and perform step S2;

[0036] Step S2: quickly cluster the license plate image by scanning row by pixel, initially segment each character in each character class, and execute step S3;

[0037] Step S3: compare the height of each character after clustering, and remove characters whose height is less than Height / 2 as noise;

[0038]Step S4: Determine whether the number of character classes (the number of all characters) is less than 7 or the width of a certain character is greater than the width of other characters and exceeds the threshold A (the width of Chinese characters on a license plate is generally 60mm), if the number of character classes is less than 7 or a certain character is...

Embodiment 2

[0043] Embodiment two: on the basis of embodiment one, the specific steps of determining the upper, lower, left, and right borders of the license plate in step S2 are

[0044] Step S11: Since the license plate has 7 characters, it can be divided into 8 segments from the left border to the right border, and the license plate image is scanned row by row with the threshold L1=Width / 8. If there is a license plate image whose length is greater than L1 then is the upper and lower borders of the license plate, where Width is the pixel width of the license plate.

[0045] Step S12: Since the license plate characters generally occupy 1 / 2 of the height in the height direction of the license plate, the license plate image is scanned column by column with the threshold L2=Height / 4. If the length of a certain column of the license plate image is greater than L2, it is the left and right borders of the license plate , where Height is the pixel height of the license plate.

[0046] Step S13...

Embodiment 3

[0047] Embodiment 3, on the basis of Embodiment 1, step S2 scans the license plate image row by pixel to quickly cluster the specific steps as follows:

[0048] Step S21: When the distance d between two pixels is less than or equal to d_const, it can be considered that the two pixels belong to the same character, where d_const is the distance d_const between the two pixel coordinates A(x1, y1), B(x2, y2) (A,B)=|x1-x2|+|y1-y2|;

[0049] Step S22: Generally, Width / 7 in front of the license plate image is the Chinese character part (the license plate image is from left to right, and Width / 7 is the Chinese character part from the left), which can be obtained from the d_const calculation formula in step S21. According to the characteristics of Chinese characters, binarization The distance between pixels that are not 0 in the same Chinese character after that may be greater than 1, so take d_const=2 or 3 clustering, at this time, the Chinese characters are segmented, and the Chinese...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention belongs to the field of license plate character image processing, in particular to a license plate character segmentation method after the license plate is accurately positioned by an intelligent transportation system. Aiming to solve the problem that the character radical is considered to be the background and is cut due to the local minimum caused by a horizontal projection method, the invention provides a novel method for analyzing and segmenting the license plate characters; and according to the principle that elements belonging to the same character form a connected domain, and combined with the fixed proportion relation of the fixed height and space of the license characters, the segmentation problem of the characters of the automobile license plate under complex condition is better solved, thereby lowering the requirement for accurately positioning the license plate.

Description

technical field [0001] The invention relates to the field of license plate character image processing, in particular to a license plate character segmentation method after the license plate is accurately positioned by an intelligent transportation system. Background technique [0002] The method that is often used at present is the horizontal projection method: that is, along the horizontal direction, calculate the number of primes belonging to the license plate characters in each column. In this way, the projection of the character block in the horizontal direction will obtain a local minimum value at the gap between the characters, so the correct segmentation position of the character should be near the above local minimum value. Since the gap between the 2nd and 3rd characters is the largest, it is generally to find the end position of the 2nd character, and then find out the starting positions of 1, 3, 4, 5, 6, and 7 in sequence. However, in actual processing, due to th...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/34
Inventor 王德鹏范高生
Owner SICHUAN JIUZHOU ELECTRIC GROUP
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap