Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

License plate region extraction method and system based on vertical gradient, terminal and readable storage medium

A license plate area and vertical gradient technology, applied in the field of image processing, to achieve the effect of dense edge distribution, accurate and reliable elimination, and strong consistency of visual features

Pending Publication Date: 2021-08-13
世邦通信股份有限公司
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The difficulty of license plate recognition in this scenario far exceeds that in restricted scenarios, and the license plate recognition algorithm in restricted scenarios is no longer applicable

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 region extraction method and system based on vertical gradient, terminal and readable storage medium
  • License plate region extraction method and system based on vertical gradient, terminal and readable storage medium
  • License plate region extraction method and system based on vertical gradient, terminal and readable storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0070] Embodiments of the present invention provide a license plate area extraction method based on vertical gradient, including the following steps:

[0071] Step 1: Obtain the image I to be identified, and convert the image I to be identified into a grayscale image GI and calculate the vertical gradient to obtain a gradient image.

[0072] In this embodiment, the image 1 to be identified is not limited to the source of image i to be identified as the image i to be identified as the image i to be identified. As 1 shown in 1 is a high-definition capture picture of 2160 * 4096, including a license plate from far more than a distance, including new energy license plates. The Sobel Operator Gx is used to convolute the grayscale image GI in the horizontal direction, obtain the gradient image Gx represented by the vertical gradient of the image I, as follows:

[0073]

[0074] Symbol "*" represents the convolution. As shown in 2 as a gradient image, where the vertical edge is very ob...

Embodiment 2

[0111] In the manner of the description of the first embodiment, the vertical gradient is calculated to obtain a gradient image after the gradient image and in step 2, based on the gradient image search gradient concentration region as a candidate area of ​​the license plate, execute: excluding the gradient The vertical edge in the image is specifically as follows:

[0112] Step 1: The gradient image is divided into zone; where the dimension of the grayscale image Gi is assumed to be M × N, that is, the height is m, the width is n. The Gi is uniformly divided into WDC × HTC non-overlapping area. The specific division value is an empirical value based on the characteristics of the capture image. Taking the image of 2160 × 4096 as an example, the image can be divided into 8 x 8 non-overlapping regions, and the dimension of each region is 270 × 512.

[0113] STEP 2: The vertical gradient of each region exceeds the pixel point and its number gcount, which calculates the gradient mean ...

Embodiment 3

[0118] In the manner of the description of the first embodiment, the vertical gradient is calculated to obtain a gradient image after the gradient image and in step 2, based on the gradient image search gradient concentration region as a candidate area of ​​the license plate, execute: excluding the gradient The image is too long and the excessive vertical edge, the specific:

[0119] S1: The continuous edge is calibrated based on the gradient image. Among them, a continuous edge is calibrated according to the gradient image GX, and the standard can be calibrated using an 8 neighborhood.

[0120] S2: If the length of the continuous edge is greater than the excessive edge point threshold Tlong or less than the threshold Tshort of the excess short edge, all the pixels on the continuous edge do not belong to the license plate area, which belongs to the excessive vertical edge or over Short vertical edges and eliminate excessive vertical edges or excessive vertical edges;

[0121] Amon...

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 discloses a license plate region extraction method and system based on a vertical gradient, a terminal and a readable storage medium, and the method comprises the steps: 1, obtaining a to-be-recognized image I, converting the to-be-recognized image into a grayscale image gI, and calculating the vertical gradient to obtain a gradient image; 2, searching for a gradient concentration area as a candidate area of a license plate based on the gradient image, and combining overlapped areas based on an overlapped range; and 3, selecting the candidate areas after combination in the step 2 to extract a license plate area. According to the method, random vertical edges and too long and too short vertical edges are eliminated by further utilizing the vertical gradient, so that the accuracy and reliability of license plate recognition are improved, and the method can be suitable for license plate recognition in a limited scene.

Description

Technical field [0001] The present invention belongs to the field of image processing, and more particularly to a license plate area extraction method, system, terminal, and readable storage medium based on vertical gradient. Background technique [0002] At present, with the high-definition camera's popularization, the capture range of the bayonet electronic police system can be covered to the distance to the installation projection position, and it can be covered to the distance installation projection position of about 15m. Unlike similar parking charges, this capture scene is generally called non-restricted scenes. [0003] Under the non-restricted scene, the background environment, illumination conditions, etc. are affected by image shooting. The collected license plate images include not only a variety of complex texture interference, such as flowers, billboards, isolation railings, etc., also have license plate area position, size , The direction and brightness change is l...

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): G06K9/32G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06V20/62G06V20/625G06N3/047G06N3/045G06F18/241
Inventor 孙志强刘波霍亚
Owner 世邦通信股份有限公司
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
Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com