Inclined license plate correction and variable-length license plate recognition method based on deep learning

A tilt correction and deep learning technology, applied in the field of license plate image recognition, which can solve the problems of limited recognition accuracy of license plate numbers and license plate deformation.

Active Publication Date: 2019-11-08
ZHEJIANG UNIV
View PDF12 Cites 15 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, problems such as license plate size, shooting angle, license plate deformation, and different lengths of license plate numbers

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
  • Inclined license plate correction and variable-length license plate recognition method based on deep learning
  • Inclined license plate correction and variable-length license plate recognition method based on deep learning
  • Inclined license plate correction and variable-length license plate recognition method based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment

[0084] The present invention adopts such as Figure 4 and as Figure 5 The license plate area picture and the license plate number area picture shown are used as training license plate tilt correction network and variable length license plate number recognition network, wherein the license plate area picture not only includes the complete license plate area, but also includes part of the background area, and the license plate area picture set includes not only the following The picture of the license plate area with a lower slope in Figure 6(a) also includes the picture of the license plate area that is severely deformed due to the shooting angle as shown in Figure 7(a); the picture of the license plate number area only contains the license plate number and does not contain any background Part, the sample image set of the license plate number area contains images of the license plate number area with variable lengths.

[0085] In this embodiment, 14051 license plate area pict...

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 an inclined license plate correction and variable-length license plate recognition method based on deep learning. The method comprises two steps of training and prediction, andthe training stage comprises the following steps: constructing a license plate inclination correction network, and carrying out training by using license plate region sample pictures and normalized coordinates of corresponding angular points; constructing an uncertain-length license plate number recognition network, and performing training by using the license plate number region sample picture and the corresponding license plate number; the prediction stage comprises the following steps: sending a to-be-identified license plate picture to a license plate inclination correction network to predict license plate corner information and perform inclination correction; positioning by using a target detection method to obtain a license plate number area in the license plate area after inclination correction; and sending the license plate number area to an indefinite-length license plate number identification network and carrying out post-processing to obtain a license plate number. According to the invention, the inclined license plate can be corrected, and license plate number identification can be directly carried out from the whole license plate number area; and the method has extremely high robustness in a complex real scene.

Description

technical field [0001] The invention relates to the field of license plate image recognition, in particular to a method for rectifying inclined license plates and recognizing indefinite length license plates based on deep learning. Background technique [0002] The application of license plate recognition is an important application of image processing and computer vision in the field of intelligent transportation. It plays an important role in vehicle monitoring, assisting traffic law enforcement and tracing traffic accidents. However, problems such as license plate size, shooting angle, license plate deformation, and different lengths of license plate numbers are still important factors that limit the current license plate recognition technology to achieve extremely high recognition accuracy. Contents of the invention [0003] In order to solve the problems in the background technology, the present invention provides a method for rectifying inclined license plates and id...

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
IPC IPC(8): G06K9/32G06N3/08
CPCG06N3/08G06V30/1478G06V20/63G06V20/625G06V30/10
Inventor 张三元祁忠琪涂凯吴书楷
Owner ZHEJIANG UNIV
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