Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

End-to-end license plate recognition method and device based on lightweight network

A license plate recognition and network-side technology, which is applied in the field of end-to-end license plate recognition methods and devices based on lightweight networks, can solve the problems of low efficiency and accuracy of license plate number recognition

Active Publication Date: 2021-01-12
BEIJING ITARGE TECH CO LTD
View PDF3 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

As a result, the efficiency and accuracy of license plate number recognition are low

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
  • End-to-end license plate recognition method and device based on lightweight network
  • End-to-end license plate recognition method and device based on lightweight network
  • End-to-end license plate recognition method and device based on lightweight network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0054] figure 1 A schematic flowchart of an embodiment of the light-weight network-based end-to-end license plate recognition method of the present invention, figure 2 It is a schematic diagram of the organizational structure of a lightweight license plate recognition network model according to an embodiment of the present invention.

[0055] like Figure 1 to Figure 2 As shown, this embodiment provides an end-to-end license plate recognition method based on a lightweight network, the lightweight network is a lightweight license plate recognition network model, and the lightweight license plate recognition network model includes a feature extraction network, a differentiated character classification network and license plate number encoding / decoding module; the method may include:

[0056] Step 101, acquiring a license plate image.

[0057] The acquisition of the license plate image may be captured by a surveillance camera located at an intersection, a bayonet, a parking l...

Embodiment 2

[0092] see figure 2 and Figure 7 As shown, the device of this embodiment can include: a lightweight license plate recognition network model, including a feature extraction network, a differentiated character classification network, and a license plate number compilation / decoding module; the functions realized by the license plate number compilation / decoding module can be programmed The method is solidified in the differentiated classification network.

[0093] Wherein, the feature extraction network is used to input the license plate image into the feature extraction network, perform shallow-level processing on the license plate image, and extract the shallow-level texture features of the license plate image;

[0094] performing deep-level processing on the shallow-level texture features, and extracting deep-level features of the license plate image;

[0095] Obtaining advanced semantic features based on the deep-level features to obtain multiple feature maps; the size spe...

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 embodiment of the invention discloses an end-to-end license plate recognition method and device based on a lightweight network, and the method and device can improve the license plate recognitionefficiency and precision to a certain degree. The method comprises the steps of inputting a license plate image into a feature extraction network, performing shallow processing on the license plate image, and extracting shallow texture features of the license plate image; performing deep processing on the shallow texture features, and extracting deep features of the license plate image; obtainingadvanced semantic features based on the deep features to obtain a plurality of feature maps; inputting the feature maps into a differentiated character classification network, classifying the corresponding feature maps based on different classifiers, and outputting a classification prediction result, wherein the differentiated character classification network is provided with a plurality of classifiers, and each classifier corresponds to a corresponding position character in the feature map; inputting the classification prediction result into a license plate number encoding / decoding module fordecoding to obtain a license plate number, the license plate number encoding / decoding module stores a corresponding relationship between the classification prediction result and the character.

Description

technical field [0001] The invention relates to the field of intelligent transportation, in particular to an end-to-end license plate recognition method and device based on a lightweight network. Background technique [0002] With the development of 5G and the Internet of Things era, more and more image acquisition terminals are deployed at traffic intersections, parking lots, gas stations, etc., and more and more scenarios require intelligent monitoring of the status of motor vehicles. Among them, the license plate is the only ID card of a motor vehicle, and the accuracy and timeliness of license plate number recognition is an important factor reflecting the performance of intelligent monitoring equipment. [0003] In some existing license plate recognition solutions, the monitoring image captures each vehicle in real time, determines the position of the vehicle in the picture and cuts it out, detects the position of the license plate based on the obtained vehicle image and...

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): G06K9/34G06K9/46G06K9/62
CPCG06V30/153G06V10/462G06V20/625G06F18/24Y02T10/40
Inventor 杨云飞隋立林
Owner BEIJING ITARGE TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products