A license plate super-resolution processing method and system based on deep learning

A technology of super-resolution and processing methods, which is applied in neural learning methods, image data processing, image analysis, etc., and can solve problems such as high-compression images and less video research

Active Publication Date: 2020-08-11
ZHEJIANG UNIV
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the existing super-resolution technology based on deep learning is known to consider the super-resolution of ideal images and videos, and there are few studies on high-compression images and videos.

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
  • A license plate super-resolution processing method and system based on deep learning
  • A license plate super-resolution processing method and system based on deep learning
  • A license plate super-resolution processing method and system based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0052] In order to make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be described clearly and completely in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of the embodiments of the present invention, not all the embodiments.

[0053] First, a brief introduction to the license plate super-resolution method provided by the embodiment of the present invention: first obtain n frames of images containing license plate information from the original compressed surveillance video {J i }, then for each frame of image J i Perform target detection, obtain the target area of ​​each frame of image, and then select interest points in each target area to obtain the interest point trajectory, then select n interest points located on the interest point trajectory, and take the i...

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 super-resolution processing method and system based on deep learning, which includes obtaining a series of images containing license plate information from the original compressed surveillance video; then tracking the objects of interest; selecting several points of interest and Capture images and register these images; use the deep learning training library to obtain the corresponding deep network weights, and use the deep network weights to perform super-resolution processing on the registered multi-frame images to obtain a clear high-resolution license plate I. The present invention uses a super-resolution deep network model and a gradient-guided network model to jointly optimize the strategy to learn the internal relationship between degraded characters and clear characters in high-compression surveillance video, and realize super-resolution and deblocking effect through a common deep network The effect is to improve the resolution of the degraded characters and remove the block effect, which solves the problem that the license plate characters cannot be seen clearly in the high-compression surveillance video.

Description

Technical field [0001] The invention relates to a digital video image processing method, in particular to a method and system for processing license plate super-resolution based on deep learning. Background technique [0002] In recent years, my country has entered a period of rapid social development, and the domestic security situation is severe. In order to maintain social stability and protect the safety of people's lives and properties, the state has invested 300 billion yuan in implementing safe city video surveillance projects in many cities across the country, and established a relatively complete video surveillance system. The surveillance camera captures and records various scenes that have occurred in the surveillance area in the past and present. On the one hand, it leaves clues for various future criminal investigations and evidence for civil disputes; on the other hand, it deterred various criminals and prevented various criminal activities. The license plate is t...

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 Patents(China)
IPC IPC(8): G06N3/04G06N3/08G06K9/20G06T5/20G06T7/269G06T7/30H04N19/42H04N19/44
CPCH04N19/42H04N19/44G06N3/08G06T5/20G06T7/269G06T7/30G06V10/23G06V2201/07G06N3/045Y02T10/40
Inventor 王维东黄露
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