A Fast and Robust License Plate Location Method in Surveillance Scenarios

A highly robust, license plate location technology, applied in the field of pattern recognition and computer vision, which can solve the problems of reduced image clarity and contrast, the camera is easily covered with a layer of dust, and the robustness of license plate detection is not enough.

Active Publication Date: 2019-12-31
FUZHOU UNIV
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] 2) License plate detection is not robust enough
The license plate image acquisition system works in a variety of complex environments, affected by weather, light changes, smudges, shooting positions and angles, etc., and it is easy to cover the lens of the lens due to the long-term lack of regular maintenance of the camera Dust, which leads to a decrease in image clarity and contrast, affects the accuracy of license plate detection

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 Fast and Robust License Plate Location Method in Surveillance Scenarios

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0052] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0053] This embodiment provides a highly robust fast license plate location method in a monitoring scene, such as figure 1 As shown, in view of the problem that license plate recognition based on high-resolution video poses a huge challenge to the real-time performance of the system, we propose a motion detection method using Epanechikov kernel and frame adaptation. The method first uses the background difference method based on the Epanechikov kernel to detect whether there is motion from the video stream, and then adopts frame-adaptive moving object filtering according to the motion situation, thereby speeding up the speed of license plate recognition. This method uses the size of the license plate to dynamically estimate the size of the vehicle, filters out the connected areas that do not conform to the rules, and reduces the detection time of the li...

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 present invention relates to a fast license plate location method with high robustness in a monitoring scene, comprising the following steps: Step S1: performing semantic grouping on license plate training images, that is, dividing the license plate images into two semantic groups of license plate and non-license plate, and grouping Input the training image into the support vector machine for training; step S2: adopt the adaptive fast vehicle detection method based on Epanechikov kernel density estimation to quickly detect the vehicle area; step S3: use the frame sliding algorithm based on AFRD to filter the video frame, Obtain the frame image of interest; step S4: adopt the method combining morphology, color and MSER to detect the license plate in the vehicle area in the frame image of interest; step S5: apply the training in step S1 to the candidate license plate obtained in step S4 A good SVM does binary classification and gets the correct license plate. The positioning method of the invention can improve the robustness of the license plate recognition system in harsh environments such as low contrast and low light.

Description

technical field [0001] The invention relates to the fields of pattern recognition and computer vision, in particular to a highly robust and rapid license plate location method in a monitoring scene. Background technique [0002] With the acceleration of the construction of Ping An China, the distribution of monitoring cameras is becoming wider and wider, and the resolution of the cameras is getting higher and higher. Compared with using standard bayonet vehicle image acquisition devices, direct use of these monitoring environments all over the streets and alleys will bring more benefits. Here comes some new challenges. Traditional license plate recognition systems are widely used in road tolls, parking lots, etc., but most application scenarios require the support of some special hardware devices, such as ground sensing coils, which can only recognize one vehicle at a time, and the recognition speed is slow. Moreover, most of the image transmissions use analog signals, the ...

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 & AuthorityPatents(China)
IPC IPC(8): G06K9/00G06K9/32G06K9/62
CPCG06V20/52G06V20/63G06V20/625G06F18/2411
Inventor柯逍张毓峰陈羽中
OwnerFUZHOU UNIV