A kind of license plate location based on mathematical morphology and edge detection

Author(s):  
Hu Hongping ◽  
Bai Yanping
Author(s):  
Weifang Zhai ◽  
Terry Gao ◽  
Juan Feng

The license plate recognition technology is an important part of the construction of an intelligent traffic management system. This paper mainly researches the image preprocessing, license plate location, and character segmentation in the license plate recognition system. In the preprocessing part of the image, the edge detection method based on convolutional neural network (CNN) is used for edge detection. In the design of the license plate location, this paper proposes a location method based on a combination of mathematical morphology and statistical jump points. First, the license plate area is initially located using mathematical morphology-related operations and then the location of the license plate is accurately located using statistical jump points. Finally, the plate with tilt is corrected. In the process of character segmentation, the border and delimiter are first removed, then the character vertical projection method and the character boundary are used to segment the character for actually using cases.


2013 ◽  
Vol 671-674 ◽  
pp. 2851-2854 ◽  
Author(s):  
Hong Ping Hu ◽  
Yan Ping Bai

The proposed method of the license plate location is based on mathematical morphology, color feature and edge detection. It consists of two parts on raw location of license plate and refined location of license plate. In the raw location part, resize the image into the same size, consider the relation of R, G, B components in the image to deal with the noise, detect the edge and use the operations of mathematical morphology repeatedly to obtain the license plate candidate, obtain the raw location by the up-down bound and left-right bound of the license plate. In the refined location part, for the obtained raw location of the license plate, use Radon transformation, edge detection, use the operations of mathematical morphology, find the up-down bound and left-right bound so that the refined location of the license plate is obtained. It is shown that the proposed method is suit for the Chinese license plate with white character and blue background.


Sign in / Sign up

Export Citation Format

Share Document