Vehicle License Plate Detection and Recognition Based on Contour Extraction in Various Environments
Sung-Kook Pyo1, Sang-Hun Lee2, Gang-Seong Lee3, Young-Soo Park4
1Sung-Kook Pyo, Department of Plasma biodisplay, Kwang Woon University, Korea, East Asian.
2Sang-Hun Lee, Department of Ingenium College of Liberal Arts, Kwangwoon University, Korea, East Asian.
3Gang-Seong Lee, Department of Ingenium College of Liberal Arts, Kwangwoon University, Korea, East Asian.
4Young-Soo Park, Department of Ingenium College of Liberal Arts, Kwangwoon University, Korea, East Asian.
Manuscript received on 08 June 2019 | Revised Manuscript received on 14 June 2019 | Manuscript Published on 22 June 2019 | PP: 77-83 | Volume-8 Issue-8S2 June 2019 | Retrieval Number: H10150688S219/19©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open-access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In this paper, we have tried to detect the area of the license plate in the environment of the vehicle and recognize the characters in the detected license plate area. We propose a method to detect the license plate based on contour extraction that adapts to the surrounding environment and a method of recognizing characters by template matching in detected license plate area. The proposed method is divided into the detection of license plate area and the character recognition in the license plate area. DoG(Difference of Gaussian) and Morphology operation were used to emphasize the character part outline, and the license plate was detected by determining the aspect ratio of the characters. And the character recognition process was performed with the detected license plate area. Through noise remove and normalization, characters were segmented by vertical histogram. The template was matched with the divided characters to recognize the characters. In this study, we used 130 different vehicle image data such as vehicle license plates, which are inclined in front of the vehicle, and license plates with changes in the environment around the vehicle. In the detection plate area, the character recognition rate was 96% in the case of the slanted plate, 93% in the various background environment, and 97% in the plate image of the front face.
Keywords: License Plate Detection, Character Recognition, Normalization, Vertical Histogram, Template Matching.
Scope of the Article: Software Engineering Tools and Environments