본문 바로가기
  • Home

Fusion Methods of License Plate Detection and Super Resolution for Improving License Plate Recognition

  • Journal of The Korea Society of Computer and Information
  • Abbr : JKSCI
  • 2011, 16(4), pp.55-62
  • Publisher : The Korean Society Of Computer And Information
  • Research Area : Engineering > Computer Science

송태엽 1 이영현 1 Min-Jae Kim 1 Bonhwa Ku 1 Hanseok Ko 1

1고려대학교

Accredited

ABSTRACT

This paper proposes fusion methods of license plate detection and super-resolution for improving license plate recognition in low-resolution images. In the proposed method, we apply the license plate detection based on local structure pattern feature and the sequential super-resolution based on Kalman filter. The proposed fusion methods are divided into two according to whether the license plate is detected or not in the input image : (i) performing license plate detection after restoring whole image through super resolution, and (ii) restoring only the detected region through super-resolution after detecting the license plate. We demonstrated effectiveness of the proposed methods in various environments.

Citation status

* References for papers published after 2023 are currently being built.