DOI QR코드

DOI QR Code

Text Area Detection of Road Sign Images based on IRBP Method

도로표지 영상에서 IRBP 기반의 문자 영역 추출

  • Chong, Kyusoo (SOC Research Institute, Korea Institute of Civil Engineering and Building Technology)
  • 정규수 (한국건설기술연구원 SOC성능연구소)
  • Received : 2014.10.16
  • Accepted : 2014.11.19
  • Published : 2014.12.31

Abstract

Recently, a study is conducting to image collection and auto detection of attribute information using mobile mapping system. The road sign attribute information detection is difficult because of various size and placement, interference of other facilities like trees. In this study, a text detection method that does not rely on a Korean character template is required to successfully detect the target text when a variety of differently sized texts are present near the target texts. To overcome this, the method of incremental right-to-left blob projection (IRBP) was suggested as a solution; the potential and improvement of the method was also assessed. To assess the performance improvement of the IRBP that was developed, the IRBP method was compared to the existing method that uses Korean templates through the 60 videos of street signs that were used. It was verified that text detection can be improved with the IRBP method.

최근 Mobile Mapping System을 활용한 영상의 수집과 도로표지 속성정보의 자동 인식을 위한 연구가 진행되고 있다. 도로표지는 판의 규격, 글씨크기 및 배치가 다양하고 가로수 등 타 시설물의 간섭으로 인해 일정한 패턴을 찾아 정보를 추출하기 어렵다. 본 연구에서는 다양한 크기의 한글 문자가 있거나, 한글문자 주변에 심벌이 위치한 도로표지에 대해서도 국문지명을 성공적으로 검출하기 위해서는, 한글문자 템플릿에 의존하지 않는 새로운 국문지명 검출 방법이 필요하다. 그 새로운 한 방법으로서, 점진적 좌측방향으로의 블럽 투사(incremental right-to-left blob projection, IRBP)를 제시하고, 그 가능성과 개선 정도를 평가하였다. 성능 평가하기 위하여, 60개의 도로표지 영상 데이터로 기존의 한글 템플릿을 사용하는 경우와 비교하여 성능을 평가하였다. 전반적으로, IRBP 방법으로 국문지명 검출 성능을 개선할 수 있음을 확인하였다.

Keywords

References

  1. Kyusoo Chong, "Gradation Image Processing for Text Recognition in Road Sign Using Image Division and Merging", Korea Society of ITS, vol. 13, no. 2, pp.27-33, 2014.
  2. Lloyd A. Flectcher and Rangacher Kasturi, "A Robust Algorithm for Text String Separation from Mixed Text/Graphics Images", IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 10, no. 6, pp.910-918. 1988 https://doi.org/10.1109/34.9112
  3. Kyun Koh, Jungshin Park, Sukjung Yoon, Sungwoo Hong, Mansuk Song, Yillbyung Lee, "Implementation of a Korean Text Recognition System", Journal of KIISE, vol. 14, no. 2, pp.167-169, 1987.
  4. In Dong Lee, Oh Seok Kwon, Tae Kyun Kim, "A Method to Extract Characters and Non-characters Seperatery from Document Image", Journal of KIISE, vol. 17, no. 3, pp.247-258, 1990.
  5. Gyo-Hyoun Gwon, Jong-Cheon Park, Byeong-Min Jeon, "HanGeul Text Region Detection Using Edge-Based Method in Mobile Images", Journal of KOEN, vol. 8, no. 1, 2011.
  6. Gyo-Hyoun Gwon, Jong-Cheon Park, Byeong-Min Jeon, "Text Region Detection Using Morphological Reconstruction Method in Book-cover Images", Journal of KEIA, vol. 5, no. 2, 2011
  7. Do-Hyung Kim, Seon-Hwa Lee, Mi-Sook Kim, Eui-Young Cha, "A Study on Character Extraction in Vehicle Number Plate and Character Recognition", Journal of KIISE, vol. 27, no. 2, 2000.
  8. Hi-Seok Kim, "Design of an Licence Plate Pre-processing Localization Algorithm using FPGA", Journal of KIISE, vol. 12, no. 2, 2014. https://doi.org/10.14801/kiitr.2014.12.2.23
  9. Sheen Lew, ByeongJun Jeong, Hyunchul Kang, "Recognition of License Plates Using a Hybrid Statistical Feature Model and Neural Networks", Journal of KIISE, vol. 36, no. 12, 2009.
  10. Sang Ok Koo, Soon Ki Jung, " Intermediate Font Generation based on Shape Analysis of Hangul Glyph", Journal of KIISE, vol. 36, no. 4, pp.311-325, 2009.

Cited by

  1. 위치기반 유사도 검증을 이용한 도로표지 안내지명 자동인식 개선방안 연구 vol.18, pp.6, 2014, https://doi.org/10.12815/kits.2019.18.6.241