Multi-stage with neuro-fuzzy approach for efficient on-road speed sign detection and recognition

Hsin Han Chiang, Yen Lin Chen*, Tsu Tian Lee

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

6 Citations (Scopus)


This paper presents an automatic speed sign recognition system to provide speed limit awareness functions for driver assistance. To reduce the influence of digital noise caused by ambient lighting conditions and sign pollutions, an efficient segmentation method based on pan-red color information was applied to extracting the shape of a speed sign. Based on the edge gradient information of a circular shape, a radial symmetry detection strategy is proposed for fast detection of the speed sign candidates from various urban road scenes. The recognition of the information of a speed sign is achieved through the proposed fuzzy adaptive-kernel-based learning vector quantization (FAKLVQ) approach, which also verifies each candidate to eliminate non-target blobs. Experiments demonstrated the feasibility and effectiveness of the proposed system under a wide variety of outdoor conditions.

Original languageEnglish
Pages (from-to)2919-2939
Number of pages21
JournalInternational Journal of Innovative Computing, Information and Control
Issue number7
Publication statusPublished - 2013
Externally publishedYes


  • Color segmentation
  • Fuzzy
  • Learning vector quantization (LVQ)
  • Radial symmetry detection
  • Speed sign recognition

ASJC Scopus subject areas

  • Software
  • Theoretical Computer Science
  • Information Systems
  • Computational Theory and Mathematics


Dive into the research topics of 'Multi-stage with neuro-fuzzy approach for efficient on-road speed sign detection and recognition'. Together they form a unique fingerprint.

Cite this