Extracting characters from real vehicle licence plates out-of-doors

Bing-Fei Wu*, S. P. Lin, C. C. Chiu

*Corresponding author for this work

Research output: Contribution to journalReview articlepeer-review

43 Scopus citations


Most applications of automatic vehicle identification systems are outdoors and are consumer-orientated. The recognition rate, the system reliability and the processing speed are very important. A system for extracting characters from licence plates has been presented. The system can successfully extract characters from licence plates in any part of a captured image without the need to consider the luminance of the surroundings or the size and inclination of the licence plate or the colour of the vehicle. Moreover, the processing time of this system is satisfactory for several applications. The proposed system depends only on the completeness of a character and so is tolerant of skewed plates and complex backgrounds. In a real parking lot, an extraction rate of 95.6% was obtained by applying the system to 228 vehicles.

Original languageAmerican English
Pages (from-to)2-10
Number of pages9
JournalIET Computer Vision
Issue number1
StatePublished - 9 Aug 2007


Dive into the research topics of 'Extracting characters from real vehicle licence plates out-of-doors'. Together they form a unique fingerprint.

Cite this