Recognition of Car License Plates Using Morphological Information and SOM Algorithm
Kwang-Baek Kim*, and Dae-Su Kim**
*Department of Computer Engineering, Silla University, 1-1 San, Gwaebop-dong, Sasang-gu, Busan 617-736, Korea
**Department of Computer Science, Hanshin University, Korea
In this paper, we propose car license plate recognition using morphological information and a self-organizing map (SOM) algorithm. Morphological information on horizontal and vertical edges was used to extract the license plate from a car image. A 4-directional contour tracking algorithm was applied to extract the specific area, including characters, from an extracted plate. Recognition of extracted character strings was studied using the SOM algorithm. We used 50 car images to evaluate performance. Extraction for character strings by the proposed method showed better results than that from conventional color information on RGB and HSI. License plate recognition using the SOM algorithm was very efficient.