Vehicle License Plate Detection using Local Fractal Dimension and Morphological Analysis

Vehicle number plate detection is widely used for detecting speeding cars and in electronic toll collection. This paper makes significant contribution by proposing the use of text detector. It was assumed that all plates contain text. The text can solely be considered as a special texture pattern for analysis. To extract text information from the picture, an adaptive threshold, fractal filtering and morphological analysis has been used. Furthermore, the paper presents a novel technique using multi-resolution fractal filtering. This expelled all non-textual patterns and made analysis versatile. The results have showed the possibility of using fractal dimension to highlight text regions and morphological analysis to filter non-plate regions.