Discontinuous Character Recognition Method in Variable Message Signs for Intelligent Transportation Systems
Wahyono Wahyono, Han-Sung Park, Kang-Hyun Jo · International Journal of Machine Learning and Computing · 2013
Variable Message Signs (VMSs) are a key element of Intelligent Transportation Systems.The VMSs generally displays dynamic text information that each character consists of matrix of a light-emitting diodes lamp, named LED character.Compared with general printed text, the LED character is quite difficult to be recognized due to its discontinuity.Hence, this paper addresses a novel method for recognizing the LED character.Firstly, color segmentation is performed in order to distinguish foreground (i.e.text) and background.The supporting points (SP) represented the center of a segment are then acquired.For generating character region, these points are grouped based on their properties.Lastly, feature extraction is applied by dividing component into several blocks and counting the number of SP in the block.Afterwards, character template matching approach is implemented for recognizing component as certain alphanumeric.Experimental results evince that the proposed method is a robust to recognize the LED text.