Blind signal extraction algorithm for the license plate matching of vehicle positioning system

In this paper, we use ICALAB tool with our algorithm to extract the desired image of license plate captured by sensors, which will be essential used by vehicle positioning system (VPS). The use of the blind least mean square algorithm (BLMS) can often provide a useful method to extract the sources and independent component. The BLMS learning rule has a self-normalizing property, such that the unit length of the weight vector w/sub 1/ is approximately maintained. Simulation results show that the proposed BLMS significantly extract the license plate number when the vehicle is driving from Lane 1 to Lane 2 with moving blur.