Abstract :
Previously, D.W. Tindall (1995) discussed some of the challenges associated with automatic licence plate recognition (ALPR) and technical solutions based on digital signal processing, including the use of neural network recognition algorithms. Since that time, Racal has installed and commissioned systems worth over £1M and has gained considerable practical experience. Contracts from overseas are now also being awarded. In nearly every instance, the customer requires recognition of licence plates from countries other than the native format, and this places further challenges upon the DSP algorithms and the available processing power. Racal has continued to develop the ALPR product family and associated subsystems, and the paper presents some of the experiences gained so far and how we believe ALPR systems will develop in the next few years. The first part of the paper concentrates on the variability of licence plates in a multinational environment and solutions to some of these challenges. The second part discusses the development of practical systems which offer customers an effective and automatic means for monitoring traffic in both controlled and open road scenarios
