A single processor implementation of video character recognition for the identification of aircraft and road vehicles

There has recently been an increased interest in the use of digital processing for automatic recognition of characters from video. The requirements for video-rate character recognition place severe demands on processing power. This has led many to recommend dedicated hardware solutions for real-time operation. This paper reports on a single general-purpose processor implementation of automatic video character recognition algorithms. After outlining the basic principles of the processing method, the results of application to civil aircraft recognition at London Heathrow Airport and automatic vehicle recognition are summarised. The advantages of a single processor solution translate to significant hardware savings, offering cost benefits, simplified software structure and ease of code portability, between different processors. The importance of efficient data pre-selection is underlined.