The Application of Colour Filtering to Real-Time Person Tracking

We present results from multiple experiments with colour filtering methods in order to improve robustness in an integrated surveillance system to track people in a subway station. The system is designed to operate in real-time in a distributed local network of off-the-shelf computers, resulting in practical constraints not found in developmental systems. We show that the quality of colour information is degraded by electrical interference and image compression to such an extent that it is no longer useful for local edge detection. We give a recommendation as to what methods can be used to filter out most of the image noise influencing local edge detection and show how using these methods increases robustness of tracking. A previous version of this article appeared in the Proceedings of the 2nd European Workshop on Advanced Video-Based Surveillance Systems (AVBS’2001), pages 227–234, September 2001. This work is funded by the European Union, grants ADVISOR (IST-1999-11287) and Smart II (FMRXCT96-0052). Introduction With the recent advances of computer technology real-time automated visual surveillance has become a popular area for research and development. Surveillance cameras are installed in many public areas to improve safety, and computer-based image processing is a promising means to handle the image databases generated by large networks of cameras. The task of an integrated surveillance system is to warn an operator when it detects events which may require human intervention, for example to avoid possible accidents. The warnings can only be reliable if the system can detect and understand human behaviour and for this it must locate and track people reliably. A frequent problem with surveillance systems is the poor quality of the video images. Installed cameras are often old, and images are transmitted down long analogue cables which are vulnerable to interference and signal degradation. There are many unsolved problems when it comes to processing this noisy data by means of a computer. Many methods to make image analysis more