An enhanced behaviour recognition system for remote supervision of vulnerable individuals

The use of technology to remotely supervise vulnerable individuals is wide-ranging. The type of remote supervision that inspired the research in this paper is position monitoring (also referred to as tracking); this is a technique commonly proposed for the remote supervision of children. Current, commercially-available systems depend upon the explicit configuration of parameters such as a coordinate-based range of permissible motion; this is inflexible and takes no account of other information that could affect one´s interpretation of the results. This paper presents an approach that can incorporate data from multiple sources, one of which is a position sensor. The use of fuzzy logic builds flexibility into the system. The output of the system is mapped to a graduated scale that can be presented in various ways, depending on the needs of the audience.