Vision system tracks car thieves

Vision system tracks car thieves
Melanie Reynolds An artificial intelligence system that can identify the suspicious movements of car thieves is undergoing a six month trial in the UK. The integrated traffic and pedestrian model-based vision system combines a vehicle tracking system from Reading University with one that tracks humans from Leeds University. The system, comprising a Unix workstation and a domestic video camera, works by building a model from observing people’s behaviour in a car park over several hours. The speed of approach and the distance of a person from a car is measured and added to others to build up a probability distribution model. This is then compared with the behaviour of an individual entering the car park and assessed to see if it matches. “The behaviour may be abnormal in a completely uninteresting way. But the supposition is that anything that doesn’t conform with normality may be of interest,” said Professor David Hogg of Leeds University, who is involved in the project. Preliminary results of the trial have proved encouraging, claims Hogg: “However, there is still more to be done.”


Leave a Reply

Your email address will not be published. Required fields are marked *

*