Real time crime prediction
What is it about?
The paper presents an approach to crime prediction based on video analysis, neuro-fuzzy inference and density mapping. It is our believe crime has indicator event that can be modeled and used predict new events based on similarity of events patterns. To do this, we first have to define our crime indicator events, model the indicator events using pattern recognition methods and then make crime event predictions in new videos collected in real time from a wide area surveillance network.
Why is it important?
The advantages of the approach include: the ability to predict crime in real time due to the use of video based events, the ability to generate fuzzy rules from data, the ability to optimize fuzzy rule-base by learning and the ability of weighting different crime variables. The framework has prospects for developing a police field decision support system, policing performance monitoring or policing tactics evaluation system. The logic is that since the approach allows you to visualize crime indicator events (either increasing or decreasing) in a thematic display, it can be used to support informed policing decisions in the field or monitor performance crime counter measures in real time