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.
Featured Image
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
Perspectives
The experiments were mainly focused on depicting regions of high crime imminence and identifying crimes types with average imminence levels greater than a set threshold. But in real world deployment, further explorations can be done to determine the behavior of crime concentrated regions (hotspots), which could be acute or chronic. Determining the behavior and underlying factors in crime imminence regions could be of great importance because it informs the kind of policing intervention that should be taken. For example, there can be situations where visible police presence can rather provoke or trigger crime than prevent it. Such situations may involve gangsters and psychopaths who may simply take pleasure in committing crimes just to defy the presence of law enforcement officers. So depending on the dynamics and the underlying factors police may decide to go undercover or not.
Nurudeen Mohammed
Central South University
Read the Original
This page is a summary of: Crime prediction and mapping based on real time video analysis, Journal of Ambient Intelligence and Smart Environments, March 2018, IOS Press,
DOI: 10.3233/ais-180476.
You can read the full text:
Contributors
The following have contributed to this page
