Police Eyes: Real world automated detection of traffic violations

Dangerous lane changing, illegal overtaking, and driving in the wrong lane account for a high percentage of the total accidents that occur on the road, second only to accidents due to over-speeding. Automated traffic applications typically encompass the detection and segmentation of moving vehicles as a crucial process. Background subtraction and shadow detection are amongst the most challenging tasks involved in the segmentation of foreground blobs in dynamic environments. An effective balance between accuracy and speed is required to process a continuous feed of high resolution images from multiple cameras. Police Eyes is a mobile, real-time traffic surveillance system we have developed to enable automatic detection of traffic violations. Police Eyes would be useful to police for enforcing traffic laws and would also increase compliance with traffic laws even in the absence of police. The system detects illegal crossings of solid lines using image processing and efficient computer vision techniques on image sequences acquired from IP cameras. The automatic solid line crossing detection system can be used at locations where the traffic violations are notoriously high and are known to create traffic congestion and avoidable accidents. The system can be installed on an embankment, at an intersection area, at a lane change restriction area, at a no parking area or anywhere there is an observed pattern of drivers intentionally violating traffic laws. We have installed the system in an industrial grade embedded PC and deployed it in a police mannequin. Results of an empirical field evaluation show that the system performs well in a variety of real-world traffic scenes.