Context-based driver support system development: Methodology and case study

摘要

In modem world count of vehicle in the roads grown every year that causes increasing traffic accidents. In this case a system that can recognize a dangerous situation caused by a driver drowsiness and / or distraction and can help avoid part of these accidents. ADAS systems that use different sensors and integrated to vehicles are popular for luxury segment and, therefore, less accessible for end mass consumer. This paper aims at methodology development for vehicle incidents alerting based on mobile measurements of a driver behavior in the cabin using the personal smartphone mounted in the vehicle windshield. Methodology includes the reference model for the driver support system, dangerous situation recognition method, and context- based models aimed at recommendation generation for the driver based on recognized dangerous situation and current situation around the vehicle.