Almost 30% of accidents are due to carelessness to forward vehicle. Over the recent years, manufacturers have tried to develop Advanced Driving Assistance Systems (ADAS) to reduce this rate. Such systems may work based on different sensors in which camera is one of the most desired one due to its low cost and size.

But there are several challenges with vision-based ADAS in terms of efficiency and robustness to different lighting and weather conditions.  We, at Rahbin Sanat Nasir Company, have developed a monocular forward collision warning (FCW) system that works real-time on a low-end embedded processor. The robustness of the system can be seen in some urban videos.