A real-time rangefinding system that can be used as a safe driving support system on expressways has been developed using a real-time stereo vision method that works stably while taking into account both environmental factors, such as changes in the amount of sunlight, the presence of road signs, and shadows on the road, and shock and vibration to cameras mounted on the vehicle. The method was implemented and evaluated with the high-speed video imaging processing system ISHTAR (Image Sequencing Hardware for Temporal Analysis in Real Time). The experimental results indicate that the system is capable of recognizing various types of vehicles and determining ranges in real time, whatever the variations in environmental factors, traffic conditions, and vehicle vibration are.
Abstract