Vision-based side vehicle detection from a moving vehicle

Abstract

In this paper, a side vehicle detection system with real-time on-board vision is proposed. The system employs camera vision to detect side moving vehicles, and provides necessary warning to drivers and passengers when the vehicle is making lane change. Based on SURF, feature detection and comparison are applied to find the same features between two consecutive images and calculate the feature vectors. Further image analysis is performed based on the feature vectors to determine whether there are side moving vehicles. If any side moving vehicle is detected, the system will apply inverse perspective projection technique to estimate the positions of surrounding vehicles in Cartesian space. Then, the system will generate on-screen signs of safety, warning, or danger to inform drivers and passengers based on the estimated distance to side moving vehicles. The approach has been successfully validated in real traffic environments by performing experiments with two CCD cameras mounted on top of side mirrors of a roadway moving vehicle.