Following vehicle detection based on shift of feature plane using affine transform
Shintaro Arai, Osamu Inoue, Shinji Ozawa · 2010
This paper proposes a novel vehicle detection system based on shift of a feature plane, which consisted of feature points on a front surface of vehicle, using in-vehicle camera for following vehicles' observation. We find that the feature plane of the vehicle's front surface shifts in accordance with the affine transform. We perform experiments and evaluate performances of our system. As results, we confirm that our system can locate and track the rear and side vehicles accurately and robustly.