3-9 An Approach to Vehicle Recognition using Supervised Learning

To enhance safety and traffic efficiency, a driver assistance system and an autonomous vehicle system are being developed. A preceding vehicle recognition method is important to develop such systems. In this paper, the vision-based preceding vehicle recognition method, based on supervised learning from sample images is proposed. And the improvement for Modified Quadratic Discriminant Function (MQDF) classifier that is used in the proposed method is also shown. In the case of the road environment recognition including the preceding vehicle recognition, there are many reports, in which evaluations are done with a few images, but a quantitative evaluation with large number of images has rarely been done. We prepare over 1,000 sample images for passenger vehicles, which are recorded on a highway at daytime, and evaluate the proposed method with those images. The evaluation result shows that the performance in a low order case is improved from the ordinary MQDF. The feasibility of the proposed method is proved, due to the result that the proposed method indicates over 98% as classification rate.