単眼車載カメラ画像を利用した自動車運転支援

概要

論文の詳細を見る
In recent years, an in-vehicle camera is spreading as a drive-recorder. We expect the development of the driving support system for the car, which combines an invehicle camera with image processing technology. In this paper, we described a few key algorithms for driving support system. Mainly, we proposed three methods as follows. ① Detection of lane lines using Hough transform, and generation of search area for car ahead based on lane lines. ② Detection of car ahead using symmetry of car shape. ③ Measurement of the distance between cars using the object which size is known beforehand like a number plate. We clarified experimentally the usefulness of the proposed methods. These methods can be expected as an additional function to an in-vehicle monocular camera.
広島工業大学の論文