首页> 外文会议>Asian Control Conference >A study on detection method of vehicle based on lane detection for a driver assistance system using a camera on highway
【24h】

A study on detection method of vehicle based on lane detection for a driver assistance system using a camera on highway

机译:基于车道检测的基于摄像头的驾驶员辅助系统车辆检测方法研究

获取原文

摘要

In this work, we introduce an approach to detect information about lanes and vehicles for driver assistance, or warning system. Most previous research works could only detect the lanes or vehicles separately. However, the combination of lane information and vehicle information is able to support the lane keeping system, or driver assistance system, and to improve the reliability results. For driver assistance system, it must detect multi-lanes (left, right side lane, and middle lane), and discover vehicles ahead of the test vehicle on these lanes. Therefore, in this study, we use a camera, it is installed on the windscreen of the test vehicle. Images from the camera are used to detect three lanes, and detect multi-vehicles. In lane detection, line detection and inverse perspective mapping are used. For the vehicle detection, we combine the horizontal and vertical edge detection, the horizontal edge is used to detect the vehicle candidates, and then the vertical edge detection is used to verify the vehicle candidates. The proposed algorithm works in an average of 42 milliseconds for each frame with resolution on a 3.30 GHz Intel CPU. The system was tested on the highway in Korea.
机译:在这项工作中,我们引入一种方法来检测有关车道和车辆的信息,以提供驾驶员帮助或预警系统。以前的大多数研究工作只能单独检测车道或车辆。然而,车道信息和车辆信息的组合能够支持车道保持系统或驾驶员辅助系统,并改善可靠性结果。对于驾驶员辅助系统,它必须检测多车道(左侧,右侧车道和中间车道),并在这些车道上发现测试车辆之前的车辆。因此,在本研究中,我们使用照相机,将其安装在测试车辆的挡风玻璃上。来自摄像机的图像用于检测三个车道,并检测多辆车。在车道检测中,使用线检测和反透视映射。对于车辆检测,我们将水平和垂直边缘检测结合在一起,使用水平边缘检测候选车辆,然后使用垂直边缘检测来验证候选车辆。所提出的算法每帧平均工作42毫秒,在3.30 GHz Intel CPU上的分辨率。该系统已在韩国的高速公路上进行了测试。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号