首页> 外文期刊>マツダ技報 >Detection algorithm for vehicles with multi-line CCD sensors
【24h】

Detection algorithm for vehicles with multi-line CCD sensors

机译:带有多线CCD传感器的车辆的检测算法

获取原文
获取原文并翻译 | 示例
           

摘要

We have been developing "side and rear warning system" and "rear-collision reduction system with rear collision prediction" to meet the increasing demands for the active safety of vehicles (Fig. 1). In realizing these systems, it is crucial to adequately detect vehicles around while driving, in addition to the establishment of appropriate interface to drivers such as warning tuning. In detecting vehicles around, white lines are often falsely detected as vehicles when there are up and down slopes on roads. Although the white lines on roads can be identified by carefully evaluating the distribution of points on a vertical plane (Fig. 9), it was difficult to get a close distance distribution with ordinary obstacle sensors. We employ multi-brie CCD sensors that make it possible to easily acquire detailed two-dimensional distance distribution and evaluate the subtle differences in data distribution on vertical plane. We have developed a detection technology to identify and eliminate the white lines effectively by defining the relative inclination between vehicle and road based on the two-dimensional distance distribution from the sensors. This detection technology enables us to detect vehicles around even with the changing pitch attitude of vehicles. We will describe the outline of multi-line CCD sensor, the details of the detection system and the results of its evaluation in applying to the detection of vehicles behind in an adjacent lane.
机译:我们一直在开发“侧面和后方预警系统”和“具有后部碰撞预测功能的后减撞系统”,以满足对车辆主动安全性日益增长的需求(图1)。在实现这些系统时,除了为驾驶员建立适当的接口(例如警告调整)外,在行驶时充分检测周围的车辆至关重要。在检测周围的车辆时,当道路上有上下坡度时,白线常常被错误地检测为车辆。尽管可以通过仔细评估垂直平面上的点分布来识别道路上的白线(图9),但使用普通的障碍物传感器很难获得近距离分布。我们采用多束CCD传感器,可以轻松获取详细的二维距离分布并评估垂直面上数据分布的细微差异。我们已经开发出一种检测技术,通过基于传感器的二维距离分布来定义车辆和道路之间的相对倾斜度,从而有效地识别和消除白线。这种检测技术使我们即使在车辆俯仰角变化的情况下也可以检测周围的车辆。我们将描述多线CCD传感器的轮廓,检测系统的详细信息以及其在检测相邻车道后面的车辆时的评估结果。

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