【24h】

Motion and Haar-like Features Based Vehicle Detection

机译:基于运动和类似Haar的特征的车辆检测

获取原文
获取原文并翻译 | 示例

摘要

Lienhart et al. have added basic and rotated haar-like features in the face detection scheme based on a boosted cascade of simple feature classifiers. There are two key contributions in the paper. The first is introduction of motion features. With the haar-like and motion features, our sample vehicle detector shows off on average a 5% lower false alarm rate at a given hit rate. Considering the speed performance, motion features are used to given the candidate regions and haar-like features are used to delect the vehicles at the above results. Secondly, the haar-like features are used for the highway except face. In the face detection, haar-like feature shows detection rate comparable to the best previous system. But little work is done in the traffic domain. And empirical analysis of vehicle detection is provided in the experiment section.
机译:Lienhart等。已基于简单特征分类器的增强级联,在人脸检测方案中添加了基本和旋转的类似哈尔的特征。本文有两个主要贡献。首先是运动功能的介绍。凭借类似哈尔(haar)和运动的功能,我们的样本车辆检测器在给定的命中率下平均可将误报率降低5%。考虑到速度性能,在上述结果中,运动特征用于指定候选区域,类哈尔特征用于偏转车辆。其次,除面部外,类似haar的特征用于高速公路。在人脸检测中,类似“哈尔”的功能显示出的检测率可与以前的最佳系统媲美。但是在流量域中几乎没有完成任何工作。实验部分提供了车辆检测的实证分析。

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