首页> 外国专利> Method and apparatus for integrating object detection information detected by respective object detectors corresponding to respective cameras located in the surroundings for collaborative driving using an application capable of V2X communication and sensor fusion by a plurality of vehicles.

Method and apparatus for integrating object detection information detected by respective object detectors corresponding to respective cameras located in the surroundings for collaborative driving using an application capable of V2X communication and sensor fusion by a plurality of vehicles.

机译:使用用于能够通过多个车辆进行V2X通信和传感器融合的应用来集成由与位于周围环境中的各个相机相对应的各个物体检测器检测到的物体检测信息以进行协作驾驶的方法和装置。

摘要

PROBLEM TO BE SOLVED: To provide an object detection method and device capable of acquiring accurate object detection information for a surrounding environment. SOLUTION: Object detection information corresponding to each of the cameras located in the periphery is integrated by deep learning-based image analysis and detected by an object detector that detects an object in each image generated from each camera. When the first to n-th object detection information are obtained from the first object detector to the n-th object detector in descending order of detection reliability, the integrated device increases k from 3 to n. Meanwhile, the (k-2)th object and the kth object are integrated by a matching operation to obtain the (k-1)th object integrated information, and the (k-1)th object integrated information is reprojected to an image. [Selection diagram] Figure 2
机译:解决的问题:提供一种能够获取周围环境的准确的物体检测信息的物体检测方法和装置。解决方案:通过基于深度学习的图像分析集成与位于外围的每个摄像头相对应的对象检测信息,并由对象检测器检测该对象检测器,该对象检测器检测从每个摄像头生成的每个图像中的对象。当以检测可靠性的降序从第一对象检测器到第n对象检测器获得第一至第n对象检测信息时,集成装置将k从3增加至n。同时,通过匹配操作将第(k-2)个对象和第k个对象进行积分以获得第(k-1)个对象积分信息,并且将第(k-1)个对象积分信息重新投影到图像上。 [选择图]图2

著录项

相似文献

  • 专利
  • 外文文献
获取专利

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