This paper shows how to use the automated driving toolbox in MATLAB by changing the lane geometry to modify the curvature of the road and to change the number of lanes present inside the road. The lanes can be solid, dashed, and colored yellow or white. Solid yellow means it is a lane for opposing traffic, and solid white means same-way traffic. In addition, dashed lane lines can be used to overtake other vehicles. In addition, the actors can be simulated to show traversing the road and also showing sensor data, in our case, from radar and the way it can detect obstacles in the path of the vehicle, for example, a pedestrian, and accordingly brake the vehicle. It also shows how to do actor classification using micro-Doppler signatures and Convolutional Neural Network and viewing their spectrograms.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Use of Radar Waveform in Actor Detection and Classification in Self-Driving Cars

  • Ananya Dutta,
  • Aradhana Misra,
  • Surajit Deka,
  • Ridip Tukaria,
  • Kandarpa Kumar Sarma

摘要

This paper shows how to use the automated driving toolbox in MATLAB by changing the lane geometry to modify the curvature of the road and to change the number of lanes present inside the road. The lanes can be solid, dashed, and colored yellow or white. Solid yellow means it is a lane for opposing traffic, and solid white means same-way traffic. In addition, dashed lane lines can be used to overtake other vehicles. In addition, the actors can be simulated to show traversing the road and also showing sensor data, in our case, from radar and the way it can detect obstacles in the path of the vehicle, for example, a pedestrian, and accordingly brake the vehicle. It also shows how to do actor classification using micro-Doppler signatures and Convolutional Neural Network and viewing their spectrograms.