ROAD RECOGNITION METHOD FOR ANTICOLLISION RADAR SYSTEM

This paper describes a method to substantially reduce false and missing alarms of a collision avoidance system (CAS) based on a range sensor (radar or lidar). The method applies road recognition techniques on data coming from a scanning seensor. The core problems to be solved by any CAS are the ability and reliability to discriminate the potential dangerous obstacles from others. An object has to be considered dangerous if it lies or will lie on the foreseen trajectory of the vehicle and if its distance is too short with respect to safety distance, related to the speeds of the vehicle and of the obstacle. In a previous work, it was demonstrated that trajectory forecast relying on the measure of vehicle yaw velocity alone is not accurate enough to guarantee safety in all highway situations. This is of particular relevance to avoid alarms on stationary obstacles lying out of the trajectory, which can cause dangerous emergency braking. To improve trajectory forecast, a road recognition method has been developed and applied on maps acquired by a scanning radar sensor. The method is derived from computer vision techniques and tailored to particular features of radar maps. The paper describes the method employed for road recognition and knowledge representation in terms of geometrical and temporal continuity constraints. The algorithm, operating in real-time on a demonstrator vehicle, has been extensively tested on real traffic scenarios. Results of processing and improvement in performances are reported.

• Supplemental Notes:
  • Five volumes of papers and one volume of abstracts comprise the published set of conference materials.
• Corporate Authors:

  VERTIS

  TORANOMOM 34 MORI BUILDING 1-25-5
  TORANOMON, MINATOKU, TOKYO 105  Japan 
• Authors:
  • Rebora, C
  • Saroldi, Andrea
  • Anerdi, G
  • Re Fiorentin, S
• Conference:
  • Steps Forward. Intelligent Transport Systems World Congress
  • Location: Yokohama, Japan
  • Date: 1995-11-9 to 1995-11-11
• Publication Date: 1995-11

Language

• English

Media Info

• Pagination: p. 1057

Subject/Index Terms

• TRT Terms: Algorithms; Crash avoidance systems; Intelligent transportation systems; Motor vehicles; Radar; Real time control; Sensors; Speed; Traffic speed; Trajectory; Vehicle trajectories
• Geographic Terms: Italy
• Old TRIS Terms: Anticollision systems; Real-time systems; Road recognition
• Subject Areas: Highways; Operations and Traffic Management; Vehicles and Equipment; I91: Vehicle Design and Safety;

Filing Info

• Accession Number: 00722038
• Record Type: Publication
• Report/Paper Numbers: Volume 3
• Files: TRIS
• Created Date: Jun 24 1996 12:00AM