Pedestrian Detection for Anti-Tampering Vehicle Protection

摘要

Vehicle survivability in the form of anti-tampering tools is an important part of the FCS community of vehicles. Autonomous or semi-autonomous vehicles traversing through unknown terrain require the ability to detect, predict, and avoid potential aggression from hostile forces. Expanding upon pedestrian detection techniques will provide a viable solution in the near term. With the current vision based pedestrian detection algorithms, an anti- tampering suite can be developed. First, the system must determine people in the scene. Once this is accomplished, the people can be checked for weapons and their movements can be tracked. Finally deciphering the tracked movements can help distinguish between friendly and unfriendly actions by people. This paper will discuss the need for an anti-tampering suite and how pedestrian detection techniques can be used to address this need.