Robust trajectory planning for target following vision-based autonomous land vehicle guidance

Ching Heng Ku, Wen-Hsiang Tsai

Research output: Contribution to journalArticlepeer-review

Abstract

We propose a robust trajectory planning method for vision-based autonomous land vehicle (ALV) guidance by target following using a visual field model. The proposed trajectory planning method aims to guide the ALV to achieve the goal that the targeted person will always appear in the image. It is found in this study that the goal can be achieved if three visual contact constraints in the visual field model are satisfied. The first constraint postulates that the target has appeared in the image. The second constraint asks that the vehicle head at the next position points straight towards to the target's current position. The third constraint sets bounds between the next position of the ALV and the current position of the target. A formula for the trajectory of the vehicle that satisfies the second visual contact constraint is derived. The steps for generating a speed and a turn angle of the ALV for realtime navigation are described as a trajectory-planning algorithm. Finally, the approach is tested on a real ALV and successful navigation sessions confirm the feasibility and the robustness of the approach.

Original languageEnglish
Pages (from-to)103-116
Number of pages14
JournalJournal of Navigation
Volume57
Issue number1
DOIs
StatePublished - 1 Jan 2004

Keywords

  • Land vehicle guidance
  • Trajectory-planning
  • Visual field model

Fingerprint Dive into the research topics of 'Robust trajectory planning for target following vision-based autonomous land vehicle guidance'. Together they form a unique fingerprint.

Cite this