Model-based guidance by the longest common subsequence algorithm for indoor autonomous vehicle navigation using computer vision

Abstract

The use of the longest common subsequence algorithm is proposed for model-based guidance of autonomous vehicles by computer vision in indoor environments. A map of the corridor contour of a building, describing the navigation environment and measured before navigation sessions, is used as the model for guidance. Two cameras mounted on the vehicle are used as the vision sensors. The wall baselines in the images taken from the two cameras are extracted and constitute the input pattern. The environment model and the input pattern are represented in terms of line segments in a two-dimensional floor-plane world. By encoding the line segments into one-dimensional strings, the best matching between the environment model and the input pattern is just the longest common subsequence of the strings. So no complicated comparison is needed and robust matching results can be obtained. The actual position and orientation of the vehicle are determined accordingly and used for guiding the navigation of the vehicle. Successful navigation sessions on an experimental vehicle are performed and confirm the effectiveness of the proposed approach.