Path Planning and Following Algorithms in an Indoor Navigation Model for Visually Impaired

Abstract

This paper describes, path planning and following algorithms for use in indoor navigation for the blind and visually impaired. Providing indoor navigational assistance for this type of users presents additional challenges not faced by conventional guidance systems, due to the personal nature of the interactions. The algorithms are part of an overall Indoor Navigation Model that is used to provide assistance and guidance in unfamiliar indoor environments. Path planning uses the A* and Dijkstra's shortest path algorithms, to operate on an "Intelligent Map", that is based on a new data structure termed "cactus tree" which is predicated on the relationships between the different objects that represent an indoor environment. The paths produced are termed "virtual hand rails", which can be used to dynamically plan a path for a user within a region. The path following algorithm is based on dead reckoning, but incorporates human factors as well as information about the flooring and furnishing structures along the intended planned path. Experimental and simulating results show that the guiding/navigation problem becomes a divergent mathematical problem if the positional information offered by the positioning and tracking systems does not reach a certain requirement. This research explores the potential to design an application for the visually impaired even when to- date 'positioning and tracking' system cannot offer reliable position information that highly required by this type of application.