1. Introduction
Finding an office or a room in a large public building, such as a university, a city hall or a shopping centre,
is often difficult for normal persons. For blind persons it is far more difficult because they must completely rely
on people passing by to ask for information. When a building is becoming more familiar, they can memorise all
locations for moving about.
There exist systems which try to alleviate the problems of blind and visually impaired persons. For the stateof-
the-art of existing systems we refer to [1-5]. Navigation is one of the fundamental problems to be solved in
creating completely autonomous mobility systems, mainly for blind persons or mobile robots. Navigation
models can be classified into local and global ones. Local models navigate by exploring information in the
environment immediately around the actual position. They are completely blind to everything beyond the reach
of the system’s sensors. In contrast, global navigation takes into account the environment as a whole, far
beyond the reach of the system’s sensors. They enable auto-localisation and optimisation of navigation routes.
There are several systems which deal with simultaneous localisation and mapping (SLAM). Morioka et al.
[6] proposed a model that allows navigation in crowded environments such as shopping centres and transport
1. IntroductionFinding an office or a room in a large public building, such as a university, a city hall or a shopping centre,is often difficult for normal persons. For blind persons it is far more difficult because they must completely relyon people passing by to ask for information. When a building is becoming more familiar, they can memorise alllocations for moving about.There exist systems which try to alleviate the problems of blind and visually impaired persons. For the stateof-the-art of existing systems we refer to [1-5]. Navigation is one of the fundamental problems to be solved increating completely autonomous mobility systems, mainly for blind persons or mobile robots. Navigationmodels can be classified into local and global ones. Local models navigate by exploring information in theenvironment immediately around the actual position. They are completely blind to everything beyond the reachof the system’s sensors. In contrast, global navigation takes into account the environment as a whole, farbeyond the reach of the system’s sensors. They enable auto-localisation and optimisation of navigation routes.There are several systems which deal with simultaneous localisation and mapping (SLAM). Morioka et al.[6] proposed a model that allows navigation in crowded environments such as shopping centres and transport
การแปล กรุณารอสักครู่..
