A vision-based system supports mapping services for visually impaired people in indoor environments

This paper describes and extensively evaluates a visual-based system that autonomously operators for both building a map and localization tasks. The proposed system is to assist mapping services to the visually impaired/blind people in small or mid-scale environments such as inside a building or campus of school, hospital. Toward this end, the proposed approaches solely rely on visual data thanks to a self-designed image acquisition system. On one hand, a robust visual odometry method is utilized to create a map of the environments. On the other hand, the proposed approaches utilize FAB-MAP algorithm that is maybe the most successful for learning places in the environments. Map building and learning places in an environment are processed in an off-line phase. Through a matching place procedure, online captured images are continuously positioned on the map. Furthermore, we utilize a Kalman Filter that combines the matching results of current observation and the estimation of robot states based on its kinematic model. We evaluate performances of the proposed system through experimental schemes. The results show that the constructed map coincides with ground truth, and matching image-to-map is high confidence. The evaluations also contain scenarios which the blind pupils move following Robot. The experimental results confirmed that proposed system feasibly navigating blind pupils in indoor environments.