Long Range LiDAR Characterisation for Obstacle Detection for use by the Visually Impaired and Blind

Obstacle detection and avoidance is a huge area of interest for autonomous vehicles and, as such, has become an important research topic. Detecting and identifying obstacles enables navigation through an ever changing environment. This work looks at the technology used in self-driving vehicles and examines whether the same technology could be used to aid in navigation for visually impaired and blind (VIB) people. For autonomous vehicles, obstacle detection relies on different sensor modalities to provide information on the vehicles surroundings. A combination of the same sensors placed on a white cane could be used to perform free-space assessment over the whole height of the user and provide additional environmental information not available from the cane alone. This provides its own challenges and advantages. The speeds are much slower when dealing with pedestrians and scanning can be achieved by the movement of the cane. However, the weight and size must be significantly reduced. The full system will be integrated into a smart cane and will consist of four main sensors as well as range sensors. The aim of this work is to report on the characterisation of a long range LiDAR (up to 10m) that will be integrated into a smart white cane developed as part of the INSPEX H2020 project.