Smartphone-like device that maps surroundings to aid blind

A smartphone-like gadget that senses an entire room's features, builds a virtual map of it and communicates this to the user may one day replace the humble white cane to help blind sense their surroundings.

Using special multi-sensor array technology, the Indoor Navigation Project will enable blind to sense their surroundings beyond the cane's tip, researchers said.

Project leader Dr Iain Murray from the Curtin University said the gadget would resemble a smartphone and would sense an entire room's features, build a virtual map of it and communicate this to the user.

"A cane is beneficial for going up and down stairs or detecting if obstacles are right in front of you, but is not capable of telling you if something is more than a metre or two away," Murray said.

"What we are developing is a multi-sensor device for people who are blind, who are also often hearing impaired, to tell them what is exactly around them from wall to wall," said Murray said researchers will take on the development of one type of sensor, whether it be sensing the change of velocity, images or noise.

They will pool the sensors into one gadget, which will ultimately contribute towards building a map of the indoor environment including both moving and non-moving objects.

"While many indoor locations already have a map that people who are blind can use to find their way around, they don't allow for change and can therefore be quite dangerous," he said.

The sensors will determine the direction and distance a user walks whilst indoors along with identifying features such as stairs and inclines.

Such devices will extract the edges of paths and obstacles using stereoscopic cameras, classify what they are, and build a map of the environment using image processing techniques.

They can pick up audio cues to help locate and track moving objects from a mobile receiver.

They would allow for efficient methods of communicating map and sensor data across networks, to ensure data is available before a user collides with an obstacle (in less than 1/6 of a second), researchers said.

The sensors will enable efficient methods of security and trust within networks, to eliminate risk of leading the blind into dangerous situations.

Once obstacles are identified, a map will be constructed using a Building Information Modelling system suitable for the vision impaired, developed in the sixth research project.