A vision system for detecting paths and moving obstacles for the blind

In this paper we present a monocular vision system for a navigation aid. The system assists blind persons in following paths and sidewalks, and it alerts the user to moving obstacles which may be on collision course. Path borders and the vanishing point are de- tected by edges and an adapted Hough transform. Opti- cal flow is detected by using a hierarchical, multi-scale tree structure with annotated keypoints. The tree struc- ture also allows to segregate moving objects, indicating where on the path the objects are. Moreover, the centre of the object relative to the vanishing point indicates whether an object is approaching or not. I. INTRODUCTION Navigation of blind people is very arduous because they must use the white cane for obstacle detection while following the front sides of houses and shops, meanwhile memorising all locations they are becom- ing familiar with. In a new, unfamiliar setting they completely depend on people passing by to ask for a certain shop or the closest post office. Crossing a street is a challenge, after which they may be again disoriented. In a society in which very sophisticated technology is available, from tracking GPS-RFID equipped containers in an area of hundreds of meters to GPS-GIS car navigation to Bluetooth emitting the sound of movie trailers to mobile phones in front of cinemas, one can question what it may cost to provide the blind with the most elementary technology to make life a little bit easier. This technology may not replace the cane, but should complement it: alert the user to obstacles a few metres away and provide guid- ance for going to a specific location in town or in a shopping centre. Different approaches exist to help the visually im- paired. One system for obstacle avoidance is based on a hemispherical ultrasound sensor array (1). It can detect obstacles in front and unimpeded directions are obtained via range values at consecutive times. The system comprises an ARM9 embedded system, the sensor array, an orientation tracker and a set of pager motors. Talking Points is an urban orientation system (5) based on electronic tags with spoken (voice) mes- sages. These tags can be attached to many landmarks like entrances of buildings, elevators, but also bus stops and busses. A push-button on a hand-held device is used to activate a tag, after which the spoken mes- sage is made audible by the device's small loud- speaker. iSONIC (10) is a travel aid complementing the cane. It detects obstacles at head-height and alerts by vibration or sound to dangerous situations, with an algorithm to reduce confusing and unnecessary detec- tions. iSONIC can also give information about object