作者: Wasiq Khan , Abir Hussain , Bilal Khan , Raheel Nawaz , Thar Baker
关键词: Process (computing) 、 Data stream mining 、 Feature extraction 、 Computer science 、 Reliability (computer networking) 、 Auditory display 、 Wearable computer 、 Obstacle 、 Human–computer interaction 、 Spatial analysis
摘要: Outdoor mobility of Visually Impaired People (VIPs) has always been challenging due to the dynamically varying scenes and environmental states. Variety systems have introduced assist VIPs’ that include sensor mounted canes use machine intelligence. However, these are not reliable when used navigate VIPs in changing environments. The associated challenges robust sensing avoiding diverse types obstacles, modelling states (e.g. moving objects, road-works), effective communication interpret hazards. In this paper, we propose an intelligent wearable auditory display framework will process real-time video multi-sensor data streams to: a) identify type b) recognize surrounding scene/objects corresponding attributes geometry, size, shape, distance from user), c) automatically generate descriptive information about recognized obstacle/objects attributes, d) produce accurate, precise spatial instructions audio-visual form safely with or without assistance traditional means.