Object detection and sensory feedback techniques in building smart cane for the visually impaired

Abstract

People with vision impairment often encounter difficulties in situations that normal sighted people take for granted. Crossing roads and streets in busy environments with high traffic density is a major safety challenge. Investigation of proper technologies such as a smart cane that can provide navigational assistance to people with vision impairment has been an ongoing research topic of great interest. This paper provides a review of the current technology of smart canes and associated research topics. Specifically, the paper identifies a number of technical and application-oriented research challenges in designing smart cane systems. This includes the limitations of sensors, signal processing techniques, single sensors versus multiple sensors, adaptability standards, as well as pragmatic issues such as cost, size and usability. The paper then presents a survey of recent research focusing on various methods of object detection and sensory feedback systems, as well as the integration of other technology components that may be implemented in a smart cane system. Furthermore, the paper describes a summary of reported smart cane prototypes and categorizes them into various types such as how the object detection system is embedded on or separate from the cane, whether the system is able to detect static or dynamic objects, and what type of feedback is implemented into the system. In addition to defining common technical and application challenges involved in smart cane technology and developing an overview for both research and prototypes, the paper also provides an insight regarding the importance and need for safer and more effective object detection technology for the visually impaired community.