Image sensing system for navigating visually impaired people

About 285 million people are visually impaired around the world. In the United States, this number is over 1.3 million, which represents 0.4% of the population. However, the cost is about 4 billion dollars a year to build infrastructures and provide proper health care to meet the needs of these people. This paper proposes a system that would help the visually impaired adapt to the environment rather than trying to adapt the environment to them. The development of this system will enable the visually impaired to maintain efficient daily activities in many different environments. This system uses multiple sensors and analyzes the information acquired in real-time. In this paper, we particularly emphasize image processing to provide effective obstacle avoidance, object recognition and 3D reconstruction of the environment.