A Navigation System based on Event Camera for People Living with Blindness and Vision Impairment

Neuromorphic sensing has drawn significant interest in automated sensing for its low latency, low power consumption, and high throughput signal processing. Leveraging these advantages, in this work, we propose a proof-of-concept navigation system for people with vision impairment based on an event camera with physical and optical filtering that greatly helps our real-time denoising and backend signal processing. We used a 940±2 nm filter directly on the event camera with our own controlled series of 940 nm LEDs for scene exposure when and where required. Our event processing algorithm is developed to potentially provide real-time directional feedback to the person through bone-conduction vibrations that do not interfere with the normal hearing pathway; hence, the person is able to continue to hear the surrounding environment around them.