FEATURE BASED OPTICAL FLOW MODEL FOR OBSTACLE DETECTION ON A CAMERA PHONE

Abstract

A wide variety of assistive tools has failed to address some or all of the challenges faced by people with visual disability. For instance, obstacle avoidance poses a big challenge to visually impaired persons especially during the navigation of unknown places. The high resolution and low cost camera phone provides an alternative to record video and capture high quality images, which simplifies the computation of optical flow vectors for obstacle detection. This study, explore the benefit of camera phone in the capturing and processing of video frame into flow vectors for obstacle detection.