Visual Scene Prediction for Blind People Based on Object Recognition

Vision is one of the very essential human senses which is playing the most important role in human perception about surrounding environment. Hence, over the published papers on this subject, a variety of computer vision products and services are proposed by developing new methods for scene understanding. This work aims to introduce an approach that restores a central function of the visual system which is the visual scene interpretation. The proposed pipeline is based on the identification of surrounding objects in dissimilar frames from video scene. Otherwise, the experimental results using feature extractions showed good accuracy for detecting object in natural conditions. Features extraction and Bag of Visual Words correspondence are exploited for scene type understanding.