Object Discrimination Using Stereo Vision for Blind through Stereo Sonification

In this paper, a development of image processing, stereovision methodology and a sonification procedure for image sonification system for vision substitution are presented. The hardware part consists of a sunglass fitted with two mini cameras, laptop computer and stereo earphones. The image of the scene in front of blind people is captured by stereo cameras. The captured image is processed to enhance the important features in the scene in front of blind user. The image processing is designed to extract the objects from the image and the stereo vision method is applied to calculate the disparity which is required to determine the distance between the blind user and the objects. The processed image is mapped on to stereo sound for the blind’s understanding of the scene in front. Experimentations were conducted in the indoor environment and the proposed methodology is found to be effective for object identification, and thus the sound produced will assists the visually impaired for their collision free navigation.