Open AccessStereo Vision-based Path Finder for Visually Impaired
Author(s) -
Shrugal Varde,
M. S. Panse
Publication year - 2020
Publication title -
international journal of recent technology and engineering
Language(s) - English
Resource type - Journals
ISSN - 2277-3878
DOI - 10.35940/ijrte.c4603.099320
Subject(s) - computer science , obstacle , computer vision , visually impaired , path (computing) , artificial intelligence , segmentation , stereopsis , stereo camera , real time computing , human–computer interaction , political science , law , programming language
This paper introduces a novel electronic mobility aid for visually impaired users that helps them navigate in any given environment and avoid knee level to head height obstacles. The mobility aid uses stereo imaging system to capture the images of the area in front of the user. The processing unit generates a disparity map and a segmentation algorithm extracts information about the relative distance of obstacles from the user. This information is relayed to the user in simplified vibration pattern feedback to inform the user of the path to be taken to avoid collision with the obstacle. Special hardware was designed to make the system portable and cost effective.The mobility aid was validated on 55 visually impaired users. The subjects walked in a controlled test environment with a varying number of obstacles placed in their path. The accuracy of the device to help the user avoid obstacles and the average speed of walking of the user were determined. The results obtained were satisfactory and the device has the potential for use in standalone mode as well as in conjunction with a white cane and thus help visually impaired people counter mobility problems.