An Efficient Smart Goggle For Blind People
Neha Shinde, Priti Panchal, Omkar Saswade, Ashwini Pandagale · Zenodo (CERN European Organization for Nuclear Research) · 2020
The smart talking goggles prototype could be a small, cheap and simply wearable navigation aid for visually impaired persons. Its functionality addresses navigation and identification of objects for guiding the user to some and native navigation for sidewalks and corridors, object like human, non-human and solid objects like vehicle – car, bus, tempo, bike etc. With their approximate distance from an individual wearing our prototype. This prototype aims at achieving accuracy in detection of those objects, alerting user by outputting tone and voice stating object classification and sort of distance with avoidance of static further as moving obstacles. During this prototype, we specialize in local navigation and identification which applies to both in- and outdoor situations. The detection of path borders and obstacles before of the user and just beyond the reach of the white cane, such the user are often assisted in centring on the trail and alerted to looming hazards. Employing a stereo camera worn at goggles height, a conveyable computer in an exceedingly shoulder-strapped pouch or pocket and just one earphone or small speaker. The vision[13] algorithms are optimized such the system can work some frames per second . For achieving real time analysis, we used machine learning and neural networking to classify and process incoming dataset with stored procedures with the assistance of cloud computing platform, because node MCU or arduino boards don