Design and Implementation of Smart Assistance using Image Segmentation and Reorganization

Abstract

The most valuable present we can give someone is their eyes, which are also very important to our daily existence. For persons who have vision impairment, Long-term independent walking on the road is made possible by the ingenious support hardware design. The goal of this project is to create a prototype of smart assistance that will allow blind people to navigate and move safely. A smart assistance system for the visually impaired is a system for blind persons built using a vibration motor combined with a raspberry pi board. The camera captures live streaming video which will be further processed for segmentation using Region Growing Algorithm(RGA) and object reorganization by support vector machine(SVM), since the device selects the object's name, translates the speech and then produces an audio output in the form of commands when it detects an obstruction. The visually impaired individual receives our assistance, enabling the blind person to reach his destination independently. The proposed system aids in accident prediction for the blind boosts self-assurance in those with visual impairments, and assists in obstacle detection for those with visual impairments while they walk so that the user can easily navigate the impediment.