Eigenfaces Based Face Recognition Device for Visually Impaired Persons

Abstract

This paper proposes to develop an embedded system based instrument for visually impaired, it is a wearable device in the form of badge holder of the patent which is designed using Raspberry Pi in addition with visual and hearing sensor. This system works to detect the human faces. With this help user can identify the person in front of the vision sensor. Open CV computer vision and machine learning libraries were used to simulate a system that aids the blind person to recognize his family and friend’s facial images that are stored in a database, and if a match is found in the database, the system will announce the name of the person via sound sensor to the blind person. Further for this system modelling we used various algorithms. Viola–Jones object detection framework as known as Haar-features for face detection, Eigenfaces for face recognition and python text to speech library used for audio output. The simulation considered the recognition of a static facial image (photo) and a live facial image. This system is beneficial in terms of its portability, low cost, low power consumption.