Developing A System For Visually Impaired Persons

Abstract

In order to help visually impaired, an emolyzer called emotion analyzer can be used to detect the emotion of the person opposite to them and helps them for an effective interaction with other people. In this proposed method, visually challenged people use spectacles in which a camera is inserted for capturing the video of person who is communicating with them and earphones are used so that the detected emotion by the device, can be communicated to blind people who is using the device. This device can identify some basic emotions like happiness, sadness, a neutral face and an angry face so this makes emolyzer a helpful and an effective tool which can be used by visually impaired. The emolyzer is made using Raspberry Pi and is made portable for ease of use. Here emolyzer is trained using FER-2013 dataset which contains more than 35,000 gray scale images. A free source Open CV python library is used for detection of faces. The detected emotion is finally transformed into speech by installing python libraries like play sound.