Friendly Supporting System for Improvement of Disabled People's Comfortable Feeling Using Sound and Image Processing

Abstract

In this paper, we report a friendly supporting system for disabled people that is able to have comfortable feeling for both disabled and their assisting people using sound and image processing for being enhanced their Kansei. In short, this supporting system employs multi-modal interface adapter, which is able to operate easily with voice recognition, and a new combination of sound and image signals processing, which detect abnormal condition of disabled people. The multi-modal interface adapter consists of mouse emulator for operating all client equipments needed for daily life, and voice recognition for manipulating these apparatus. The sound processing judges unordinary state by detection of abnormal sounds from medical devices. At the same time, this device catches the client voice in surrounding noisy sounds. The image processing judges the abnormal states of disabled person and medical electric devices by images through the cameras. Some experiments have been performed and results show that each processing is effective key technology for a system configuration of supporting system. From these results, we find much possibility for a simple system, which is able to secure the privacy for disabled person and lighten the burdens of their helpers, and get comfortable feeling both disabled and assisting people of which assists and enhance their Kansei interface.