A Computer Vision Approach to Personalized Light Dimming for Elderly and Bedridden Care

The use of Internet of Things (IoT) solutions for home automation, shopping centers, and other public spaces has increased in the latest years. Albeit, using these devices can be a challenge for immobile, mute, and individuals with limited digital literacy. To serve their needs, a computer vision (CV) based gesture-controlled home automation system can be outperforming. Nevertheless, existing solutions for CV-based home automation systems are mainly restricted to basic load switching and they offer limited control over fan speed and light intensity control. This paper explores the establishment of a CV algorithm of pinch and spread gesture class using a MediaPipe library to modulate the light brightness and fan speed with the fingers proximity. A digital solid-state relay (SSR) has also been developed to interact with a CV algorithm using an Arduino Mega 2560 microcontroller board. The proposed study can lay the foundation for smart and inclusive solutions while also helping disenfranchised people.