Machine Vision Enabled Fall Detection System for Specially Abled People in Limited Visibility Environment

In this study, a unique fall detection system with machine vision capabilities is presented. It is intended to help people with special needs in environments with low visibility. People with disabilities are far more likely to fall, so it is important to spot such accidents right once for their safety and wellbeing. However, current fall detection technologies sometimes have trouble operating effectively in dimly lit environments. Our suggested method uses cutting-edge computer vision algorithms to detect falls in real-time, even in environments with poor sight, in order to solve this problem. The device collects visual data using infrared sensors and a low-light camera, and uses a deep learning algorithm to identify falls. The deep learning model can learn complicated patterns and traits related to falls because it was trained on a huge dataset of annotated fall and non-fall samples. The model takes the collected photos and extracts the pertinent elements, then categorises them as autumn events or non-fall events. The system also uses a threshold-based methodology to distinguish between falls and other behaviours that could cause false alarms. To assess the system's performance, numerous tests were run, including ones in controlled low lighting and in real-world situations. The outcomes show that the suggested system can detect falls reliably even in difficult lighting situations, attaining high detection accuracy and a low false alarm rate. The machine vision enabled fall detection system has a great deal of promise to improve the safety and independence of people with disabilities by offering prompt help in the event of falls, especially in dimly light areas. The system provides a dependable and efficient way to meet the particular requirements of people with disabilities in low-light situations and can be deployed in a variety of venues, including homes, hospitals, or care facilities.