Human Fall Detection Using Vision Sensors
by Benson Isaac — last modified 2012-04-19 09:48
This is an ongoing project which focuses on developing a non-invasive sensor based fall detection system to address the growing concerns for developing a smart home system for the care of elderly. This work is part of an innovative paradigm of research aimed at a synergistic exploitation of distributed devices via techniques that are at the intersection of data mining, inferencing and learning, and distributed decision making for developing a co-robotics based solution for personalized care for elderly with cognitive impairment. The initial part of this work, involves the implementation of a human motion tracking system for monitoring the activities of one person using the Microsoft Kinect sensor. The human gait information obtained thus would be used to determine fall by matching the gait patterns with that of the patterns associated with the fall. The overall objectives of this work are directed towards realizing the larger goal of developing a robust system for fall detection for the elderly with measures in place against false detections.