Remote Fall Tracking using Multiple Extraction Methods and Supervised Learning
Hephzibah Thomas1, Thyla B2
1Hephzibah Thomas, Department of Electronics and Communication, KCG College of Technology, India.
2Thyla B. Department of Electronics and Communication, KCG College of Technology, India.
Manuscript received on 01 May 2019 | Revised Manuscript received on 15 May 2019 | Manuscript published on 30 May 2019 | PP: 206-211 | Volume-8 Issue-7, May 2019 | Retrieval Number: G5120058719/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Owing to the psychological and physical aftermath post mishaps, healthy environment has become a general passion. Mishaps compose of falls which have become a grave distress for the elderly and diseased living alone. Researchers have involved in finding the optimal alternatives. This includes wearable sensors, artificial intelligence, etc. What is this paper about? It was to find the best approach to efficiently detect a fall with fewer false alarms and was implemented by finding the Histogram of Oriented Gradients along with statistical methods which extracts relevant features and compared it with the trained videos. A supervised learning technique is exploited, where the database is trained with videos that contain both fall and quotidian activities (QA). Support Vector Machine (SVM) is utilized in distinguishing fall and daily events. The doctor/caretaker is intimated via email on detection of fall.
Keyword: GMM, HOG, Real-time processing, Supervised Learning, Statistical feature, SVM.
Scope of the Article: Smart Learning Methods and Environments.