A rule based method for fall and inactivity detection using spatiotemporal human pose features

Modern computer vision techniques can provide valuable and objective support for monitoring the daily functioning of elderly people and individuals living alone, particularly in the context of the growing role of telemedicine. This paper presents a prototype solution whose hardware layer can be based on readily available cameras transmitting video streams to a computer. The software layer consists of a set of developed decision rules in which spatial and temporal features relevant to the detection of falls and inactivity are defined on the basis of human body keypoints. Such inactivity may result from fainting or loss of consciousness. Threshold values for the defined parameters were determined experimentally through the analysis of video material containing recordings of normal daily activities performed by participants, as well as simulated fall and inactivity events. The results of numerical experiments demonstrated that falls could be detected with a sensitivity of 91% and a specificity of at least 97.7%. For the detection of inactivity occurring either as a consequence of a fall or independently of it, the corresponding values were 93.9% and 100%, respectively.

Authors

Institutions

Publication Details

Journal
Journal of Automation Electronics and Electrical Engineering
Published
2026-11-16
DOI
https://doi.org/10.24136/jaeee.2026.011
Primary Topic
Context-Aware Activity Recognition Systems
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

A rule based method for fall and inactivity detection using spatiotemporal human pose features

Jacek Jakubowski, Jakub Korzeniewski
Journal of Automation Electronics and Electrical Engineering
Context-Aware Activity Recognition Systems
article

A rule based method for fall and inactivity detection using spatiotemporal human pose features

Jacek Jakubowski, Jakub Korzeniewski
article en

Abstract

Modern computer vision techniques can provide valuable and objective support for monitoring the daily functioning of elderly people and individuals living alone, particularly in the context of the growing role of telemedicine. This paper presents a prototype solution whose hardware layer can be based on readily available cameras transmitting video streams to a computer. The software layer consists of a set of developed decision rules in which spatial and temporal features relevant to the detection of falls and inactivity are defined on the basis of human body keypoints. Such inactivity may result from fainting or loss of consciousness. Threshold values for the defined parameters were determined experimentally through the analysis of video material containing recordings of normal daily activities performed by participants, as well as simulated fall and inactivity events. The results of numerical experiments demonstrated that falls could be detected with a sensitivity of 91% and a specificity of at least 97.7%. For the detection of inactivity occurring either as a consequence of a fall or independently of it, the corresponding values were 93.9% and 100%, respectively.

Journal of Automation Electronics and Electrical EngineeringVol. 8(3)
Military University of Technology in Warsaw (PL)
Good health and well-being
Context-Aware Activity Recognition Systems
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon