Abstract

Existing video monitoring techniques for sleep apnoea require clinicians to analyze substantial amounts of video data. Analysis of the covered human body from video is a challenging task as traditional computer vision methods such as correlation, template matching, background subtraction, contour models and related techniques for object tracking become ineffective because of the large degree of occlusion for long periods. To the authors’ best knowledge, there is no previously published method to estimate pose from persistently covered human body. This paper presents an automated monocular video monitoring approach to recover the human pose in conditions with persistently heavy obscuration, allowing for further analysis of covered human activity. In evaluation, we demonstrate that the proposed technique is able to identify human configurations with various poses and occlusion levels in two different environments