摘要:
A method and a system for unobtrusively measuring a person's respiration, wherein at least two microphones positioned at either side of a person's head are utilized based on a distance of the persons head to the microphones and an identification of which microphone is in a line of sight and which is out of the line of site of the person's breathing as computed on the basis of images captured by a video camera.
摘要:
The invention relates to methods and apparatus for monitoring movement and breathing of two or more subjects occupying common bedding. Particularly, the invention relates to a method for monitoring movement of subjects located in a common bedding, the method comprising the steps of: —imaging the bedding by an optical sensor; —performing a motion estimation by producing motion vectors indicating the local displacement of corresponding image blocks between consecutive images, or images that are several frames apart, received from said optical sensor; —calculating motion clusters by measuring spatial and temporal correlations of the motion vectors; —segmenting the calculated motion clusters by assignment of each motion cluster to a corresponding subject, wherein the assignment of the motion clusters to the corresponding subject is based on the spatial and/or temporal similarity of the motion clusters among each other and on previous segmentation results.
摘要:
The invention relates to methods and apparatus for monitoring movement and breathing of two or more subjects occupying common bedding. Particularly, the invention relates to a method for monitoring movement of subjects located in a common bedding, the method comprising the steps of:—imaging the bedding by an optical sensor;—performing a motion estimation by producing motion vectors indicating the local displacement of corresponding image blocks between consecutive images, or images that are several frames apart, received from said optical sensor;—calculating motion clusters by measuring spatial and temporal correlations of the motion vectors;—segmenting the calculated motion clusters by assignment of each motion cluster to a corresponding subject, wherein the assignment of the motion clusters to the corresponding subject is based on the spatial and/or temporal similarity of the motion clusters among each other and on previous segmentation results.