Device, system and method for automated detection of orientation and/or location of a person

Patent metadata

Field	Value
Publication number	US-10121062-B2
Application number	US-201514930212-A
Country	US
Kind code	B2
Filing date	Nov 2, 2015
Priority date	Nov 3, 2014
Publication date	Nov 6, 2018
Grant date	Nov 6, 2018

Abstract

A device, system and method automatically detects orientation and/or location of a person. To increase the robustness and accuracy, the device includes an image data interface (20) for obtaining image data of a person (110), which image data includes a sequence of image frames over time, a motion detector (21) for detecting motion within said image data, a motion intensity detector (22) for identifying motion hotspots representing image areas showing frequently occurring motion, and a person detector (23) for detecting the orientation and/or location of at least part of the person (110) based on the identified motion hotspots.

First claim

The invention claimed is:

1. A device for automated detection of orientation and/or location of a person, comprising:
an image data interface configured to obtain image data of a person, said image data comprising a sequence of image frames over time,
a motion detector configured to detect motion within said image data,
a motion intensity detector configured to identify motion hotspots, wherein the motion hotspots show areas of the image frames with frequently occurring motion,
a person detector configured to identify the motion hotspot showing the most intensive motions over a predetermined time period as a first boundary of the person.

2. The device as claimed in claim 1 , wherein said motion detector is further configured to determine motion images representing motion detected within said image data from one image frame to another image frame and wherein said motion intensity detector is configured to identify motion hotspots by accumulating said motion images.

3. The device as claimed in claim 2 , wherein said person detector is further configured to detect a left and/or right boundary of the person in an image by edge strength analysis in one or more accumulated images in a direction substantially parallel to the longitudinal axis of the person.

4. The device as claimed in claim 3 , wherein said person detector is further configured to cluster edges detected in several accumulated images and to identify the largest cluster of edges to the left and right, respectively, as the respective boundary.

5. The device as claimed in claim 1 , wherein the first boundary is an upper boundary of the person in an image.

6. The device as claimed in claim 1 , wherein said person detector is further configured to identify the motion hotspot located farthest away from the first boundary towards an edge of the image opposite the first boundary as a second boundary of the person in an image, wherein the first boundary is an upper boundary and the second boundary is a lower boundary.

7. The device as claimed in claim 6 , wherein said person detector is further configured to identify locations of body parts of the person based on the detected upper and/or lower boundary and known or average body proportions of the person.

8. The device as claimed in claim 1 , wherein said person detector is further configured to detect the presence, orientation and/or location of one or more additional persons or at least parts thereof based on the identified motion hotspots and to use the result of this detection in the detection of the orientation and/or location of at least part of the person.

9. The device as claimed in claim 8 , wherein said person detector is further configured to use the intensity, location, pattern and/or duration of motion of said one or more additional persons in the detection of the presence, orientation and/or location of one or more additional persons or at least parts thereof and/or in the orientation and/or location of at least part of the person.

10. The device as claimed in claim 1 , further comprising:
an analysis unit configured to analyze movements of the person or one or more body parts of the person over time based on the detected orientation and/or location of at least part of the person as detected regularly, continuously or from time to time, and
an evaluation unit configured to classify the detected movements into natural and unusual movements, determining a disease severity score estimation and/or issuing an information signal in case the analyzed movements exceed predetermined criteria for issuing an information signal.

11. A device for automated detection of orientation and/or location of a person, comprising:
an image data interface configured to obtain image data of a person, said image data comprising a sequence of image frames over time and to determine motion images representing motion detected within said image data from one image frame to another image frame
a motion detector configured to detect motion within said image data,
a motion intensity detector configured to identify motion hotspots which show areas of the image frames with frequency occurring motion and to accumulate motion images covering a predetermined time period, to obtain one or more accumulation images by summing up binary occurrences of motion pixels or motion vector information within two or more motion images,
a person detector configured to detect an orientation and/or location of at least part of the person based on the identified motion hotspots.

12. The device as claimed in claim 11 , wherein said motion intensity detector is further configured to identify image areas as motion hotspots, which have a size above a size threshold or the largest sizes within one or more accumulation images and which show motion above a motion threshold or the most intensive motions.

13. A system for automated detection of orientation and/or location of a person, comprising:
one or more video cameras configured to acquire image data of a person, said image data comprising a sequence of image frames over time;
the device as claimed in claim 1 for automated detection of orientation and/or location of at least part of the person based on the acquired image data, and
an output interface configured to output information related to the detected orientation and/or location of at least part of the person.

14. A method for automated detection of orientation and/or location of a person, comprising:
obtaining image data of a person, said image data comprising a motion estimation in a sequence of accumulated image frames over time,
detecting motion within said accumulated image data,
determining motion images representing motion from one image frame to another image frame in the accumulated image data,
identifying motion hotspots representing image areas showing most intense motion over a predetermined period of time, taking into account large motion areas and thresholding the accumulated image data with an adaptive threshold,
identifying a boundary of the person based on the identified most intense motion hotspots.

15. A non-transitory computer readable medium with a computer program with executable instructions causing a computer to carry out the steps of the method as claimed in claim 14 when said computer program is carried out on the computer.

16. A device for automated detection of orientation and/or location of a person, comprising:
one or more video cameras configured to generate image data of a person including a sequence of video frames of the person;
one or more computer processors programmed to:
receive the image data from the one or more video cameras;
detect motion within the image data from the one or more video cameras;
identify motion hotspots within the image data from the one or more video cameras by accumulating image data over a plurality of rames, wherein the motion hotspots show areas with frequently occurring motion;
detect a left and/or right boundary of the person in an image by edge strength analysis in one or more accumulated images in a direction substantially parallel to the longitudinal axis of the person based on the identified motion hotspots; and

a user interface configured to output information related to the detected left and/or right boundary of the person.

Assignees

• Koninklijke Philips Nv

Inventors

• Heinrich Adrienne
• Xie Yingrong
• Van Der Heide Esther Marjan
• Falck Thomas

Classifications

• G16H30/40

  for processing medical images, e.g. editing · CPC title

• A61B5/002 Primary

  Monitoring the patient using a local or closed circuit, e.g. in a room or building (A61B5/0017 takes precedence) · CPC title

• G06V40/20

  Movements or behaviour, e.g. gesture recognition (recognition of facial expressions G06V40/16) · CPC title

• G06F18/23

  Clustering techniques · CPC title

• G06F18/2431

  Multiple classes · CPC title