Device for characterizing the actimetry of a subject in real time

Patent metadata

Field	Value
Publication number	US-12153126-B2
Application number	US-202017766760-A
Country	US
Kind code	B2
Filing date	Oct 7, 2020
Priority date	Oct 7, 2019
Publication date	Nov 26, 2024
Grant date	Nov 26, 2024

Abstract

The invention discloses a device ( 1 ) for characterizing in real time the actimetry of a subject, having: a radar ( 2 ) emitting and receiving radar signals, and having a software interface for configuring the shape of the signal emitted; processing and computing means ( 3 ) coupled to the radar ( 2 ), having a trained classifier ( 3 a ) using a database, said processing and computing means ( 3 ) being configured to perform in real time: —a capture of color micro-Doppler images ( 6 ) having several color channels (R, V, B), each having micro-Doppler signatures ( 6 a ) with color pixels the value of which is a function of a reflectivity and a speed of the subject; —a processing of the micro-Doppler images ( 6 ) for: computing a so-called monochromatic image having monochromatic pixels, each having a given monochromatic intensity, on the basis of the color pixels of each color micro-Doppler image; transforming the monochromatic image into a binary image by segmentation, according to a binary luminous intensity threshold, of the monochromatic pixels, producing binary pixels, the value of which is dependent on the chromatic intensity of the monochromatic pixel associated with the binary pixel, with respect to the threshold.

First claim

The invention claimed is:

1. A device for characterizing in real time actimetry of a subject, having:
a radar emitting and receiving radar signals, and having a software interface for configuring a shape of the signal emitted;
means for processing and computing coupled to the radar, having a classifier trained using a database,
said means for processing and computing being configured to perform in real time:
a capture of color micro-Doppler images having several color channels, each having micro-Doppler signatures with color pixels;
a processing of micro-Doppler images for:
computing a monochromatic image having monochromatic pixels, each having a given monochromatic intensity, on the basis of the color pixels of each color micro-Doppler image;
transforming the monochromatic image into a binary image by segmentation, according to a binary luminous intensity threshold, of the monochromatic pixels,

by producing binary pixels, a binary value of which for each binary pixel is a function of a value of chromatic intensity of the monochromatic pixel associated with the binary pixel, with respect to the binary luminous intensity threshold,
by forming, on a surface of the binary image, segmented areas which have binary pixels of the same binary value, and which result from the transformation of each micro-Doppler signature;
computing geometrical shape parameter values on each segmented area, each geometrical shape parameter being solely a parameter that characterizes a geometrical shape of segmented areas,
classifying each binary image in a class pertaining to the actimetry of the subject, as a function of the values of the parameters computed for all the segmented areas of the binary image, using the classifier trained with these geometrical shape parameters computed over segmented areas of binary images of test subjects having a known actimetry,

means for storing coupled to the means for processing and computing for storing the classifier, the micro-Doppler images, the monochromatic images, the binary images and the classification obtained from the binary images.

2. The device according to claim 1 , wherein the geometrical shape parameters of each segmented area of binary pixels are chosen from among the following list:
surface,
perimeter,
first-degree Centroid,
second-degree Centroid,
orientation,
Computing of the zeroth-, first- and second- to nth-order moments,
Bounding square, and/or
Bounding ellipse.

3. The device according to claim 1 , said binary luminous intensity threshold being variable as a function of the luminance.

4. The device according to claim 1 , wherein the segmentation of the pixels is done using said binary luminous intensity threshold according to the Otsu method.

5. The device according to claim 1 , said binary luminous intensity threshold being fixed.

6. The device according to claim 1 , wherein the classifier is a binary classifier and/or a cascade of binary classifiers.

7. The device according to claim 1 , wherein the classifier is a multi-class classifier.

8. The device according to claim 1 , wherein the device is configured to continuously characterize the actimetry of the subject, the database being continuously supplied with the classification obtained from the binary images of the subject with the geometrical shape parameters applied to each segmented area and performed by the classifier on the basis of the micro-Doppler images.

9. The device according to claim 1 , wherein the means for processing and computing are configured to filter out the pixels of a same binary value as the pixels of the segmented areas, but located at a distance from the segmented areas.

10. The device according to claim 1 , wherein the device is onboard.

11. The device according to claim 1 , wherein processing of the micro-Doppler images scans each of the binary images in real time with a sliding sub-window;
and for each position of said sliding sub-window, geometrical shape parameters are extracted to classify each sub-image extracted from the binary image under consideration, in a class relating to the actimetry of the subject.

12. The device as claimed in claim 11 , wherein the processing of the micro-Doppler images makes it possible, with the sliding sub-window, to carry out tracking to modify capturing parameters as a function of the segmented areas and of expected classes, to modify the next segmented areas to arrive.

13. The device according to claim 1 , wherein the computing of the monochromatic image is performed in shades of gray, said shades of gray being a function of a value of the color channels of the color pixels.

14. The device according to claim 1 , wherein the computing of the monochromatic image is performed with the color gray which is a function of a value of the color channels of the color pixels, following the formula: Gray=0.299*Red+0.587*Green+0.144*Blue, for each pixel of the color micro-Doppler images which has a red intensity value, a green intensity value and a blue intensity value.

15. The device according to claim 1 , wherein the class is chosen from among the following list:
walking,
running
standing up,
sitting,
TUG class,
walking with an object carried in both hands
falling,
sitting on a chair,
tying of shoelaces,
Parkinsons gait,
sitting on the ground,
picking up an object,
searching for an object under a chair,
standing up from a chair and walking.

16. The device according to claim 1 , wherein the means for storing are configured to store successive sequences of classes obtained, and to determine an activity map of the subject.

17. The device according to claim 1 , wherein the radar emits the radar signals in a frequency band between 6 MHz and 250 GHz.

18. A system for controlling actimetry of a subject, said system comprising:
a building with walls;
a device for characterizing in real time the actimetry of said subject, said device being incorporated into a wall of said building and having:
a radar emitting and receiving radar signals, and having a software interface for configuring a shape of the signal emitted;
means for processing and computing coupled to the radar, having a classifier trained using a database,
said means for processing and computing being configured to perform in real time:

a capture of color micro-Doppler images having several color channels, each having micro-Doppler signatures with color pixels;
a processing of micro-Doppler images for:
computing “monochromatic image” having monochromatic pixels, each having a given monochromatic intensity, on the basis of the color pixels of each color micro-Doppler image;
transforming the monochromatic image into a binary image by segmentation, according to a binary luminous intensity threshold, of the monochromatic pixels,
by producing binary pixels, a binary value of which for each binary pixel is a function of a value of a chromatic intensity of the monochromatic pixel associated with the binary pixel, with respect to the binary luminous intensity threshold,
by forming, on the surface of the binary image, segmented areas which have binary pixels of the same binary value, and which result from the transformation of each micro-Doppler signature;
computing geometrical shape parameter values on each segmented area, each geometrical shape parameter being solely a parameter that characterizes the geometrical shape of segmented areas,
classifying each binary image in a class pertaining to the actimetry of the subject, as a function of the values of the parameters computed for all the segmented ar

Assignees

• Ecole Nat Superieure De Lelectronique Et De Ses Applications
• Centre Nat Rech Scient
• Univ Glasgow Court
• Cy Cergy Paris Univ
• Cy Energy Paris Univ

Inventors

• Romain Olivier
• Lekernec Julien
• Lorandel Jordane
• Fioranelli Francesco

Classifications

• G08B21/0446

  worn on the body to detect changes of posture, e.g. a fall, inclination, acceleration, gait · CPC title

• G06V10/764

  using classification, e.g. of video objects · CPC title

• A61B5/1117

  Fall detection · CPC title

• G01S7/417

  involving the use of neural networks · CPC title

• A61B5/7264

  Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems · CPC title