Obtainment of cleaned sequences relating to a center of gravity

Patent metadata

Field	Value
Publication number	US-11633155-B2
Application number	US-202016745131-A
Country	US
Kind code	B2
Filing date	Jan 16, 2020
Priority date	Mar 21, 2016
Publication date	Apr 25, 2023
Grant date	Apr 25, 2023

Abstract

Methods and systems are provided for obtaining cleaned sequences showing trajectories of movement of a center of gravity and for estimating a biometric information pattern or value of a target. One of the methods includes removing noises from initial sequences showing trajectories of movement of a center of gravity to obtain the cleaned sequences. Another one of the methods includes reading cleaned sequences of the target into a memory, extracting features from the cleaned sequences, and estimating a biometric information pattern or value of the target from the extracted features, using a classification or regression model of biometric information patterns or values. The biometric information pattern may be a pattern derived from respiratory or circulatory organs of a target.

First claim

What is claimed is:

1. A method for estimating biometric information, the method comprising:
selectively removing one or more noises from one or more data sequences not associated with biometric information of an individual;
rotating one or more trajectories in the one or more data sequences to overlap a center of gravity of the individual; and
estimating the biometric information from the one or more data sequences.

2. The method according to claim 1 , wherein the one or more data sequences show a biometric information pattern which is repeated at regular intervals.

3. The method according to claim 2 , wherein the biometric information pattern is a pattern derived from respiratory or circulatory organs of a target.

4. The method according to claim 1 , wherein selectively removing the one or more noises further comprises:
removing, from the one or more data sequences, one or more noises which are not one or more trajectories relating to a biometric information pattern or value which is repeated at regular intervals.

5. The method according to claim 4 , wherein the probabilistic model is estimated by the density estimation of a particular model, a Kernel density estimation, a neural network, time-series models, or state space models.

6. The method according to claim 1 , wherein the one or more data sequences are calculated based on data obtained from one or more load sensors, one or more visual sensors, one or more infrared sensors, or a combination of these.

7. The method according to claim 6 , wherein the one or more data sequences are calculated based on data obtained from load sensors, and the load sensors are attached to a bed or chair on which a target is lying, sleeping or sitting.

8. The method according to claim 1 , wherein selectively removing the one or more noises further comprises:
calculating a physical quantity from the one or more data sequences, finding one or more outliers in the physical quantity, and then removing one or more trajectories corresponding to the one or more outliers from the one or more data sequences to obtain first sequences.

9. The method according to claim 8 , wherein the physical quantity is a location of a center of gravity or a location vector of a center of gravity, a direction vector, a velocity vector, an acceleration vector, or a combination of these.

10. The method according to claim 8 , wherein the physical quantity is a vector or matrix which is selected from the group consisting of various statistics in a certain time window of a location of a center of gravity or a location vector of a center of gravity, a direction vector, a velocity vector, an acceleration vector, or a combination of these.

11. The method according to claim 8 , the one or more trajectories corresponding to the one or more outliers is one or more trajectories of abnormal movement of the center of gravity.

12. The method according to claim 8 , wherein a probabilistic model for the physical quantity is estimated, the one or more outliers are the corresponding physical quantity of which the probability or probabilistic density is lower than a predetermined threshold, and the removal of noises is carried out by removing one or more trajectories corresponding to the one or more outliers from the one or more data sequences.

13. The method according to claim 8 , the method further comprising rotating the first sequences so as to take one common direction to obtain second sequences.

14. The method according to claim 13 , the method further comprising calibrating positions of the second sequences to obtain third sequences.

15. A method for estimating a biometric information pattern or value of a target, the method comprising:
reading, into a memory, cleaned sequences of the target obtained byl the method according to claim 1 , wherein the removal of noises is carried out by selectively removing one or more noises from one or more data sequences not associated with biometric information of an individual;
extracting features from the cleaned sequences; and
estimating a biometric information pattern or value of the target from the extracted features using biometric information patterns.

16. The method according to claim 15 , further comprising:
obtaining a classification or regression model of the biometric information patterns by reading training data sequences;
extracting features from the training data sequences; and
training the classification or regression model, using the features extracted from the training data together with manually labeled biometric information patterns or values or using a rule which uses previous knowledge of experts.

17. A system, comprising:
a processor; and
a memory storing a program, which, when executed on the processor, performs an operation comprising:
selectively removing one or more noises from one or more data sequences not associated with biometric information of an individual;
rotating one or more trajectories in the one or more data sequences to overlap a center of gravity of the individual; and
estimating the biometric information from the one or more data sequences.

18. The system according to claim 17 , wherein selectively removing the one or more noises further comprises:
calculating a physical quantity from the one or more data sequences;
finding one or more outliers in the physical quantity; and
removing one or more trajectories corresponding to the one or more outliers from the one or more data sequences to obtain first sequences.

19. The system according to claim 17 , wherein the one or more data sequences are calculated based on data obtained from one or more load sensors, one or more visual sensors, one or more infrared sensors, or a combination of these.

20. The system according to claim 17 , wherein the biometric information pattern is a pattern derived from respiratory or circulatory organs of a target.

Assignees

• IBM

Inventors

• Katsuki Takayuki
• Morimura Tetsuro

Classifications

• A61B5/6892

  Mats · CPC title

• A61B5/1128

  using image analysis (A61B5/1127 takes precedence) · CPC title

• A61B2560/0223

  of calibration, e.g. protocols for calibrating sensors · CPC title

• A61B5/6891

  Furniture · CPC title

• A61B5/02

  Detecting, measuring or recording for evaluating the cardiovascular system, e.g. pulse, heart rate, blood pressure or blood flow · CPC title