US2021279452A1PendingUtilityA1

Action-estimating device

62
Assignee: ASILLA INCPriority: May 27, 2018Filed: May 19, 2021Published: Sep 9, 2021
Est. expiryMay 27, 2038(~11.9 yrs left)· nominal 20-yr term from priority
Inventors:Daisuke Kimura
G06V 40/23A61B 5/107G06Q 50/10A61B 5/1128G16H 40/20G06T 2207/10016G16H 50/20G06T 2207/30196G06T 7/75G06T 2207/20081G16H 80/00G16H 40/67G06T 2207/20076G06T 7/73G06K 9/00342
62
PatentIndex Score
0
Cited by
0
References
0
Claims

Abstract

[Problem]To provide an action-estimating device which is capable of estimating an action of a subject appearing in a plurality of time-series images with high precision.[Solution]Provided is an action-estimating device 1 comprising: an estimating-side obtaining unit 12 for obtaining a plurality of time-series images Y in which a subject Z appears; an estimating-side detecting unit 13 for detecting a plurality of articulations A appearing in each time-series image Y; an estimating-side measuring unit 14 for measuring coordinates of the detected plurality of articulations A in each time-series image Y; an estimating unit 15 for estimating an action of the subject Z based on displacement of the coordinates of the measured plurality of articulations A in the plurality of time-series images Y; and a storing unit 3 for storing a plurality of choices of the action to be estimated. The estimating-side detecting unit 13 further detects a background appeared in each time-series image Y. In order to estimate the action of the subject Z, the estimating unit 15 calculates a probability of each of the plurality of choices based on the displacement of the coordinates of the measured plurality of articulations A in the plurality of time-series images Y, and corrects the calculated probability of each of the plurality of choices based on the detected background.

Claims

exact text as granted — not AI-modified
1 . A non-transitory computer-readable medium storing an action-estimating program for a computer storing a plurality of choices of action to be estimated, the action-estimating computer program comprising:
 a step for obtaining a plurality of time-series images in which a subject appears;   a step for detecting a plurality of articulations appearing in each time-series image;   a step for measuring coordinates of the detected plurality of articulations in each time-series image;   a step for estimating an action of the subject based on displacement of the coordinates of the measured plurality of articulations in the plurality of time-series images; and   a step for detecting a background appeared in each time-series image,   wherein, in the estimating step, a probability of each of the plurality of choices is calculated based on the displacement of the coordinates of the measured plurality of articulations in the plurality of time-series images, and the calculated probability of each of the plurality of choices is corrected based on the detected background.   
     
     
         2 . The action-estimating program according to  claim 1 , wherein, in the estimating step, one or more choices are excluded from among the plurality of choices based on the detected background. 
     
     
         3 . The action-estimating program according to  claim 1 , wherein the choices whose actions have a relation greater than a prescribed value between each other are associated with each other in the computer,
 wherein, in the estimating step, when any one of the plurality of choices associated with each other was excluded or the probability of any one of the plurality of choices associated with each other was decreased, the probability of the associated choice which was not excluded or whose probability was not decreased is increased.   
     
     
         4 . The action-estimating program according to  claim 2 , wherein the choices whose actions have a relation greater than a prescribed value between each other are associated with each other in the computer,
 wherein, in the estimating step, when any one of the plurality of choices associated with each other was excluded or the probability of any one of the plurality of choices associated with each other was decreased, the probability of the associated choice which was not excluded or whose probability was not decreased is increased.   
     
     
         5 . An action-estimating device comprising:
 an estimating-side obtaining unit configured to obtain a plurality of time-series images in which a subject appears;   an estimating-side detecting unit configured to detect a plurality of articulations appearing in each time-series image;   an estimating-side measuring unit configured to measure coordinates of the detected plurality of articulations in each time-series image;   an estimating unit configured to estimate an action of the subject based on displacement of the coordinates of the measured plurality of articulations in the plurality of time-series images;   a setting unit configured to set a purpose or application for the estimating of the action of the subject; and   a storing unit configured to store a plurality of choices of the action to be estimated,   wherein, in order to estimate the action of the subject, the estimating unit calculates a probability of each of the plurality of choices based on the displacement of the coordinates of the measured plurality of articulations in the plurality of time-series images, and corrects the calculated probability of each of the plurality of choices based on the set purpose or application.   
     
     
         6 . The action-estimating device according to  claim 5 , wherein the estimating unit excludes one or more choices from among the plurality of choices based on the set purpose or application in order to estimate the action of the subject. 
     
     
         7 . The action-estimating device according to  claim 5 , wherein the choices whose actions have a relation greater than a prescribed value between each other are associated with each other in the storing unit, and
 wherein, when any one of the plurality of choices associated with each other was excluded or the probability of any one of the plurality of choices associated with each other was decreased, the action estimating unit increases the probability of the associated choice which was not excluded or whose probability was not decreased, in order to estimate the action of the subject.   
     
     
         8 . The action-estimating device according to  claim 6 , wherein the choices whose actions have a relation greater than a prescribed value between each other are associated with each other in the storing unit, and
 wherein, when any one of the plurality of choices associated with each other was excluded or the probability of any one of the plurality of choices associated with each other was decreased, the action estimating unit increases the probability of the associated choice which was not excluded or whose probability was not decreased, in order to estimate the action of the subject.   
     
     
         9 . A non-transitory computer-readable medium storing an action-estimating program for a computer storing a plurality of choices of action to be estimated and in which a purpose or application for estimating of action of a subject is set, the action-estimating computer program comprising:
 a step for obtaining a plurality of time-series images in which the subject appears;   a step for detecting a plurality of articulations appearing in each time-series image;   a step for measuring coordinates of the detected plurality of articulations in each time-series image; and   a step for estimating an action of the subject based on displacement of the coordinates of the measured plurality of articulations in the plurality of time-series images,   wherein, in the estimating step, a probability of each of the plurality of choices is calculated based on the displacement of the coordinates of the measured plurality of articulations in the plurality of time-series images, and the calculated probability of each of the plurality of choices is corrected based on the set purpose or application.   
     
     
         10 . The non-transitory computer-readable medium according to  claim 9 , wherein, in the estimating step, one or more choices are excluded from among the plurality of choices based on the set purpose or application. 
     
     
         11 . The non-transitory computer-readable medium according to  claim 9 , wherein the choices whose actions have a relation greater than a prescribed value between each other are associated with each other in the computer,
 wherein, in the estimating step, when any one of the plurality of choices associated with each other was excluded or the probability of any one of the plurality of choices associated with each other was decreased, the probability of the associated choice which was not excluded or whose probability was not decreased is increased.   
     
     
         12 . The non-transitory computer-readable medium according to  claim 10 , wherein the choices whose actions have a relation greater than a prescribed value between each other are associated with each other in the computer,
 wherein, in the estimating step, when any one of the plurality of choices associated with each other was excluded or the probability of any one of the plurality of choices associated with each other was decreased, the probability of the associated choice which was not excluded or whose probability was not decreased is increased.   
     
     
         13 . An action-estimating method executed on a computer storing a plurality of choices of action to be estimated and in which a purpose or application for estimating of action of a subject is set, the method comprising:
 a step for obtaining a plurality of time-series images in which the subject appears;   a step for detecting a plurality of articulations appearing in each time-series image;   a step for measuring coordinates of the detected plurality of articulations in each time-series image; and   a step for estimating an action of the subject based on displacement of the coordinates of the measured plurality of articulations in the plurality of time-series images,   wherein, in the estimating step, a probability of each of the plurality of choices is calculated based on the displacement of the coordinates of the measured plurality of articulations in the plurality of time-series images, and the calculated probability of each of the plurality of choices is corrected based on the set purpose or application.   
     
     
         14 . The action-estimating program according to  claim 13 , wherein, in the estimating step, one or more choices are excluded from among the plurality of choices based on the set purpose or application. 
     
     
         15 . The action-estimating program according to  claim 13 , wherein the choices whose actions have a relation greater than a prescribed value between each other are associated with each other in the computer,
 wherein, in the estimating step, when any one of the plurality of choices associated with each other was excluded or the probability of any one of the plurality of choices associated with each other was decreased, the probability of the associated choice which was not excluded or whose probability was not decreased is increased.   
     
     
         16 . The action-estimating program according to  claim 14 , wherein the choices whose actions have a relation greater than a prescribed value between each other are associated with each other in the computer,
 wherein, in the estimating step, when any one of the plurality of choices associated with each other was excluded or the probability of any one of the plurality of choices associated with each other was decreased, the probability of the associated choice which was not excluded or whose probability was not decreased is increased.

Cited by (0)

No later patents cite this yet.

References (0)

No backward citations on record.