Gesture detecting method, gesture detecting apparatus, and recording medium

• US 7,650,017 B2
• Filed: 06/03/2005
• Issued: 01/19/2010
• Est. Priority Date: 06/08/2004
• Status: Expired due to Fees

• Current Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Inventors: Yamada, Miki
• Primary Examiner(s): Mehta; Bhavesh M
• Assistant Examiner(s): MACKOWEY, ANTHONY M
• Application Number:
  US11/143,719
• Publication Number:
  US 20050271252A1
• Time in Patent Office: 1,691 Days
• Field of Search: 382/103, 382/107, 382/202, 715/863
• US Class Current: 382/107
• CPC Class Codes: G06V 40/165 using facial parts and geometric relationships
• CPC Class Codes: G06V 40/20 Movements or behaviour, e.g. gesture recognition recognition of facial expressions G06V40/16

Abstract

A gesture detecting apparatus prepares time series data of coordinates of specific part of object gesturing, calculates velocity vector of the specific part from the time series data of coordinates, to obtain time-series velocity vectors, detects dynamic periods when displacement per unit time of the specific part is larger than predetermined value, based on the time-series velocity vectors, acquires each representative velocity vector from velocity vectors in each dynamic period, stores the each representative velocity vector corresponding to the each dynamic period in memory, calculates first intersection angle of two representative velocity vectors stored in the memory, the two representative velocity vectors corresponding to two successive dynamic periods, and detects that the motion of the specific part is reciprocating motion when the first intersection angle is larger than first threshold value.

Claims

1. A gesture detecting method implemented by a computer programmed as a gesture-detecting apparatus, the method comprising:

calculating, by the computer, a velocity vector of a specific part of an object gesturing, from time series data of coordinates of the specific part, to obtain a time-series velocity vectors;

detecting dynamic periods when displacement per unit time of the specific part is larger than a displacement threshold value, based on the time-series velocity vectors;

acquiring a representative velocity vector from velocity vectors in each dynamic period;

storing each representative velocity vector corresponding to each dynamic period in a storage unit;

calculating a first intersection angle of two representative velocity vectors stored in the storage unit, the two representative velocity vectors corresponding to two successive dynamic periods; and

detecting that a motion of the specific part is a reciprocating motion when the first intersection angle is larger than a first threshold value.

2. The method according to claim 1, wherein the detecting the dynamic periods detects the dynamic periods when magnitudes of the velocity vectors are larger than a velocity threshold value.

3. The method according to claim 1, further comprising:

calculating a second intersection angle of two representative velocity vectors stored in the storage unit, the two representative velocity vectors corresponding to two dynamic periods every two dynamic periods; and

detecting that the motion of the specific part is a repetitive motion which is a motion repeating the reciprocating motion, when the first intersection angle is larger than the first threshold value and the second intersection angle is smaller than a second threshold value.

4. The method according to claim 1, wherein the representative velocity vector is one of the velocity vectors in each dynamic period, a magnitude of the representative velocity vector being the largest one of the velocity vectors in the each dynamic period.

5. The method according to claim 1, wherein the representative velocity vector is an average velocity vector of the velocity vectors in the each dynamic period.

6. The method according to claim 1, wherein the representative velocity vector is a velocity vector at a central time of the each dynamic period.

7. The method according to claim 1, wherein the time series data of coordinates is time series data of coordinates of a head of the object gesturing.

8. The method according to claim 1, wherein the velocity vector is an angular velocity vector.

9. The method according to claim 1, wherein the detecting each dynamic period includes:

detecting a time point when a magnitude of one of the time series velocity vectors is larger than a velocity threshold value and a magnitude of another of the time series velocity vectors immediately prior to the one of the time series velocity vectors is smaller than the velocity threshold value; and

acquiring the dynamic period which starts at the time point and is a predetermined amount of time.

10. The method according to claim 1, wherein the detecting each dynamic period includes detecting each static period when magnitudes of the velocity vectors are smaller than or equal to a velocity threshold value;

and

detecting the dynamic period which is between two successive static periods.

11. A gesture detecting method implemented by a computer programmed as a gesture-detecting apparatus, the method comprising:

calculating, by the computer, a velocity vector of a specific part of an object gesturing, from time series data of coordinates of the specific part, to obtain time-series velocity vectors;

detecting a representative velocity vector whose magnitude is larger than a velocity threshold value and a maximum, from the time-series velocity vectors, to obtain time-series representative velocity vectors;

storing the time-series representative velocity vectors in a storage unit;

calculating a first intersection angle of two successive representative velocity vectors stored in the storage unit; and

detecting that a motion of the specific part is a reciprocating motion when the first intersection angle is larger than a first threshold value.

12. The method according to claim 11, further comprising:

calculating a second intersection angle of two representative velocity vectors every two representative velocity vectors stored in the storage unit; and

detecting that the motion of the specific part is a repetitive motion, which is a motion repeating the reciprocating motion, when the first intersection angle is larger than the first threshold value and the second intersection angle is smaller than a second threshold value.

13. A computer readable storage medium storing a computer program for causing a computer to perform a method having steps of:

calculating a velocity vector of a specific part of an object gesturing, from time series data of coordinates of the specific part, to obtain time-series velocity vectors;

detecting dynamic periods when displacement per unit time of the specific part is larger than a displacement threshold value, based on the time-series velocity vectors;

acquiring a representative velocity vector from velocity vectors in each dynamic period;

storing each representative velocity vector corresponding to each dynamic period in a storage unit;

calculating a first intersection angle of two representative velocity vectors corresponding to two successive dynamic periods stored in the storage unit; and

detecting that a motion of the specific part is a reciprocating motion when the first intersection angle is larger than a first threshold value.

14. The computer readable storage medium of claim 13, the method further comprising:

calculating a second intersection angle of two representative velocity vectors stored in the storage medium, the two representative velocity vectors corresponding to two dynamic periods every two dynamic periods; and

detecting that the motion of the specific part is a repetitive motion, which is a motion repeating the reciprocating motion, when the first intersection angle is larger than the first threshold value and the second intersection angle is smaller than a second threshold value.

15. A gesture detecting apparatus, comprising:

a first calculating unit configured to calculating a velocity vector of a specific part of an object gesturing, from time series data of coordinates of the specific part, to obtain time-series velocity vectors;

a first detecting unit configured to detect dynamic periods when displacement per unit time of the specific part is larger than a displacement threshold value, based on the time-series velocity vectors;

an acquiring unit configured to acquire a representative velocity vector from velocity vectors in each dynamic period;

a storage unit to store each representative velocity vector corresponding to each dynamic period;

a second calculating unit configured to calculate a first intersection angle of two representative velocity vectors corresponding to two successive dynamic periods; and

a second detecting unit configured to detect that a motion of the specific part is a reciprocating motion when the first intersection angle is larger than a first threshold value wherein the above units are implemented in hardware or hardware/software combination.

16. The apparatus according to claim 15, further comprising:

a third calculating unit configured to calculate a second intersection angle of two representative velocity vectors corresponding to two dynamic periods every two dynamic periods;

a third detecting unit configured to detect that the motion of the specific part is a repetitive motion, which is a motion repeating the reciprocating motion, when the first intersection angle is larger than the first threshold value and the second intersection angle is smaller than a second threshold value.