SYSTEM AND METHOD FOR EFFECTIVE SECTION DETECTING OF HAND GESTURE

• US 20140169626A1
• Filed: 12/10/2013
• Published: 06/19/2014
• Est. Priority Date: 12/18/2012
• Status: Active Grant

• Current Assignee: Hyundai Motor Company (Hyundai Motor Group)
• Original Assignee: Hyundai Motor Company (Hyundai Motor Group)
• Inventors: Kim, Sung Un, Joo, Sihyun
• Granted Patent: US 9,171,223 B2
• Time in Patent Office: Days
• US Class Current: 382/103
• CPC Class Codes: G06V 40/20 Movements or behaviour, e.g. gesture recognition recognition of facial expressions G06V40/16

Abstract

A system is provided for detecting an effective section of a gesture by recognizing the gesture, pose information and motion information included in the gesture from an acquired image. In addition, a controller determines whether a pose has been recognized based on the pose information and when the pose has been recognized, an effective section is detected based on a start point and an end point of the pose. Further, when the effective section for the pose is detected, the gesture is recognized based on the motion information.

Claims

1. A method for detecting an effective section of a gesture, the method comprising:

acquiring, by a controller, an image that includes the gesture and a background;

extracting, by the controller, pose information and motion information included in the gesture as features;

determining, by the controller, whether a pose has been recognized based on the pose information;

in response to determining that a pose has been recognized, detecting, by the controller, the effective section of the gesture based on a start point and an end point of the pose; and

in response to detecting the effective section of the gesture, recognizing, by the controller, the gesture based on the motion information.

2. The method of claim 1, wherein the pose information includes information regarding the pose of a fist with a thumb or a forefinger stretched out, a palm open upside, or clenched fingers.

3. The method of claim 1, ,wherein the motion information includes information regarding various motions including moving a hand horizontally or vertically, rotating the hand, and waving the hand.

4. The method of claim 1, wherein the determining of whether the pose has been recognized includes:

determining, by the controller, stored feature information that corresponds to the extracted pose information; and

in response to determining corresponding stored feature information, recognizing, by the controller, the extracted pose information as a pose.

5. The method of claim 1, wherein the detecting of the effective section of the gesture includes:

detecting, by the controller, the start point of the pose;

detecting, by the controller, the end point of the pose;

determining, by the controller, whether the section from the start point to the end point continues for a predetermined time or more; and

in response to determining that the section continues for the predetermined time or more, determining, by the controller, the section between the start point to the end point as the effective section.

6. The method of claim 1, wherein the recognizing of the gesture includes:

determining, by the controller, stored feature information that corresponds to the extracted motion information; and

in response to determining corresponding stored feature information, outputting, by the controller, the extracted motion information as a gesture recognition result.

7. The method of claim 1, wherein the extracting of pose information and motion information as features includes:

extracting, by the controller, the pose as pose information using a template matching algorithm or extracting and then modeling the features of a pose image, and

extracting, by the controller, the motion as motion information by recognizing the motion using signal processing or learning algorithm,

8. The method of claim 1, wherein the acquiring of the image includes:

removing, by the controller, the background from the acquired image.

9. A system that detects an effective section of a hand gesture, the system comprising:

a controller configured to;

acquire an image having a background and the gesture that includes pose information and motion information;

detect the pose information and the motion information from the gesture;

extract the pose information by determining whether information corresponding to the motion information is in stored feature information;

detect a start point of the pose information, in response to detecting the pose information;

detect an end point of the pose information, in response to detecting the pose information; and

detect the effective section for the pose based on the start point and the end point; and

recognize a gesture by determining whether the extracted motion information is included in the stored feature information, in response to detecting the effective section.

10. The system of claim 9, wherein the controller is further configured to:

remove the background from the acquired image; and

store and provide feature information used to detect the pose information and the motion information.

11. The system of claim 9, wherein the pose information includes information regarding the pose of a fist with a thumb or a forefinger stretched out, a palm open upside, or clenched fingers.

12. The system of claim 9, wherein the motion information includes information regarding various motions including moving a hand horizontally or vertically, rotating the hand, and waving the hand.

13. A non-transitory computer readable medium containing program instructions executed by a processor or controller, the computer readable medium comprising:

program instructions that acquire an image having a background and the gesture that includes pose information and motion information;

program instructions that detect the pose information and the motion information from the gesture;

program instructions that extract the pose information by determining whether information corresponding to the motion information is in stored feature information;

program instructions that detect a start point of the pose information, in response to detecting the pose information;

program instructions that detect an end point of the pose information, in response to detecting the pose information; and

program instructions that detect the effective section for the pose based on the start point and the end point; and

program instructions that recognize a gesture by determining whether the extracted motion information is included in the stored feature information, in response to detecting the effective section.

14. The non-transitory computer readable medium of claim 13, further comprising:

program instructions that remove the background from the acquired image; and

program instructions that store and provide feature information used to detect the pose information and the motion information.