Apparatus and method for recognizing hand rotation

• US 8,824,732 B2
• Filed: 01/24/2012
• Issued: 09/02/2014
• Est. Priority Date: 01/24/2011
• Status: Expired due to Fees

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Grzesiak, Grzegorz Pawel
• Primary Examiner(s): Abdi, Amara
• Application Number:
  US13/357,092
• Publication Number:
  US 20120189163A1
• Time in Patent Office: 952 Days
• US Class Current: 382/103
• CPC Class Codes: G06V 40/28 Recognition of hand or arm movements, e.g. recognition of deaf sign language static hand signs G06V40/113

Abstract

An apparatus and a method are provided that can intuitively and easily recognize hand rotation. The apparatus for recognizing a hand rotation includes a camera for photographing a plurality of hand image data, a detector for extracting circles through fingers of the hand image data and a controller for recognizing hand rotation through changes in positions and sizes of the circles extracted from each of the plurality of hand image data.

Claims

1. An apparatus for recognizing hand rotation, the apparatus comprising:

a camera for photographing a plurality of hand image data;

a detector for extracting circles from fingers of the hand image data; and

a controller for recognizing the hand rotation through changes in positions and sizes of the circles extracted from each of the plurality of hand image data.

2. The apparatus as claimed in claim 1, wherein the fingers from which the circles are extracted are in a bent state.

3. The apparatus as claimed in claim 1, wherein the plurality of hand image data photographed through the camera comprise sequential hand rotation movements.

4. The apparatus as claimed in claim 1, wherein the detector extracts edges of the hand image data, extracts circles from the edges of the fingers of the hand image data, and detects center values and radius values of the extracted circles.

5. The apparatus as claimed in claim 1, wherein the detector extracts circles from the fingers by using a Hough transform method.

6. The apparatus as claimed in claim 1, wherein the controller receives center values and radius values of the circles for each of the plurality of hand image data from the detector, detects position changes of the circles from a center of a hand through the center values of the circles, detects size changes of the circles through the radius values of the circles, and recognizes the hand rotation when the detected position changes and size changes of the circles correspond to patterns enabling the hand rotation to be recognized.

7. The apparatus as claimed in claim 6, wherein, when the patterns in which positions of the circles are moved to an outside of the center of the hand image and sizes of the circles become smaller are generated, the controller recognizes the hand rotation.

8. The apparatus as claimed in claim 7, wherein Equation 1 the sizes of the circles are determined as becoming smaller when the hand rotates, according to:

9. A method of recognizing a hand rotation, the method comprising:

extracting circles from fingers of a plurality of hand image data when the plurality of hand image data are received; and

recognizing the hand rotation through changes in positions and sizes of the circles detected from each of the plurality of hand image data.

10. The method claimed in claim 9, wherein the fingers from which the circles are extracted are in a bent state.

11. The method claimed in claim 9, wherein the plurality of hand image data photographed through a camera comprise sequential hand rotation movements.

12. The method as claimed in claim 9, wherein extracting the circles comprises:

extracting edges for the plurality of hand image data;

extracting circles from the edges of the fingers of the plurality of hand image data; and

detecting center values and radius values of the extracted circles.

13. The method as claimed in claim 9, wherein circles are extracted from the fingers by using a Hough transform method.

14. The method as claimed in claim 9, wherein recognizing the hand rotation comprises:

detecting position changes of the circles from a hand through center values of the extracted circles;

detecting size changes of the circles through radius values of the extracted circles; and

recognizing the hand rotation when the position changes and the size changes of the circles for the plurality of hand image data correspond to patterns for a hand rotation recognition.

15. The method as claimed in claim 14, wherein, the hand rotation is recognized when patterns in which positions of the circles are moved to an outside of a center of the hand image and sizes of the circles become smaller are generated.

16. The method as claimed in claim 15, wherein the sizes of the circles are determined as becoming smaller when the hand rotates, according to: