APPARATUS AND METHOD FOR RECOGNIZING HAND SHAPE USING FINGER PATTERN

• US 20130259384A1
• Filed: 03/27/2013
• Published: 10/03/2013
• Est. Priority Date: 03/28/2012
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Kwon, Mu-Sik, Kim, Moo-Rim
• Granted Patent: US 9,443,138 B2
• Time in Patent Office: Days
• US Class Current: 382/199
• CPC Class Codes: G06V 40/107 Static hand or arm

Abstract

An apparatus and a method for recognizing a hand shape by using a finger pattern are disclosed. The throughput of data used to recognize a hand shape is reduced and the hand shape can be quickly recognized by setting in advance a finger pattern in accordance with one or more hand shapes, calculating edges for each pixel line of an input image, detecting finger feature data by using a preset feature reference value, detecting a finger candidate pattern by using the detected finger feature data, and recognizing a hand shape corresponding to a finger pattern that matches the detected finger candidate pattern.

Claims

1. An apparatus for recognizing a hand shape by using a finger pattern, the apparatus comprising:

a memory for storing a preset finger pattern according to each hand shape in accordance with one or more hand shapes; and

a controller for calculating edges for each pixel line within an input image, detecting finger feature data according to a preset feature reference value by using the calculated edges, detecting a finger candidate pattern by using the detected finger feature data, and outputting a hand shape corresponding to a finger pattern that matches the detected finger candidate pattern among finger patterns stored in the memory as a recognition result.

2. The apparatus of claim 1, wherein the preset feature reference value includes at least one of a length of a pixel group between the calculated edges, an average pixel value of the pixel group, a ratio of lengths between adjacent pixel groups, and a direction of the calculated edges.

3. The apparatus of claim 2, wherein the controller detects pixel groups having similar lengths and similar average pixel values among pixel groups between the calculated edges as the finger feature data.

4. The apparatus of claim 2, wherein, when the ratio of the lengths between the adjacent pixel groups is equal to a preset threshold ratio, the controller detects the adjacent pixel groups as the finger feature data.

5. The apparatus of claim 2, wherein the controller determines pixel groups having similar lengths, similar average pixel values, and the same direction of the calculated edges among the pixel groups between the calculated edges as the finger feature data.

6. The apparatus of claim 1, wherein, when there is a slope difference between a direction of the detected finger feature data and a vertical direction of the input image by comparing the direction of the detected finger feature data with the vertical direction of the input image, the controller determines that a hand shape is rotated.

7. A method of recognizing a hand shape by using a finger pattern in a hand shape recognition apparatus, the method comprising:

storing a preset finger pattern according to each hand shape in accordance with one or more hand shapes;

calculating edges for respective pixel lines within an input image;

detecting finger feature data according to a preset feature reference value by using the calculated edges;

detecting a finger candidate pattern by using the detected finger feature data; and

outputting a hand shape corresponding to a finger pattern that matches the detected finger candidate pattern among finger patterns stored in a memory as a recognition result.

8. The method of claim 7, wherein the preset feature reference value includes at least one of a length of a pixel group between the calculated edges, an average pixel value of the pixel group, a ratio of lengths between adjacent pixel groups, and a direction of the calculated edges.

9. The method of claim 8, wherein detecting the finger feature data comprises detecting pixel groups having similar lengths and similar average pixel values among pixel groups between the calculated edges as the finger feature data.

10. The method of claim 8, wherein detecting the finger feature data comprises detecting the adjacent pixel groups as the finger feature data when the ratio of the lengths between the adjacent pixel groups is equal to a preset threshold ratio.

11. The method of claim 8, wherein detecting the finger feature data comprises determining pixel groups having similar lengths, similar average pixel values, and an equal direction of the calculated edges among the pixel groups between the calculated edges as the finger feature data.

12. The method of claim 7, further comprising:

comparing a direction of the detected finger feature data with a vertical direction of the input image; and

when there is a slope difference between the direction of the detected finger feature data and the vertical direction of the input image, determining that a hand shape is rotated.