Gesture recognition method and interactive system using the same

• US 8,867,791 B2
• Filed: 02/14/2014
• Issued: 10/21/2014
• Est. Priority Date: 09/02/2009
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: PixArt Imaging Incorporated
• Inventors: Su, Tzung-Min, Chen, Hsin-Chia
• Primary Examiner(s): Carter, Aaron W
• Application Number:
  US14/181,055
• Publication Number:
  US 20140160053A1
• Time in Patent Office: 249 Days
• Field of Search: 382/100, 382/103, 345156-158, 345173-175, 715/863
• US Class Current: 382/103
• CPC Class Codes: G06F 3/017 Gesture based interaction, e.g. based on a set of recognized hand gestures interaction based on gestures traced on a digitiser G06F3/04883
• CPC Class Codes: G06F 3/042 by opto-electronic means
• CPC Class Codes: G06F 3/04883 for inputting data by handwriting, e.g. gesture or text

Abstract

A gesture recognition method for an interactive system includes the steps of: capturing image windows with an image sensor; obtaining information of object images associated with at least one pointer in the image windows; calculating a position coordinate of the pointer relative to the interactive system according to the position of the object images in the image windows when a single pointer is identified according to the information of object images; and performing gesture recognition according to a relation between the object images in the image window when a plurality of pointers are identified according to the information of object images. The present invention further provides an interactive system.

Claims

1. A gesture recognition method for an interactive system, the interactive system comprising at least one light source, a reflector, and an image sensor configured to capture an image window comprising object images of at least one pointer caused by the pointer occluding light emitted from the at least one light source or the reflector, the gesture recognition method comprising:

capturing the image window with the image sensor;

obtaining information of the object images in the image window, wherein the information of the object images comprises an average number of the object images, an average distance between the object images, and a greatest distance between the object images;

determining whether there is a plurality of pointers according to the information of the object images; and

executing, when it is determined that there is a plurality of pointers, a gesture recognition according to a relation between the object images in successive image windows.

2. The gesture recognition method as claimed in claim 1, further comprising:

calculating, when it is determined that there is a single pointer, position coordinates of the pointer relative to the interactive system according to a position of the object image in the image window; and

controlling a motion of a cursor on an image display according to a change of the position coordinates of the pointer in the successive image windows.

3. The gesture recognition method as claimed in claim 1, wherein the executing comprises:

comparing the average number of the object images and the average distance between the object images with a threshold value;

executing a gesture recognition of up, down, left, right, zoom in, or zoom out when the average number of the object images and the average distance between the object images are not greater than the threshold value;

executing a gesture recognition of rotation when the average number of the object images or the average distance between the object images is greater than the threshold value; and

refreshing a display of the image display according to the recognized gesture.

4. The gesture recognition method as claimed in claim 3, wherein the refreshing includes any one selected from the group consisting of scrolling a screen, scaling an object, rotating an object, switching a display, and showing a menu.

5. The gesture recognition method as claimed in claim 3, wherein when the average number of the object images and the average distance between the object images are not greater than the threshold value, the gesture recognition method further comprises:

dividing the object images into two groups.

6. The gesture recognition method as claimed in claim 5, wherein the object images are divided into a real-image group and an imaginary-image group.

7. The gesture recognition method as claimed in claim 5, further comprising:

executing a gesture recognition of left or right when both the two groups of the object images move in the same direction in the image window;

executing a gesture recognition of up or down when the average distance between the two groups of the object images changes; and

executing a gesture recognition of zoom in or zoom out when the greatest distance between the object images in the image window changes.

8. The gesture recognition method as claimed in claim 3, wherein the executing a gesture recognition of rotation comprises:

identifying a direction along which a majority of the object images move to recognize the orientation of rotation.

9. The gesture recognition method as claimed in claim 1, wherein the executing comprises recognizing up, down, left, right, zoom in, or zoom out.

10. The gesture recognition method as claimed in claim 1, wherein the executing comprises recognizing a rotation.

11. A gesture recognition method for an interactive system, the interactive system comprising a light-emitting unit and an image sensor configured to capture an image window comprising object images of a plurality of pointers caused by the pointers occluding light emitting from the light-emitting unit, the gesture recognition method comprising:

capturing image windows with the image sensor; and

executing a gesture recognition according to relations between a plurality of object images in successive image windows,

wherein the relations between the object images comprise a change of an average distance between the object images, a change of a greatest distance between the object images, and a direction of displacement.

12. The gesture recognition method as claimed in claim 11, wherein

the light-emitting unit is a passive light source having a reflecting surface,

the image windows captured by the image sensor comprise real object images and imaginary object images, and

the relations between the object images comprise a change of an average distance between the real and imaginary object images, a change of a greatest distance between the real and imaginary object images, and a direction of displacement.

13. The gesture recognition method as claimed in claim 12, further comprising:

dividing the object images into a real-image group and an imaginary-image group.

14. The gesture recognition method as claimed in claim 11, further comprising:

refreshing a display of an image display according to the recognized gesture.