TW201419050A - Gesture recognition system and gesture recognition method - Google Patents

Abstract

A gesture recognition system is provided. The gesture recognition system includes an image capturing module, an image analyzing module, a judging module and a determining module. The image capturing module is to obtain an image of a plane before an electronic device captured by a camera from the camera. The camera and a light source are respectively arranged on different side of the electronic device. The light source is to project lights to the plane. The image analyzing module is to determine whether there are two objects including similar character in the obtained image. The judging module is to determine whether the position of any one point of one object is the same as the position of one point of the other object when there are two objects including the similar character in the obtained image. The determining module is to determine that the plane is touched when the position of any one point of one object is the same as the position of one point of the other object, and determine that the same position in the two objects is the touched position.

Description

Gesture recognition system and gesture recognition method

The present invention relates to a gesture recognition system and a gesture recognition method, and more particularly to a gesture recognition system and a gesture recognition method capable of determining whether a plane is touched.

As shown in FIG. 1, the current gesture recognition is to capture the motion of the object 200 on a plane 300 by a camera 100, and determine the symbol on the touched plane 300 according to the image, and perform the corresponding operation of the symbol. . However, current gesture recognition determines the touched symbol by the sign of the plane 300 corresponding to the position at which the object 200 is staying, while the object 200 may not actually touch the symbol of the plane 300 at this time, resulting in misjudgment.

In view of the above, it is necessary to provide a gesture recognition system and a gesture recognition method for determining whether an object has touched a plane.

A gesture recognition system, comprising: an image acquisition module, configured to acquire, from an image capturing device, an image of a plane in front of an electronic device captured by the camera device, wherein the camera device and the camera device The light source of the planar projection light is respectively disposed on different sides of the electronic device; an image analysis module is configured to determine whether two objects having similar feature points exist in the acquired image; and a determining module is configured to When there are two objects having similar feature points in the acquired image, determining whether the two objects in the acquired image have the same position; and a determining module for when the two objects have the same position At the point of time, it is determined that the plane in front of the electronic device is touched and the point at which the same position is determined is the touched position.

A gesture recognition method, comprising: acquiring, from an image capturing device, an image of a plane in front of an electronic device captured by the camera device, wherein the camera device and the light source that projects light to the plane are respectively disposed on Different sides of the electronic device; determining whether there are two objects having similar feature points in the acquired image; and when there are two objects having similar feature points in the acquired image, determining the acquired image Whether the two objects have points of the same position; and when there are points of the same position among the two objects, it is determined that the plane in front of the electronic device is touched and the point of the same position is determined to be the touched position.

According to the present invention, an image of a plane in front of an electronic device captured by the camera device is acquired from an image pickup device, and two objects having similar feature points are present in the acquired image, and the two objects are When there is a point of the same position, it is determined that the plane in front of the electronic device is touched, so that it can be determined whether the object has touched the plane.

Please refer to FIG. 2, which is a block diagram of a gesture recognition system 10 according to an embodiment of the present invention. In the present embodiment, the gesture recognition system 10 is installed and operated in an electronic device 1. The electronic device 1 is connected to a light source 2 and an imaging device 3. The gesture recognition system 10 acquires an image of a plane 4 in front of the electronic device 1 taken by the camera device 3, determines a touched symbol based on the image captured by the camera device 3, and performs a corresponding operation. The electronic device 1 further includes a central processing unit (CPU) 20 and a memory 30.

In the first embodiment of the present invention, the light source 2 is only used to generate light, and the plane 4 in front of the electronic device 1 is a plane 4 on which a physical keyboard is placed, wherein the physical keyboard can be a physical keyboard or a keyboard drawn on paper. Etc., and can be placed on a tabletop or a wall plane. In the second embodiment of the present invention, the plane 4 is a projection plane, and the light source 2 is a projection device for emitting projection light onto the plane 4 in front of the electronic device 1 and projecting on the plane 4. Generate a virtual keyboard. When the light generated by the light source 2 is blocked by the object in front of the plane 4, a shadow is formed on the plane 4 in front of the electronic device 1, the shadow having a similar shape to the corresponding object, only the shadow and the corresponding object The size is different, that is, the shadow has similar feature points to the corresponding object.

The imaging device 3 is for taking an image of the plane 4 on the front side of the electronic device 1. The light source 2 and the imaging device 3 are respectively disposed on different sides of the electronic device 1, so that the shadow formed by the object and the object can be taken by the camera device 3, and when the object contacts the plane 4, In the image taken by the imaging device 3, the object has a point at the same position as the shadow of the object on the plane 4; and when the object does not touch the plane 4, the object captured by the imaging device 3 A point that does not have the same position as the shadow of the object on plane 4. As shown in FIG. 3, when the object X contacts the plane 4, the end a of the object X contacting the plane 4 has the same position as the shadow b corresponding to the end a. Wherein the position is represented by a two-dimensional coordinate on the image, and the object is in the same coordinate system as the shadow of the object, and when the object has a point with the same coordinate as the shadow of the object, determining the object and the object The shadow has the same position as the point. In the present embodiment, the light source 2 and the imaging device 3 are respectively disposed on the left and right sides or the right and left sides of the top end of the electronic device 1. In other embodiments, the light source 2 and the imaging device 3 are respectively disposed on the upper half of the left and right sides or the upper half of the right and left sides of the electronic device 1.

In the embodiment, the gesture recognition system 10 includes an image acquisition module 11 , an image analysis module 12 , a determination module 13 , and a determination module 14 . The module referred to in the present invention refers to a series of computer blocks that can be executed by the central processing unit 20 of the electronic device 1 and that can perform specific functions, and are stored in the memory 30 of the electronic device 1. The memory 30 also stores a plurality of character templates, wherein the character templates can be 1, 2, ..., A, B, ..., F1, F2, ..., Enter, and the like.

The image acquisition module 11 is configured to acquire an image captured by the imaging device 3 from the imaging device 3.

The image analysis module 12 is configured to determine whether two objects having similar feature points exist in the acquired image.

When there are two objects having similar feature points in the acquired image, the determination module 13 determines whether the two objects have points of the same position.

When the two objects have points of the same position, the determining module 14 determines that the plane 4 in front of the electronic device 1 is touched and determines that the point of the same position is the touched position.

In the embodiment, the gesture recognition system 10 further includes an execution module 15 . The determining module 14 further determines image content corresponding to the touched position in the acquired image, compares the determined image content with the stored plurality of character templates, and determines the determined image. Whether the content matches a character template. The execution module 15 is configured to determine, when the determining module 14 determines that the image content matches a character template, the image content is a character corresponding to the character template, and execute the determined character Operation. The matching of the determined image content with the character template means that the similarity between the determined image content and the character template is greater than a predetermined value. For example, when the execution module 15 determines that the image content matches the character template of a blank key, the image content is determined to be the blank key, and the function corresponding to the blank key is executed.

Please refer to FIG. 4 , which is a flowchart of a gesture recognition method according to an embodiment of the present invention.

In step S401, the image acquisition module 11 acquires an image taken by the imaging device 3 from the imaging device 3.

In step S402, the image analysis module 12 determines whether there are two objects having similar feature points in the acquired image. When there are two objects having similar feature points in the acquired image, step S403 is performed. When there are two objects having similar feature points in the acquired image, step S401 is performed.

In step S403, the determination module 13 determines whether the two objects have points of the same position. When the two objects have points of the same position, step S404 is performed. When the two objects do not have the same position, the process returns to step S401.

In step S404, the determining module 14 determines that the plane 4 in front of the electronic device 1 is touched and determines that the point of the same position is the touched position.

In this embodiment, after step S404, the gesture recognition method further includes the steps of:

The determining module 14 determines image content corresponding to the touched position in the acquired image, compares the determined image content with the stored plurality of character templates, and determines the determined image content. Whether it matches a character template. When the determining module 14 determines that the determined image content matches a character template, the execution module 15 determines that the image content is a character corresponding to the character template, and executes the determined symbol. Operation. Wherein, when the determined image content is different from any of the character templates, step S401 is continued.

It is to be understood by those skilled in the art that the present invention may be made in accordance with the present invention.

10. . . Gesture recognition system

1. . . Electronic device

2. . . light source

3. . . Camera

4. . . flat

20. . . CPU

30. . . Memory

11. . . Image acquisition module

12. . . Image analysis module

13. . . Judging module

14. . . Determine module

15. . . Execution module

FIG. 1 is a schematic diagram of a conventional gesture recognition recognition of a touched symbol.

2 is a block schematic diagram of a gesture recognition system in accordance with an embodiment of the present invention.

FIG. 3 is a schematic diagram of the recognition of the gesture recognition system shown in FIG. 1.

FIG. 4 is a flowchart of a gesture recognition method according to an embodiment of the present invention.

1. . . Electronic device

2. . . light source

3. . . Camera

10. . . Gesture recognition system

11. . . Image acquisition module

12. . . Image analysis module

13. . . Judging module

14. . . Determine module

15. . . Execution module

20. . . CPU

30. . . Memory

Claims (8)

A gesture recognition system, the gesture recognition system comprising:
An image acquisition module is configured to acquire, from an image capturing device, an image of a plane in front of an electronic device captured by the camera device, wherein the camera device and the light source that projects light to the plane are respectively disposed on the electronic device Different sides
An image analysis module is configured to determine whether two objects having similar feature points exist in the acquired image;
a determining module, configured to determine, when the two objects having similar feature points exist in the acquired image, whether the two objects in the acquired image have the same position; and a determining module, When the two objects have points of the same position, it is determined that the plane in front of the electronic device is touched and the point at which the same position is determined is the touched position. The gesture recognition system of claim 1, wherein the plane is a plane on which a physical keyboard is placed or a projection plane on which a virtual keyboard is projected. The gesture recognition system of claim 2, wherein the gesture recognition system further includes an execution module, wherein the determination module is further configured to determine an image corresponding to the touch position in the acquired image. And comparing the determined image content with a stored plurality of character templates to determine whether the determined image content matches a stored character template, the execution module is configured to determine in the determining module When the image content matches a stored character template, the image content is determined to be a character corresponding to the character template, and an operation corresponding to the determined character is performed. The gesture recognition system of claim 1, wherein the location is represented by a two-dimensional coordinate on the image, and the two objects in the acquired image are in the same coordinate system. A gesture recognition method includes:
Obtaining, from an image capturing device, an image of a plane in front of an electronic device captured by the camera device, wherein the camera device and a light source that projects light to the plane are respectively disposed on different sides of the electronic device;
Determining whether there are two objects having similar feature points in the acquired image;
When there are two objects having similar feature points in the acquired image, determining whether the two objects in the acquired image have points of the same position; and when the two objects have points of the same position, determining The plane in front of the electronic device is touched and the point at the same location is determined to be the touched location. The gesture recognition method of claim 5, wherein the plane is a plane on which a physical keyboard is placed or a projection surface on which a virtual keyboard is projected. The gesture recognition method of claim 6, wherein the gesture recognition method further comprises:
Determining image content corresponding to the touch position in the acquired image, comparing the determined image content with a stored plurality of character templates, and determining whether the determined image content is related to a stored word Meta-class matching; and when the determined image content matches a stored character template, determining that the image content is a character corresponding to the character template, and performing an operation corresponding to the determined symbol. The gesture recognition method of claim 5, wherein the position is represented by a two-dimensional coordinate on the image, and the two objects in the acquired image are in the same coordinate system.