KR100916836B1 - Method and apparatus for recognizing pointing gesture - Google Patents

Abstract

A method and an apparatus for recognizing pointing gesture are provided to recognize pointing gesture of user in consideration of a structural feature of a human body. A human body detecting unit(130) is the image inputted through the image data input unit(120), anatomy is detected. The human body detecting unit extracts location information of the human body element. A hand location correcting unit(140) amends the location information of the hand. A pointing gesture detecting unit(150) detects a generation of a pointing gesture. A pointing gesture recognizing unit(160) calculates the pointing direction of the pointing gesture based on the location information of the pointing gesture recognition apparatus and location information of the corrected hand as described above. The location information of the control object stored with the calculated pointing direction as described above in advance is used. The pointing gesture recognizing unit recognizes clearly the function in which the detected pointing gesture is intended.

Description

METHOD AND APPARATUS FOR RECOGNIZING POINTING GESTURE}

The present invention relates to a method and apparatus for recognizing a directed gesture, and more particularly, to a method and apparatus for detecting a gesture of a person in a mobile device equipped with a stereo camera and recognizing a direction of the gesture.

The most common user interface for controlling a computer is using a keyboard and mouse, and directed gestures and click actions as the WIMP (Windows, Icons, Menus and Pointing devices) paradigm has become the mainstream for modern graphical user interfaces (GUIs). It is possible to fully control the computer alone.

In particular, due to the development of a mobile robot, a directed gesture technology in a dynamic environment is required in various application fields. A directed gesture is a user's expression of intention through an operation of pointing a direction using a hand, an object or a direction. It provides an interface to point to or move the mouse on the display screen.

Among them, research has been actively conducted in the fields of human computer interaction (HCI), human robot interaction (HRI), smart home, and the like, regarding a technology of recognizing a gesture of a user using a camera.

HCI is a study on the natural interaction between humans and computers. Speech Recognition technology, Speech Synthesis or Text-to-Speech technology that attempts to connect humans and computers through voice, Natural Language Processing technology, which makes it possible to handle everyday language or pseudo.

HRI is a technology that forms an interaction and communication link between two objects, human and robot, through various communication channels, and includes voice recognition technology, gesture recognition technology, tactile / force recognition and emotion recognition technology. It can be divided into recognition technology for recognizing the intention of human beings, interface technology for remote operation for controlling the operation of robot located at a remote place, and cognitive and emotional interaction technology for implicit and interactive communication between human and robot. Can

As described above, researches for recognizing directed gestures have also been conducted in the related art, but have the following problems in common.

First of all, it is assumed that the camera is fixed, so it is difficult to use it as a technical problem in a mobile mobile device environment or a human moving environment. There are possible drawbacks.

Secondly, the structural characteristics of the human body when performing the directed gesture are not considered.

In order to obtain the direction indicating the best intention of the user, it is essential to calculate the structural characteristics of the human body when performing the directed gesture. Difficult to identify and likely to cause recognition errors.

Third, although various functions can be performed by using an indication gesture, the related arts of the related art have a disadvantage in that application fields are limited because most of them can perform only limited functions.

Accordingly, Korean Patent Publication No. 2007-16794 (system and method for gesture-based control system) and Korean Patent Application Publication No. 2002-25198 (targeting method and device, image processor, and target and target point identification method) Although a technique for accurately recognizing a user's gesture is disclosed, the disadvantages of the related arts described above are still not solved, and in particular, there is no consideration or lack of dynamic environment and three-dimensional space. It does not provide a solution to the problem of gesture detection that accurately detects the error.

An object of the present invention is to accurately recognize a user's directed gesture in consideration of structural features of a human body even in an environment in which the user is movable.

Another object of the present invention is to further expand the range of use of the directed gesture that was available only in limited places by accurately recognizing the directed gesture of the user even in a mobile environment.

Still another object of the present invention is to accurately recognize a user's directed gesture, thereby accurately performing a control function corresponding to the directed gesture.

The objects of the present invention are not limited to the above-mentioned objects, and other objects that are not mentioned will be clearly understood from the following description.

In order to achieve the above object, the directed gesture recognition method of a mobile device equipped with a stereo camera according to an aspect of the present invention, (a) receiving an image captured by the stereo camera, (b) the input Detecting a human body region from the detected image, and extracting position information of a human component required for indicating a gesture gesture from the detected human body region, and (c) obtaining position information of a hand among the extracted position information of the human component. Correcting, (d) detecting whether the directed gesture is generated through the corrected position information of the hand, and (e) when the directed gesture occurs, the position information of the mobile device and the corrected hand. Calculating an indication direction of the directed gesture based on the position information of (f) positioning of the mobile device and positioning of a control object previously stored; Projecting a beam to a three-dimensional space model to calculate a pointed position according to the pointing direction; (g) analyzing a function intended by the detected directed gesture when there is a control object at the pointed position; h) controlling the control object to execute a function intended by the directed gesture.

In order to achieve the above object, a moveable directed gesture recognition apparatus equipped with a stereo camera according to another aspect of the present invention, the image data input unit for receiving data for the image captured by the stereo camera, the input image A human body detecting unit for detecting a human body region and extracting position information of a human component required for the recognition of the directive gesture in the detected human body region, and a hand for correcting the position information of the hand among the extracted position information of the human component A position correcting unit, an indication gesture detecting unit detecting whether the indication gesture is generated based on the position information of the corrected hand, and when the indication gesture occurs, the position information of the indication gesture recognition device and the corrected hand Calculate an instruction direction of the directed gesture based on the positional information of the; The control object performs the function on the basis of the recognized gesture and the indication gesture recognition unit for recognizing a function intended by the detected indication gesture by using the indication direction and position information of the control object stored in advance. It includes a control unit for controlling to.

In order to achieve the above object, a recording medium having recorded thereon a program for recognizing an indication gesture of a mobile device equipped with a stereo camera according to another aspect of the present invention, the program is (a) by the stereo camera; Receiving a photographed image, (b) detecting a human body region from the input image, and extracting position information of a human component required for indicating gesture recognition from the detected human body region, (c) extracting Correcting the position information of the hand among the position information of the human body component; (d) detecting whether the directed gesture is generated through the corrected position information of the hand, (e) generating the directed gesture Calculating a direction of indication of the directed gesture based on the position information of the mobile device and the position information of the corrected hand, (f) Calculating the position of the indication according to the direction of the instruction by projecting the positional information of the mobile device and the positional information of the control object previously stored on the three-dimensional space model, (g) when the control object exists at the position of the indication, Program code for implementing a step of analyzing a function intended by the directed gesture and (h) controlling the control target to execute a function intended by the directed gesture.

Specific details for achieving the above object will be apparent with reference to the embodiments described below in detail with the accompanying drawings.

However, the present invention is not limited to the embodiments disclosed below, but may be configured in different forms, and the present embodiments are intended to complete the disclosure of the present invention and to provide general knowledge in the technical field to which the present invention belongs. It is provided to fully inform those who have the scope of the invention.

According to one of the problem solving means of the method and apparatus for recognizing the directed gesture of the present invention described above, there are one or more of the following effects.

Even in an environment in which the user is movable, there is an advantage of accurately recognizing the user's directed gesture in consideration of structural features of the human body.

In addition, by accurately recognizing the user's directed gesture even in an environment in which the user is movable, there is an advantage in that the range of the place of use of the directed gesture that has been available only in limited places can be further extended.

In addition, by accurately recognizing the user's directed gesture, there is an advantage that the control function corresponding to the directed gesture can be accurately performed.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention.

As those skilled in the art would realize, the described embodiments may be modified in various different ways, all without departing from the spirit or scope of the present invention.

In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and like reference numerals designate like parts throughout the specification.

Throughout the specification, when a part is "connected" to another part, this includes not only "directly connected" but also "electrically connected" with another element in between. .

In addition, when a part is said to "include" a certain component, it means that it may further include other components, except to exclude other components unless specifically stated otherwise.

Hereinafter, with reference to the accompanying configuration diagram or processing flow chart, it will be described in detail for the practice of the present invention.

1 is a block diagram illustrating a configuration of an apparatus for recognizing an indication gesture according to an embodiment of the present invention.

According to an embodiment of the present invention, a movable directed gesture recognition apparatus 100 equipped with a stereo camera includes an image data input unit configured to receive data about an image captured by the stereo camera 110 and the stereo camera 110 ( 120, when it is determined that a person exists in the input image data, the human body detecting unit 130 detects a human body region and extracts position information of a human component necessary for indicating a gesture gesture from the detected human body region. The hand position corrector 140 which corrects the position information of the hand among the position information of the component, determines whether the directed gesture is generated based on the corrected position information, and when the directed gesture occurs, Indicative gesture detection unit 150 to detect, in the case where an indication gesture occurs in the position information of the indication gesture recognition apparatus and the position information of the corrected hand By calculating the direction of indication of the indication gesture detected by the indication gesture detection unit 150 in seconds, and using the calculated indication direction and position information of the control target stored in advance, the detected indication gesture is intended. It may include a directive gesture recognition unit 160 for recognizing the control unit 170 and a control unit 170 for controlling the control target to perform the function based on the recognized function.

1 according to an embodiment of the present invention refers to software or a hardware component such as a field programmable gate array (FPGA) or an application specific integrated circuit (ASIC), and performs predetermined roles.

However, 'components' are not meant to be limited to software or hardware, and each component may be configured to be in an addressable storage medium or may be configured to reproduce one or more processors.

Thus, as an example, a component may include components such as software components, object-oriented software components, class components, and task components, and processes, functions, properties, procedures, and subs. Routines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables.

Components and the functionality provided within those components may be combined into a smaller number of components or further separated into additional components.

For example, the moveable directed gesture recognition apparatus 100 equipped with a stereo camera according to an embodiment of the present invention may move through a predetermined driving device, and track a moving object through the mounted stereo camera. The robot may be photographed.

First, the image data input unit 120 of the apparatus 100 shown in FIG. 1 receives data about an image captured by the stereo camera 110.

In this case, the received data includes a color image and a depth image. In the directed gesture recognition apparatus 100 illustrated in FIG. 1, both the directed gesture recognition apparatus 100 and the user move. In a possible environment, it is preferable to use a stereo camera capable of acquiring a depth image in addition to the color image because the user should recognize that the user indicates the three-dimensional space by using the directed gesture.

Meanwhile, the human body detector 130 detects a human body region from an image input through the image data input unit 120, and extracts a human body component required for directed gesture recognition from the detected human body region.

To this end, the human body detector 130 analyzes whether a person exists in the input image by using the depth image, and when it is determined that the person exists, detects the human body region from the depth image, and indicates the gesture in the detected human region. Detect human body components including face, hands, shoulders, etc. required for recognition.

FIG. 2 is a block diagram illustrating a detailed configuration of the human body detecting unit 130. The human body identifying unit 131 which identifies the presence or absence of a person using a color image and a depth image input from the image data input unit 120, and a person. If it is determined to exist, the human body region detection unit 132 for detecting the human body region, the human body component extraction unit 133 for identifying the human body components necessary for the indication gesture recognition, and extracts the position of the human body components; Include.

The human body identification unit 131 identifies whether or not a person is using a variety of geometric information about the object of the human size in the depth image, which is irrelevant to the change in lighting and background even in an environment where the user can move by using the depth image There is an advantage that can be detected reliably.

For example, as geometric information used to identify the presence of a person, information such as an aspect ratio of an area, a location where an area exists, and an actual size of the area in a three-dimensional space may be used. 131 identifies whether the person is present using one or more of the above-described information.

The human body detection unit 132 detects a human body area when the human body identification unit 131 determines that a person exists, and preprocesses using a morphology operation to detect the detected human body area as one combined area. Then, the detected human body region is detected as one combined region using a region combining algorithm.

The human body component extractor 133 extracts positions of human components such as hands, face, and shoulders, which are required for the directed gesture recognition, in the human body region detected by the human body region detector 132.

On the other hand, the hand position correction unit 140 calculates the position information of the human component extracted by the human body detection unit 130, and corrects the position information of the hand that is most important for recognizing the directed gesture of the calculated position information.

The reason for correcting the position information of the hand is to recognize the direction of the directed gesture more accurately. For this purpose, the hand position corrector 140 uses the structural characteristics of the human body when performing the directed gesture. Correct the position information.

The directed direction of the directed gesture uses a method of recognizing the center point and the fingertips of both eyes, or the shoulder and the fingertip. The conventional pointing direction recognition methods cannot recognize the correct pointing direction at a certain angle. In addition, when the user performs the directed gesture, the direction indicated by the user is different from the direction indicated by the system.

3 is a block diagram showing a detailed configuration of the hand position correction unit 140 according to an embodiment of the present invention. Hereinafter, the configuration and operation of the hand position correction unit 140 will be described in detail with reference to FIG. 3.

The hand position correction unit 140 includes a training unit 141, a correction unit 142, and a training data DB 143, and the result corrected by the correction unit 142 is again input to the training unit 141 to correct the correction unit ( 142) can be used to increase the performance.

The training unit 141 may generate a calibration model by using the training data stored in the training data DB 143. The techniques used for the calibration may include SVM (Support Vector Machine), Bayesian Classification, Various classifiers based on probability, such as Hidden Markov Model (HMM) and Neural Network (NN), may be used.

The training data DB 143 stores the position information of the human body components and training data which stores the direction indicated by the current user and the recognized indication direction, and thus can be used to reduce the recognition error of the indication direction.

Through the operation of each of the above-described components, the correction unit 142 performs a function of correcting the position of the hand of the position information of the human body components, through this process the direction error of the indication direction of the directed gesture is reduced, As a result, the direction of indication can be recognized more accurately.

On the other hand, the directed gesture detection unit 150 determines whether the directed gesture occurs through the position information of the human body components and the position information of the corrected hand, and detects the directed gesture.

4 is a block diagram showing a detailed configuration of the directed gesture detector 150 according to an embodiment of the present invention. Hereinafter, the configuration and operation of the directed gesture detector 150 will be described in detail with reference to FIG. 4.

The directed gesture detection unit 150 includes a training unit 151, a detection unit 152, and a training data DB 153, and the results detected by the detection unit 152 are again input to the training unit 151 to detect the detection unit 152. Can be used to increase performance.

The training unit 151 may generate a calibration model for detection by using the training data stored in the training data DB 153. Examples of techniques used for the calibration may include SVM (Support Vector Machine) and BEIJIAN classification technique (Bayesian). Various classifiers based on probabilities, such as Classification, Hidden Markov Model (HMM), and Neural Network (NN), may be used.

The training data DB 153 contains training data necessary for the directed gesture detection and may be used to increase the accuracy of the detector 152.

The detector 152 performs a function of immediately detecting the directed gesture. In this case, the detection may be performed by being divided into an operation detection for detecting whether the directed gesture has occurred and a state detection for detecting in what state the generated indication gesture is present.

The occurrence and stop of the directed gesture are detected through motion detection, and whether the detected directed gesture is indicating the current direction or moving to indicate the direction can be detected through the state detection.

For reference, the state detection is a function necessary to perform all the functions of the various directed gestures, and there is a disadvantage in that the use of the technology is limited because the state detection is not considered in the prior art.

On the other hand, when the directed gesture recognition unit 160 confirms that the directed gesture has occurred through the directed gesture detection unit 150, the position information and the hand position correcting unit 140 of the apparatus 100 shown in FIG. The direction of the directed gesture detected by the directed gesture detector 150 is calculated based on the corrected position information of the hand.

Thereafter, the directed gesture recognition unit 160 three-dimensionally displays the position information of the apparatus 100 shown in FIG. 1 and the position information of a control object (for example, a digital device such as a digital TV or an audio) that is stored in advance. Project to the spatial model to calculate the pointing position with respect to the pointing direction.

Then, by using the instruction position calculated as described above and the position information of the control target stored in advance, it is possible to recognize a function intended by the detected instruction gesture of the user.

The controller 170 controls the control target to perform a preset function based on the directive gesture recognized by the directive gesture recognition unit 160.

To this end, the control unit 170 generates a command that can be actually performed by a control target using a control signal that is previously defined for a function recognized by the directive gesture recognition unit 160. The wireless control signal may be a wireless or wired control signal, and the wireless control signal may control a control target through infrared transmission, Bluetooth, or the like, or may control the device 100 itself shown in FIG. 1.

If the control object does not exist in the instruction position, the controller 170 generates a proper control command and performs a movement function of moving the apparatus 100 shown in FIG. 1 to a specific position.

In addition, in order to perform a control target selection and a function of controlling the control target, it is determined whether the selected control target is an object requiring control, and generates a necessary control command when the control target is required.

For example, when a user's directed gesture indicates a cup, a separate control command is not generated. However, when a user's directed gesture indicates a home appliance such as a digital TV or audio, a control command is generated to operate a corresponding control object.

If the indicated control target is a device having a screen, a virtual mouse function may be performed. To perform the virtual mouse function, a control command is transmitted to a computer currently outputting a screen on a screen to perform an operation like a mouse. You can do that.

5 and 6 are flowcharts illustrating a directed gesture recognition process of a mobile device equipped with a stereo camera according to an embodiment of the present invention.

For example, a mobile device equipped with a stereo camera according to an embodiment of the present invention may be a robot capable of moving through a predetermined driving device and tracking and photographing a moving object through the mounted stereo camera.

For convenience of description, the description will be made with reference to the apparatus shown in FIG. 1.

First, the moving object in the stereo camera 110 is photographed (S501). Here, the moving object may be a user who wants to control the control object through the directed gesture.

After operation S501, the image data input unit 120 receives color image data and depth image data of the photographed object (S502).

Since the device 100 illustrated in FIG. 1 should recognize that the user indicates a three-dimensional space by using a directed gesture, it is preferable to use a stereo camera that can acquire a depth image in addition to the color image.

After step S502, when it is determined that a person is present in the input image, the human body detecting unit 130 detects a human body region and extracts position information of a human body component necessary for indicating a gesture gesture from the detected human body region.

In more detail, the human body identification unit 131 of the human body detection unit 130 identifies the presence or absence of a person using the color image and the depth image of the image input to the image data input unit 120 (S503).

At this time, the depth image identifies whether the person is a person using various geometric information about the object of human size, and by using the depth image, the person can be reliably detected regardless of the lighting change and the background even in an environment where the user can move. There is an advantage to this.

For example, as geometric information used to identify the presence or absence of a person, information such as an aspect ratio of an area, a location where an area exists, and an actual size of the area in a three-dimensional space may be used.

After the step S503, when the human body identification unit 131 determines that a person exists, the human body area detection unit 132 detects the human body area (S504).

In this case, the human body detecting unit 132 performs preprocessing using a morphology operation to detect the detected human body region as one combined region, and then combines the detected human body region using a region combining algorithm. The detected area.

After the step S504, the human body component extractor 133 extracts the positions of the human body components required for the directive gesture recognition in the human body region detected by the human body region detector 132 (S505).

After step S505, the hand position correction unit 140 calculates the position information of the human component extracted by the human body detection unit 130 and corrects the position information of the hand which is most important for recognizing the directed gesture among the calculated position information. (S506).

The reason for correcting the positional information of the hand among the calculated positional information of the human body components is to more accurately recognize the direction of instruction of the directed gesture, and for this purpose, the hand position corrector 140 performs The positional information of the hand is corrected using the structural characteristics of the human body.

After the step S506, the directed gesture detection unit 150 determines whether the directed gesture occurs based on the corrected position information of the hand and detects the directed gesture (S507).

After step S507, that is, when the directed gesture occurs, the directed gesture recognition unit 160 is based on the position information of the device 100 shown in FIG. 1 and the position information of the hand corrected by the hand position correcting unit 140. In operation S508, the direction of the directed gesture detected by the directed gesture detector 150 is calculated.

After step S508, the directed gesture recognition unit 160 projects the position information of the apparatus 100 shown in FIG. 1 and the position information of the surrounding object previously stored in the three-dimensional space model to indicate the position of the instruction direction. Calculate (S509).

After step S509, the directed gesture recognition unit 160 determines whether a control object exists at the indicated position (S510), and if so, recognizes the type of the control object (S511).

After step S511, the directed gesture recognition unit 160 recognizes a function intended by the directed gesture detected by the directed gesture detection unit 150 by using the calculated direction of instruction and position information of a control target previously stored. (S512).

After step S512, the control unit 170 generates a command that can be actually executed by the control target by using a control signal that is previously defined for the function recognized by the directed gesture recognition unit 160. The signal may be a wireless or wired control signal, and the wireless control signal controls an object to be controlled through infrared transmission or Bluetooth, that is, executes an intended function of the recognized directed gesture (S513).

If, as a result of step S510, the control object does not exist in the instruction position, the controller 170 generates a suitable control command and performs a movement function of moving the apparatus 100 shown in FIG. 1 to a specific position (S514). ).

Through the present invention, the user is able to recognize a natural and intuitive directional gesture that enables interaction with a computer, a robot or an intelligent home appliance.

The scope of the present invention is shown by the following claims rather than the above description, and all changes or modifications derived from the meaning and scope of the claims and their equivalents should be construed as being included in the scope of the present invention. do.

1 is a block diagram illustrating a configuration of an apparatus for recognizing an indication gesture according to an embodiment of the present invention.

2 is a block diagram showing a detailed configuration of the human body detecting unit according to an embodiment of the present invention.

3 is a block diagram showing the detailed configuration of the hand position correction unit according to an embodiment of the present invention.

4 is a block diagram showing the detailed configuration of the directed gesture detector according to an embodiment of the present invention.

5 and 6 are flowcharts illustrating a directed gesture recognition process of a mobile device equipped with a stereo camera according to an embodiment of the present invention.

<Explanation of symbols on main parts of the drawings>

110: stereo camera

120: image data input unit

130: human body detection unit

140: hand position correction unit

150: directed gesture detection unit

160: directed gesture recognition unit

170: control unit

Claims (16)

A moveable directed gesture recognition device equipped with a stereo camera, Image data input unit for receiving data about the image taken by the stereo camera, A human body detecting unit detecting a human body region from the input image and extracting position information of a human body component necessary for recognizing an indication gesture from the detected human body region; Hand position correction unit for correcting the position information of the hand of the extracted position information of the human component, An indication gesture detection unit detecting whether the indication gesture is generated through the corrected position information of the hand; When the directed gesture occurs, the direction of the directed gesture is calculated based on the position information of the directed gesture recognition device and the position information of the corrected hand, and the calculated direction and the prestored control. An indication gesture recognition unit for recognizing a function intended by the detected indication gesture using the location information of the object; A controller for controlling the control target to perform the function based on the recognized function Directed gesture recognition device comprising a. The method of claim 1, The human body detecting unit may identify a human body using a color image and a depth image of the input image. A human body detection unit detecting the human body area when it is determined that the human body is present in the human body identification unit; Human body component extraction unit for extracting the position of the human body components for the directed gesture recognition in the human body area detected by the human body area detection unit Directed gesture recognition device comprising a. The method of claim 2, And the body region detector is configured to perform a preprocessing process using a morphology operation to detect the detected body region as one combined region. The method of claim 2, The human body identifying unit identifies the presence or absence of a person by using one or more of an aspect ratio of an area, a location where an area exists, and an actual size of the area in a three-dimensional space. The method of claim 1, The directed gesture recognizing unit is configured to project the position information of the directed gesture recognition device and the pre-stored position information of the control target to a three-dimensional space model to calculate an indication position with respect to the pointing direction. . The method of claim 5, wherein And the control unit performs a preset function according to the type of the control target when the control target exists in the instruction position. The method of claim 6, The controller generates a wireless control signal for controlling the control target, the directed gesture recognition device. The method of claim 1, The directed gesture detection unit performs an operation detection for detecting whether the directed gesture has occurred and a state detection for detecting a current state of the generated directed gesture. The method of claim 1, And the hand position correcting unit corrects the position information of the hand based on training data storing position information of human body components, a direction indicated by a user, and a recognized indication direction. In the directed gesture recognition method of a mobile device equipped with a stereo camera, (a) receiving an image captured by the stereo camera, (b) detecting a human body region from the input image, and extracting position information of a human body component necessary for indicating a gesture gesture from the detected human body region, (c) correcting position information of a hand among position information of the extracted human body components; (d) detecting whether the directed gesture is generated based on the corrected position information of the hand, (e) calculating an indication direction of the indication gesture based on the position information of the mobile device and the corrected hand position information when the indication gesture occurs; (f) calculating the indication position according to the indication direction by projecting the position information of the mobile device and the position information of the control target previously stored in the three-dimensional space model; (g) analyzing a function intended by the detected directed gesture when there is a control object at the indicated position; and  (h) controlling the control target to execute a function intended by the directed gesture Including a directed gesture recognition method. The method of claim 10, In step (b), (b-1) identifying the presence or absence of a person using a color image and a depth image of the input image; (b-2) detecting the human body region when it is determined in step (b-1) that a person exists; and (b-3) extracting the positions of the human body components required for the directive gesture recognition in the human body region detected in the step (b-2). The method of claim 10, The step (b) is a directed gesture recognition method for performing a preprocessing process using a morphology (morphology) operation to detect the detected human body region as one combined region. The method of claim 11, In the step (b-1), the presence or absence of a person is identified using one or more of an aspect ratio of an area, a location where an area exists, and an actual size of the area in a three-dimensional space. . The method of claim 10, (i) moving the mobile device to a specific position if there is no control object at the indicated position; Directed gesture recognition method further comprising. The method of claim 10, In the step (h), when the control target is a device having a screen, performing the virtual mouse function based on the directed gesture, directed gesture recognition method. A recording medium having recorded thereon a program for recognizing a directed gesture of a mobile device equipped with a stereo camera, 16. The computer program product of claim 10, wherein the program comprises program code for implementing the respective steps of the methods as claimed in any one of claims 10-15.

Images