KR101525011B1 - tangible virtual reality display control device based on NUI, and method thereof - Google Patents

tangible virtual reality display control device based on NUI, and method thereof Download PDF

Info

Publication number
KR101525011B1
KR101525011B1 KR1020140134925A KR20140134925A KR101525011B1 KR 101525011 B1 KR101525011 B1 KR 101525011B1 KR 1020140134925 A KR1020140134925 A KR 1020140134925A KR 20140134925 A KR20140134925 A KR 20140134925A KR 101525011 B1 KR101525011 B1 KR 101525011B1
Authority
KR
South Korea
Prior art keywords
gesture
screen
user
hands
coordinates
Prior art date
Application number
KR1020140134925A
Other languages
Korean (ko)
Inventor
조경은
엄기현
이성조
이병권
Original Assignee
동국대학교 산학협력단
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 동국대학교 산학협력단 filed Critical 동국대학교 산학협력단
Priority to KR1020140134925A priority Critical patent/KR101525011B1/en
Application granted granted Critical
Publication of KR101525011B1 publication Critical patent/KR101525011B1/en

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/01Input arrangements or combined input and output arrangements for interaction between user and computer
    • G06F3/011Arrangements for interaction with the human body, e.g. for user immersion in virtual reality
    • G06F3/013Eye tracking input arrangements
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/01Input arrangements or combined input and output arrangements for interaction between user and computer
    • G06F3/017Gesture based interaction, e.g. based on a set of recognized hand gestures

Landscapes

  • Engineering & Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • User Interface Of Digital Computer (AREA)

Abstract

The present invention relates to an NUI-based real-time virtual space display control device and a control method for controlling a display device through a user gesture, and more particularly, The chain code is generated using the coordinates projected on the divided two-dimensional screen, and a different gesture model is applied to each screen to recognize the user gesture to change the screen or change the setting of the virtual object constituting the screen to display again .

Description

TECHNICAL FIELD [0001] The present invention relates to a virtual reality display control apparatus based on NUI,

BACKGROUND OF THE INVENTION 1. Field of the Invention [0002] The present invention relates to a virtual space display control apparatus for providing realistic contents, and more particularly, to a real-time virtual space display control apparatus based on NUI for controlling a display apparatus through a user gesture and a control method thereof.

As the computing function has evolved recently, the ubiquitous era has come. Therefore, research on ubiquitous computing is under way. The key of ubiquitous computing is that users use computers and networks easily and conveniently. In this context, there are various studies related to NUI (Natural User Interface) / NUX (Natural User Experience), which allow users to interact with computers using gestures that are natural movements of human beings, out of universal input devices such as a mouse or a keyboard have.

Apparatuses for displaying a virtual space including realistic contents provide a realistic screen to a user and utilize a user gesture as a way of interacting with a user. A user gesture is a person-centered interface that is natural and intuitive to a person, which means that a user gesture is directly related to a function provided by the display device. Therefore, there is a need to recognize the gesture and to control the display device by reflecting this.

Further, a technique for recognizing a gesture needs to be processed at a high speed in order to provide quick feedback to the user, and it is necessary to identify whether the gesture is a gesture for recognizing the function accurately and controlling the function. That is, there is a need to process information on the user's movement inputted in real time within a time as fast as it is recognized by the user in real time.

 Choi, H., "Kinect-Based Motion Recognition Model for 3D Content Control", The Journal of the Korea Contents Association, Vol 14 No.1 pp 24-29, 2014.1.

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and it is an object of the present invention to provide a device for controlling a display device through a natural and intuitive gesture operation of a user in order to solve a disadvantage that a device-dependent interface requires a simple memorization and a learning process through education . It is also an object of the present invention to provide means for recognizing gestures directly associated with the screens displayed by the virtual space display device and the virtual objects constituting the screens. Further, the present invention provides a means for displaying and outputting a result screen according to a gesture taken by the user in a time as fast as the user can recognize in real time.

In order to solve the above-mentioned problems, the virtual space display control apparatus according to an embodiment of the present invention extracts motion information on the user's face and body joints from the image, ; A sensing unit for generating an event signal indicating a screen switching according to a movement of a user when a change in one or more joints related to a rotation angle of a face or a user's step exceeds a preset threshold value; The grip state of the user's hands is determined and the trajectory of the movement of both hands of the user is projected on a two-dimensional screen perpendicular to the user's gaze direction according to the type of the gripped hand, A gesture model representing a gesture learning model is applied to the grip state of the hands, the two-dimensional coordinate, and the chain code, thereby recognizing a gesture corresponding to the movement of both hands of the user A gesture recognition unit; A control unit for changing a setting of any one of a screen, a screen menu, and a virtual object configuring the screen according to the event signal and the recognized gesture; And an output unit for outputting a setting value changed by the control unit.

In the virtual space display control apparatus according to an embodiment, the sensing unit may determine whether the user is walking according to the relative change value of the joints, wherein the joints related to the user's walking are the shoulders, waist, and joints of both hands .

In the sensing unit of the virtual space display control apparatus according to an embodiment, the event signal includes final coordinate values of the face or the joint and screen information currently being displayed.

In the virtual space display control apparatus according to an embodiment, the gesture recognition unit can recognize the movements of both hands while both hands or one hand are released from the grip state as one gesture.

In the gesture recognition unit of the virtual space display control apparatus according to an embodiment, the gesture model is provided for each screen, and further includes a gesture model for the left and right hands on the same screen.

In the above-described embodiment, a gesture model corresponding to a currently displayed screen is applied among the gesture models, and a gesture model corresponding to each hand can be applied when both hands are in a grip state.

In the virtual space display control apparatus according to an embodiment, the gesture recognizing unit divides the two-dimensional screen into a grid space of a predetermined size, and sequentially stores coordinates of the grid space so as to correspond to three-dimensional coordinates constituting the locus For each of the stored coordinates, a relative vector of the corresponding coordinates and the previous coordinates may be calculated and classified into the unit direction vectors.

In the above-described embodiment, the gesture recognition unit changes the diagonal vector to the previous coordinate and the next coordinate of the corresponding coordinate in a diagonal relationship, and stores the diagonal vector when the most recently stored direction vector is not a diagonal vector.

In the virtual space display control apparatus according to an exemplary embodiment, the gesture recognition unit can recognize the two chain codes having the same combination of direction vectors as different gestures when the start coordinates and the end coordinates of the movement of the hands are different.

In the virtual space display control apparatus according to an embodiment, the controller changes a screen setting by selecting a next screen in accordance with the recognized gesture or the final two-dimensional coordinate among screens linked to a currently displayed screen.

In the virtual space display control apparatus according to an exemplary embodiment, the controller selects a virtual object corresponding to a start coordinate of the two-dimensional coordinates according to the recognized gesture, and changes the state, position, or size of the selected virtual object.

According to another aspect of the present invention, there is provided a method for controlling a virtual space display device, the method comprising: inputting a captured image of a user by tracking the user, Extracting information; Generating an event signal indicating a screen change according to a movement of a user when a change in one or more joints related to a rotation angle or a step of the face exceeds a preset threshold value in the extracted motion information; Changing a screen according to the event signal when the event signal is present; If there is no event signal, the gesture recognition unit analyzes the extracted motion information to determine a grip state of the user's hands, and determines a trajectory of the movement of both hands of the user according to the type of the gripped hand, A chain code composed of a unit direction vector is generated using two-dimensional coordinates projected on a two-dimensional screen perpendicular to the visual line direction, and a gesture learning model for the grip state, the two-dimensional coordinate, Recognizing a gesture corresponding to a movement of both hands of the user by applying a gesture model; Changing a setting of any one of a screen, a menu of a screen or a virtual object constituting a screen according to the recognized gesture on the basis of a screen currently being displayed by the controller; And outputting a setting value changed by the control unit to an output unit.

In the control method of the virtual space display apparatus according to another embodiment, in the step of generating the event signal, the joints related to the user's walking correspond to the joints of the shoulders, the waist and both hands and the user walks according to the relative change value of the joints And the event signal includes final coordinate values of the face or the joint and screen information currently being displayed.

The step of recognizing the gesture in the control method of the virtual space display apparatus according to another embodiment recognizes the movements of both hands while both hands or one hand are released from the grip state as a single gesture.

In the method of controlling a virtual space display apparatus according to another embodiment of the present invention, in the step of recognizing the gesture, the gesture model is provided for each screen and further comprises a gesture model for each of the left and right hands on the same screen, A gesture model corresponding to a currently displayed screen is applied, and a gesture model corresponding to each hand can be applied when both hands are in a grip state.

The step of recognizing the gesture of the control method of the virtual space display apparatus according to another embodiment may include dividing the two-dimensional screen into a grid space of a predetermined size, and outputting coordinates of the grid space corresponding to the three- Are stored in order, and relative vectors of the coordinates and the previous coordinates are calculated for all the stored coordinates and classified into the unit direction vectors.

In the above-described embodiment, the recognition of the gesture may be performed by changing diagonal vectors when the previous coordinates and the next coordinates of the corresponding coordinates correspond to a diagonal relation, and the most recently stored direction vector is not a diagonal vector.

The step of recognizing the gesture of the control method of the virtual space display apparatus according to another embodiment recognizes the two chain codes having the same combination of direction vectors as a different gesture when the start coordinate and the end coordinate of the movement of the both hands are different .

The embodiments according to the present invention can provide a virtual space display device by changing a screen to be displayed, a menu of a screen, a virtual object constituting the screen, and a result screen through a gesture taken by the virtual space display device, Can be controlled. Further, by providing a gesture model for each of a plurality of screens that can be displayed by the virtual space display device, the gesture directly related to the screen and the virtual object constituting the screen is recognized, and the user's motion is analyzed using the two- It is possible to improve the accuracy of the gesture recognition and to feedback on the gesture in real time.

1 is a block diagram illustrating components of a virtual space display control apparatus according to an embodiment of the present invention.
Fig. 2 is a diagram showing a schematic function of the model house simulation system.
3B is a table showing a type of a gesture in which a right hand is gripped for setting an internal interior menu, and FIG. 3C is a table showing a type of a gesture in which a left hand is gripped for setting a system menu. FIG. 5 is a table showing a gesture in which both hands are gripped to change and set a virtual object to be configured.
4 is a diagram illustrating mapping of a user gesture to a lattice space screen by a virtual space display control apparatus according to an embodiment of the present invention.
5 is a diagram showing an 8-direction unit vector.
FIG. 6A shows a process of storing a two-dimensional coordinate by a virtual space display control apparatus according to an embodiment of the present invention with respect to a user gesture, FIG. 6B shows a process of generating a chain code using the two- Fig.
7 is a flowchart illustrating a method of controlling a virtual space display apparatus according to another embodiment of the present invention.
FIG. 8 is a flowchart showing the detailed steps of step S740 of FIG.

Prior to the description of the concrete contents for carrying out the present invention, for the convenience of understanding, the outline of the solution of the problem to be solved by the present invention or the core of the technical idea will be given first.

NUI technology allows users to accurately recognize gestures and reproduce them as they are, or to recognize and react quickly when a user takes a predefined gesture. When the user desires to control the device, the device corresponds to the gesture taken by the user among the predefined gestures and allows the device to perform a predetermined operation according to the gesture. Just as a computer command is defined, a gesture is predefined and the user controls the device through defined gestures. Unlike keyboards, mice, etc., NUI has the advantage of switching from a machine-centric interface to a human-centric interface because it is natural and intuitive to use. However, if you draw a circle with your right hand, it will be drawn differently for each person. Therefore, the user gesture is difficult for the user to recognize the same gesture as the same gesture as the device.

NUI technology is mainly used for virtual reality game or virtual reality simulation device. Conventional games or simulation devices have to learn how to operate input devices such as a keyboard, a mouse, and a remote control, and thus have a disadvantage in that the reality becomes inferior especially in a game providing a virtual reality. Therefore, we use the user gesture as an interface to make the user feel the reality about the provided virtual reality. In addition, in order to maintain the reality, the device must be able to react immediately to the user gesture.

Therefore, the present invention controls a display device through an intuitive and natural user gesture and displays a different learning model for each screen in order to accurately recognize a gesture taken by a user among gestures directly related to a screen provided by the display device and a virtual object constituting the screen And provides a virtual space display control apparatus and a control method thereof that ensure real-time processing by simply processing input data.

1 is a block diagram illustrating components of a virtual space display device according to an exemplary embodiment of the present invention. The virtual space display control device 1 receives the color image including the depth information through the RGB-D camera 2, changes the setting of the screen etc. according to the identified gesture, and outputs the changed information to the display device 3 A gesture recognition unit 30, a gesture model 35, a control unit 40, a DB 45, and an output unit 50. The input unit 10, the sensing unit 20, the gesture recognition unit 30, For the sake of understanding, the interior simulation system will be described as an example of the display device 3, and Fig. 2 is a diagram showing a schematic function of the model house simulation system. Referring to FIG. 2, the model house simulation system selects the size and structure of a house, arranges floor, wallpaper, furniture and the like in the interior of the house, and then observes and stores every corner of each house. The model house simulation system can display the model house as a real size three-dimensional image so that the user can experience the same as the real world house. The user can interact with the model house displayed through the gesture while changing the setting or moving the place.

Hereinafter, the detailed operation of each component will be described in detail.

The input unit 10 tracks the user from the RGB-D camera 2, receives the captured image, and extracts information on the user's face and body joints from the image. The RGB-D camera 2 can photograph a color image including depth information. The RGB-D camera 2 photographs the movement of the user interacting with the screen displayed by the virtual space display control device 1, Thereby generating an image. Therefore, the input unit 10 extracts information about the user's face and body joints related to the gesture from the input image information. Various products have been developed for the RGB-D camera 2, and the embodiments of the present invention have recently used the most widely used Kinetics of Microsoft. The kinetics traces the user and analyzes the captured 3D image to provide the user's skeleton information. The embodiment of the present invention extracts and uses necessary information from the skeleton information. It is determined which gesture of the user is to be used according to the function provided by the virtual space display control device 1, and the joint coordinates of the body related to the gesture . At this time, the coordinates are three-dimensional coordinates including depth information. For example, in embodiments of the present invention, in order to use a face rotation gesture to change a screen according to a change in a line of sight, coordinates related to a face are extracted, and a gesture for walking is changed to change a screen according to movement. And extracts the coordinates for

The sensing unit 20 generates an event signal indicating a screen transition according to the movement of the user when the rotation angle of the face or the change of one or more joints related to the user's walking exceeds a predetermined threshold value. When a user moves his or her gaze and changes an object to be viewed, an event signal is generated to display a screen centered on the object. The movement of the line of sight can be known by the rotation angle of the face, and when a preset threshold value is exceeded, a new screen is set. The threshold value can be determined in relation to the virtual object constituting the screen. For example, if a user looks at a virtual object positioned at the right end, it is possible to display a screen centering on the virtual object. However, if a virtual object does not exist at a point of view of the user on the currently displayed screen, It is not necessary to change and display again. Therefore, the threshold value for the change in the face rotation angle may be set differently for each screen.

The screen switching according to the movement of the user may occur when the user takes a gesture operation to walk. Therefore, when the change of the joint related to the stepping exceeds the predetermined threshold value, an event signal indicating the screen change is generated. Among the joints of the body, the joints related to the steps can be the joints of the shoulders, the waist and both hands. It is to judge whether the user is walking based on the relative difference between the shoulder and waist joints moving forward and backward and the change value according to the shaking of both hands when walking. The threshold according to the step may be determined according to the screen currently being displayed as in the case of the change in the line of sight. The virtual space display control apparatus 1 can select one of a plurality of predefined screens and display it. The user can select a result screen from among the screens associated with the current screen in the same direction as the direction of the gait. Therefore, the event signal may include final coordinate values in which the face or joint is changed and the screen information currently being displayed. The screen information includes both the screen itself currently being displayed and the configuration object information constituting the screen.

When the sensing unit 20 senses the user's gaze change gesture and the gesture and generates an event signal, the control unit 40 processes the gaze change gesture and the gesture gesture and changes the setting of the screen for displaying the result screen accordingly. The control unit 40 will be described later in detail.

The gesture recognition unit 30 analyzes the information on the joints of the body, particularly both hands, to determine which gesture the user has taken from among the predefined gestures. Specifically, the grip state of the user's hands is judged, and the two-dimensional coordinate obtained by projecting the trajectory of the movement of both hands on the two-dimensional screen perpendicular to the user's gaze direction is used according to the type of the gripped hand And generates a chain code composed of unit direction vectors. Then, a gesture model 35 indicating a gesture learning model is applied to the grip state of the both hands, the two-dimensional coordinate and the chain code, and a gesture corresponding to the movement of both hands of the user is recognized.

Before describing the order in which the gesture recognition unit 30 operates, a description will be given first of what is a gesture model that is a gesture learning model and a criterion for dividing a gesture in units of a single command in a user's continuous motion.

In reality, the user distinguishes the gesture because it is instantaneously recognizable as to which gesture it is at the same time as the recognition. However, there is a problem that the device needs to be unitized at the input stage in order to distinguish the gesture. It is a matter of what information should be input to the input unit in order to produce a result that can be known only after the device identifies the input object through analysis and processing. In other words, it is a problem that the user can determine the start and end of the gesture based on the moving information.

In the embodiments of the present invention, the grip of the user's hand is used as a criterion for classifying the type of the gesture in the movement of the user. That is, the movement of both hands during the release of both hands or one hand from the grip state can be recognized as one gesture. Since a person takes a gesture through the hand, the grip state of the hand is closely related to the gesture action. Actually, it is more natural to hold the hand and draw a circle or to push it to the right, as it is in the open hands. In particular, when interacting with a device displaying a virtual reality, a gesture represents a movement for controlling the virtual objects displayed on the screen. Therefore, using the grip state of the hand can easily solve the problem of determining a single instruction unit.

Next, the gesture needs to be defined differently for each displayed screen, and it is necessary to classify the type of the gesture according to the type of the function provided by the virtual-space display control device 1. [ Therefore, in the embodiments of the present invention, the gesture defines the type and number of gestures according to the screen, and classifies the gestures according to the types. For example, since the gestures that can change or set the system menu and the gestures that change the objects constituting the screen need to be distinguished from each other, they are classified according to the types of hands that are in a grip state.

For example, in the interior simulation system of FIG. 2, the types of gestures can be defined by dividing categories such as a system menu setting for an interior, an interior menu setting, and a configuration object of a screen. 3A is a table showing a type of a gesture in which the left hand grips a system menu, FIG. 3B is a table showing a type of a gesture in which a right hand is set to grip an interior menu, and FIG. A table showing a gesture in which both hands are gripped to change and set a virtual object to be constructed. 3A and 3B, all the gestures having the same shape are swiped to the left. However, since both gestures correspond to the left-hand grip state and the right-hand grip state, Is set. This classification also allows for more variety of device control with a simple gesture.

When a gesture is taken with both hands gripped as shown in FIG. 3C, both movements of both hands can be identified by a gesture. In the interior simulation system, it is possible to define a gesture as if the object is actually scaled up or down (scale down) in order to change the virtual object on the screen. This natural and intuitive operation allows the user to control the device by inputting the same feeling as the real world.

If the type of the gesture is classified as described above, the virtual space display control device 1 functionally requires a means for recognizing the gesture differently on each screen, and embodiments of the present invention use the different gesture model 35 . Therefore, the gesture model 35 is provided for each screen in the gesture recognition unit 30, and the gesture model 35 is further provided for the left and right hands in the same screen.

Now, the operation of the gesture recognition unit 30 will be described in order.

First, in the case where the gesture recognition unit 30 determines the grip state of both hands, if the image is input using the KNECK as in the embodiment of the present invention, the grip information can be directly known from the image information. In other cases, various implementations may be applied, such as the implementation being different depending on the input device, the shape of both hands being extracted by the color extraction method in the image, and the grip state being judged.

After determining the grip state, the gesture recognition unit 30 projects the trajectory of the movement of both hands according to the type of the gripped hand onto the two-dimensional screen perpendicular to the direction of the user's gaze, and stores the two-dimensional coordinate. The two-dimensional coordinate is for converting into a unit direction vector, and generates a chain code composed of permutations of unit direction vectors for one locus. The projection of the three-dimensional trajectory on the two-dimensional screen can be embodied by dividing the two-dimensional screen into a grid space of a predetermined size and sequentially storing the coordinates of the grid space so as to correspond to the three-dimensional coordinates constituting the trajectory extracted from the image . The degree of division may be determined depending on the degree of possible movement of the display device and the user to be implemented.

FIG. 4 is a diagram illustrating sequential recording of corresponding grid space information along a user's motion in a two-dimensional screen divided into a grid space. In the case of Fig. 4, the screen is divided into 10 pieces and 6 pieces. The lattice space can be composed of two-dimensional coordinates. In the drawing, the lattice space is represented by a sequence number for each lattice space.

The coordinates corresponding to the lattice space are stored along the movement of the hand, and the relative vectors of the coordinates and the previous coordinates are calculated for all the stored coordinates and classified into unit direction vectors. The unit direction vector can basically use an 8-direction vector, as shown in FIG. A directional vector that is more subdivided than an eight-directional vector may be used, which may be determined by the size of the screen to be displayed or the complexity of the gesture to be used.

Since user gestures include both straight and curved movements, it is not easy to recognize them at once. Specifically, since the embodiments of the present invention utilize the lattice space, there arises a problem that the unit vectors in the direction of 2, 4, 6, and 8 in the diagonal direction are recognized as combinations of unit vectors in directions 1, 3, 5, and 7. Therefore, in order to solve this problem, the gesture recognition unit 30 changes the diagonal vector to the diagonal vector when the most recently stored direction vector is not a diagonal vector when the previous coordinates and the next coordinates correspond to the diagonal coordinates. This is for recognizing the diagonal unit vector accurately and diagonally processing the part where the movement of the hand is bent as in the case where the unit vector in the direction of (3) and the unit vector in the direction of (1) are consecutively made. Therefore, when the direction of the movement of the hand changes, the gesture is further simplified by looking diagonally. As described above, since the diagonal processing is performed in the embodiments of the present invention, it is possible to recognize and simplify the gesture of the linear movement and the gesture of the curved movement at a time.

6A shows an example of generating a chain code for hand motion using a two-dimensional screen divided into nine lattice spaces. The movement of the hand corresponds to the unit vector of direction (4) which falls down from the center to the lower right to judge by the human eye. However, human movement becomes a form of sinuous line, not a straight line, because of slight differences. First, as in each step of 6a, save the coordinates according to the movement of the hand, remove the overlapping coordinates, and if you simply do, the final result will be the coordinates of 5-8-9.

6B is an example in which the stored coordinates are divided into 8-direction unit vectors. For the 5-8-9 coordinates stored in FIG. 6A, a relative vector is calculated for the respective coordinates and for the previous coordinates. 5-8 corresponds to the unit vector in the direction of the ⑤ downward from the top, and 8-9 corresponds to the unit vector in the direction of the ③ swipe to the right. In this case, the previous coordinate 5 and the next coordinate 9 correspond to the lattice space located diagonally. Since the previous unit vector is not one of the unit vectors of direction 2, 4, 6, and 8, The unit vector ⑤, which is the previous unit vector, is changed to the unit vector ④, and the unit vector is stored. As shown in the example, if there is fine movement at the beginning of the movement of the hand but it is included in one grid space, it is regarded as redundant data and removed, and the gesture which is sweeping from the center to the right in the defined data is identified.

The gesture recognition unit 30 recognizes the gesture by applying the gesture model 35 to the grip state, coordinates, and chain code after the chain code is generated. The gesture model (35) is a gesture learning model. Due to the nature of a gesture, it is necessary to actually recognize various motions as a gesture. When people take a gesture, the actual movements are physically taken differently, and even the same person moves differently for every action. The gesture model 35 can use a hidden markov model (HMM). Hidden Markov model is a kind of automata, which is suitable for processing sequential data because the transition between states constituting the model is made through a certain probability value. Therefore, it is widely used for recognizing sequential data such as voice data. In the embodiment of the present invention, the gesture model 35 is implemented using the hidden Markov model and the Baum-Welch algorithm. The type of state is determined in the HMM, and the accuracy is determined according to the number of stored learning data, so the value is experimentally determined.

As described above, the gesture model 35 is provided for each screen differently, and a different gesture model 35 is provided for the left and right hands on the same screen. It is to distinguish gesture types by providing different functions for each screen. In addition, since the displayed screens may include different real-sensible contents, the types of gestures required for the respective screens are different, and even if they are the same gesture, they need to be treated differently according to the screen.

When applying the gesture model 35 among the plurality of gesture models 35, a gesture model 35 corresponding to the currently displayed screen is applied. When both hands are in a grip state, a gesture model 35 corresponding to each hand (35) is applied. This allows different gestures to be perceived on the screen, and enables the identification of both hand movements on a single screen.

The number of states of the HMM is set according to the pattern of the gesture, and the learning data is stored for each gesture. In the embodiment of the present invention, 2 to 8 states are set, and 40 pieces of data are used for each gesture.

The likelihood probability for the gesture model calculated by the chain code generated by the user's hand movements is calculated as shown in Equation (1), and the gesture model 35 having the maximum value is classified as a gesture (gesture q) taken by the user have.

Figure 112014095579778-pat00001

A description of the variables in Equation (1) is as follows.

Figure 112014095579778-pat00002
Is a vector chain code for hand action,
Figure 112014095579778-pat00003
Is a learned gesture model,
Figure 112014095579778-pat00004
Is the prior probability of a gesture. In the case of recognizing the gesture according to the movement of the hand as shown in Equation 1 and recognizing both hand movements as the gesture in a state in which the both hands are gripped with the hands, the chain code generated according to the movements of the left hand and right hand, Model 35, thereby recognizing the gesture in the same manner as the one-handed gesture recognition. Therefore, embodiments of the present invention may further include a gesture model 35 for the left hand and the right hand, respectively, when both hands are in the grip state, and recognize gestures for the movements of both hands.

Further, the gesture recognition unit 30 recognizes the two chain codes having the same combination of direction vectors as different gestures when the start coordinates and the end coordinates of the gesture are different. Since the motions for drawing the circles and the letters M use only the relative vectors, the changes in the slopes are similar, and the same chain code is generated. In order to solve this problem, the start and end coordinates of the start and end of the motion are referred to and the two are distinguished.

The control unit 40 changes the value of the virtual object constituting the screen or the screen according to the event signal and the recognized gesture, or selects a menu provided on the screen. If there is an event signal, a screen is changed by selecting one screen among a plurality of screens. If there is no event signal, a new screen is selected according to the recognized gesture based on the screen currently being displayed, Select the menu provided or change the settings of the virtual objects that make up the screen. The function according to the gesture is generally stored in the DB 45 by defining the corresponding relationship in advance among the functions provided by the virtual space display control device 1 by the developer in advance. Accordingly, the control unit 40 controls the display screen, the configuration, and the menu to be displayed by referring to the correspondence relationship of the DB 45 according to two factors, that is, the recognized gesture and the generated event signal. Referring to FIGS. 3A, 3B, and 3C as a specific example, corresponding functions for a specific gesture are determined and stored in the DB 45 in advance.

A plurality of screens that can be displayed by the virtual space display control device 1 are determined in advance, and a plurality of screens have a relationship connected to each other because the virtual space is formed by projecting the real time clock. In the interior simulation system, for example, the entrance and the living room are connected to form a screen. Accordingly, the next screen can be selected according to the recognized gesture or final coordinates among the plurality of screens connected to the currently displayed screen. The setting of the virtual object constituting the screen may be performed by selecting a virtual object corresponding to the start coordinates of the gesture and changing the state, position or size of the virtual object selected by the predefined function according to the recognized gesture . For example, if the living room screen is currently displayed in the interior simulation system and the user grasps both hands while pointing both hands to the sofa on the screen and takes the small direction adjusting gesture of FIG. 3c, the function of changing the position of the sofa by identifying the corresponding gesture .

The output unit 50 outputs the setting value changed by the control unit 40 to the display device 3. [ The display device 3 refers to a device for displaying a two-dimensional or three-dimensional image to a user, such as an intra-simulation system. The display device 3 may be a means capable of reproducing images such as a TV, a monitor, and the like.

7 is a flowchart illustrating a method of controlling a virtual space display apparatus according to another embodiment of the present invention. Each step of FIG. 7 corresponds to each component of the virtual space display control apparatus 1 of FIG. 1, and description of the correspondence is mainly described without describing the overlapping description.

In operation S710, the input unit captures an image of the user and extracts motion information on the user's face and body from the RGB-D camera. And corresponds to the input unit 10 of Fig.

In step S720, when the rotation angle of the extracted face or the change of one or more joints related to the user's step exceeds a predetermined threshold, the sensing unit generates an event signal indicating a screen change according to the movement of the user. The joints related to the user's step are corresponding to the joints of the shoulders, the waist and both hands, and judge whether the user is walking according to the relative change value of the joints. The event signal may include final coordinate values of the face or joint and screen information currently being displayed. And corresponds to the sensing unit 20 of Fig.

It is determined in step S730 whether an event signal has been generated. If an event signal has been generated, the process proceeds to step S750. Otherwise, the process proceeds to step S740.

The step S740 recognizes the gesture by analyzing the information of the body joint, in particular, the information of the joint of the hand. Fig. 8 is a flowchart showing the detailed steps of the above step.

Specifically, in the above step, the gesture recognition unit analyzes the read motion information to determine a grip state of the user's hands (S742), and determines a locus of movement of both hands according to the type of the gripped hand, Dimensional coordinate system (step S744) on the two-dimensional screen which is perpendicular to the direction (S744). Then, a gesture model representing a gesture learning model is applied to the grip state of the both hands, the two-dimensional coordinate and the chain code, and the gesture corresponding to the movement of both hands of the user is recognized. At this time, the movement of the both hands is recognized as one gesture while both hands or one hand are released from the grip state. In addition, the gesture model is provided for each screen, and further includes a gesture model for the left and right hands on the same screen. A gesture model corresponding to a currently displayed screen is applied from among the gesture models, and when both hands are in a grip state, a gesture model corresponding to each hand is applied, thereby recognizing the gestures with respect to the movements of both hands.

Generating the chain code divides the two-dimensional screen into a grid space of a predetermined size, stores coordinates of the grid space in order so as to correspond to the three-dimensional coordinates constituting the locus, and stores the coordinates in the coordinates and the previous coordinates Can be generated by calculating a relative vector for a unit direction vector. In this case, if the previous coordinates and the next coordinates correspond to the diagonal coordinates, the diagonal vectors are stored if the most recently stored direction vector is not a diagonal vector, and two chain codes having the same combination of direction vectors are stored in the gesture If the starting and ending coordinates are different, it is recognized as another gesture. And corresponds to the gesture recognition unit 30 of Fig.

In step S750, when the event signal is received, the control unit selects one of the plurality of screens to change the screen. If not, the new screen is selected according to the gesture recognized in step S740 based on the currently displayed screen Change the screen, select a menu provided on the screen, or change the setting of a virtual object constituting the screen. The selection of a screen to be changed is determined in advance according to the correspondence relationship between the gesture and the function according to the function provided by the virtual display device. Specifically, among the plurality of screens connected to the currently displayed screen, the gesture or the next screen And can select a virtual object corresponding to the start coordinates of the gesture according to the recognized gesture and change the state, position or size of the selected virtual object. This corresponds to the control unit 40 of Fig.

In step S760, the output unit outputs the set value changed by the control unit. The display device will receive the output information and display it to the user. This corresponds to the output section 50 of Fig.

Meanwhile, the embodiments of the present invention can be embodied as computer readable codes on a computer readable recording medium. A computer-readable recording medium includes all kinds of recording apparatuses in which data that can be read by a computer system is stored.

Examples of the computer-readable recording medium include a ROM, a RAM, a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device and the like, and also a carrier wave (for example, transmission via the Internet) . In addition, the computer-readable recording medium may be distributed over network-connected computer systems so that computer readable codes can be stored and executed in a distributed manner. In addition, functional programs, codes, and code segments for implementing the present invention can be easily deduced by programmers skilled in the art to which the present invention belongs.

As described above, the present invention has been described with reference to particular embodiments, such as specific elements, and specific embodiments and drawings. However, it should be understood that the present invention is not limited to the above- And various modifications and changes may be made thereto by those skilled in the art to which the present invention pertains.

Accordingly, the spirit of the present invention should not be construed as being limited to the embodiments described, and all of the equivalents or equivalents of the claims, as well as the following claims, belong to the scope of the present invention .

1: Virtual space display control device
10: Input unit
20:
30: Gesture recognition section 35: Gesture models
40: control unit 45: DB
50: Output section
2: RGB-D camera
3: Display device

Claims (18)

An input unit for inputting an image captured by tracking a user and extracting motion information on joints of the user's face and body from the image;
A sensing unit for generating an event signal indicating a screen switching according to a movement of a user when a change in one or more joints related to a rotation angle of a face or a user's step exceeds a preset threshold value;
The grip state of the user's hands is determined and the trajectory of the movement of both hands of the user is projected on a two-dimensional screen perpendicular to the user's gaze direction according to the type of the gripped hand, A gesture model representing a gesture learning model is applied to the grip state of the hands, the two-dimensional coordinate, and the chain code, thereby recognizing a gesture corresponding to the movement of both hands of the user A gesture recognition unit;
A control unit for changing a setting of any one of a screen, a screen menu, and a virtual object configuring the screen according to the event signal and the recognized gesture; And
And an output unit configured to output a setting value changed by the control unit.
The method according to claim 1,
Wherein the joint related to the user's walking in the sensing unit is a shoulder, a waist and a joint of both hands, and determines whether the user is walking according to a relative change value of the joints.
The method according to claim 1,
Wherein the event signal includes a final coordinate value of the face or the joint and screen information currently being displayed in the sensing unit.
The method according to claim 1,
Wherein the gesture recognition unit recognizes the movements of both hands as one gesture while both hands or one hand are released from the grip state.
The method according to claim 1,
Wherein the gesture recognition unit applies a gesture model corresponding to a currently displayed screen among a plurality of different gesture models and applies a gesture model to each of the left and right hands on the same screen, .
6. The method of claim 5,
A gesture model corresponding to a currently displayed screen is applied from among the gesture models, and a gesture model corresponding to each hand is applied when both hands are in a grip state.
The method according to claim 1,
Wherein the gesture recognition unit divides the two-dimensional screen into a grid space of a predetermined size, sequentially stores the coordinates of the grid space so as to correspond to three-dimensional coordinates constituting the locus, Calculating a relative vector with respect to the coordinates, and classifying the vector into the unit direction vector.
8. The method of claim 7,
Wherein the gesture recognizing unit changes the diagonal vector to a diagonal line when the previous coordinate and the next coordinate of the corresponding coordinate correspond to a diagonal line and the most recently stored direction vector is not a diagonal vector.
The method according to claim 1,
Wherein the gesture recognition unit recognizes the two chain codes having the same combination of direction vectors as different gestures when the start coordinates and the end coordinates of the movement of the hands are different.
The method according to claim 1,
Wherein the control unit changes the setting of the screen by selecting a next screen according to the recognized gesture or the final two-dimensional coordinate among the screens associated with the currently displayed screen.
The method according to claim 1,
Wherein the control unit selects a virtual object corresponding to a start coordinate of the two-dimensional coordinates according to the recognized gesture, and changes the state, position, or size of the selected virtual object.
Extracting motion information on joints of the user's face and body from the image by inputting an image of the user traced by the input unit;
Generating an event signal indicating a screen change according to a movement of a user when a change in one or more joints related to a rotation angle or a step of the face exceeds a preset threshold value in the extracted motion information;
Changing a screen according to the event signal when the event signal is present;
If there is no event signal, the gesture recognition unit analyzes the extracted motion information to determine a grip state of the user's hands, and determines a trajectory of the movement of both hands of the user according to the type of the gripped hand, A chain code composed of a unit direction vector is generated using two-dimensional coordinates projected on a two-dimensional screen perpendicular to the visual line direction, and a gesture learning model for the grip state, the two-dimensional coordinate, Recognizing a gesture corresponding to a movement of both hands of the user by applying a gesture model;
Changing a setting of any one of a screen, a menu of a screen or a virtual object constituting a screen according to the recognized gesture on the basis of a screen currently being displayed by the controller; And
And outputting a setting value changed by the control unit to the output unit.
13. The method of claim 12,
Wherein the joint related to the user's walking in the step of generating the event signal corresponds to the joint of the shoulder, waist and both hands and determines whether the user is walking according to the relative change value of the joints, Or the last coordinate value of the joint and the screen information currently being displayed.
13. The method of claim 12,
Wherein the step of recognizing the gesture recognizes the movement of both hands while the hands are released from the grip state as one gesture.
13. The method of claim 12,
Wherein a gesture model corresponding to a currently displayed screen is applied from among a plurality of different gesture models in the recognition of the gesture and further comprises a gesture model for the left and right hands on the same screen,
And a gesture model corresponding to each hand is applied when both hands are in a grip state.
13. The method of claim 12,
The step of recognizing the gesture may include dividing the two-dimensional screen into a grid space of a predetermined size, sequentially storing the coordinates of the grid space so as to correspond to three-dimensional coordinates constituting the locus, And calculating a relative vector with respect to the coordinates and the previous coordinates, and classifying the vector into the unit direction vector.
17. The method of claim 16,
Wherein the step of recognizing the gesture is performed by changing a diagonal vector to a previous coordinate and a next coordinate of the corresponding coordinate and a diagonal vector when the most recently stored direction vector is not a diagonal vector. / RTI >
13. The method of claim 12,
Wherein the step of recognizing the gesture recognizes the two chain codes having the same combination of direction vectors as a different gesture when the start coordinates and the end coordinates of the movement of the both hands are different.
KR1020140134925A 2014-10-07 2014-10-07 tangible virtual reality display control device based on NUI, and method thereof KR101525011B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR1020140134925A KR101525011B1 (en) 2014-10-07 2014-10-07 tangible virtual reality display control device based on NUI, and method thereof

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR1020140134925A KR101525011B1 (en) 2014-10-07 2014-10-07 tangible virtual reality display control device based on NUI, and method thereof

Publications (1)

Publication Number Publication Date
KR101525011B1 true KR101525011B1 (en) 2015-06-09

Family

ID=53503903

Family Applications (1)

Application Number Title Priority Date Filing Date
KR1020140134925A KR101525011B1 (en) 2014-10-07 2014-10-07 tangible virtual reality display control device based on NUI, and method thereof

Country Status (1)

Country Link
KR (1) KR101525011B1 (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108243293A (en) * 2016-12-23 2018-07-03 炬芯(珠海)科技有限公司 A kind of method for displaying image and system based on virtual reality device
KR20180131507A (en) * 2017-05-31 2018-12-10 충남대학교산학협력단 Motion direction search apparatus based on binary search
CN116501234A (en) * 2023-06-26 2023-07-28 北京百特迈科技有限公司 User coupling intention rapid acquisition method, device, equipment and storage medium

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20130111248A (en) * 2010-06-29 2013-10-10 마이크로소프트 코포레이션 Skeletal joint recognition and tracking system
KR20140028064A (en) * 2011-06-06 2014-03-07 마이크로소프트 코포레이션 System for recognizing an open or closed hand

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20130111248A (en) * 2010-06-29 2013-10-10 마이크로소프트 코포레이션 Skeletal joint recognition and tracking system
KR20140028064A (en) * 2011-06-06 2014-03-07 마이크로소프트 코포레이션 System for recognizing an open or closed hand

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108243293A (en) * 2016-12-23 2018-07-03 炬芯(珠海)科技有限公司 A kind of method for displaying image and system based on virtual reality device
KR20180131507A (en) * 2017-05-31 2018-12-10 충남대학교산학협력단 Motion direction search apparatus based on binary search
KR102095235B1 (en) * 2017-05-31 2020-04-01 충남대학교산학협력단 Motion direction search apparatus based on binary search
CN116501234A (en) * 2023-06-26 2023-07-28 北京百特迈科技有限公司 User coupling intention rapid acquisition method, device, equipment and storage medium

Similar Documents

Publication Publication Date Title
US10664060B2 (en) Multimodal input-based interaction method and device
US10761612B2 (en) Gesture recognition techniques
US10394334B2 (en) Gesture-based control system
Rautaray Real time hand gesture recognition system for dynamic applications
US8457353B2 (en) Gestures and gesture modifiers for manipulating a user-interface
KR101956325B1 (en) System for finger recognition and tracking
US20110289455A1 (en) Gestures And Gesture Recognition For Manipulating A User-Interface
US20130335318A1 (en) Method and apparatus for doing hand and face gesture recognition using 3d sensors and hardware non-linear classifiers
US10416834B1 (en) Interaction strength using virtual objects for machine control
AU2012268589A1 (en) System for finger recognition and tracking
WO2018000519A1 (en) Projection-based interaction control method and system for user interaction icon
CN109145802B (en) Kinect-based multi-person gesture man-machine interaction method and device
KR101525011B1 (en) tangible virtual reality display control device based on NUI, and method thereof
Rehman et al. Two hand gesture based 3d navigation in virtual environments
KR20160141023A (en) The method of dynamic and static gesture recognition using depth camera and interface of immersive media contents
KR20130066812A (en) Method, apparatus, and computer readable recording medium for recognizing gestures
Abdallah et al. An overview of gesture recognition
KR101447958B1 (en) Method and apparatus for recognizing body point
CN113807280A (en) Kinect-based virtual ship cabin system and method
Leite et al. A system to interact with CAVE applications using hand gesture recognition from depth data
Piumsomboon Natural hand interaction for augmented reality.
Bernardes et al. Comprehensive model and image-based recognition of hand gestures for interaction in 3D environments
Zhang et al. Free-hand gesture control with" touchable" virtual interface for human-3DTV interaction
Kolhekar et al. A reliable hand gesture recognition system using multiple schemes
Prema et al. Gaming using different hand gestures using artificial neural network

Legal Events

Date Code Title Description
E701 Decision to grant or registration of patent right
GRNT Written decision to grant
FPAY Annual fee payment

Payment date: 20180430

Year of fee payment: 4

FPAY Annual fee payment

Payment date: 20190430

Year of fee payment: 5