JP4409545B2 - Three-dimensional position specifying device and method, depth position specifying device - Google Patents

Description

  The present invention relates to a three-dimensional position specifying technique for shooting an object operated by a player and specifying its position.

2. Description of the Related Art Conventionally, it is known that a player's action is shot using an imaging device such as a video camera, and a moving image of the player is displayed on a screen so that a command can be input or a game can be played (for example, patents) Reference 1) . In such an image processing apparatus, a command can be input when a moving image of a player contacts a menu screen or object arranged on the screen. That is, the player's moving image itself functions as an input interface.
JP 2002-196855 A

  In an application such as a game that uses a player's moving image as an input interface as described above, it is important to derive the behavior of the player who operates the application in a natural manner by means of on-screen effects, sounds, and the like. Requesting unnatural movement from a player can result in a loss of player interest in the application.

  The present invention has been made in view of these problems, and an object of the present invention is to provide a technique for realizing an interface that is easy for a player to use in an apparatus that uses the operation of the player as an input interface.

  One embodiment of the present invention is an apparatus for specifying a three-dimensional position of an object. This apparatus includes a camera that captures an object operated by a player, a reference image storage unit that stores a reference image for identifying the object, and a reflected image of a predetermined entry area that the object is supposed to enter. Reflecting means for projecting toward the object and causing the camera to capture a direct image and a reflected image of the object, and a reflecting surface area specifying unit for specifying a reflecting surface area corresponding to the reflected image from the frame shot by the camera And matching using the reference image to the frame to detect the reflected image from the in-frame localization part that identifies the position of the object in the frame and the reflective surface area, along the optical axis of the camera A depth localization unit that identifies the position of the object in the depth direction.

  Here, “object” is a generic term for what is operated by the player within the shooting range of the camera, and part of the body such as the player's head, arms, hands, feet, mouth, and part of the player's body. Included are objects such as bars, sheets, boxes, etc. operated by (eg hands, feet, mouth) and devices such as controllers.

  According to this aspect, an action in the depth direction, such as pushing or pulling out an object by the player, is detected by detecting the entry of the object into a predetermined entry area using a reflection image by the reflector. In the conventional object detection based on the inter-frame difference, it is very difficult to detect the movement of the object along the depth direction, that is, the optical axis direction of the camera. Since the reflected image from the intersecting direction is used, the movement of the object in the depth direction can be accurately detected.

  Another aspect of the present invention is a three-dimensional position specifying method. In this method, a direct image of an object and a reflected image obtained by reflecting the object from a predetermined depth position are photographed by a camera, and an area corresponding to the reflected image and other areas including the direct image are specified in the frame. Then, matching is performed on the direct image to specify the position of the object in the frame, and detecting whether or not the object is at a predetermined depth position from the difference in the reflected image between the plurality of frames.

  It is also effective as an aspect of the present invention to replace the components and expressions of the present invention among methods, systems, computer programs, recording media storing computer programs, and the like.

  According to the present invention, it is possible to detect the movement of an object in the depth direction using a frame photographed by a camera.

Embodiment 1 FIG.
FIG. 1 shows an overall configuration of a three-dimensional position specifying apparatus 10 according to an embodiment of the present invention. The three-dimensional position specifying apparatus of the present embodiment captures an object operated by a player with a single camera, specifies the three-dimensional position of the object by image processing, and displays a screen corresponding to the specified three-dimensional position on the display. indicate. A typical example of an application using the three-dimensional position specifying apparatus 10 is an action game in which a character or the like displayed on the screen is operated by a player's action, but other forms of games, simple business applications, digital photographs, etc. The present invention can also be applied to album display and music data playback applications.

  The three-dimensional position specifying device 10 includes a display 40, a camera 20 installed on the upper side of the display, an image processing device 30, and a reflector 50.

  The display 40 is preferably arranged in front of the player 72. The player 72 operates the object while viewing his / her image taken by the camera 20.

  The camera 20 captures the object 70 operated by the player 72 and outputs a frame to the image processing device 30 at a predetermined frame rate. In order to speed up the responsiveness of image detection, the frame rate is preferably as high as possible. The camera 20 is installed on the upper side of the display 40. The shooting range 26 of the camera 20 is set so as to capture at least the object 70 operated by the player 72. As a result, the player 72 can operate the object 70 with the front facing the display 40 side. However, depending on the characteristics of the application executed by the three-dimensional position specifying apparatus 10, the camera 20 is installed below or on the side of the display 40, or in a place different from the direction in which the player 72 looks at the display 40. May be installed.

  The frame output from the camera 20 is displayed on the display 40 via the image processing device 30. In this case, it is preferable that the captured frame is subjected to mirror processing by the image processing device 30 and a mirror image of the player 72 is displayed on the display 40. By displaying the mirror image, for example, when the player raises the hand, the image in the screen raises the same hand as if it was reflected in the mirror, so that the player can easily recognize his / her movement. However, the image processing device 30 may display the screen as it is captured on the display 40 without performing the mirror surface processing. Further, a screen that is inverted up and down by the image processing device 30 may be displayed on the display 40 according to the characteristics of the application executed by the three-dimensional position specifying device 10.

  The image processing apparatus 30 has a function of loading and executing application software stored in an external storage medium. The image processing device 30 performs the above-described mirror processing on the frame output from the camera 20, detects an object image in the frame, displays a predetermined image superimposed on the object, or performs a player action. Performs processing such as giving instructions to the application. The specular image that has undergone predetermined processing by the image processing device 30 is output to the display 40. The image processing apparatus 30 is typically a dedicated machine such as a game console, but may be a general-purpose personal computer or server having an image input / output function. A more detailed function and configuration of the image processing apparatus 30 will be described later.

  The display 40 may include a speaker 42. The speaker 42 reproduces sound, accompaniment, and the like output from the image processing device 30 in accordance with objects displayed on the display 40 and other images. The speaker 42 is preferably configured integrally with the display 40 and is disposed in the vicinity of the display 40. However, the speaker 42 and the display 40 may not be integrated and may be disposed at positions separated from each other.

  The reflector 50 is installed between the player 72, the display 40, and the camera 20, and has a role of causing the camera 20 to take a reflected image of the object 70. In this specification, the “object” is a generic term for what is operated by the player 72 within the shooting range 26 of the camera 20, and a part of the body such as the player's head, arms, hands, feet, mouth, and the like. Included are objects such as sticks, sheets, boxes and the like that are manipulated by body parts (eg hands, feet, mouth), and devices such as controllers. In this specification, the expression that the object is moved by the player's intention is expressed as “an object operated by the player”, including the case where the object is a part of the body. In FIG. 1, as an example, the finger of the player is shown as the object 70.

A direct image of the object 70 is taken by the camera 20, and a reflected image by the reflector 50 is also taken by the camera 20. In other words, the camera 20 includes both a direct image and a reflected image of the object 70 in one frame. In this way, by capturing the object 70 as an image viewed from two directions, a direct image and a reflected image, the three-dimensional position of the object 70 can be specified only by an image from a single camera, as will be described later. it can.
In order to simplify the description, in the following description, the number of objects 70 operated by the player 72 is one, but it goes without saying that the same processing can be performed even if two or more objects exist.

  The reflector 50 includes two reflecting surfaces 52 and 54, each of which reflects the object 70, and these reflected images are taken by the camera 20. Accordingly, the reflection surfaces 52 and 54 are given a predetermined angle so that the reflected image of the object 70 is connected by the lens of the camera 20. Further, the installation location of the reflector 50 is limited to a position separated from the camera 20 by a predetermined distance.

  As shown in FIG. 1, entry areas 62 and 64, which are areas where the reflection image of the object 70 can be projected toward the camera 20, are spread above the reflection surfaces 52 and 54. The spread of the entry areas 62 and 64 is determined by the degree of inclination of the reflection surfaces 52 and 54, and is a range where the object 70 is assumed to enter. In the example of FIG. 1, the entry areas 62 and 64 are set so as not to cross each other. Therefore, when the object 70 exists in the entry area 62, a reflected image reflected by the reflection surface 52 is taken by the camera 20, and when the object 70 exists in the entry area 64, it is reflected by the reflection surface 54. The reflected image is taken by the camera 20. However, when the object 70 has a certain length in the depth direction of the reflector 50 like a finger or a stick, the object 70 exists simultaneously in both the entry areas 62 and 64.

  In general, when an object motion is to be detected based on the difference between frames, a motion substantially parallel to the direction along the optical axis of the camera (z direction in FIG. 1) Therefore, it is difficult to detect. Therefore, in the present embodiment, an image from a direction different from the direct image of the object is acquired by using the reflection by the reflector 50, and the motion of the object in the z direction is reliably detected using the reflected image. I was able to do it. Hereinafter, the direction along the optical axis of the camera is simply referred to as “depth direction”.

  FIG. 2 is a diagram illustrating a simplified hardware configuration of the camera 20 and the image processing apparatus 30. The camera 20 includes an image sensor 22 as an image sensor and an image processing unit 24. The image sensor 22 is generally a CCD sensor or a CMOS sensor, and records an image by capturing an image formed by a lens (not shown) with a light receiving element. The captured image is temporarily stored in a memory (not shown) such as a RAM. Since the configuration of the camera 20 is well known, further detailed description is omitted.

  The image processing unit 24 is composed of a circuit such as an ASIC, and performs image data output from the image sensor 22 among A / D conversion, demosaic, white balance processing, noise removal, contrast enhancement, color difference enhancement, and gamma processing. Implement what you need. The image data processed by the image processing unit 24 is transferred to the image processing device 30 via a communication interface (not shown). In the following description, for simplicity, it is assumed that the image data passed from the image processing unit 24 to the image processing device 30 is RAW data obtained by digitizing the output signal from the image sensor 22 as it is. Format, for example, compressed data such as JPEG. In the latter case, the image processing apparatus 30 includes an image decoding unit that decodes the compressed data before the processing unit 32.

  The image processing apparatus 30 includes a processing unit 32, an image output unit 34 that outputs image data passed from the processing unit 32 to the display 40, and an audio output unit that outputs audio data passed from the processing unit 32 to the speaker 42. 36. In addition to this, the image processing apparatus 30 executes a predetermined application in accordance with the load unit that reads application software stored in an arbitrary recording medium such as a CD-ROM, DVD-ROM, and flash memory. An application execution unit is provided. Since these functions are naturally provided in a dedicated machine such as a game console or a personal computer, further detailed description is omitted.

  FIG. 3 is a plan view showing the structure of the reflector 50. The reflector 50 has a thin plate shape as a whole, and has the first reflecting surface 52 and the second reflecting surface 54 that are spaced apart in the depth direction as described above. The reflecting surfaces 52 and 54 are mirrors in one example, and may be a mirror-finished metal, plastic, glass on which metal is deposited, or the like. The first reflecting surface 52 and the second reflecting surface 54 are arranged in parallel, and are arranged so that their long axes are substantially perpendicular to the optical axis of the camera 20. As shown in FIG. 1, the first reflecting surface 52 and the second reflecting surface 54 are set to angles that reflect an object above the reflecting surface and project the reflected image toward the lens of the camera 20.

  Markers 56 for causing the image processing device 30 to recognize the position of the reflector 50 are disposed at both ends of the reflector 50 in the long axis direction. The marker 56 may be a colored portion, may be provided with a predetermined pattern such as a check, or may be a two-dimensional code. Light sources such as LEDs may be embedded at both ends. In short, as long as the information necessary for specifying the position of the reflector 50 in the frame output from the camera 20 can be given, the form is not limited.

  Since the reflector 50 has a predetermined width in the depth direction and includes a plurality of reflecting surfaces in the depth direction, a plurality of entry areas 62 and 64 can be set in the depth direction. Each of the reflecting surfaces 52 and 54 projects a reflected image of a different approach area where the object is supposed to enter toward the camera 20 and causes the camera 20 to capture a reflected image of the object. By doing so, as will be described later, the movement of the object in the depth direction can be detected.

  FIG. 4 is a diagram showing a detailed configuration of the processing unit 32. These configurations are realized by a CPU, a memory, a program loaded in the memory, and the like, but are illustrated here as functional blocks realized by their cooperation. Accordingly, those skilled in the art will understand that these functional blocks can be realized in various forms by hardware only, software only, or a combination thereof.

  The image acquisition unit 102 acquires the frames output from the camera 20 one by one and sends them to the image inversion unit 104 and the image linkage audio control unit 150.

  The image reversing unit 104 performs a mirror surface process (that is, an image reversal process) on the frame received from the image acquisition unit 102 to generate a mirror image. The specular image is sent to the three-dimensional localization unit 110 and the on-screen display unit 144.

  The three-dimensional localization unit 110 specifies the three-dimensional position of the object using the frame captured by the camera 20 received from the image inversion unit 104. The three-dimensional position means a two-dimensional position corresponding to the position of the object in the frame, that is, the position in the screen, and the position in the depth direction of FIG. In the present embodiment, the object in the screen is recognized by specifying the two-dimensional position of the object, and the specific action of the player is detected by specifying the depth position of the object.

  The three-dimensional localization unit 110 includes a reflective surface area specifying unit 112, a depth localization unit 122, an in-frame localization unit 114, and a reference image storage unit 120.

  The reflection surface area specifying unit 112 specifies a reflection surface area that is an area corresponding to the first reflection surface 52 and the second reflection surface 54 of the reflector 50 from the frame photographed by the camera 20. The reflective surface area specifying unit 112 detects two markers 56 from within the frame, and specifies an area sandwiched between them as a reflective surface area.

  The depth localization unit 122 specifies the position of the object in the depth direction by detecting the reflected image from the reflection surface region specified by the reflection surface region specification unit 112. Specifically, the depth localization unit 122 compares the reflection surface areas between a plurality of frames and detects the difference. If there is no reflection image in the reflection surface area in a certain frame and a reflection image is reflected in the reflection surface area in the subsequent frame, it can be determined that the object is located in the entry area corresponding to the reflection surface.

  In order to execute this determination with certainty, it is necessary to clearly distinguish the reflected image of the object reflected on the first reflecting surface 52 and the second reflecting surface 54 from the other images. Therefore, in one embodiment, the depth localization unit 122 acquires a default image of the reflecting surface area before starting the three-dimensional position specifying process. When a difference is detected between the default image and the reflection surface area of an arbitrary frame, it may be determined that the object is located in the entry area.

  The depth localization unit 122 performs the same processing on the reflective surface areas corresponding to the first reflective surface 52 and the second reflective surface 54, so that the object enters the first approach area 62 and the second approach area 64, respectively. It is determined whether or not. The result of this determination is sent to the input control unit 130.

  The in-frame localization unit 114 specifies the position of the object in the frame. The in-frame localization unit 114 includes an object detection unit 116. The object detection unit 116 performs well-known pattern matching using a reference image (template) of the object on the frame received from the image inversion unit 104, and specifies the position of the object in the frame. The target to be matched may be the frame itself received from the image inverting unit 104, or may be the one obtained by removing the reflective surface area specified by the reflective surface area specifying unit 112 from the frame.

  The reference image storage unit 120 stores a reference image for specifying an object. A reference image prepared in advance for an object whose position in the frame should be specified may be stored. However, as will be described later, the object to be specified is captured by the camera 20, and the object will exist from the frame. A region may be cut out and stored in the reference image storage unit 120 as a reference image. In the former case, for example, if the object is a hand, a reference image created by taking an average of images of tens or thousands of hands may be stored, and the age, gender, physique, etc. of the player may be stored. A plurality of reference images classified accordingly may be stored. Any matching technique using a reference image can be used. Since these are well known to those skilled in the art, further detailed description is omitted.

  Information on the in-frame position of the object specified by the in-frame localization unit 114 is given to the input control unit 130.

  The input control unit 130 gives an instruction based on information obtained by image processing on a frame captured by the camera 20 to an application execution unit (not shown) that executes an application such as a game. The input control unit 130 includes an action specifying unit 132, a display control unit 134, and an image storage unit 136.

  The action specifying unit 132 detects the movement of the object 70 in the depth direction between the first approach region 62 and the second approach region 64 based on the determination of the depth position by the depth localization unit 122, and the action of the player Is identified. The action specifying unit 132 may specify the movement of the object 70 toward the camera 20 in the depth direction and the movement away from the camera 20 as different player actions. The action specifying unit 132 gives the specified action to the application execution unit (not shown) and the display control unit 134. The application execution unit receives a given action as an input and gives a predetermined function.

  The display control unit 134 superimposes the direct image of the object photographed by the camera 20 on the display to display an image to be displayed in order to achieve a predetermined purpose. The display control unit 134 has different display modes when the object is located in the first entry region 62 corresponding to the first reflection surface 52 and when the object is located in the second entry region 64 corresponding to the second reflection surface 54. An image may be displayed. The display control unit 134 searches the image storage unit 136 for an image corresponding to the position of the object, and outputs the image to the on-screen display unit 144.

  The image storage unit 136 stores the above-described image displayed so as to be superimposed on the direct image of the object. Examples of this image include characters used in games, pointers such as cursors, instruments such as musical instruments and weapons, marks such as stars and the sun, images of parts of the body such as hands and feet, or inputs such as keyboards and calculators There is an image of the device. These images may be input images that the OS has in order to accept input from the user, or may be application images read from the application software being executed. In order for the display control unit 134 to display images of different modes according to the position of the object in the frame, the image storage unit 136 may hold images of a plurality of modes, or to change a certain image The necessary data may be held.

  The on-screen display unit 144 displays the image output from the display control unit 134 on the mirror image obtained from the image inverting unit 104 and sends the image to the image output unit 34. The image output unit 34 displays a screen superimposed on the mirror image of the player on the display 40.

  The image linkage audio control unit 150 controls the audio output unit 36 so as to output audio linked to the position of the object detected by the three-dimensional localization unit 110 and the action of the player detected by the input control unit 130. . The specific configuration of the image linkage audio control unit 150 will be described in detail in the third and fourth embodiments.

  The arrangement instruction unit 142 displays on the display 40 a display for instructing the player where to place the reflector 50 together with the image of the player taken by the camera 20. As described above, since the reflector 50 must cause the camera 20 to capture a reflection image of the object 70, the position where the reflector 50 is to be disposed is limited to a certain range. Therefore, in order to make the player place the reflector 50 in the correct position, for example, a frame line is displayed on the display 40, and the reflector 50 photographed by the camera 20 is placed inside the frame line. Have the position adjusted.

  The arrangement confirmation unit 140 refers to the frame photographed by the camera 20 and confirms whether or not the reflector 50 is installed at an appropriate position. Specifically, the reflection surface area specifying unit 112 detects the positions of the markers 56 in the frame at both ends of the reflector 50, and the arrangement confirmation unit 140 displays the position of the marker 56 on the frame line displayed by the arrangement instruction unit 142. It is determined whether it is inside. If the marker is inside the frame line, a display indicating that the reflector is appropriately arranged is displayed on the display 40, and the arrangement instruction unit 142 is instructed to stop displaying the frame line. As long as the marker does not enter the inside of the frame line, the localization processing by the three-dimensional localization unit 110 may not be started.

  Next, an example of an application using the three-dimensional position specifying device according to the present embodiment will be described with reference to FIGS. This application is a calculator that allows players to input numbers by pushing the calculator key image displayed on the screen.

  FIG. 5A shows a state in which the object 70 operated by the player 72, that is, the player's finger, is positioned on the near side of the first entry area 62 that extends above the first reflecting surface 52 of the reflector 50. This is a state in which an application executed on the image processing apparatus 30 is waiting for some action of the player. FIG. 5B shows a screen 44 displayed on the display 40 and recognized by the player at this time. As shown in the figure, direct images of the player 72, the object 70 and the reflector 50 are displayed on the screen 44. The reflection surface area specifying unit 112 specifies the reflection surface area 50 ′ by detecting the marker 56 from the frame.

  When the camera 20 and the image processing device 30 are switched on and are in a standby state, a default image of the reflecting surface region 50 ′ may be stored. In the standby state, there is nothing other than the background above the first entry area 62 and the second entry area 64. Therefore, if the default image is stored, the difference when the object enters the first entry area 62 and the second entry area 64 can be easily taken, and therefore the reflected image of the object in the reflection surface area can be obtained. The detection process is robust.

  In the conventional method for detecting an object based on a difference in motion between frames, if the object stops on the screen, the difference disappears and nothing can be recognized. On the other hand, if the default image is stored in advance as in the present embodiment, even when the object is still in the state of entering the entry area, Therefore, the depth position of the object can be continuously recognized.

  FIG. 6A shows a state where the object 70 operated by the player 72 has entered the first entry area 62. FIG. 6B shows a screen 44 displayed on the display 40 and recognized by the player at this time. As the object 70 enters the first entry area 62, a reflection image 70 a of the object 70 is displayed in an area corresponding to the first reflection surface 52 of the screen 44. The depth localization unit 122 detects the reflected image 70a by taking the difference of the reflecting surface area between frames.

  When the action specifying unit 132 knows that the object 70 has entered the entry area 62, the action specifying unit 132 instructs the display control unit 134 to display an application image 80 for executing the calculator application on the display 40. The application image 80 includes a plurality of key areas for inputting numbers or symbols. The application image 80 is preferably a line drawing or semi-transparent so as not to hinder the visual recognition of the player's motion to be superimposed, but may be an opaque image. In addition, the action specifying unit 132 instructs an application execution unit (not shown) to start the calculator application. Thereafter, the in-frame localization unit 114 continues to track the object 70 by specifying the position of the object 70 in the frame by matching. In FIG. 6B, the object 70 is at the key position corresponding to “5” in the application image 80.

  FIG. 7A shows a state in which the player 72 moves the object 70 in the first entry area 62, that is, in a plane perpendicular to the depth direction. FIG. 7B shows a screen 44 displayed on the display 40 and recognized by the player at this time. As shown in FIG. 7B, the player 72 has moved the object 70 from the position “5” of the application image 80 to the position “1”. Along with this, the position of the reflection image 70a of the object in the reflection surface area also changes. However, since the object remains in the first entry area 62, an instruction by the action specifying unit 132 is not generated, and therefore the aspect of the application image 80 does not change. The in-frame localization unit 114 keeps track of which key of the application image 80 the object 70 is on.

  FIG. 8A shows a state in which the object 70 operated by the player 72 has entered the second entry area 64 beyond the first entry area 62. FIG. 8B shows a screen 44 displayed on the display 40 and recognized by the player at this time. As shown in FIG. 8B, when the object 70 has entered the second entry area 64, a portion of the reflection surface area corresponding to the first reflection surface 52 is reflected by the first reflection surface 52. Reflected images 70a by the second reflecting surface 54 are respectively displayed on portions corresponding to the second reflecting surface 54. The depth localization unit 122 detects the reflected images 70a and 70b by taking the difference of the reflecting surface area between the frames.

  Based on the information from the depth localization unit 122, the action specifying unit 132 recognizes that the object 70 has moved from the first entry region 62 to the second entry region 64, and the object 70 becomes a camera in the depth direction. It is determined that the action of moving to the opposite side has been executed by the player 72. Accordingly, the action specifying unit 132 notifies the application execution unit that the key of the application image 80 corresponding to the position in the frame where the current object 70 exists is input, that is, pushed. At the same time, the action specifying unit 132 instructs the display control unit 134 to change the display mode of the key of the application image 80 corresponding to the position in the frame where the current object 70 exists. In the example of FIG. 8B, the color of the key “1” corresponding to the position of the object 70 in the frame is changed (see 80a in the figure). The change in display mode may be a change in color, flashing or lighting of a key, a mode in which a key is pressed, or the like. In this way, the player can input numbers into the calculator application by operating the object. Similarly, a keyboard may be displayed as an application image and used as an input device for a word processor.

  Thereafter, when the player 72 pulls the object 70 back to the position of the first entry area 62, the action specifying unit 132 detects this action and cancels the selection of the key corresponding to “1” of the application image 80. judge. The action specifying unit 132 instructs the display control unit 134 to return the display mode of the key corresponding to “1” to the original state.

  When the player 72 returns the object 70 to a position before the first entry area 62, that is, the state shown in FIG. 5A, the action specifying unit 132 detects this action and determines that the player has stopped operating the calculator application. To do. The action specifying unit 132 informs the application execution unit to stop the calculator application and instructs the display control unit 134 to hide the application image 80. Thus, the screen returns to the screen as shown in FIG.

  FIG. 9 is a flowchart for executing the application described with reference to FIGS. 5 to 8 in the three-dimensional position specifying apparatus 10 according to the present embodiment.

  After the reflector 50 is installed at an appropriate position, the object 70 and the reflector 50 are photographed by the camera 20, and the image acquisition unit 102 acquires a frame including a direct image and a reflected image of the object 70 (S10). The reflective surface area specifying unit 112 specifies the reflective surface area by detecting the marker 56 in the frame given from the image reversing unit 104 (S12). The depth localization unit 122 identifies the position of the object in the depth direction by detecting a difference between frames within the reflection surface area (S14). The action specifying unit 132 determines whether or not the object 70 has entered the first entry area 62 according to the information from the depth localization unit 122 (S16). As long as the object 70 does not enter the first entry area 62 (N in S16), the application is not executed.

  When the object 70 enters the first entry area 62 (Y in S16), the action specifying unit 132 instructs the application execution unit to start the application. The in-frame localization unit 114 specifies the position of the object in the frame by matching, and the display control unit 134 displays a predetermined application image superimposed on the position of the object in the frame (S18). The in-frame localization unit 114 continues to track the object as long as the object 70 is inside the first approach area 62 (S20). The depth localization unit 122 also detects a difference between frames in the reflection surface area corresponding to the first reflection surface 52 and the second reflection surface 54, and specifies the position of the object in the depth direction (S22).

  The action specifying unit 132 determines whether or not the object 70 has entered the second entry region 64 according to the information from the depth localization unit 122 (S24). Unless the object 70 enters the second entry area 64 (N in S24), the processes of S18 to S22 are repeated. When the object 70 enters the second entry area 64 (Y in S24), the action specifying unit 132 determines that the key of the application image 80 has been pushed, and transmits the information to the application execution unit and the display control unit 134. Accordingly, processing corresponding to the position of the object 70 in the frame is executed by the application, and the display mode of the application image 80 changes (S26).

  As described above, in this embodiment, by detecting the entry of an object into a predetermined entry area using a reflection image by a reflector, an action in the depth direction such as pushing and pulling out the object by the player is performed. To detect. In the conventional object detection based on the inter-frame difference, it is very difficult to detect the movement of the object along the depth direction, that is, the optical axis direction of the camera. Since the reflected image from the intersecting direction is used, the movement of the object in the depth direction can be accurately detected.

  If a function such as displaying a character or the like at the position in the frame where the object exists and outputting sound is added, the character display and sound output will always continue while the player moves the object. Will be. Unless other input devices were used together, it was difficult to turn off their display and sound according to the player's will. On the other hand, in this embodiment, by moving the object 70 between the first entry area 62 and the second entry area 64, the function of a specific application is turned on / off, and the image is displayed / hidden. Switching between these functions can be easily realized by only operating the object.

  Furthermore, as described with reference to FIGS. 5 to 8, in the present embodiment, the player's action can have a plurality of meanings. That is, while the object 70 is in the first entry area 62, the operation of the object 70 by the player 72 corresponds to a “select” operation in the application image 80. Therefore, by moving the object, the key selected in the application image 80 moves. Then, when the player 72 further pushes the object 70 in a state where a desired key is selected, the object 70 enters the second entry area 64, thereby giving a “decision” operation. As described above, the present embodiment is characterized in that the stroke of the object can be detected.

  Various applications can be considered by utilizing this stroke. For example, by catching the movement of pushing the hand toward the camera and making the movement of pulling the hand out of the camera corresponding to the release, it is possible to realize a game of catching or releasing the character appearing on the screen by the hand. .

  Further, the movement of pushing the hand toward the camera may correspond to turning on a specific function, and the movement of pulling out the hand from the camera may correspond to turning off the specific function. Using this, for example, in paint software, when you push your hand, the shape of the cursor displayed on the screen changes, and if you move your hand in that state, you can draw a line on the screen and pull out your hand. It is also possible to apply such that the cursor shape returns to its original shape and lines cannot be drawn on the screen even if the hand is moved. In a conventional similar application, once an object is recognized, a line is written every time the hand is moved. On the other hand, according to the present embodiment, the player can easily use functions on and off through simple actions.

  In the first embodiment, it has been described that the reflector 50 includes the two reflecting surfaces 52 and 54. However, the number of reflecting surfaces may be one or three or more. When there is only one reflecting surface, actions such as pushing and pulling out the object by the player cannot be specified, but it is possible to specify whether or not the object is located at least in the entry area corresponding to the reflecting surface. Even when there are three or more reflecting surfaces, entry areas are set corresponding to each of the reflection surfaces, and the depth localization unit 122 determines whether or not an object has entered each entry area as described above. Increasing the number of reflective surfaces makes it possible to identify more complex player actions and thus provide more diverse instructions to the application.

Embodiment 2. FIG.
In the first embodiment, it has been described that the application image of the calculator is displayed so as to overlap the position in the frame of the object. However, in the second embodiment, an example in which a character that can be operated by a player is displayed will be described.

  FIG. 10A shows the overall configuration of the three-dimensional position specifying device 12 in the second embodiment. The arrangement of the camera 20, the image processing device 30, the display 40, and the reflector 50 is the same as that in the first embodiment. In FIG. 10A, the object 76 operated by the player is the entire player's hand. When the object 76 enters the first entry area 62, processing by each functional block is performed as described above, and the action specifying unit 132 specifies the action of the player.

  In the second embodiment, the object detection unit 116 uses two reference images that represent a state in which the palm is opened and a reference image that represents a state in which the palm is closed, as the reference images that match the object 76. Matching with the reference image is executed. By doing so, the object detection unit 116 can detect not only the position of the object in the frame but also the opening and closing of the hand. The action specifying unit 132 displays the character image with the mouth open at the position in the frame of the object when the palm is open, and displays the mouth when the hand is closed. Command the closed character image to be displayed at the position in the frame of the object.

  FIG. 10B shows the screen 44 in a state where the character image 82 corresponding to the state shown in FIG. Since the hand that is the object 76 in FIG. 10A is closed, the character image 82 with the mouth closed is displayed superimposed on the object 76 in FIG. 10B.

  As shown in FIG. 11A, when the player 72 opens the hand that is the object 76 in the first entry area 62, the object detection unit 116 detects a state in which the hand is opened. In response to this, as shown in FIG. 11B, a character image 82 with a closed mouth is superimposed on the object 76 and displayed.

  The action specifying unit 132 may cause the audio output unit 36 to output a sound in accordance with the change in the mouth of the character. For example, a voice may be emitted when the mouth is opened, and a voice may not be emitted when the mouth is closed. Thus, an application that allows the player to speak the character by opening and closing the hand in the first entry area 62 can be realized.

  The object detection unit 116 may have a plurality of stages of reference images from when the hand is closed to when the hand is opened, and may detect the degree of opening of the hand by matching using them. In this case, the action specifying unit 132 may instruct the display control unit 134 to change the opening degree of the mouth of the character image according to the degree of opening of the hand. Further, the action specifying unit 132 may instruct the voice output unit 36 to change the loudness, pitch, and voice color according to the opening degree of the character's mouth. In this case, a plurality of audio data is prepared in an audio data storage unit (not shown), and the audio output unit 36 searches for and outputs appropriate audio data in accordance with an instruction from the action specifying unit 132.

  Also in the second embodiment, the first entry area 62 and the second entry area 64 may be used for on / off of a specific function. As an example, when the object 76 enters the first entry area 62, the display of the character image 82 is started, and only when the object 76 is located in the second entry area 64, a sound is emitted in accordance with the opening and closing of the hand. May be. Even if the player 72 opens and closes the hand which is the object 76 when the object 76 is positioned in the first entry area 62, the mouth of the character image 82 in the screen is opened and closed in accordance with the movement, but no sound is produced.

  The object on which the character image is superimposed may be a part of the other body of the player 72, for example, the mouth. FIGS. 12A and 12B show such a state. Within the frame photographed by the camera 20, the object detection unit 116 detects the player's lips 78 by matching and specifies the position in the frame. Moreover, the object detection part 116 detects the opening degree of a mouth by performing a matching using a some reference image. The display control unit 134 displays a character image 84 having upper and lower lip shapes superimposed on the player's lips. The display control unit 134 changes the interval between the upper and lower lips of the character image 84 according to the opening degree.

  The action specifying unit 132 instructs the audio output unit 36 to output a sound in accordance with the opening / closing timing of the mouth. By using this, it is possible to realize an application that outputs a voice different from the player's voice, such as an animal cry or a famous person's voice, in accordance with the movement of the player's mouth.

  Another character image may be displayed at a position that does not overlap with the mirror image of the player, and the player may move the mouth by imitating how the character moves the mouth. If the movement of the mouth of the character can be matched, the voice of the character may be output.

  In the example of FIG. 12, once the player's mouth is detected by the object detection unit 116, sound is output every time the player moves the mouth thereafter. Therefore, by using depth localization using the reflector 50, sound is output only when an object such as a player's hand or finger is present in the first entry area 62 or the second entry area 64. Also good. Alternatively, as described above, when the object enters the first entry area 62, the display of the character image superimposed on the player's mouth is started, and when the object enters the second entry area 64, the movement is matched with the movement of the player's mouth. Audio output may be started. As described above, in the present invention, since the depth position is specified not by matching but by taking the difference in the reflection surface area between frames, the object used for depth localization and the object for specifying the position in the frame by matching are used. And may be different. Therefore, the depth localization using the reflector can be used only as a switch for turning on and off a specific function using object matching.

Embodiment 3 FIG.
In the first and second embodiments, the technology for specifying the three-dimensional position of an object using the reflector 50 in which the two reflecting surfaces are spaced apart in the depth direction has been described. In the first and second embodiments, matching using a reference image is performed in order to specify the position of the object in the frame. For this reason, it is necessary to store the reference image of the object in the reference image storage unit 120 in advance. Although a reference image may be stored in advance, when a part of the player's body is used as an object, it is desirable to acquire a reference image of the object for each player in order to improve recognition accuracy.

  Therefore, in the conventional technology, a reference screen of an object is acquired by displaying a predetermined screen before executing an application and prompting the player to perform an operation of photographing the object with a camera. However, this not only causes the player to perform useless operations, but also has a problem that applications such as games cannot be executed quickly.

Therefore, in the third embodiment, a technique for acquiring a reference image of an object using the same configuration as the three-dimensional localization of the object without holding the reference image of the object in advance will be described. This allows the player to quickly execute the application after installing the camera, display, and reflector.
Instead of the reflector having two reflecting surfaces arranged in the depth direction apart from each other used in the first and second embodiments, in the third embodiment, the normal of each surface is the side where the object exists. The difference is that a reflector having a first reflecting surface and a second reflecting surface that are angled so as to intersect each other and simultaneously reflect an object is used.

  FIG. 13 shows a three-dimensional position specifying device 14 according to the third embodiment. FIG. 14 shows a screen 44 displayed on the display 40 and recognized by the player in the state shown in FIG. FIG. 15 is a cross-sectional view of a plane perpendicular to the depth direction of the reflector 170, that is, the z direction.

  In FIG. 13, the basic functions and arrangement of the camera 20, the image processing device 30, and the display 40 are the same as those described in FIG. In the three-dimensional position specifying device 14, the configuration of the reflector 170 is different from that of the first embodiment. The reflector 170 has a first reflecting surface 172 and a second reflecting surface 174. As shown in FIG. 15, the first reflecting surface 172 and the second reflecting surface 174 are angled so that the normals 172d and 174d of the respective surfaces intersect on the side where the object exists, and two reflections of the object It arrange | positions so that an image may be reflected toward the camera 20 simultaneously.

  As shown in FIG. 14, the first reflecting surface 172 and the second reflecting surface 174 are composed of a plurality of strip-like reflecting surfaces 178a to 178d arranged in the depth direction. In FIG. 14, one reflecting surface is composed of four strip-like reflecting surfaces. Similar to the reflector 50, identification markers 176 are also provided at both ends of the reflector 170 in the long axis direction.

The first reflecting surface 172 and the second reflecting surface 174 may be made of a mirror, mirror-finished metal, plastic, glass on which metal is vapor-deposited, etc., as with the reflector 50. It is preferable to be configured by a microprism mirror on a plane arranged in a plane. By configuring the reflective surface with a microprism mirror, the thickness of the reflector 170 can be suppressed, and installation is easy and space-saving.
In FIGS. 13 and 15, the first reflecting surface 172, the second reflecting surface 174, and the strip-shaped reflecting surfaces 178 a to 178 d are drawn at an angle in order to make the direction of reflection easy to understand. It should be noted that the microprism mirror can provide such a reflection angle even on a substantially flat surface.

  Since the first reflecting surface 172 and the second reflecting surface 174 are arranged at an angle, as shown in FIG. 14, the object output image 70 c by the first reflecting surface 172 is displayed on the frame output from the camera 20. Then, two reflected images 70d of the object by the second reflecting surface 174 appear. That is, the state is the same as when the object 70 is captured in stereo by a plurality of cameras. Therefore, it is possible to specify the three-dimensional position of the object 70 from the two reflected images 70c and 70d using a well-known stereo photography technique.

  FIG. 16 shows a configuration of the image processing apparatus 30 according to the second embodiment. The functions of the image acquisition unit 102, the image inversion unit 104, the arrangement confirmation unit 140, the arrangement instruction unit 142, and the on-screen display unit 144 are the same as those in FIG.

  The reflecting surface area specifying unit 112 specifies the reflecting surface area based on the position of the marker 176 in the frame received from the image inverting unit 104.

  The in-frame localization unit 114 includes a stereo image analysis unit 118 in addition to the object detection unit 116. The stereo image analysis unit 118 uses the two reflection images 70c and 70d in the reflection surface area specified by the reflection surface area specification unit 112 to specify the position of the object 70 in the frame according to a known technique. The in-frame position of the object 70 can be roughly determined from the difference between the position where the reflected images 70c and 70d are reflected and the size of the reflected images 70c and 70d.

  The reference image storage unit 120 cuts out an image of a predetermined range centered on the position in the frame of the object 70 specified by the stereo image analysis unit 118 from the frame and stores it as a reference image. As shown in FIG. 15, the object 70 should be in the vicinity of the position where the normal lines of the first reflection surface 172 and the second reflection surface 174 that extend from the place where the two reflection images of the object 70 exist in the frame. is there. Therefore, a reference image of the object can be acquired by cutting out an area in the frame corresponding to the circle 180 in FIG.

  Although the accuracy of the position of the object in the frame determined by the stereo image of the reflected image is not so high, the low accuracy can be covered by cutting out an image in a wider range than the object to be detected. The size of the range for cutting out the reference image may be set to an appropriate value through experiments. The stereo image analysis unit 118 stores the image in the predetermined range in the reference image storage unit 120 as a reference image.

  When these series of processes are completed, the object detection unit 116 can perform localization and tracking of the object 70 using the reference image in the reference image storage unit 120.

  The depth localization unit 122 detects, based on the difference between frames, which of the plurality of strip-like reflection surfaces 178a to 178d reflects the reflection images 70c and 70d of the object 70 on the reflection surface. The position in the depth direction of 70 can be specified.

  FIG. 17 is a flowchart showing a procedure for executing a calculator application similar to that shown in FIGS. 5 to 8 in the third embodiment.

  After the reflector 170 is placed at an appropriate position, the object 70 and the reflector 170 are photographed by the camera 20, and the image acquisition unit 102 acquires a frame including a direct image and a reflected image of the object 70 (S40). The reflecting surface area specifying unit 112 specifies the reflecting surface area by detecting the marker 176 in the frame given from the image inverting unit 104 (S42). The reflection surface area specifying unit 112 may acquire an image of the reflection surface area before the object 70 enters the entry area 182 as a default image for taking a difference from the reflection image of the object 70. Good. The depth localization unit 122 detects that an object has entered the entry area 182 by detecting a difference between frames in the reflection surface area. In response to this, the stereo image analysis unit 118 in the frame of the object 70 based on the reflection images 70c and 70d of the two objects displayed on the reflection surface area specified by the reflection surface area specifying unit 112. A rough position is specified (S44). The stereo image analysis unit 118 cuts out an image within the predetermined range 180 centered on the specified position in the frame from the frame as a reference image to be used for matching, and stores it in the reference image storage unit 120 (S46). The subsequent steps are the same as those after S14 in FIG. 9, and the in-frame localization unit 114 specifies the position of the object in the frame using the reference image stored in the reference image storage unit 120, and the depth localization unit 122 The position in the depth direction of the object is specified by detecting a difference between frames in the reflection surface area.

  If the stereo image analysis unit 118 cannot accurately specify the position of the object in the frame, the reference image is not properly cut out, and the object detection unit 116 cannot detect the object by matching. In this case, the player may be notified that the object is to be cut out again.

  As described above, according to the third embodiment, by using a reflector including a first reflection surface and a second reflection surface that are angled so that their normals intersect on the side where the object exists. Obtain a stereo image of the reflected image of the object. By detecting the difference between the default image of the background in the reflection surface area and the image when the object enters, the timing for acquiring the stereo image of the object for cutting out the reference image can be measured. In addition, by storing the default image, the robustness of the difference detection is increased. By analyzing the stereo image, it is possible to specify a rough position in the frame of the object without executing matching of the object, so that the portion can be taken out as a reference image.

  Thus, since the operation of storing the reference image by the player is omitted at the stage of cutting out the reference image, it contributes to the quick start of the application. In other words, the player is not forced to perform a specific action, and the procedure for acquiring the reference image is hidden from the player. After the reference image is cut out, the position of the object in the frame is detected with high accuracy by matching. As described above, one of the features of the third embodiment is that both the rapid start of the application and the high position accuracy by matching are achieved.

  If the above-described reflector 170 is used, the three-dimensional position of the object can be specified only by the reflected image even if there is no direct image of the object in the frame photographed by the camera. However, in the third embodiment, in order to increase the accuracy of localization within a frame, after the reference image is successfully cut out, the position of the object in the frame is specified by matching using the reference image.

  By increasing the number of strip-like reflecting surfaces of the reflector 170 and increasing the detection accuracy of the object movement in the depth direction, a more complicated application can be realized. As an example, virtual surgery can be considered. A three-dimensional image of the surgical site is displayed on the liquid crystal stereoscopic display, and the player operates by holding a bar-shaped object with his hand instead of a surgical tool such as a scalpel. A three-dimensional position of the object is specified by the three-dimensional localization unit 110, and a three-dimensional image of the surgical site displayed on the liquid crystal stereoscopic display is changed according to the position. As an example, an image is displayed in which the surgical site is incised when the object is moved in a certain direction. At this time, LEDs may be mounted at a plurality of locations of the object, and the locus of the LED when the object is moved is detected in the frame to obtain the motion vector of the object. By doing so, it is also possible to output a predetermined sound effect from the speaker in synchronization with the movement of the object by using a fourth embodiment described later.

  If a microprism mirror is used as the reflecting surface, the angle of view can be adjusted by controlling the curvature of the uneven surface of the mirror. Therefore, the entry area for determining the entry of the object is not limited to the vertically upper side of the reflecting surface as shown in FIG. 13, and can be widened in a fan shape or conversely narrowed. If the approach area is widened, the position accuracy can be lowered, but the range in which the movement of the object in the depth direction can be detected can be widened.

Embodiment 4 FIG.
In Embodiments 1 to 3, it has been described that the three-dimensional position of the object operated by the player is specified using the reflector, and the action of the application is specified by using this to operate the function of the application. In any of these, by changing the display mode of the application image displayed on the screen, it is possible to notify the player that the action is recognized and the specific function is enabled or disabled.

  However, not only a visual notification through a change in the display mode of the application image in the display but also a sound according to the action is output from the speaker, it is advantageous that the player can be notified audibly. In this case, if the sound is output after the recognition of the object image, the recognition of the player through vision and the timing at which the player listens to the sound may be shifted, which may give the player a sense of incongruity. Therefore, in the fourth embodiment, the velocity vector of the object is detected by image processing, the expected movement time until the object reaches the virtual or real contact surface is calculated, and before the object reaches the contact surface, A technique for executing audio output will be described.

  FIG. 18 shows a configuration of the three-dimensional position specifying device 16 according to the fourth embodiment. In the figure, the camera 20, the image processing device 30, the display 40, and the reflector 170 are the same as those shown in FIG. Here, the calculator application described with reference to FIGS. 5 to 8 is considered.

  The player 72 operates the object 70. The depth localization unit 122 detects that the object 70 has entered the entry area corresponding to the strip-shaped reflection surface 178d, and the action specifying unit 132 specifies the movement of the object 70 in the camera direction, and the application execution unit and display This is notified to the control unit 134. Thereby, as described above, the display mode of the selected area of the application image is changed, and a number corresponding to the selected area is input to the calculator application.

  In the fourth embodiment, a predetermined sound effect is output from the speaker 42 together with the change in the display mode of the application image. The player 72 can have a stronger sense of operating the application via the object by listening to the sound as the display mode of the application image changes. Moreover, it can be made to recognize that the virtual contact surface (assumed contact surface) W exists in the part corresponding to the strip-shaped reflective surface 178d shown in FIG.

  FIG. 19 shows a configuration of the image linkage audio control unit 150 in the image processing apparatus 30 according to the fourth embodiment. Obviously, these functional blocks can also be realized by hardware, software, firmware, or a combination thereof. Hereinafter, each functional block will be described with reference to FIG.

  The velocity vector calculation unit 160 calculates a velocity vector of the movement of the object 70 operated by the player 72 toward the assumed contact surface W using the frame photographed by the camera 20. Specifically, the velocity vector of the object is calculated based on the difference between the reflected images between a plurality of frames. A frame in which the depth localization unit 122 determines that the object 70 has entered the entry area corresponding to the strip-shaped reflection surface 178a, and a frame in which the object has been determined to have entered the entry area corresponding to the strip-shaped reflection surface 178b or 178c. The time difference tf is calculated with reference to the frame rate of the camera 20. Also, assuming that the distance between the strip-shaped reflecting surface 178a and the strip-shaped reflecting surface 178b or 178c is ls, the velocity vector calculation unit 160 sets the depth v of the object 70, that is, the velocity v in the z direction to v = ls / tf. Calculated by

  The movement time calculation unit 156 calculates the movement time tm = li / v until the object 70 reaches the assumed contact surface W using the speed v and the distance li between the object and the assumed contact surface W.

  The delay time acquisition unit 154 acquires a delay time td until the sound emitted from the speaker 42 arranged away from the player reaches the player 72. Actually, the exact distance L from the speaker 42 to the player 72 is unknown because it differs depending on the player, but since the position where the reflector 170 is to be placed is determined, the distance L is practically a constant. Absent. If the sound speed Vs is also a constant, the delay time td is given by a constant. In this case, the delay time acquisition unit 154 may take a constant td. In another embodiment, the distance L between the speaker 42 and the player 72 may be input by the player. In this case, the delay time acquisition unit 154 calculates the delay time td using td = L / Vs.

  The sound synchronization unit 158 outputs sound that is synchronized with the action of the player from the speaker 42 with reference to the movement time tm and the delay time td. Specifically, the audio synchronizer 158 outputs a predetermined audio after the elapse of the time obtained by subtracting the delay time td from the moving time tm, starting from the shooting time of the frame used to calculate the velocity v. Accordingly, the player listens to the sound emitted from the speaker 42 substantially simultaneously with the arrival of the object on the assumed contact surface W.

  In the above example, the sound is output in accordance with the change in the display mode of the application image displayed on the display. However, the application image is not used, and the presence of the virtual contact surface is simply output to the player by the sound. It can also be recognized.

  When the assumed contact surface is a real surface, the output sound may be a type of sound that is completely different from the sound actually generated when the object and the surface come into contact with each other.

  As described above, according to the fourth embodiment, the time for the object to reach the contact surface is calculated before the object reaches the virtual or real contact surface, and the sound is considered in consideration of the audio delay. To advance. Thereby, it is possible to inform the player that the action has been recognized through both visual and auditory senses.

  Note that the audio delay at the distance between the speaker and the player is not a problem in practice, so that the audio synchronization unit 158 uses the delay time acquisition unit 154 particularly when the moving speed of the object is relatively slow. The delay time td need not be considered. In this case, when the action identifying unit 132 identifies a player's action, for example, a selected action, the voice synchronization unit 158 causes the speaker 42 to output a corresponding click sound or sound effect. Moreover, in order to improve game property, the audio | voice synchronizer 158 may output an audio | voice at timing earlier than the measured movement time.

  In this way, by outputting sound in accordance with the detection of the player action, it is possible to enhance the player's sensibility. In other words, by outputting an appropriate sound effect as an object is moved in and out of a certain entry area, the entry of the object into that entry area is detected and has a different meaning from the other areas. You can let the player know. In addition, when the player performs trial and error on the operation method of the application by the object, it is possible to guide the user to learn the operation method by outputting a sound when the object exists in an appropriate area.

  In the application shown in FIG. 12, the speed vector calculation unit 160 may calculate the opening / closing speed of the mouth using the difference information of the mouth between a plurality of frames. Then, the audio synchronization unit 158 may adjust the audio output timing so that the opening / closing of the player's mouth and the sound emitted from the speaker are synchronized using the opening / closing speed and the delay time.

Embodiment 5 FIG.
In the fourth embodiment, it is described that the velocity vector in the depth direction of the object operated by the player is calculated using the width in the depth direction of the reflector. In the fifth embodiment, a technique for estimating the velocity vector of an object from only a frame photographed by a camera without using a reflector will be described. However, in this embodiment, the moving component in the depth direction of the object cannot be detected, and only the moving component in the frame is targeted.

  FIG. 20 shows a configuration of the three-dimensional position specifying device 18 according to the fifth embodiment. The arrangement of the camera 20, the image processing device 30, and the display 40 is the same as in the above-described embodiment. In the fifth embodiment, a reflector is not used, and instead, a player operates an object 74 to which a light emitting element such as an LED is attached.

  FIG. 21 is a diagram for explaining the principle of a method for calculating the velocity vector of an object from one frame captured by the camera 20. An image sensor such as a CCD or CMOS outputs a signal according to the amount of light stored in the element, but a certain time is required to scan the entire element. More specifically, each pixel of the image sensor starts from the top row of elements and moves to the bottom row one by one, and then the readout of the accumulated light quantity starts. As with daylighting, reading starts from the top row and proceeds in sequence one row at a time to the bottom at the same speed. Accordingly, information having a time difference between the pixel column from which lighting is started and the pixel column from which readout is started is included in one frame.

  For example, in a CMOS sensor, since there is a difference in the daylighting start time in each column, if the movement of the object is fast, the image is distorted at the top of the previously read image and the bottom of the last read image (moving body distortion). Since the CMOS reads out one line at a time, if one screen is read in 1/15 seconds, a difference of 1/15 seconds appears at the beginning and end of reading.

  Therefore, when an object that emits light is moved at a high speed, a portion where the light has moved in the frame may appear as a locus 75 as shown in FIG. Therefore, if the time required to scan the first pixel and the last pixel in the image sensor is known, it is possible to calculate the speed of the object that created this locus.

  FIG. 22 shows a configuration of the image linkage audio control unit 150 according to the fifth embodiment.

  The daylighting time acquisition unit 152 acquires the daylighting time te of the image sensor 22 employed in the camera 20. This information may be input in advance or may be acquired by communicating with the camera 20.

  The trajectory measuring unit 164 receives the frame in which the trajectory 75 remains from the image inverting unit 104, and measures the length p and the direction of the trajectory included therein. The velocity vector calculation unit 160 calculates the velocity v = p / te of the object using the trajectory length p and the daylighting time te.

  Similarly to the fourth embodiment, the movement time calculation unit 156 uses the calculated speed v and the distance li between the object 74 and the assumed contact surface W to calculate the movement time tm of the object 74 to the assumed contact surface W. calculate. In the example of FIG. 21, the assumed contact surface W is virtually set in the frame, and the distance li between the object 74 and the assumed contact surface W is calculated by analyzing the frame.

  The delay time acquisition unit 154 and the audio synchronization unit 158 are the same as those in the fourth embodiment.

  FIG. 23 is a flowchart of processing for outputting sound in cooperation with an image in the fifth embodiment. First, the camera 20 takes a picture of operating the object 74 having a light emitting element such as an LED (S60). The captured frame is transferred from the image acquisition unit 102 to the trajectory measurement unit 164 in the image linkage audio control unit 150. The trajectory measurement unit 164 detects the trajectory 75 generated by the LED in the frame, measures the length and direction of this trajectory, and passes the result to the velocity vector calculation unit 160 (S62). The velocity vector calculation unit 160 calculates the velocity v of the object using the length and direction of the trajectory and the daylighting time of the image sensor (S64). Subsequently, the movement time calculation unit 156 calculates the movement time tm of the object to the contact surface using the distance li to the assumed contact surface W and the speed v (S66). The audio synchronizer 158 calculates the audio output timing by subtracting the delay time td from the moving time tm, starting from the shooting time of the frame used to calculate the velocity vector (S68). And according to an output timing, a predetermined audio | voice is output from the audio | voice output part 36 (S70). Thereby, the player 72 listens to the sound emitted from the speaker 42 substantially at the same time when the object reaches the assumed contact surface W.

  As described above, according to the fifth embodiment, a state in which an object is moved by attaching a light emitter is photographed, the lighting time in the image sensor of the camera, and the light emitter in the frame output from the image sensor. The speed of the object can be calculated using the information of the trajectory. The fifth embodiment has a feature that the velocity information of the object can be obtained by measuring the trajectory in a single frame regardless of the difference between a plurality of frames. However, it is premised that the illuminant attached to the object is illuminated and that the locus remains in the frame.

  The present invention has been described above based on the embodiment. It is to be understood by those skilled in the art that these embodiments are exemplifications, and various modifications are possible for each component or combination of processes, and such modifications are also within the scope of the present invention. In addition, any combination of the constituent elements described in the embodiment and a representation of the present invention converted between a method, an apparatus, a system, a computer program, a recording medium, and the like are also effective as an aspect of the present invention.

  In the embodiment, the application that displays the mirror image of the player and the object on the display has been described. However, the moving image captured by the camera may not be displayed on the display.

  In order to effectively execute the application as described in the above-described embodiment, a camera capable of shooting a video at a sufficiently high frame rate, and a calculation capability and a drawing capability capable of processing such a high frame rate are provided. It is desirable to use a combination of an image processing apparatus and a display capable of displaying an image at a high frame rate.

1 is a diagram illustrating an overall configuration of a three-dimensional position specifying device according to Embodiment 1. FIG. It is the figure which simplified and demonstrated the hardware constitutions of the camera and the image processing apparatus. It is a top view which shows the structure of a reflector. It is a figure which shows the detailed structure of a process part. (A) is a figure which shows the positional relationship of an object and an approach area | region, (b) is a figure which shows the screen displayed on a display and recognized by the player. (A) is a figure which shows the positional relationship of an object and an approach area | region, (b) is a figure which shows the screen which is displayed on a display and is recognized by the player. (A) is a figure which shows the positional relationship of an object and an approach area | region, (b) is a figure which shows the screen displayed on a display and recognized by the player. (A) is a figure which shows the positional relationship of an object and an approach area | region, (b) is a figure which shows the screen displayed on a display and recognized by the player. 9 is a flowchart for executing the application described with reference to FIGS. 5 to 8 in the three-dimensional position specifying apparatus according to the first embodiment. (A) is a figure which shows the positional relationship of an object and an approach area | region, (b) is a figure which shows a mode that a character image is displayed according to the hand which is an object. (A) is a figure which shows the positional relationship of an object and an approach area | region, (b) is a figure which shows a mode that a character image is displayed according to the hand which is an object. (A), (b) is a figure which shows the application example which displays a character image according to the mouth which is an object. FIG. 10 is a diagram illustrating a configuration of a three-dimensional position specifying device according to a third embodiment. It is a figure which shows the screen which is displayed on a display and is recognized by the player in the state shown in FIG. It is sectional drawing of the plane perpendicular | vertical to the depth direction of a reflector. 6 is a diagram illustrating a configuration of an image processing apparatus according to Embodiment 3. FIG. 10 is a flowchart showing a procedure for executing a calculator application similar to that shown in FIGS. 5 to 8 in the third embodiment. It is a figure which shows the structure of the three-dimensional position identification apparatus which concerns on Embodiment 4. FIG. FIG. 10 is a diagram illustrating a configuration of an image linkage audio control unit in an image processing apparatus according to a fourth embodiment. FIG. 10 is a diagram illustrating a configuration of a three-dimensional position specifying device according to a fifth embodiment. It is a figure explaining the principle of the method of calculating the speed vector of an object from one flame | frame image | photographed with the camera. FIG. 10 is a diagram illustrating a configuration of an image linkage audio control unit in a fifth embodiment. In Embodiment 5, it is a flowchart of the process which outputs an audio | voice in cooperation with an image.

Explanation of symbols

  10 three-dimensional position specifying device, 20 camera, 22 image sensor, 24 image processing unit, 30 image processing device, 34 image output unit, 36 audio output unit, 40 display, 42 speaker, 50 reflector, 52 first reflecting surface, 54 second reflective surface, 56 marker, 62 first entry area, 64 second entry area, 70 object, 70a reflected image, 70b reflected image, 72 player, 74 object, 75 locus, 76 object, 80-84 application image, 110 Three-dimensional localization unit, 112 Reflection surface area specifying unit, 114 In-frame localization unit, 120 Reference image storage unit, 122 Depth localization unit, 132 Action specifying unit, 134 Display control unit, 150 Image linkage audio control unit, 154 Delay time Acquisition part, 1 6 traveling time period calculation unit, 158 audio synchronization unit, 160 velocity-vector calculating unit, 170 reflector, 172 a first reflecting surface, 174 second reflecting surface, 176 markers, 180 cut-out region, 182 entry regions.

Claims (16)

A camera that shoots an object operated by the player;
A reference image storage unit for storing a reference image for specifying the object;
Reflecting means for projecting a reflected image of a predetermined entry area assumed to be entered by the object toward the camera, and causing the camera to capture a direct image and a reflected image of the object;
A reflection surface region specifying unit for specifying a reflection surface region that is a portion corresponding to the reflection image from a frame photographed by the camera;
An in-frame localization unit that performs matching using the reference image for the frame and identifies the position of the object in the frame;
A depth localization unit that detects the reflected image from the reflective surface region and identifies the position of the object in the depth direction along the optical axis of the camera;
An apparatus for specifying a three-dimensional position of an object.   2. The three-dimensional image according to claim 1, wherein the depth localization unit detects a difference between the reflected images between a plurality of frames and determines whether or not the object is located in the entry area. 3. Positioning device. The reflecting means has a marker for notifying the reflecting surface area specifying unit of the position of the reflecting means,
The three-dimensional position specifying device according to claim 1, wherein the reflecting surface area specifying unit specifies the reflecting surface area by detecting the marker from within the frame. The reflection means has a first reflection surface and a second reflection surface that are angled so that the normals of the respective surfaces intersect on the side where the object exists and reflect the object at the same time. Arranged to project towards the camera,
The in-frame localization unit identifies the position of the object in the frame using the two reflected images,
The said reference image memory | storage part cuts out the image of the predetermined range centering on the position specified by the said in-frame localization part from the said frame, and memorize | stores it as the said reference image. 3D positioning device.   The three-dimensional position specifying apparatus according to claim 4, wherein the in-frame localization unit tracks the object by executing matching using a reference image acquired in the frame. The reflecting means has a predetermined width in the depth direction, and a plurality of entry areas are set in the depth direction,
The depth localization unit determines whether the object has entered the plurality of entry areas,
5. The three-dimensional position specification according to claim 4, further comprising an action specifying unit that detects a movement of the object in the depth direction based on the determination of the depth localization unit and specifies a player's action. apparatus. The reflection means has a first reflection surface and a second reflection surface that are arranged apart from each other in the depth direction, and is arranged so as to project reflected images of different entry areas onto the camera,
3. The three-dimensional position specification according to claim 1, wherein the depth localization unit determines whether or not the object is located in an entry area corresponding to the first reflection surface or the second reflection surface, respectively. apparatus.   8. The apparatus according to claim 7, further comprising an action specifying unit that detects a movement of the object in the depth direction between the first reflection surface and the second reflection surface based on the determination of the depth localization unit. The described three-dimensional position identification device.   When the object enters an entry area corresponding to the first reflection surface, or enters an entry area corresponding to the second reflection surface, it is superimposed on a direct image of the object photographed by the camera, The three-dimensional position specifying apparatus according to claim 7, further comprising a display control unit configured to display an image to be displayed to achieve a predetermined object on a display. A display control unit for displaying on the display an image to be displayed in order to achieve a predetermined purpose, superimposed on a direct image of the object photographed by the camera;
The display control unit displays the images in different display modes depending on whether the object is located in an entry area corresponding to the first reflection surface and an entry area corresponding to the second reflection surface. The three-dimensional position specifying device according to claim 7. The action specifying unit specifies the movement of the object in the depth direction toward the camera and the movement away from the camera as a player action,
The three-dimensional position specifying apparatus according to claim 6, further comprising an input control unit that gives different functions to each action. An arrangement instructing unit that displays on the display a display instructing the player where to place the reflecting means together with the image of the player taken by the camera;
An arrangement confirmation unit that confirms whether or not the reflection means is arranged at a position to be arranged with reference to a frame photographed by the camera;
The three-dimensional position specifying device according to claim 1, further comprising: A velocity vector calculation unit that calculates a velocity vector of the object based on a difference between the reflected images between a plurality of frames;
A delay time acquisition unit for acquiring a delay time until a sound emitted from a speaker arranged away from the player reaches the player;
An audio output unit that outputs audio synchronized with the player's action with reference to the velocity vector and the delay time from a speaker;
The three-dimensional position specifying device according to claim 1, further comprising: A camera that shoots an object operated by the player;
Reflecting means for projecting a reflected image of a predetermined entry area assumed to be entered by the object toward the camera;
A reflection surface region specifying unit for specifying a reflection surface region that is a portion corresponding to the reflection image from a frame photographed by the camera;
A depth localization unit that detects the reflected image from the reflection surface region and determines whether the object has entered a portion defined by the entry region in a depth direction along the optical axis of the camera;
An apparatus for specifying a depth position of an object. A direct image of the object operated by the player and a reflected image obtained by reflecting the object by the reflecting means arranged in a predetermined depth direction along the optical axis of the camera are photographed by the camera.
An image processing device performs matching using a reference image prepared in advance for specifying the object on a frame photographed by the camera, specifies the position of the object in the frame, and wherein identifying the reflective surface area is a portion corresponding to the reflected image, detects the reflected image from the reflective surface area, the depth position of the from the difference of the reflected image object Jo Tokoro between a plurality of frames from A depth position specifying method that detects whether or not there is a depth . A reflecting means for projecting a reflected image of a predetermined entry area assumed to be entered by an object operated by a player toward the camera is arranged, and then the direct image of the object and the projected image are projected by the reflecting means. Take a reflection image with the camera ,
An image processing device performs matching using a reference image prepared in advance for specifying the object on a frame photographed by the camera, specifies the position of the object in the frame, and A reflection surface area corresponding to the reflection image is identified , the reflection image is detected from the reflection surface area , and the position of the object in the depth direction along the optical axis of the camera is identified. A method for specifying a three-dimensional position of an object.

Images