JP2012108849A - Recognition device and recognition method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable a device without a touch panel to recognize input operation on the basis of the motion of a user with lighter processing loads.SOLUTION: A recognition device 1 recognizing input operation on the basis of the motion of a user includes: a motion determining part 23 for determining a motion of the user on the basis of image data obtained by capturing the motion of the user; and an output part 25 for outputting information on the input operation on the basis of the determined motion of the user. A device without a touch panel can recognize input operation on the basis of the motion of the user, accordingly. The recognition device 1 includes an extraction part 21 for extracting image data represented by one-dimensional data from image data composed of one- or higher dimensional data. The motion determining part 23 determines the motion of the user on the basis of the image data represented by the one-dimensional data, thereby reducing processing loads in the determination.

Description

  The present invention relates to a recognition device and a recognition method for recognizing an input operation based on a user's operation.

  In recent years, some terminals such as a portable terminal and an electronic book terminal have a function of recognizing a page turning operation of turning a book page by sliding a finger left and right on a display. Such a function is realized, for example, by configuring the display of the terminal with a touch panel. In addition, a technique is known in which a part where a hand is captured is extracted from an image acquired by a camera and a recognized predetermined hand movement is used as input information to the apparatus (for example, see Patent Document 1).

JP 2001-56861 A

  However, a terminal not equipped with a touch panel cannot recognize a page turning operation or the like based on a finger movement with respect to the display. Since the display is mounted on the terminal in advance, it is not possible to make a terminal that is not mounted with a touch panel recognizable for page feed operation or the like. Moreover, since the display comprised with the touch panel is expensive, the terminal carrying a touch panel also becomes expensive. Further, in the technique described in Patent Document 1, since the hand is recognized by determining the color of each pixel from the entire image acquired by the camera, the processing load is large.

  Therefore, the present invention has been made in view of the above problems, and a recognition device and a recognition device that can recognize an input operation based on a user's operation with a light processing load even in a device without a touch panel. It aims to provide a method.

  In order to solve the above-described problems, a recognition apparatus according to the present invention is a recognition apparatus that recognizes an input operation based on a user's action, and a plurality of data elements represented by one-dimensional or more data are arranged in two dimensions. The acquisition means for acquiring the two-dimensional data that captures the user's action, and the data elements represented by the predetermined one-dimensional data among the one-dimensional or higher data from the two-dimensional data acquired by the acquisition means Extraction means for extracting two-dimensional data, analysis means for analyzing a time series change of one-dimensional data in the data element based on the two-dimensional data extracted by the extraction means, and one-dimensional data analyzed by the analysis means Based on the time-series change of the data, a motion determination unit that determines the user's motion, and outputs information related to the input operation based on the user's motion determined by the motion determination unit Output means.

  In order to solve the above-described problem, a recognition method of the present invention is a recognition method in a recognition device that recognizes an input operation based on a user's action, and a data element represented by one or more dimensions is two-dimensional. The acquisition step for acquiring two-dimensional data that is arranged in multiples and captures the user's action, and the two-dimensional data acquired in the acquisition step is represented by predetermined one-dimensional data of one or more dimensions. An extraction step for extracting two-dimensional data composed of data elements, an analysis step for analyzing a time series change of the one-dimensional data in the data element based on the two-dimensional data extracted in the extraction step, and an analysis in the analysis step A motion determination step for determining a user's action based on a time-series change of the one-dimensional data, and a motion determination step And an output step for outputting information related to the input operation based on the user action determined in step S1.

  According to the recognition apparatus and the recognition method of the present invention, the user's action is determined based on the two-dimensional data that captures the user's action, and information related to the input operation is output based on the determined user action. Even on-board devices, it is possible to recognize input operations based on user actions. Further, since the user's action is determined based on the data element represented by the one-dimensional data extracted from the two-dimensional data composed of the data elements represented by one or more dimensions, the processing load in the determination process is as follows. light.

  In the recognition apparatus of the present invention, the acquisition unit acquires image data including pixels represented by one-dimensional or more data and captures the user's action, and the extraction unit uses the image data acquired by the acquisition unit. Extracting image data composed of pixels represented by predetermined one-dimensional data out of one or more dimensions, and analyzing means based on the image data extracted by the extracting means, a time series of one-dimensional data It is characterized by analyzing the change of.

  In this case, the user's action is determined based on the image data that captures the user's action, and information about the input operation is output based on the determined user action. The input operation based on the user's operation can be recognized from the image acquired by the above. In addition, since the user's action is determined based on the image data represented by the one-dimensional data extracted from the image data composed of the pixels represented by the one-dimensional or more data, the processing load in the determination process is light.

  In the recognition apparatus of the present invention, the acquisition unit acquires image data represented by a plurality of dimension data including a luminance value, and the one-dimensional data is a luminance value.

  There is a high possibility that a portion where a part of the body such as the user's hand is captured in the image data has a luminance value different from that of other portions such as the background. According to the above configuration, the image data represented by the brightness value is extracted from the image data represented by the data of a plurality of dimensions including the brightness value, so that a part of the body such as the user's hand is appropriately recognized. It becomes possible to do.

  In the recognition apparatus of the present invention, the analysis unit sets the first and second coordinate axes that intersect with each other in the image data, and for each first coordinate value on the first coordinate axis along the direction of the second coordinate axis. The luminance value of the pixel is integrated, and a change in time series of the integrated luminance value for each first coordinate value is analyzed.

  In this case, since the integral value of the luminance value for each first coordinate value is calculated, the user's action is determined based on the time-series change of the one-dimensional data for each first coordinate value. Thereby, a user's operation | movement can be determined with a light processing load.

  In the recognition apparatus of the present invention, the analyzing means analyzes a change in a region on the first coordinate axis where the integrated luminance value for each first coordinate value is less than a predetermined threshold value. .

  A portion where a part of the body such as a user's hand is photographed in the image data is likely to have a lower luminance value than other portions. According to the above configuration, since the change in the region where the integrated luminance value for each first coordinate value is less than the predetermined threshold is analyzed, the movement of a part of the body such as the user's hand is appropriately determined. It becomes possible.

  In the recognition apparatus of the present invention, the analysis means reflects the user's hand based on the number of areas on the first coordinate axis whose integrated luminance value is less than a predetermined threshold value. The motion determination means determines the motion of the user's hand based on the motion of the region where the integrated luminance value is less than a predetermined threshold on the first coordinate axis.

  In this case, when the user's hand is captured in the image data, the number of regions on the first coordinate axis whose integrated luminance value is less than a predetermined threshold is a number depending on the shape of the hand. . According to the above configuration, since it is determined whether or not the area reflects the user's hand based on the number of areas whose luminance values are less than the predetermined threshold, the user's hand can be recognized appropriately.

  The recognition apparatus of the present invention further includes a learning unit that learns a correspondence between a time-series change of the one-dimensional data analyzed by the analysis unit and a user's action, and the motion determination unit includes the analysis unit. Based on the time-series change of the analyzed one-dimensional data, the user's action corresponding to the time-series change of the one-dimensional data is determined with reference to the association learned by the learning means. And

  In this case, the correspondence between the time-series change of the one-dimensional data in the acquired image data and the user's action is learned in advance, and the analysis of the one-dimensional data analyzed with reference to the correspondence is performed. Since the user's action is determined based on the time series change, the user's action can be reliably determined.

  The recognition apparatus according to the present invention further includes a content processing unit that rewrites content to be displayed according to information related to an input operation output from the output unit, and a display unit that displays the content rewritten by the content processing unit. It is characterized by providing.

  In this case, since the content to be displayed is rewritten according to the information related to the input operation that reflects the user's operation, it is possible to display the appropriate content according to the user's operation.

  In the recognition device of the present invention, the acquisition unit is configured by an imaging device that can be worn on the user's body.

  In this case, such means can be additionally provided to an apparatus that does not have a means for imaging the user's body or the like in advance. In addition, by attaching the acquisition unit in the vicinity of the determination target portion of the user's body, it is possible to appropriately capture the determination target portion.

  Even a device without a touch panel can recognize an input operation based on a user's operation with a light processing load.

It is a block diagram which shows the functional structure of a recognition apparatus. It is a hard block diagram of a recognition device. It is a figure which shows the example of mounting | wearing of the camera which comprises an acquisition part. It is a figure which shows the example of the image acquired by the acquisition part. It is a figure which shows the integration direction of the luminance value in image data. It is a figure which shows the change along the horizontal-axis direction of the example of the image which consists of the pixel represented with a luminance value, and the image of the integrated luminance value. It is a figure which shows the example of the change along the horizontal-axis direction of the image of the integrated luminance value. It is a figure which shows the example of the preferable arrangement position of a camera. It is a flowchart which shows the processing content of the recognition method in a recognition apparatus.

  Embodiments of a recognition apparatus and a recognition method according to the present invention will be described with reference to the drawings. If possible, the same parts are denoted by the same reference numerals, and redundant description is omitted.

  FIG. 1 is a block diagram illustrating a functional configuration of the recognition apparatus 1. The recognition device 1 of the present embodiment is a device that recognizes an input operation based on a user's operation. As shown in FIG. 1, the recognition apparatus 1 functionally includes an acquisition unit 10 (acquisition unit), a determination unit 20, a content processing unit 30 (content processing unit), and a display unit 40 (display unit). The determination unit 20 includes an extraction unit 21 (extraction unit), an analysis unit 22 (analysis unit), a motion determination unit 23 (motion determination unit), a learning unit 24 (learning unit), and an output unit 25 (output unit). . Further, the content processing unit 30 includes a content rewriting unit 31 and a content storage unit 32.

  FIG. 2 is a hardware configuration diagram of the recognition apparatus 1. As shown in FIG. 2, the recognition apparatus 1 physically includes a CPU 101, a RAM 102 and a ROM 103 which are main storage devices, a communication module 104 which is a data transmission / reception device, an auxiliary storage device 105 such as a hard disk and a flash memory, and an input device. The computer system includes an input device 106 such as a keyboard and an output device 107 such as a display. Each function shown in FIG. 1 has a communication module 104, an input device 106, and an output device 107 under the control of the CPU 101 by loading predetermined computer software on the hardware such as the CPU 101 and the RAM 102 shown in FIG. This is realized by reading and writing data in the RAM 102 and the auxiliary storage device 105.

  The recognition device 1 includes, for example, a personal computer that is wired to the acquisition unit 10 and the display unit 40, a server that is wirelessly connected to the acquisition unit 10 and the display unit 40, or a terminal device. The terminal device is exemplified by a mobile terminal and an electronic book terminal, for example. Again, with reference to FIG. 1, each function part of the recognition apparatus 1 is demonstrated in detail.

  The acquisition unit 10 is a part that acquires two-dimensional data that captures a user's action by arranging a plurality of data elements represented by two or more dimensions in a one-dimensional manner. In the present embodiment, the acquisition unit 10 acquires, for example, image data that includes pixels represented by one-dimensional or more data and captures a user's action. In the present embodiment, image data is acquired by the acquisition unit 10 as two-dimensional data that captures the user's action, and the two-dimensional data that captures the user's action is composed of pixels arranged in two dimensions. However, the present invention is not limited to this embodiment. The two-dimensional data that captures the user's movement is the two-dimensional data acquired by a sensor that can capture information on the presence and movement of a part of the user's body in the three-dimensional space on a two-dimensional plane. For example, it may be two-dimensional data in which only luminance is plotted on a plane.

  Specifically, the acquisition unit 10 is a sensor that can acquire one-dimensional or more information, and can be, for example, a camera that acquires an image by visible light or infrared light, or a distance image camera. Moreover, the image data acquired by the acquisition part 10 can consist of a pixel represented by multidimensional data, such as HSV format and RGB format, for example. By configuring the acquisition unit 10 with such a sensor, a means for acquiring an operation of the user's body or the like for an input device such as a touch panel or a device that does not have a device for imaging the user's body or the like in advance. Can additionally be provided.

  FIG. 3 is a diagram illustrating a mounting example of the camera constituting the acquisition unit 10. The acquisition unit 10 is preferably mounted at a position where the user's finger can be photographed. As illustrated in FIG. 3, for example, the acquisition unit 10 is attached to the user's wrist. Thereby, it is possible to appropriately capture an image of a finger which is a determination mode of the user's motion.

  FIG. 4 is a diagram illustrating an example of image data acquired by the acquisition unit 10. As shown in FIG. 4, image data including a user's finger is acquired by the acquisition unit 10. The acquisition unit 10 sends the acquired image data to the extraction unit 21.

  The determination unit 20 is a part that outputs information related to an input operation based on the user's action based on the image data acquired by the acquisition unit 10. Hereinafter, each function part with which the determination part 20 is provided is demonstrated in detail.

  The extraction unit 21 is a part that extracts image data composed of pixels represented by predetermined one-dimensional data of one or more dimensions from the image data acquired by the acquisition unit 10. More specifically, the extraction unit 21 can extract, for example, image data represented by a luminance value from the image data acquired by the acquisition unit 10.

  The portion of the image data where a part of the body such as the user's hand is photographed is likely to have a brightness value different from other parts such as the background, so the image data represented by the brightness value is extracted. By doing so, it becomes possible to appropriately recognize a part of the body such as the user's hand in the image data.

  In the present embodiment, the image data represented by the luminance value is extracted by the extraction unit 21, but the image data represented by data other than the luminance value may be extracted. The extraction unit 21 sends the extracted image data to the analysis unit 22.

  The analysis unit 22 is a part that analyzes a time-series change of one-dimensional data based on the image data extracted by the extraction unit 21. More specifically, the analysis unit 22 sets the first and second coordinate axes that intersect with each other in the image data, and sets the pixel value for each first coordinate value in the first coordinate axis along the direction of the second coordinate axis. The luminance value is integrated, and the change in the time series of the integrated luminance value for each first coordinate value is analyzed. With reference to FIG. 5, the processing content in the analysis part 22 is demonstrated concretely.

As illustrated in FIG. 5, the analysis unit 22 sets a first coordinate axis a _H and a second coordinate axis a _V in the image data P. In the example shown in FIG. 5, for the image data P having a rectangular shape, the first coordinate axis a _H is set as the horizontal axis, and the second coordinate axis a _V is set as the vertical axis. The method of setting the first and second coordinate axes for the image data P is not limited to the above setting, and the coordinate axes can be set along an arbitrary direction. Subsequently, analyzing unit 22, the direction of the second coordinate axis a _V, that is, along the direction of arrow R, integrating the luminance values of the pixels for each coordinate values in the first coordinate axis a _H.

FIG. 6A is a diagram illustrating an example of an image including pixels represented by luminance values. Also, FIG. 6 (b) is a diagram illustrating a change V of the integrated luminance value with respect to the coordinate values definitive first coordinate axis a _H. In FIG. 6B, symbols T and C indicate intersections with a predetermined threshold value in the integrated luminance value and a line indicating the change V in the integrated luminance value, respectively.

As shown in FIG. 6 (a) and 6 (b), on the first coordinate axis a _H, the luminance value of the region corresponding to the position where the finger is photographed is low. That is, as shown in FIG. 6A, the portion of the image data where the finger is imaged has a lower luminance value than the portion such as the background of the finger. Thus, in FIG. 6 (b), the accumulated luminance value V of the region corresponding to the finger on the first coordinate axis a _H is low.

Furthermore, the analysis unit 22 integrates the luminance value of the pixel for each coordinate value on the first coordinate axis a _H from each of the image data acquired in time series via the acquisition unit 10 and the extraction unit 21, to analyze changes in time series of the accumulated luminance value for each first coordinate value definitive axes a _H.

More specifically, the analysis unit 22, the region first coordinate axis a _H accumulated luminance value for each first coordinate value on V is less than the predetermined threshold value T, the first coordinate axis a _H direction Get the move. The predetermined threshold T may be a preset value, or may be an average value ± predetermined value of luminance values of all pixels included in the image data P extracted by the extraction unit 21. Good.

Further, the analysis unit 22, the movement of the first coordinate axis a _H direction area integrated luminance value V is less than the predetermined threshold value T, it can be obtained by various methods as shown below. For example, as shown in FIG. 6B, the analysis unit 22 detects the variation of the point C, which is located at the extreme end on the first coordinate axis a _{H and} at which the integrated luminance value V is equal to or less than a predetermined threshold T. It is good also as analyzing by acquiring. Further, the analysis unit 22, may be analyzed by comparing the variation of the first coordinate axis a _H accumulated luminance value for each first coordinate value on V over time. Furthermore, the analysis part 22 is good also as analyzing the fluctuation | variation of the integrated luminance value V using a wave function. Then, the analysis unit 22 sends the analysis result to the motion determination unit 23 and the learning unit 24.

Further, the analysis unit 22, the first coordinate axis a _H, based on the number and width of the area under the integrated luminance value is a predetermined threshold value, whether the area is a reflection of the user's hand Can be determined. With this determination process, the user's hand can be properly recognized. This determination process will be described with reference to FIG.

Figure 7 is a diagram showing an example of a variation along on the first coordinate axis a _H of the integrated luminance value V. For example, as illustrated in FIG. 7A, when there is a region where the integrated luminance value V is equal to or less than a predetermined threshold T, the analysis unit 22 reflects the user's hand (finger). It is determined that On the other hand, as illustrated in FIG. 7B, when there is no region where the integrated luminance value V is equal to or less than the predetermined threshold T, the analysis unit 22 has the user's hand (finger) in the image data P. It is determined that it is not captured.

Further, as shown in FIG. 7 (c), when the accumulated luminance value V is the width W ₁ of the region equal to or less than a predetermined threshold value T is the size of the predetermined range, the analysis unit 22, the It is determined that the area reflects the user's hand (finger). On the other hand, as shown in FIG. 7D, when the width W ₂ and the width W _{3 of the} region where the integrated luminance value V is equal to or smaller than the predetermined threshold T are larger or smaller than the predetermined range, the analysis is performed. The unit 22 determines that the area does not reflect the user's hand (finger).

  Further, since there is a possibility that the input operation is performed by the operation of any one of 1 to 5 fingers among the 5 fingers, as shown in FIG. 7E, the integrated luminance value V When the number of areas where the value is equal to or less than the predetermined threshold T is 1 to 5, the analysis unit 22 determines that the area reflects the user's hand (finger). On the other hand, as shown in FIG. 7F, when the number of regions where the integrated luminance value V is equal to or less than a predetermined threshold T is 6 or more, the analysis unit 22 determines that the region is a user's hand ( (Finger) is not reflected.

The motion determination unit 23 is a part that determines a user's action based on a time-series change of the one-dimensional data analyzed by the analysis unit 22. In the present embodiment, the motion determination unit 23, luminance values V integration in the first coordinate axis a _H is based on the movement of the region of less than a predetermined threshold value T, and determines the motion of the user's hand.

For example, when the analysis result indicating that the region where the integrated luminance value V is less than the predetermined threshold T has moved in the positive direction on the first coordinate axis a _H is acquired from the analysis unit 22, the motion determination unit 23 determines that the user's hand has moved to the right. When an analysis result indicating that the region where the integrated luminance value V is less than the predetermined threshold T has moved in the negative direction on the first coordinate axis a _H is acquired from the analysis unit 22, the motion determination unit 23 determines that the user's hand has moved leftward.

In addition, the motion determination unit 23 refers to the association learned by the learning unit 24 based on the change in the time series of the one-dimensional data analyzed by the analysis unit 22, and determines the time series of the one-dimensional data. The user's action corresponding to the change can be determined. As described below, the learning section 24, the movement and the first coordinate axis a _H direction area integrated luminance value V is less than the predetermined threshold value T, the correspondence between the left and right directions of the operation of the user's hand Since learning is performed and the association is stored, the motion determination unit 23 refers to the association stored in the learning unit 24, and based on the analysis result obtained from the analysis unit 22, Movement can be determined. The movement determination unit 23 sends a determination result regarding the movement of the user's hand to the output unit 25.

The learning unit 24 is a part that learns the correspondence between the time-series change of the one-dimensional data analyzed by the analysis unit 22 and the user's action. Specifically, the learning unit 24 acquires the movement in the first coordinate axis a _H direction of the region where the integrated luminance value V acquired from the analysis unit 22 is less than the predetermined threshold T, and the horizontal direction of the user's hand. Learning the association with the operation. For example, when the recognition device 1 is operated, the learning unit 24 performs the learning process by explicitly inputting to the recognition device 1 whether the movement is in the left or right direction of the user's hand. it can. In addition, the learning unit 24 can perform a learning process using a known learning algorithm such as SVM (support vector machine) or boost. The learning unit 24 stores and stores the learned association as a table, for example.

  The output unit 25 is a part that outputs information related to an input operation based on the user action determined by the motion determination unit 23. Specifically, for example, when the movement in the right direction of the user's hand is acquired from the movement determination unit 23, the output unit 25 outputs information related to the input operation such as “page feed of content”. When the movement of the user's hand in the left direction is acquired from the movement determination unit 23, the output unit 25 outputs information related to the input operation such as “return page of content”. In addition, the output unit 25 stores information on the input operation based on the user's operation acquired from the motion determination unit 23 by storing the correspondence between the user's operation and information about the predetermined input operation in advance in the output unit 25. Can be output.

  The content processing unit 30 is a part that rewrites content to be displayed in accordance with information related to the input operation output from the output unit 25, and includes a content rewriting unit 31 and a content storage unit 32.

  The content rewriting unit 31 rewrites the content based on the information regarding the input operation acquired from the output unit 25. Specifically, for example, when information such as “page feed of content” is acquired from the output unit 25, the content rewriting unit 31 stores the content of the page next to the content displayed on the display unit 40 as content storage. The content extracted from the unit 32 and displayed on the display unit 40 is rewritten with the extracted content. In addition, when information such as “content page return” is acquired from the output unit 25, the content rewriting unit 31 extracts the content of the page before the content displayed on the display unit 40 from the content storage unit 32. Thus, the content to be displayed on the display unit 40 is rewritten with the extracted content. The content storage unit 32 is a storage unit that stores content to be displayed on the display unit 40.

  The display unit 40 is a part that displays content, and includes, for example, a video projection device such as a display and a projector, and a device such as a head mounted display (HMD). The HMD may be an optical transmission type or a video transmission type. In the present embodiment, the display unit 40 displays the content, and when the content is rewritten by the content rewriting unit 31, the rewritten content is displayed.

  Here, with reference to FIG. 8, a preferable arrangement position of the sensor such as a camera in the present embodiment will be described. In the present embodiment, the acquisition unit 10 configured by a camera or the like may be disposed at a position A where external light is blocked by the hand H, or a surface B where the hand H contacts when the user performs an input operation. preferable. By arranging the acquisition unit 10 at such a position, the acquisition unit 10 acquires an image such that a portion where the user's hand is imaged has a smaller luminance value than the background portion of the user's hand. By arranging the acquisition unit 10 in this way, a part of the body such as the user's hand can be appropriately recognized from the image acquired by the acquisition unit 10.

  Next, the operation of the recognition apparatus 1 in the recognition method of the present embodiment will be described with reference to FIG. FIG. 9 is a flowchart showing the processing contents of the recognition method.

  First, the acquisition unit 10 acquires image data that includes pixels represented by one-dimensional or more data and captures a user's action (S1, acquisition step). Subsequently, the extraction unit 21 extracts only the luminance value from the image data acquired by the acquisition unit 10, and generates image data composed of pixels represented by the luminance value (S2, extraction step).

Next, the analysis unit 22 sets the first coordinate axis (horizontal axis) and the second coordinate axis (vertical axis) that intersect each other in the image data, and sets the pixel value for each coordinate value on the horizontal axis along the direction of the vertical axis. The luminance value is integrated (S3, analysis step). Subsequently, analyzing unit 22, the region first coordinate axis a _H accumulated luminance value for each first coordinate value on V is less than the predetermined threshold value T, the movement of the first coordinate axis a _H direction Analyze (S4, analysis step).

Next, the determination unit 20 determines whether or not the learning mode is set (S5). Learning mode, the learning learning section 24, the movement and the first coordinate axis a _H direction area integrated luminance value V is less than the predetermined threshold value T, the correspondence between the left and right directions of the operation of the user's hand It is a mode to do. If it is determined that the learning mode is set, the processing procedure proceeds to step S9. On the other hand, if it is not determined that the learning mode is set, the processing procedure proceeds to step S6.

In step S6, the motion determination unit 23, luminance values V integration in the first coordinate axis a _H is based on the movement of the region of less than a predetermined threshold value T, and determines the motion of the user's hand (S6, the motion determination Step). Subsequently, the output unit 25 outputs information related to the input operation based on the movement of the user's hand determined by the movement determination unit 23 to the content processing unit 30 (output step). The content rewriting unit 31 of the content processing unit 30 rewrites the content to be displayed on the display unit 40 based on the information regarding the input operation acquired from the output unit 25 (S7). Then, the display unit 40 displays the content rewritten by the content rewriting unit 31 (S8).

On the other hand, in step S9, the learning section 24, the correspondence of the integrated luminance value V movement and the first coordinate axis a _H direction area is less than a predetermined threshold value T, a horizontal direction of movement of the user's hand Is learned (S9). In this way, the process of this embodiment is complete | finished.

  In the recognition device 1 and the recognition method of the present embodiment described above, the motion determination unit 23 determines the user's motion based on the image data that captures the user's motion, and information on the input operation based on the determined user's motion. Since it is output by the output unit 25, it is possible to recognize an input operation based on the user's operation even in a device not equipped with a touch panel. Further, image data represented by one-dimensional data is extracted from image data composed of pixels represented by one-dimensional data or more, and the user's action is based on the image data represented by the one-dimensional data. Therefore, the processing load in the determination process is light.

  The present invention has been described in detail based on the embodiments. However, the present invention is not limited to the above embodiment. The present invention can be variously modified without departing from the gist thereof.

DESCRIPTION OF SYMBOLS 1 ... Recognition apparatus, 10 ... Acquisition part, 20 ... Determination part, 21 ... Extraction part, 22 ... Analysis part, 23 ... Determination part, 24 ... Learning part, 25 ... Output part, 30 ... Content processing part, 31 ... Content rewriting Part, 32 ... content storage part, 40 ... display part, a _H ... coordinate axis, a _V ... coordinate axis, P ... image data.

Claims (10)

A recognition device that recognizes an input operation based on a user's action,
An acquisition means for acquiring two-dimensional data in which a plurality of data elements represented by one or more dimensions of data are arranged in two dimensions and capturing a user's action;
Extraction means for extracting, from the two-dimensional data acquired by the acquisition means, two-dimensional data composed of the data elements represented by predetermined one-dimensional data of the one-dimensional or higher data;
Analysis means for analyzing a time-series change of the one-dimensional data in the data element based on the two-dimensional data extracted by the extraction means;
A motion determination unit that determines a user's action based on a time-series change of the one-dimensional data analyzed by the analysis unit;
And an output unit that outputs information related to an input operation based on a user action determined by the movement determination unit. The obtaining unit obtains image data that captures a user's action and includes pixels represented by one-dimensional or more data.
The extraction unit extracts image data including pixels represented by predetermined one-dimensional data among the one-dimensional or higher data from the image data acquired by the acquisition unit,
The recognition apparatus according to claim 1, wherein the analysis unit analyzes a time-series change of the one-dimensional data based on the image data extracted by the extraction unit. The acquisition means acquires image data represented by data of a plurality of dimensions including luminance values,
The recognition apparatus according to claim 2, wherein the one-dimensional data is a luminance value. The analysis unit sets first and second coordinate axes that intersect each other in the image data, and calculates a luminance value of a pixel for each first coordinate value in the first coordinate axis along a direction of the second coordinate axis. The recognition apparatus according to claim 3, wherein the time series change of the integrated luminance value for each of the first coordinate values is analyzed. 5. The analysis unit according to claim 4, wherein the analysis unit analyzes a change in a region on the first coordinate axis in which an integrated luminance value for each of the first coordinate values is less than a predetermined threshold value. Recognition device. The analysis means determines whether or not the region reflects the user's hand based on the number of regions whose integrated luminance values are less than a predetermined threshold on the first coordinate axis.
The said movement determination means determines the movement of a user's hand based on the movement of the area | region where the integrated brightness | luminance value on a said 1st coordinate axis is less than a predetermined | prescribed threshold value. Recognition device. Learning means for learning a correspondence between a time-series change of the one-dimensional data analyzed by the analysis means and a user's action;
The motion determination means refers to the correspondence learned by the learning means based on the time series change of the one-dimensional data analyzed by the analysis means, and determines the time series of the one-dimensional data. The recognition apparatus according to any one of claims 1 to 6, wherein a user action corresponding to the change is determined. Content processing means for rewriting content to be displayed according to information relating to the input operation output from the output means;
The recognition apparatus according to claim 1, further comprising display means for displaying the content rewritten by the content processing means.   The recognition apparatus according to claim 1, wherein the acquisition unit includes an imaging device that can be worn on a user's body. A recognition method in a recognition device that recognizes an input operation based on a user's action,
An acquisition step of acquiring two-dimensional data in which a plurality of data elements represented by one or more dimensions of data are arranged in two dimensions and capturing a user's action;
An extraction step of extracting, from the two-dimensional data acquired in the acquisition step, two-dimensional data composed of the data elements represented by predetermined one-dimensional data among the one-dimensional or higher data;
An analysis step of analyzing a time series change of the one-dimensional data in the data element based on the two-dimensional data extracted in the extraction step;
A motion determination step of determining a user's action based on a time-series change of the one-dimensional data analyzed in the analysis step;
An output step of outputting information related to an input operation based on the user action determined in the movement determination step.