JP4807879B2 - Image recognition apparatus, method, program, and information recording medium for extracting at least one recognition target image from input image data - Google Patents

Description

  The present invention relates to an image recognition apparatus, method, program, and information recording medium for extracting at least one recognition target image from input image data.

2. Description of the Related Art Conventionally, a pattern processing apparatus that extracts a recognition target image such as a tarot card from an image captured by a camera is known (see, for example, Patent Document 1).
JP-A-2005-242600

  However, in the pattern processing apparatus, a recognition target image corresponding to the tarot card is extracted from image data obtained by arranging the tarot cards on a single color sheet member and the background of the recognition target image is a single color. It is assumed that.

  Therefore, when an image is captured without using the single color sheet member, it is impossible to remove an image other than the recognition target image in the background of the recognition target image, and the recognition target image cannot be accurately extracted. There was a problem.

  The present invention has been made in view of the problems as described above, and an object thereof is to effectively remove images other than the recognition target image from the image data and reliably extract the recognition target image. It is an object to provide an image recognition apparatus, method, program, and information recording medium.

  (1) A computer-executable program for extracting a recognition target image from input image data, the contour of a pixel group consisting of continuous pixels having pixel information satisfying a predetermined pixel condition from the image data A contour extraction unit that performs a contour extraction process, a contour determination unit that determines whether or not a contour trajectory extracted by the contour extraction unit satisfies a predetermined trajectory condition, and a predetermined trajectory by the contour determination unit A pixel information changing unit that changes pixel information of each pixel constituting the contour extracted by the contour extracting unit to pixel information other than the pixel information satisfying the predetermined pixel condition when it is determined that the condition is not satisfied; When the contour determination unit determines that a predetermined trajectory condition is satisfied, image data corresponding to the inner region of the contour extracted by the contour extraction unit is obtained. A recognition target image extracting unit that extracts a serial recognition target image, and characterized by causing a computer to function as a.

  The present invention also relates to an image recognition apparatus including the above-described units. The present invention is also a computer-readable information storage medium, and relates to an information storage medium storing a program that causes a computer to function as each of the above-described units. The present invention also relates to an image recognition method for performing the processing of each of the above parts.

  Pixel information is information held by a pixel, for example, color data of three colors (RGB), grayscale gradation data, binary data representing white or black, luminance data, color difference data, and the like.

  Whether the pixel information satisfies a predetermined pixel condition may be determined based on whether the value of the image information is a specific value, or whether the value of the image information is equal to or greater than a predetermined threshold (or It may be determined based on whether or not:

  Whether or not the contour locus satisfies a predetermined locus condition may be determined based on whether or not the shape of the contour locus is a predetermined shape.

  A pixel group composed of continuous pixels refers to a pixel group in which each pixel belonging to the pixel group is adjacent to any other pixel in the vertical direction, the horizontal direction, or the diagonal direction.

  According to the present invention, it is possible to effectively remove images other than the recognition target image from the image data and reliably extract the recognition target image. Even if the recognition target image is surrounded by images other than the recognition target image, the recognition target image can be reliably extracted.

  (2) In the program and the information storage medium according to the present invention, the contour extraction unit detects a pixel having pixel information satisfying a predetermined pixel condition from the image data, and uses the detected pixel as a start point and sets it as a start point again. The adjacent pixels having the pixel information satisfying the predetermined pixel condition are traced according to a predetermined rule until reaching the predetermined pixel, and the contour extraction processing is performed based on the coordinate value of each traced pixel.

  An adjacent pixel refers to a pixel adjacent in the vertical direction, the horizontal direction, or the diagonal direction.

  The predetermined rule is a rule used to determine a pixel to be traced next among neighboring pixels having pixel information satisfying a predetermined pixel condition. Priority order may be used.

  According to the present invention, it is possible to effectively remove images other than the recognition target image from the image data and reliably extract the recognition target image.

  (3) In the program and the information storage medium according to the present invention, the contour extraction unit scans each pixel constituting the image data in a predetermined order, and has pixel information satisfying a predetermined pixel condition in the scanning. In this case, the contour extraction process is performed when the pixel is detected, and the process of restarting scanning from the pixel to be scanned next to the detected pixel after the contour extraction process is repeated until the scanning of all the pixels is completed.

  The process of resuming scanning is performed after the processes by the contour determination unit, the pixel information change unit, and the recognition target image extraction unit performed subsequent to the contour extraction process.

  According to the present invention, even when there are a plurality of recognition target images in the image data, all the recognition target images can be extracted.

  (4) In the program and the information storage medium according to the present invention, the contour determination unit determines whether or not the pixel sequence that forms the contour extracted by the contour extraction unit forms a vertex of a polygon. A vertex of the contour is detected, and it is determined based on the detected vertex of the contour whether or not the contour locus satisfies a predetermined locus condition.

  The determination of the contour trajectory based on the contour vertices may be performed based on the number of detected vertices, or based on the vertex position, the vertex angle, or a combination thereof.

  According to the present invention, when the shape of the recognition target image is a polygon, the shape can be easily determined, and an image having a shape different from the recognition target image is effectively removed from the image data and recognized. The target image can be reliably extracted.

  (5) In the program and the information storage medium according to the present invention, as a recognition target image determination unit that determines whether the recognition target image extracted by the recognition target image extraction unit is a recognition target image satisfying a predetermined condition. The computer is made to function, and the pixel information changing unit is in an inner area of the contour extracted by the contour extracting unit when the recognition target image determining unit determines that the image is a recognition target image satisfying a predetermined condition. When the pixel information of each pixel is changed to pixel information other than the pixel information that satisfies the predetermined pixel condition, and the recognition target image determination unit determines that the image is not a recognition target image that satisfies the predetermined condition, the contour extraction unit The pixel information of each pixel constituting the contour extracted by the above is changed to pixel information other than the pixel information satisfying the predetermined pixel condition.

  Whether or not the recognition target image satisfies a predetermined condition may be determined based on whether or not predetermined information can be read from the recognition target image.

  According to the present invention, a pixel having pixel information satisfying a predetermined condition can be prevented from being detected again from the region where the recognition target image has been extracted, and the processing burden can be reduced. Further, even when there is an image having the same shape as the recognition target image in the image data, such an image can be effectively removed and only the regular recognition target image can be reliably extracted.

  (6) In the program and the information storage medium according to the present invention, a computer functions as a binarization processing unit that performs binarization processing on the image data, and the contour extraction unit is The contour is extracted from the image data that has been binarized.

  According to the present invention, it is possible to simplify processing and reduce processing load by using image data that has been subjected to binarization processing.

  (7) In the program and the information storage medium according to the present invention, the differential processing unit that performs differential processing on the image data, and the pixel of interest and the image data that is subjected to differential processing by the differential processing unit The computer is caused to function as a blurring processing unit that performs blurring processing that averages pixel information of surrounding pixels arranged in the periphery to obtain new pixel information of the pixel of interest, and the binarization processing unit blurs by the blurring processing unit. A binarization process is performed on the processed image data.

  According to the present invention, by performing differential processing, blurring processing, and binarization processing on image data, noise due to uneven illumination, insufficient lighting, and camera performance is suppressed, and processing after the contour extraction processing is performed. Suitable image data can be obtained.

  Hereinafter, this embodiment will be described. In addition, this embodiment demonstrated below does not unduly limit the content of this invention described in the claim. In addition, all the configurations described in the present embodiment are not necessarily essential configuration requirements of the present invention.

  FIG. 1 shows an example of a functional block diagram of a game system to which the image recognition apparatus of this embodiment is applied. Note that the game device of this embodiment may have a configuration in which some of the components (each unit) in FIG. 1 are omitted.

  The imaging unit 150 is for imaging a recognition object, and the function can be realized by a camera using a CCD as an imaging element. In addition, the imaging unit 150 performs a process of outputting image data, which is a set of grayscale gradation data (pixel information) corresponding to pixels of each coordinate in the imaging range, to the processing unit 100.

  FIG. 2A is a diagram for describing imaging of a recognition target by the imaging unit 150. The imaging unit 150 images a plurality of cards 14 (recognition objects) arranged on the stage 12 by the player from below the stage 12. The stage 12 is made of a transparent glass plate or the like so that the imaging unit 150 can image the card 14.

  FIG. 2B is a diagram for explaining a card that is a recognition object in the present embodiment. A pattern 16 for identifying each card is printed on the back surface of the card 14 (the surface in contact with the stage 12). A blank portion 18 having a predetermined width is provided around the pattern 16. In the present embodiment, the shape (rectangle) of the outer edge of the pattern 16 represented on each card 14 is determined from image data obtained by imaging a plurality of cards 14, and corresponds to the area of the pattern 16 (area surrounded by the blank portion 18). The image data to be extracted is extracted as a recognition target image.

  The operation unit 160 is used by the player to input operation data, and the function can be realized by a lever, a direction instruction key, a button, or the like.

  The storage unit 170 serves as a work area for the processing unit 100, the communication unit 196, and the like, and its function can be realized by a RAM (VRAM) or the like.

  An information storage medium 180 (a computer-readable medium) stores programs, data, and the like, and the function can be realized by an optical disk (CD, DVD), a memory card, a hard disk, a memory (ROM), or the like. . The processing unit 100 performs various processes of the present embodiment based on a program (data) stored in the information storage medium 180. That is, the information storage medium 180 stores a program for causing a computer to function as each unit of the present embodiment (a program for causing a computer to execute processing of each unit).

  The display 190 is for displaying and outputting an image generated according to the present embodiment.

  The sound output unit 192 outputs the sound generated by the present embodiment, and its function can be realized by a speaker, headphones, or the like.

The portable information storage device 194 stores player personal data, game save data, and the like. Examples of the portable information storage device 194 include a memory card and a portable game system.

  The communication unit 196 performs various controls for communicating with the outside (for example, a host device or other game system), and functions thereof include hardware such as various processors or communication ASICs, programs, and the like. Can be realized.

  Note that a program (data) for causing a computer to function as each unit of this embodiment is distributed from the information storage medium of the host device (server) to the information storage medium 180 (storage unit 170) via the network and communication unit 196. May be. Use of the information storage medium of such a host device (server) can also be included in the scope of the present invention.

  The processing unit 100 (processor) performs game calculation processing, image generation processing, sound generation processing, and the like based on operation data from the operation unit 160, a program, and the like. The processing unit 100 performs various processes using the storage unit 170 as a work area. The functions of the processing unit 100 can be realized by hardware such as various processors (CPU, DSP, etc.), ASIC (gate array, etc.), and programs.

  The processing unit 100 includes a game calculation unit 110, an image generation unit 130, and a sound generation unit 140.

  The game calculation unit 110 performs game calculation processing for generating game images and game sounds. Here, as the game calculation process, a process for determining the game content and game mode, a process for starting the game when the game start condition is satisfied, a process for proceeding with the game, and a game parameter that changes depending on the game play are calculated. There is a process or a process of ending a game when a game end condition is satisfied.

  The game calculation unit 110 includes an image recognition unit 120. The image recognizing unit 120 extracts image data corresponding to the pattern portion of the card from the image data output by the image capturing unit 150 as a recognition target image, and identifies identification information and position of the card imaged from the extracted recognition target image. Perform the acquisition process. And the game calculation part 110 performs game calculation based on the identification information, position information, etc. of the acquired card | curd.

  The image recognition unit 120 includes a differentiation processing unit 121, a blur processing unit 122, a binarization processing unit 123, a contour extraction unit 125, a contour determination unit 126, a pixel information change unit 127, a recognition target image extraction unit 128, and a recognition target image determination. Part 129.

  The differential processing unit 121 performs secondary differential processing using the gradation data (pixel information) of the pixel of interest and surrounding pixels on the image data output from the imaging unit 150.

  FIG. 3 is a diagram illustrating a specific example of a 3 × 3 secondary differential operator used when performing the secondary differential processing. In this embodiment, the weighting coefficient “8” is set for the pixel of interest at the center, and the weighting coefficient “−1” is set for each of the eight adjacent pixels, and the sum of the values obtained by multiplying these pixels by the weighting coefficient is the sum. Then, by multiplying by a predetermined gain (for example, 1), secondary differential data corresponding to this pixel of interest is obtained. It should be noted that for the target pixel in the outer peripheral portion where there are no eight adjacent pixels in the surrounding area, the calculation may be performed on the assumption that there are adjacent pixels having the same gradation data as each target pixel. The secondary differentiation process may not be performed on the pixel.

  The blurring processing unit 122 performs blurring processing on the secondary differential data that has been subjected to the secondary differential processing by the differential processing unit 121, using the secondary differential data of the pixel of interest and its surrounding pixels.

  FIG. 4 is a diagram illustrating a specific example of a blur filter used in the blur process. In the present embodiment, the same weighting is applied to each of the pixel of interest at the center and eight neighboring pixels around it. Specifically, by summing the values obtained by multiplying the secondary differential data of the pixel of interest and each of the eight adjacent pixels by the coefficient value “1”, this is multiplied by a predetermined gain (for example, 1/9). New secondary differential data of the pixel of interest is obtained. That is, a blurring process for averaging the gradation data for nine pixels including the target pixel is performed. Note that the calculation is performed on the target pixel in the outer peripheral portion where there are no eight adjacent pixels in the periphery, assuming that there is an adjacent pixel having the same gradation data as the target pixel. This blurring process may be applied multiple times.

  The binarization processing unit 123 converts the new secondary differential data output from the blurring processing unit 122 into binary data by comparing it with a predetermined threshold value STH. When secondary differential data> STH, binary data = 1 is set. When secondary differential data ≦ STH, binary data = 0 is set.

  Note that the new binarized data output from the blurring processing unit 122 may be subjected to the above binarization processing by performing type conversion (cast) to an unsigned integer. By performing type conversion to an unsigned integer, the negative value of the new secondary differential data can be made near the maximum positive value, and the contrast of the image after binarization processing can be increased.

  FIG. 5 is a diagram for explaining the contents of processing by the differentiation processing unit 121, the blurring processing unit 122, and the binarization processing unit 123 described above. As shown in FIG. 5A, the second order differential operator shown in FIG. 3 is used for the image data for the two boundary portions where the pixel color changes from white to black and from black to white. When differential processing is performed, secondary differential data corresponding to the shape shown in FIG. 5B is obtained. When the blur filter shown in FIG. 4 is applied three times to the secondary differential data, new secondary differential data corresponding to the shape shown in FIG. 5C is obtained. When the blurred second-order differential data is cast to an unsigned integer, second-order differential data in which the negative value of the boundary portion is close to the maximum positive value as shown in FIG. 5D is obtained. When binarization processing is performed with the cast secondary differential data as the threshold value STH = 0, binary data in which the boundary portion is emphasized as shown in FIG. 5E can be obtained. In FIG. 5 (A), the shape that changes from white to black and from black to white has a gentle rise to the right. Can be obtained. Comparing the binary data after the binarization processing with the original image data shown in FIG. 5A, it can be seen that the color of the pixel is reversed in black and white.

  FIG. 6 is a diagram illustrating a specific example of an image (binarized image) on which binarization processing has been performed. In FIG. 6, each square area corresponds to each pixel. In addition, pixels in the hatched area are pixels with binary data (pixel information) = 0, and other areas are pixels with binary data (pixel information) = 1. The binarized image shown in FIG. 6 is a case where the number of pixels is reduced in order to simplify the binarization processing and the contour extraction processing described later. The case of being arranged on 12 is shown. In this example, not only the card 14 but also images other than the card are mixed. In the following description, a pixel with binary data = 1 is referred to as a “white pixel”, and a pixel with binary data = 0 is referred to as a “black pixel”.

  The contour extraction unit 125 scans each pixel constituting the binarized image in a predetermined order to detect the first white pixel that reaches first, and returns other white pixels adjacent to the white pixel to the first white pixel. The process of extracting the outline of the pixel group consisting of adjacent white pixels is performed by tracing (tracing) in order until it comes.

  FIG. 7 is a diagram illustrating a specific example of the contour extraction process. In the example shown in FIG. 7, the scan start point SS is set at the upper left of the binarized image, each pixel is scanned sequentially in the right direction, and finally ends at the lower right scan end point SE. To do. In such a scanning operation, the contour extraction unit 125 extracts the first detected white pixel as the trace start point TS1, and again reaches the trace start point TS in the order indicated by the solid line arrow in FIG. A process of tracing white pixels is performed.

  Details of the contour extraction process will be described below. First, the contour extraction unit 125 examines the surrounding eight pixels centering on the position of the white pixel at that time, and determines the position of the white pixel to be traced next.

  FIG. 8 is a diagram for explaining a method of determining the next white pixel in the contour extraction process described above. In FIG. 8, “C” indicates the current pixel of interest (white pixel), and “P” indicates the position of the pixel (white pixel) immediately before the current pixel. The numbers “1” to “8” indicate the priority order of selection of white pixels to be followed next. For example, if the immediately preceding white pixel is located directly above the current pixel of interest, look at each of the upper left, left, lower left, lower, lower right, right, upper right, and upper pixels in order. The white pixel that appears is determined as the white pixel to be followed next. However, since the pixel corresponding to “1” will not be the white pixel to be traced next no matter where the immediately preceding white pixel is located, It is sufficient to examine seven pixels “2” to “8”.

  In this way, when the process of tracing the white pixels in order from the trace start point TS1 is completed, the contour extracting unit 125 extracts the contour based on the coordinate values of the white pixels traced in order.

  Further, the contour extracting unit 125 changes each pixel in the inner region of the contour to a black pixel after processing for changing each white pixel constituting the contour to a black pixel by the pixel information changing unit 127 described later. After the process is performed, the contour extraction process is restarted from the pixel (next scan target pixel) on the right side of the trace start point TS1.

  9 to 12 are diagrams for explaining the flow of the contour extraction process. In FIG. 9, since it is determined that the shape of the contour locus extracted in FIG. 7 is not a predetermined shape, the pixel information changing unit 127 described later determines the white pixels that have formed the contour starting from the trace start point TS1. It has been changed to black pixels. The contour extraction unit 125 resumes the scanning operation from the pixel immediately to the right of the trace start point TS1, extracts the next detected white pixel as the trace start point TS2, and traces the white pixels in the order indicated by the solid line arrows. A process for extracting a contour based on the coordinate value of each traced white pixel is performed.

  In FIG. 10, since it is determined that the shape of the contour locus extracted in FIG. 9 is not a predetermined shape, the pixel information changing unit 127 described later determines the white pixels that have formed the contour starting from the trace start point TS2. It has been changed to black pixels. The contour extraction unit 125 restarts the scanning operation from the pixel immediately to the right of the trace start point TS2, extracts the next detected white pixel as the trace start point TS3, and traces the white pixels in the order indicated by the solid line arrows. A process for extracting a contour based on the coordinate value of each traced white pixel is performed.

  In FIG. 11, since it is determined that the shape of the contour locus extracted in FIG. 10 is not a predetermined shape, the pixel information changing unit 127 described later determines the white pixels that have formed the contour starting from the trace start point TS3. It has been changed to black pixels. The contour extraction unit 125 resumes the scanning operation from the pixel right next to the trace start point TS3, extracts the next detected white pixel as the trace start point TS4, and traces the white pixels in the order indicated by the solid line arrows. A process for extracting a contour based on the coordinate value of each traced white pixel is performed.

  In FIG. 12, since the shape of the contour locus extracted in FIG. 11 is a predetermined shape and is determined to be a regular recognition target image, a pixel information changing unit 127 described later is extracted after the recognition target image is extracted. Thus, each pixel in the inner area of the contour is changed to a black pixel. The contour extraction unit 125 restarts the scanning operation from the pixel adjacent to the right of the trace start point TS4, scans to the scan end point SE, and ends.

  The contour determination unit 126 performs a process of determining whether or not the shape of the contour locus extracted by the contour extraction unit 125 is a predetermined shape. In this embodiment, in order to extract the pattern 16 printed on the card 12 as the recognition target image, it is determined whether or not the shape of the contour locus is a rectangle. Here, the determination as to whether or not the object is a rectangle is made by detecting the vertices of the rectangle included in the pixel row constituting the contour.

  Details of the process for determining the shape will be described below. The contour determination unit 126 moves the partial pixel sequence having a predetermined length L along the contour extracted by the contour extraction unit 125, and moves the head (leading portion) and tail (tail portion) of the partial pixel sequence. The inner product value IP of two vectors set to correspond to each is calculated.

  FIG. 13 is a diagram illustrating a specific example of a partial pixel column that moves along an outline. For example, assuming that the length L = 10, a partial pixel string composed of 10 pixels along the contour is generated. In addition, two vectors A and B having a predetermined length M (head vector A is VA and tail vector B is VB) are set in the head and tail of this partial pixel column, respectively. Yes. For example, when the length M = 3, the VA has a direction in which the first pixel of the partial pixel column is the end point and the third pixel is the start point. VB has a direction in which the last pixel in the partial pixel column is the start point and the third pixel from the back is the end point.

  By the way, each of VA and VB can be expressed as follows.

VA = (difference of X component of VA, difference of Y component of VA)
= (Vax, vay)
VB = (VB X component difference, VB Y component difference)
= (Vbx, vby)
Therefore, the inner product IP of VA and VB is
IP = VA / VB = vax × vbx + vay × vby
It becomes. Contour determining unit 126 compares the inner product value IP and a predetermined threshold TIP, | IP | in the case of ≦ TIP is a right angle crucible angle Gaho between VA and VB, the apex of the rectangular portion It is determined that it is included in the pixel column, and if | IP |> TIP, it is determined that the vertex of the rectangle is not included in the partial pixel column.

  More precisely, since the angle θ formed by VA and VB is calculated by the equation arccos (IP / (| VA | × | VB |)), the length of the vector is calculated or division is performed. Although it is necessary, since it is known that the length of the vector is included in the range of 2 to 2√2 when considering VA and VB, in this embodiment, the value of TIP is set to 2. These troublesome calculations are omitted.

  In this way, when the vertex of a rectangle is detected by moving the partial pixel row along the contour, and the four rectangular vertices are detected while the partial pixel row is moved along the contour, the contour locus is detected. It is determined that the shape of the contour is a rectangle. Otherwise, it is determined that the shape of the contour locus is not a rectangle.

  When the contour determination unit 126 determines that the shape of the contour locus is a rectangle, the recognition target image extraction unit 128 extracts binary data corresponding to each pixel in the inner region of the contour as a recognition target image. Perform the process.

  The recognition target image determination unit 129 performs a process of determining whether the recognition target image extracted by the recognition target image extraction unit 128 is a recognition target image satisfying a predetermined condition (whether it is a regular recognition target image). Do. If the identification information of the card can be read from the recognition target image extracted here (binary data corresponding to each pixel in the inner region of the contour), it may be determined that the image is a regular recognition target image.

  When the contour determining unit 126 determines that the shape of the contour locus is not rectangular, the pixel information changing unit 127 performs a process of changing each white pixel constituting the contour to a black pixel. Here, the change to the black pixel is performed by setting the binary data corresponding to each white pixel traced in order by the contour extraction unit 125 to 0.

  The pixel information changing unit 127 also extracts the contour extracted by the contour extracting unit 125 when the recognition target image determining unit 129 determines that the image is a recognition target image that satisfies a predetermined condition (a regular recognition target image). A process of changing each pixel in the inner area to a black pixel is performed. For example, as shown in FIG. 14, the minimum value Xmin and the maximum value Xmax of the X coordinate of each white coordinate constituting the contour, the minimum value Ymin and the maximum value Ymax of the Y coordinate are extracted, and each Y coordinate from Ymin to Ymax is extracted. , The binary data corresponding to each pixel from the corresponding Xmin to Xmax is set to 0.

  In addition, the pixel information changing unit 127 configures the contour extracted by the contour extracting unit 125 when the recognition target image determining unit 129 determines that the image is not a recognition target image satisfying a predetermined condition (not a regular recognition target image). A process of changing each white pixel to be a black pixel is performed.

  The image generation unit 130 performs drawing processing based on the results of various processing (game processing and image recognition processing) performed by the processing unit 100 and outputs the result to the display 190. In this case, the image generated by the image generation unit 130 may be a so-called two-dimensional image or a three-dimensional image.

  The sound generation unit 140 performs sound processing based on the results of various processes performed by the processing unit 100, generates game sounds such as BGM, sound effects, or sounds, and outputs the game sounds to the sound output unit 192.

  FIG. 15 is a flowchart for explaining an example of processing of this embodiment. Here, the processing after the binarization processing unit 123 performs the binarization processing will be described.

  Scanning of each pixel constituting the binarized image obtained by the binarization processing unit 123 is started, and it is determined whether or not all the pixels have been scanned (step S110). If all the pixels have not been scanned, the next pixel is scanned (step S120). It is determined whether the pixel information of the scanned pixel satisfies a predetermined pixel condition (whether it is a white pixel) (step S130). If not satisfied (if it is a black pixel), the process returns to step S110.

  If the pixel information of the scanned pixel satisfies a predetermined pixel condition (if it is a white pixel), the pixel is extracted as a trace start point, and the predetermined pixel condition is sequentially reached from the trace start point to the trace start point again. A pixel (white pixel) having pixel information that satisfies the above is traced (step S140). Next, a contour is extracted based on the coordinate value of each traced pixel (each white pixel) (step S150). Next, the vertex of the extracted contour is detected, and it is determined whether or not the shape of the contour locus is a predetermined shape (rectangle) based on the detected vertex (step S160). If the shape of the contour locus is not a predetermined shape (rectangular), pixel information other than pixel information satisfying a predetermined pixel condition (binary data = 0) is used for pixel information of pixels constituting the contour (traced white pixels). (Step S170), and the process returns to step S110.

  If the shape of the extracted contour locus is a predetermined shape (rectangular), a process of extracting binary data corresponding to each pixel in the inner region of the contour as a recognition target image is performed (step S180). Next, it is determined whether or not the extracted recognition target image is a regular recognition target image (step S185). If it is not a regular recognition target image, the process of step S170 is performed, and the process returns to step S110.

  When it is determined that the extracted recognition target image is a regular recognition target image, a process of acquiring a card identification number, position, and the like from the extracted recognition target image is performed (step S190). Next, a process of changing the pixel information of each pixel in the inner region of the outline to pixel information (binary data = 0) other than the pixel information that satisfies a predetermined pixel condition is performed (step S200), and the process returns to step S110. Then, when all the pixels have been scanned, the process ends.

  The present invention is not limited to that described in the above embodiment, and various modifications can be made. For example, terms cited as broad or synonymous terms in the description in the specification or drawings can be replaced with broad or synonymous terms in other descriptions in the specification or drawings.

The example of the functional block diagram of the game system to which the image recognition apparatus of this embodiment is applied. The figure for demonstrating imaging of the recognition target object by an imaging part. Specific example of 3 × 3 secondary differential operator used when performing secondary differential processing A specific example of a blur filter used in blur processing. The figure for demonstrating the content of the process by a differentiation process part, a blurring process part, and a binarization process part. A specific example of a binarized image (binarized image). A specific example of contour extraction processing. The figure for demonstrating the method of determining the next white pixel in an outline extraction process. The figure for demonstrating the flow of an outline extraction process. The figure for demonstrating the flow of an outline extraction process. The figure for demonstrating the flow of an outline extraction process. The figure for demonstrating the flow of an outline extraction process. The specific example of the partial pixel row | line | column which moves along an outline. A specific example of pixel information change processing. The flowchart for demonstrating an example of the process of this embodiment.

Explanation of symbols

100 processing unit 110 game calculation unit 120 image recognition unit 121 differentiation processing unit 122 blur processing unit 123 binarization processing unit 125 contour extraction unit 126 contour determination unit 127 pixel information change unit 128 recognition target image extraction unit 129 recognition target image determination unit 130 Image generation unit 140 Sound generation unit 160 Operation unit 170 Storage unit 180 Information storage medium 190 Display 192 Sound output unit 194 Portable information storage device 196 Communication unit

Claims (9)

A computer-executable program for extracting a recognition target image from input image data,
A contour extraction unit that performs a contour extraction process for extracting a contour of a pixel group including continuous pixels having pixel information satisfying a predetermined pixel condition from the image data;
A contour determining unit that determines whether or not a contour locus extracted by the contour extracting unit satisfies a predetermined locus condition;
Pixel information other than the pixel information satisfying the predetermined pixel condition is set as the pixel information of each pixel constituting the outline extracted by the contour extraction unit when the contour determination unit determines that the predetermined locus condition is not satisfied. A pixel information changing unit to be changed to,
A recognition target image extraction unit that extracts, as the recognition target image, image data corresponding to an inner region of the contour extracted by the contour extraction unit when the contour determination unit determines that a predetermined trajectory condition is satisfied;
Causing the computer to function as a recognition target image determination unit that determines whether or not the recognition target image extracted by the recognition target image extraction unit is a recognition target image that satisfies a predetermined condition;
The pixel information changing unit
When the recognition target image determination unit determines that the image is a recognition target image that satisfies a predetermined condition, the pixel information of each pixel in the inner region of the contour extracted by the contour extraction unit is set as the predetermined pixel condition. Change to pixel information other than satisfying pixel information,
Pixel information of each pixel constituting the contour extracted by the contour extraction unit when the recognition target image determination unit determines that the image is not a recognition target image satisfying a predetermined condition, pixel information satisfying the predetermined pixel condition A program characterized by changing to pixel information other than . In claim 1,
The contour extraction unit
A pixel having pixel information satisfying a predetermined pixel condition is detected from the image data, and an adjacent pixel having pixel information satisfying the predetermined pixel condition is determined from the detected pixel as a start point until reaching the start point again. A program that traces according to a rule and performs the contour extraction process based on the coordinate value of each traced pixel. In claim 1 or 2,
The contour extraction unit
The pixels constituting the image data are scanned in a predetermined order, and when the pixels having pixel information satisfying a predetermined pixel condition are detected in the scanning, the contour extraction process is performed, and the detected pixels after the contour extraction process A program characterized by repeatedly performing a process of restarting scanning from a pixel to be scanned next until scanning of all pixels is completed. In any one of Claims 1 thru | or 3,
The contour determination unit
The vertex of the contour is detected by determining whether or not the pixel sequence constituting the contour extracted by the contour extraction unit forms a polygonal vertex, and the locus of the contour is determined based on the detected vertex of the contour. A program for determining whether or not a predetermined locus condition is satisfied. In any one of Claims 1 thru | or 4 ,
Causing a computer to function as a binarization processing unit that performs binarization processing on the image data;
The contour extraction unit
A program for extracting the contour from image data that has been binarized by the binarization processing unit. In claim 5 ,
A differential processing unit for performing differential processing on the image data;
A blur process for performing blur processing on the image data that has been subjected to differentiation processing by the differentiation processing unit, by averaging pixel information of the pixel of interest and surrounding pixels arranged in the vicinity thereof to obtain new pixel information of the pixel of interest. The computer as a part,
The binarization processing unit
A program for performing binarization processing on image data subjected to blur processing by the blur processing section. A computer-readable information storage medium, wherein the program according to any one of claims 1 to 6 is recorded. Image recognition for extracting a recognition target image from input image data using an image recognition apparatus including a contour extraction unit, a contour determination unit, a pixel information change unit, a recognition target image extraction unit, and a recognition target image determination unit A method,
A contour extracting step in which the contour extracting unit performs a contour extracting process of extracting a contour of a pixel group composed of continuous pixels having pixel information satisfying a predetermined pixel condition from the image data;
A contour determining step for determining whether the contour trajectory extracted by the contour extracting unit satisfies a predetermined trajectory condition;
If the pixel information changing unit determines that the predetermined trajectory condition is not satisfied by the contour determining unit , the pixel information of each pixel constituting the contour extracted by the contour extracting unit is changed to the predetermined pixel condition. A pixel information changing step for changing to pixel information other than the pixel information to be satisfied;
When the recognition target image extraction unit determines that a predetermined trajectory condition is satisfied by the contour determination unit , image data corresponding to the inner region of the contour extracted by the contour extraction unit is extracted as the recognition target image. A recognition target image extraction step,
A recognition target image determination unit including a recognition target image determination step of determining whether or not the recognition target image extracted by the recognition target image extraction unit is a recognition target image satisfying a predetermined condition;
In the pixel information changing step,
Pixel information of each pixel in the inner region of the contour extracted by the contour extraction unit when the pixel information changing unit determines that the recognition target image determination unit determines that the recognition target image satisfies a predetermined condition Is changed to pixel information other than the pixel information satisfying the predetermined pixel condition, and the contour extracted by the contour extraction unit when the recognition target image determination unit determines that the image is not a recognition target image satisfying the predetermined condition The pixel information of each pixel which comprises is changed to pixel information other than the pixel information which satisfy | fills the said predetermined pixel conditions, The image recognition method characterized by the above-mentioned . An image recognition apparatus for extracting a recognition target image from input image data,
A contour extraction unit that performs a contour extraction process for extracting a contour of a pixel group including continuous pixels having pixel information satisfying a predetermined pixel condition from the image data;
A contour determining unit that determines whether or not a contour locus extracted by the contour extracting unit satisfies a predetermined locus condition;
Pixel information other than the pixel information satisfying the predetermined pixel condition is set as the pixel information of each pixel constituting the outline extracted by the contour extraction unit when the contour determination unit determines that the predetermined locus condition is not satisfied. A pixel information changing unit to be changed to,
A recognition target image extraction unit that extracts, as the recognition target image, image data corresponding to an inner region of the contour extracted by the contour extraction unit when the contour determination unit determines that a predetermined trajectory condition is satisfied;
A recognition target image determination unit that determines whether or not the recognition target image extracted by the recognition target image extraction unit is a recognition target image that satisfies a predetermined condition;
The pixel information changing unit
When the recognition target image determination unit determines that the image is a recognition target image that satisfies a predetermined condition, the pixel information of each pixel in the inner region of the contour extracted by the contour extraction unit is set as the predetermined pixel condition. Change to pixel information other than satisfying pixel information,
Pixel information of each pixel constituting the contour extracted by the contour extraction unit when the recognition target image determination unit determines that the image is not a recognition target image satisfying a predetermined condition, pixel information satisfying the predetermined pixel condition An image recognition apparatus characterized by changing to pixel information other than the above .