JP2009135720A - Image composing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To satisfactorily compose a photographed image to a predetermined image. <P>SOLUTION: A plurality of template images are stored in a flash memory 36, and a photographed image is stored in a recording medium 30. A CPU 32 detects an attribute (attribute: center coordinates and size) of a transmission area frame provided in each of the plurality of template images. The CPU 32 also detects an attribute of a face frame that the photographed image has. The CPU 32 calculates a composite adaption degree in each of the plurality of template images based on the attribute of the detected transmission area frame and the face frame, and designates each of the plurality of template images in order of a higher calculated composite adaptation degree. The CPU 32 further composites a designated template image and a photographed image in an aspect that the composite adaptation degree is increased. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image composition apparatus that creates a composite image by multiplexing a predetermined image on a photographed image.

An example of this type of device is disclosed in Patent Document 1. According to this background art, when generating a composite image, first, a plurality of material images are arranged on the background image, and then a trimming process using a randomly selected template trimming mask is executed for each material image. . Thereafter, the “attraction level” is calculated for all of the plurality of material images. If the attraction level of the specific material image does not fall within the lower order, the trimming process is executed again. The trimming process is repeated until the degree of attraction of a specific material image falls within the lower order, thereby completing a composite image with added game characteristics and enjoyment.
JP 2005-284597 A [G06T 3/00, 7/00, 11/60, H04N 1/387]

  However, in the background art, although the shape of the template trimming mask assigned to the material image is adjusted based on the degree of attraction, the relative size / position of the material image and the template trimming mask is based on the degree of attraction. There is no adjustment. For this reason, in the background art, the quality of the composite image is limited.

  Therefore, a main object of the present invention is to provide an image synthesizing apparatus capable of satisfactorily synthesizing a captured image with a default image.

  An image synthesizing apparatus according to the present invention (10: reference numeral corresponding to the embodiment; the same applies hereinafter) detects a transmission area attribute provided in each of M (M: integer greater than or equal to 2) default images. Detection means (S87 to S93), second detection means (S77, S79) for detecting the attribute of the specific object image included in the photographed image, each of the M attributes detected by the first detection means and the second detection means A calculation means (S95) for calculating a composite fitness corresponding to a small difference from the detected attribute, and designates one of M default images in descending order of the composite fitness calculated by the calculation means. There are designated means (S99 to S105), and first composition means (S111 to S123) for synthesizing the default image designated by the designation means and the photographed image.

  The attribute of the transmissive area provided in each of the M (M: integer greater than or equal to 2) default images is detected by the first detection means. The attribute of the specific object image included in the captured image is detected by the second detection unit. The calculation means calculates a combined fitness corresponding to a small difference between each of the M attributes detected by the first detection means and the attribute detected by the second detection means. The designating unit designates one of the M default images in order from the one with the highest combination fitness calculated by the calculation unit. The first synthesizing unit synthesizes the captured image and the default image designated by the designation unit.

  As described above, the composite suitability corresponds to a small difference between the attribute of the transparent area on the default image and the attribute of the specific object image on the photographed image. The default images to be combined with the photographed image are designated in order from the highest combination fitness. As a result, a good composite image can be obtained.

  Preferably, the first detection means includes first position detection means (S93) for detecting the position of the transparent area as one of the attributes, and the second detection means detects the position of the specific object image as one of the attributes. The calculation means includes a two-position detection means (S79), and increases the degree of composite suitability as the difference between the position detected by the first position detection means and the position detected by the second position detection means is smaller. As a result, a default image having a transparent area provided near the specific image is preferentially designated.

  More preferably, rotation means (S55) for rotating the photographed image so that the specific object stands upright is further provided, and the calculation means pays attention to the position of the specific object image after being rotated by the rotation means. As a result, it is possible to satisfactorily calculate the composite fitness regardless of the posture of the specific object.

  Each of the M predetermined images has one or more transmissive areas, and the first detection unit detects a size of each of the one or more transmissive areas as another one of the attributes. The first position detecting unit further includes a detecting unit (S87), and performs the position detecting process on the transmissive area having a size that satisfies the size condition among the sizes detected by the first size detecting unit. As a result, the composite fitness is calculated in consideration of the position and size of the specific object image and the position and size of the transmission area.

  Here, the second detection means further includes second size detection means (S77) for detecting the size of the specific object image as another attribute, and the size condition is different from the size detected by the second detection means. Including the condition that is the smallest. As a result, the default image with the highest combined fitness is a default image provided with a transmission area having a size and position closest to the size and position of the specific object image.

  Preferably, extraction means (S69) for extracting a default image having a time setting from N (N: integer greater than or equal to N) default images, and the date of the photographed image among the default images extracted by the extraction means. Exclusion means (S71) for excluding default images having time settings that do not match is further provided, and of the N default images, the default images remaining after exclusion processing of the exclusion means correspond to M default images. As a result, it is possible to reduce the situation in which a default image that is not suitable for a captured image is combined with the captured image.

  More preferably, the designating means assigns the order according to the synthesis suitability to the default image having the time setting, and first assignment means according to the synthesis fit to the default image having the time setting. Includes a second assigning means (S103) for assigning a lower rank than the rank assigned by. As a result, the default image having the time setting is specified in preference to the default image having no time setting.

  Preferably, the designating unit includes an updating unit (S11) that updates the default image synthesized by the first synthesizing unit every time an image updating operation is received. This improves the operability.

  Preferably, the first combining unit adjusts the magnification of the captured image so that the size of the specific object image approaches the size of the transparent area, and the position of the transparent area approaches the position of the specific object image. Position adjustment means (S117) for adjusting the composite position as described above. The composition process is executed so that the specific object image is transmitted from the transmission area.

  More preferably, first magnification correction means (S25) for correcting the magnification of the photographed image in response to the magnification correction operation, first position correction means (S19) for correcting the composite position in response to the position correction operation, and designation A second combining means (S29) is further provided for again combining the predetermined image designated by the means and the photographed image with reference to the correction result of the first magnification correcting means and / or the first position correcting means. Thereby, the position and magnification of the specific object image appearing in the transmissive area can be corrected according to the preference of the operator.

  Further, a magnification correction amount normalizing means (S27) for normalizing the magnification correction amount by the first magnification correction means based on the size of the transparent area on the default image designated by the designation means is further provided, and the first synthesis means. The second magnification correction means (S119) for correcting the magnification of the photographed image based on the size of the transparent area on the default image designated by the designation means and the magnification correction amount normalized by the magnification correction amount normalization means. In addition.

  Thus, when the default image is updated, the size relationship between the transparent area and the specific object image included in the updated default image is matched with the size relationship between the transparent area and the specific object image included in the pre-update default image. be able to.

  Furthermore, a position correction amount normalizing means (S21) for normalizing the position correction amount by the first position correcting means based on the size of the transmission area on the default image specified by the specifying means is further provided, and the first combining means. Further includes second position correction means (S121) for correcting the composite position based on the size of the transparent area on the default image designated by the designation means and the position correction amount normalized by the position correction amount normalization means. .

  Thus, when the default image is updated, the positional relationship between the transparent area and the specific object image included in the updated default image is matched with the positional relationship between the transparent area and the specific object image included in the default image before the update. be able to.

  Preferably, there is further provided a cutting means (S125) for cutting out a part of the photographed image protruding from the outer edge of the predetermined image synthesized by the first synthesizing means. The specific object image corresponds to a face image.

  Preferably, receiving means (S47) for receiving an image selection operation for selecting a photographed image to be noticed by the first detection means, and generating a warning when the photographed image selected by the image selection operation does not have a specific object image. Generation means (S53) is further provided. This improves operability.

  Preferably, the first synthesizing unit executes the synthesizing process in such a manner that the synthesis suitability calculated by the calculating unit increases. Combined with the process of designating the default image in order from the one with the highest degree of synthesis suitability, a good synthesized image can be created.

  An image composition program according to the present invention is a first detection for detecting an attribute of a transmission area provided in each of M (M: an integer of 2 or more) predetermined images in a processor (32) of an image composition apparatus (10). Steps (S87 to S93), a second detection step (S77, S79) for detecting an attribute of the specific object image included in the captured image, each of the M attributes detected by the first detection step, and detection by the second detection step A calculation step (S95) for calculating a composite fitness corresponding to a small difference from the attribute that has been set, and specifying one of the M default images in order from the highest composite fitness calculated by the calculation step This is an image synthesis program for executing steps (S99 to S105, S11) and a synthesis step (S111 to S123) for synthesizing the default image designated in the designation step and the photographed image.

  An image composition method according to the present invention is an image composition method executed by an image composition device (10), and detects the attribute of a transparent area provided in each of M (M: an integer of 2 or more) default images. A first detection step (S87 to S93), a second detection step (S77, S79) for detecting an attribute of the specific object image included in the photographed image, a second of each of the M attributes detected by the first detection step A calculation step (S95) for calculating a composite fitness corresponding to a small difference from the attribute detected in the detection step, and one of M default images having a higher composite fitness calculated by the calculation step. A designation step (S99 to S105, S11) for designating sequentially, and a synthesis step (S111 to S123) for synthesizing the default image and the photographed image designated by the designation step are provided.

  According to the present invention, the composite suitability corresponds to a small difference between the attribute of the transparent area on the default image and the attribute of the specific object image on the photographed image. The default images to be combined with the photographed image are designated in order from the highest combination fitness. As a result, a good composite image can be obtained.

  The above object, other objects, features and advantages of the present invention will become more apparent from the following detailed description of embodiments with reference to the drawings.

  Referring to FIG. 1, a digital camera 10 of this embodiment includes an optical lens 12. The optical image representing the object scene is irradiated on the light receiving surface of the imaging device 14 through the optical lens 12. On the light receiving surface, charges representing the object scene image are generated by photoelectric conversion.

  When the camera mode is selected by the key input device 38, the CPU 32 instructs the imaging device 14 to repeat pre-exposure and thinning-out reading in order to execute through image processing. Each time the vertical synchronization signal Vsync is generated, the imaging device 14 performs pre-exposure on the light-receiving surface, and reads out a part of the charge generated by the pre-exposure from the light-receiving surface in a raster scanning manner. The vertical synchronization signal Vsync is output every 1/30 seconds. As a result, a low-resolution raw image signal representing the object scene is output from the imaging device 14 at a frame rate of 30 fps.

  The camera processing circuit 16 performs processing such as A / D conversion, white balance adjustment, and YUV conversion on the raw image signal output from the imaging device 14 to create image data in the YUV format. The created image data is written into the through image area 20a (see FIG. 2) of the SDRAM 20 through the memory control circuit 18. The LCD driver 22 reads the image data stored in the through image area 20a every 1/30 seconds, and drives the LCD monitor 24 based on the read image data of each frame. As a result, a real-time moving image representing the object scene, that is, a through image is displayed on the monitor screen.

  When the shutter button 38 s on the key input device 38 is half-pressed, the CPU 32 captures the posture information output from the tilt sensor 34. The taken posture information is written into the work area 20 f of the SDRAM 20 through the memory control circuit 18.

  The CPU 32 also executes face recognition processing based on the output from the camera processing circuit 16 in response to half-pressing of the shutter button 38s. When a person's face exists in the object scene, the number of faces and the center coordinates and size of the face frame surrounding each face are detected by face recognition processing. In addition, when there are a plurality of faces, any one of the face classifications is “main”, and the remaining face classification is “subordinate”. Such face information is also written into the work area 20 f of the SDRAM 20 through the memory control circuit 18. Note that today's date information detected by a clock circuit (not shown) is also written in the work area 20f.

  When the shutter button 38s is fully pressed, the CPU 32 instructs the imaging device 14 to execute main exposure and all pixel readout once. The imaging device 14 performs main exposure on the light receiving surface, and reads out all the charges generated by the main exposure from the light receiving surface in a raster scanning manner. As a result, a high-resolution raw image signal representing the scene is output from the imaging device 14.

  The output raw image signal of one frame is subjected to the same processing as described above by the camera processing circuit 16. The YUV format image data created by the camera processing circuit 16 is written into the expanded image area 20 b of the SDRAM 20 through the memory control circuit 18.

  The CPU 32 also gives a compression command and a recording command to the JPEG codec 26 and the I / F circuit 28, respectively, in response to the full press of the shutter button 38s. The JPEG codec 26 reads the image data stored in the decompressed image area 20 b through the memory control circuit 18, performs JPEG compression on the read image data, and compresses the compressed image data, that is, JPEG data through the memory control circuit 18. Write to the compressed image area 20c.

  The I / F circuit 28 reads JPEG data from the compressed image area 20 c through the memory control circuit 18 and stores the read JPEG data in a JPEG file created in the recording medium 30. The I / F circuit 28 also reads posture information, face information, and date information from the work area 20f through the memory control circuit 18, and embeds tag data in which these pieces of information are described in the header of the same JPEG file. When such a recording process is completed, the above-described through image process is resumed.

  A JPEG file containing captured images is configured as shown in FIG. As described above, the tag information includes date information, posture information, and face information. The date information is represented by date. The posture information is represented by any one of “1”, “2”, and “3” meaning an upright state, a left 90-degree rotated state, and a right 90-degree rotated state, respectively. Specifically, the face information includes face recognition information, face number information, and face frame information. In the face recognition information, “0” indicates that no face exists, and “1” indicates that a face exists. Regarding the face number information, the numerical value N corresponds to the number of faces. The face frame information includes the center coordinates and horizontal / vertical sizes of each face frame and the division of each face. The main face classification is “1”, and the secondary face classification is “2”. Note that the master / slave of the face is determined by, for example, the degree of smile and the size of the face.

  The JPEG data represents the captured image shown in FIG. 5A or FIG. 5B, for example. According to FIG. 5A, the scene is photographed in an upright state on December 10, 2005, and a single face image exists on the upper left side of the scene. The face frame Kp1 is defined so as to surround this face image. According to FIG. 5B, the scene is captured on August 26, 2006, rotated 90 degrees to the left, and a single face image exists on the right side of the scene. The face frame Kp2 is defined so as to surround this face image.

  When the reproduction mode is selected by the key input device 38 and a desired JPEG file is designated, the CPU 32 gives a reproduction instruction and an expansion instruction to the I / F circuit 28 and the JPEG codec 26, respectively. The playback command describes the identification number of the desired JPEG file.

  The I / F circuit 28 reads JPEG data stored in a desired JPEG file from the recording medium 30 and writes the read JPEG data into the compressed image area 20 c of the SDRAM 20 through the memory control circuit 18. The JPEG codec 26 reads JPEG data stored in the compressed image area 20 c through the memory control circuit 18, performs JPEG decompression on the read JPEG data, and decompresses the decompressed image data through the memory control circuit 18. Write to 20b.

  The LCD driver 22 reads the decompressed image data stored in the decompressed image area 20b through the memory control circuit 18, and drives the LCD monitor 24 based on the read decompressed image data. As a result, the captured image is displayed on the monitor screen.

  When an image update operation is performed on the key input device 38, a reproduction command and an expansion command are issued again. The identification number described in the reproduction command indicates the next JPEG file, and as a result, the captured image stored in the next JPEG file is output from the LCD monitor 24.

  The digital camera 10 of this embodiment has an image synthesis mode for synthesizing a template image with a recorded photographed image. The template image is prepared in the flash memory 36 (or the recording medium 30) in the JPEG file format, and the JPEG file is configured by tag data and JPEG data in the same manner as described above.

  As shown in FIG. 4, the tag data includes season setting information, season information, transparent area frame setting information, transparent area frame number information, and transparent area frame information. Regarding the season setting information, “0” indicates that there is no season setting, and “1” indicates that the season setting exists. Regarding the season information, “1” indicates spring, “2” indicates summer, “3” indicates autumn, and “4” indicates winter. Regarding the transparent area frame setting information, “0” indicates that there is no transparent area frame setting, and “1” indicates that there is a transparent area frame setting. Regarding the transparent area frame number information, N indicates the number of transparent area frames. The transparent area frame information includes center coordinates and horizontal / vertical sizes of each transparent area frame.

  The JPEG data represents, for example, a template image shown in FIG. 6 (A), FIG. 6 (B), FIG. 6 (C) or FIG. 6 (D). In any of FIGS. 6A to 6D, the white area corresponds to the transmission area.

  In the template image of FIG. 6A, there is no seasonal setting, and two transparent area frames Kt1a and Kt1b are arranged on the two transparent areas, respectively. In the template image of FIG. 6B, there is no seasonal setting, and a single transparent area frame Kt2 is arranged on a single transparent area.

  For the template image in FIG. 6C, there is a season setting, and the season information indicates winter. A single transmission area frame Kt3 is arranged on a single transmission area. In the template image of FIG. 6D, neither the season setting nor the transparent area frame exists. When the transmissive area extends to the outer edge of the template image as shown in FIG. 6D, the transmissive area frame is not set.

  When the image composition mode is selected, first, the JPEG file recorded on the recording medium 30 is reproduced as described above, and the captured image is displayed on the LCD monitor 24. Further, when a captured image update operation is performed, the captured image displayed on the LCD monitor 24 is updated. When a selection operation is performed on the key input device 38 in a state where a desired photographed image is displayed, whether or not the photographed image being displayed has a face image is determined based on the description of the tag data. If no face image exists, a warning is output to prompt an image update operation.

  On the other hand, if a face image exists, the posture of the object scene is determined based on the description of the tag data. The captured image is developed in the work area 20f so that the object scene is upright. Therefore, when the photographed image shown in FIG. 5A is selected, the photographed image is developed as it is in the work area 20f. On the other hand, when the photographed image shown in FIG. 5B is selected, the photographed image is developed in the work area 20f in a state rotated 90 degrees to the left. In the following, the captured image developed in the work area 20f is defined as “attention photographed image”, and the main face frame on the attention photographed image is defined as “attention face frame”.

  Next, a template image having a transparent area frame is extracted on the list L1. However, among the template images having the season setting, template images whose seasons do not match the shooting date are excluded from extraction targets.

  The date of the captured image shown in FIG. 5A is December 10, 2005, and the date of the captured image shown in FIG. 5B is August 26, 2006. On the other hand, a template image having a transparent area frame is an image shown in FIGS. 6A to 6C, and a template image having a seasonal setting is an image shown in FIG. 6C. Further, the season information of the template image shown in FIG. 6C is winter.

  Therefore, when the photographed image shown in FIG. 5A is selected as the noticeable photographed image, the template images shown in FIGS. 6A to 6C are displayed on the list L1 in the manner shown in FIG. 7A. Extracted into On the other hand, when the photographed image shown in FIG. 5B is selected as the noticeable photographed image, the template images shown in FIGS. 6A and 6B are displayed on the list L1 in the manner shown in FIG. 7B. Extracted into

  When the list L1 is completed, a transparent area frame that satisfies the size condition is selected from the transparent area frames on the template image. The size condition is a condition that the difference from the size of the target face frame is the smallest, and a transparent area frame having a size closest to the size of the target face frame is selected from the transparent area frames on the template image. The transparent area frame selection process based on such a size condition makes sense when a plurality of transparent area frames exist on the template image.

  Subsequently, the difference between the center coordinates of the selected transparent area frame and the center coordinates of the noticeable face frame on the noticeable photographed image is calculated for each of the template images on the list L1. In both the photographed image and the template image, the coordinate origin is set at the center of the image, and the above difference is the difference between the value indicated by the center coordinate of the selected transparent area frame and the value indicated by the center coordinate of the face frame of interest. Expressed by differences. The difference in the calculated center coordinates is defined as “composite suitability”, and the composite suitability increases as the difference decreases.

  When the photographed image is rotated and developed in the work area 20f, the center coordinates of the face frame are converted according to the rotation of the photographed image, and the converted center coordinates are referred to for calculating the difference.

  When the calculation of the combined fitness is completed, first, template images having seasonal settings are moved to the list L2 in order from the highest combined fitness. When the movement of all the template images having the seasonal setting is completed, the template images not having the seasonal setting are moved to the list L2 in order from the one having the highest combination fitness.

  Therefore, when the captured image shown in FIG. 5A is selected, the template images shown in FIGS. 6A to 6C are registered in the list L2 in the manner shown in FIG. 7A. When the captured image shown in FIG. 5B is selected, the template images shown in FIGS. 6A and 6B are registered in the list L2 in the manner shown in FIG. 7B.

  When the list L2 is completed, the template image registered at the top of the list L2 is reproduced in the following manner. The CPU 32 detects the corresponding JPEG file from the flash memory 36 and writes the JPEG data stored in the JPEG file to the compressed image area 20 e of the SDRAM 20 through the memory control circuit 18. The CPU 32 also gives a decompression instruction to the JPEG codec 26. The JPEG codec 26 reads the JPEG data stored in the compressed image area 20 e through the memory control circuit 18, performs JPEG expansion on the read JPEG data, and expands the decompressed image data through the memory control circuit 18 to the expanded image area 20 d of the SDRAM 20. Write to.

  When the reproduction of the template image is completed, a process of multiplexing the reproduced template image on the noticeable photographed image (as a superordinate concept, a process of combining the reproduced template image and the noticeable photographed image) is executed on the work area 20f. . In the following, a template image that is multiplexed with a noticed photographed image is defined as “attention template image”, and a transparent area frame that satisfies the size condition among the transparent area frames on the noticeable template image is defined as “attention transparent area frame”. .

  When performing the multiplex processing, the magnification of the noticeable photographed image and the multiplex position of the noticeable template image are set so that the above-described synthesis suitability increases, that is, the center coordinates and size of the noticeable transparent area frame are the center coordinates and size of the noticeable face frame. To be matched to each of the above. When the multiplex processing is completed, a part of the captured image that protrudes from the outer edge of the template image is cut out, thereby completing the composite image. The image data representing the completed composite image is then moved from the work area 20f to the expanded image area 20b. The moved image data is then read out by the LCD driver 22, and as a result, a composite image is output from the LCD monitor 24.

  When the photographed image shown in FIG. 8A is selected as the noticeable photographed image and the template image shown in FIG. 8B is designated as the noticeable template image, the face frame Kp1 shown in FIG. Thus, the transparent area frame Kt3 shown in FIG. 8B becomes the target transparent area frame.

  The magnification of the noticeable photographed image is adjusted so that the horizontal size of the noticeable face frame Kp1 and the horizontal size of the noticeable transparent area frame Kt3 coincide with each other, and the multiple position of the noticeable template image is the center coordinate of the noticeable transparent area frame Kt3. Adjustment is made so that the center coordinates of the face frame of interest Kp1 ′ having the adjusted magnification coincide with each other (see FIGS. 8C to 8E). Regarding the size adjustment, if the horizontal size of the noticeable face frame Kp1 is “W2” and the horizontal size of the noticeable transparent area frame Kt3 is “W1”, the noticeable face frame Kp1 ′ has a horizontal size of “W1”. The attention template image is multiplexed on the attention photographed image having the adjusted magnification as shown in FIG. When the multiplex processing is completed, a part of the target photographed image that protrudes from the outer edge of the target template image is cut out, thereby completing the composite image shown in FIG.

  On the other hand, when the captured image shown in FIG. 10A is selected as the focused captured image and the template image illustrated in FIG. 10B is selected as the focused template image, the face frame Kp2 illustrated in FIG. It becomes a face frame, and the transparent area frame Kt1a shown in FIG.

  The magnification of the noticeable photographed image is adjusted so that the horizontal size of the noticeable face frame Kp2 and the horizontal size of the noticeable transparent area frame Kt1a coincide with each other, and the multiple position of the noticeable template image is adjusted with the center coordinate of the transparent area frame Kt1a. Adjustment is made so that the center coordinates of the face frame of interest Kp2 ′ having the set magnification coincide with each other (see FIGS. 10C to 10E). Regarding the size adjustment, if the horizontal size of the attention face frame Kp2 is “W2” and the horizontal size of the attention transmission area frame Kt1a is “W1”, the attention face frame Kp2 ′ has a horizontal size of “W1”. The template image of interest is multiplexed on the photographed image having the adjusted magnification in the manner shown in FIG. When the multiplex processing is completed, a part of the target captured image that protrudes from the outer edge of the target template image is cut out, thereby completing the composite image (see FIG. 10G).

When a position correction operation is performed by the key input device 38 after the composite image is completed, the multiple positions of the template image of interest are corrected. The template image of interest is multiplexed again on the photographed image of interest at the corrected multiplexed position, and the composite image obtained by this is output from the LCD monitor 24 in the same manner as described above. In addition, the multiple position correction amount is normalized according to Equation 1 and Equation 2 in consideration of the multiple processing to another template image of interest updated by the template image update operation. The normalized multiple position correction amount in the horizontal direction is calculated by Equation 1, and the normalized multiple position correction amount in the vertical direction is calculated by Equation 2. Note that Hsize shown in Equation 1 and Equation 3 described later corresponds to “W1” in the example of FIG. 8C or FIG. 10C.
[Equation 1]
NChp = Chp / Hsize
NChp: normalized multiple position correction amount in the horizontal direction Chp: multiple position correction amount in the horizontal direction Hsize: horizontal size of the transparent area frame [Equation 2]
NCvp = Cvp / Vsize
NCvp: normalized multiple position correction amount in the vertical direction Cvp: multiple position correction amount in the vertical direction Vsize: vertical size of the transmission area frame

Further, when the magnification correction operation is performed by the key input device 38 after the composite image is completed, the magnification of the noticeable photographed image is corrected. The noticeable template image is multiplexed again on the noticeable photographed image having the corrected magnification, and the composite image obtained thereby is also output from the LCD monitor 24 in the same manner as described above. In addition, the magnification correction amount is normalized according to Equation 3 in consideration of multiple processing to another template image of interest updated by the template image update operation.
[Equation 3]
NCmg = Cmg / Hsize
NCmg: Normalized magnification correction amount Cmg: Magnification correction amount Hsize: Horizontal size of transparent area frame

  When the template image update operation is performed, the next template image on the list L2 is designated as the template image of interest, and is reproduced as described above. The magnification of the noticeable photographed image and the multiplexed position of the updated noticeable template image are the same as described above, so that the degree of composite matching increases (the position and size of the noticeable transparent area frame match the position and size of the noticeable face frame, respectively) Adjusted).

  Further, the multiple position of the updated template image of interest is corrected with reference to the normalized position correction amount calculated by Equation 1 and Equation 2. Further, the magnification of the noticeable photographed image is corrected with reference to the normalized magnification correction amount calculated by Equation 3.

  When the horizontal size and vertical size of the attention transparent area frame on the updated attention template image are defined as “Hsize ′” and “Vsize ′”, respectively, the multiple matching positions of the updated attention template image have an increased degree of synthesis. The position is adjusted so that it is moved by “NChp * Hsize ′” in the horizontal direction and moved by “NCvp * Vsize ′” in the vertical direction. Further, the magnification of the noticeable photographed image is a value obtained by multiplying the magnification adjusted so as to increase the combined fitness by “NCmg * Hsize ′”.

  When the position correction operation is performed after the composite image shown in FIG. 8G is completed, the multiplex position of the template image of interest is corrected in the manner shown in FIG. 9A, for example, and as shown in FIG. A composite image is obtained.

  When the template image update operation is performed after the position correction operation, the template image shown in FIG. 9C is set as the next noticed template image, and the transparent area frame Kt2 is set as the next noticed transparent area frame. The horizontal size of the target transparent area frame Kt2 matches the horizontal size of the target face frame Kp1 (see FIG. 9D), and the magnification of the target photographed image is “1.0”. However, the multiplexed position of the template image of interest is corrected with reference to the normalized position correction amount calculated by Equation 1 and Equation 2 (see FIG. 9E).

  As a result of the multiplexing process referring to the magnification and the multiplex position determined in this way, a composite image shown in FIG. 9G is obtained. For reference, FIG. 9F shows a composite image obtained when the position where the center coordinates of the noticeable transparent area frame Kt2 coincide with the center coordinates of the noticeable face frame Kp1 is a multiple position.

  Further, when the magnification correction operation is performed after the composite image shown in FIG. 10G is completed, the magnification of the noticeable photographed image is corrected in the manner shown in FIG. 11A, and as a result, shown in FIG. A composite image is obtained.

  When a template image update operation is performed after the magnification correction operation, the template image shown in FIG. 11C is designated as the next attention template image, and the transparent area frame Kt2 is set as the next attention transparent area frame. The horizontal size of the target transparent area frame Kt2 matches the horizontal size of the target face frame Kp2 (see FIG. 11C), and the magnification of the target captured image is set to “1.0”. However, this magnification is corrected with reference to the normalized magnification correction amount calculated by Equation 3 (see FIG. 11E).

  As a result of the multiplexing process referring to the magnification and the multiplex position determined in this way, a composite image shown in FIG. 11G is obtained. For reference, FIG. 11 (F) shows a composite image obtained when the magnification of the noticeable photographed image is “1.0”.

  When an image recording operation is performed on the key input device 38 after the composite image is completed, the CPU 32 gives a compression command and a recording command to the JPEG codec 26 and the I / F circuit 28, respectively, in order to execute a recording process. The JPEG codec 26 performs JPEG compression on the composite image in the same manner as described above, and writes JPEG data in the compressed image area 20 c of the SDRAM 20. The I / F circuit 28 reads JPEG data from the compressed image area 20 c through the memory control circuit 18, and creates a JPEG file containing the read JPEG data in the recording medium 30.

  When the image composition mode is selected, the CPU 32 executes processing according to the flowcharts shown in FIGS. Note that control programs corresponding to these flowcharts are stored in the flash memory 36.

  First, in step S1, the normalized multiple position correction amount and the normalized magnification correction amount are initialized. In step S3, a captured image selection process is executed, and in step S5, a template image extraction process is executed. In step S3, a captured image having a face image is selected as a focused captured image. In addition, the lists L1 and L2 are created by the process of step S5, and the template image registered at the top of the list L2 is designated as the template image of interest. In step S7, an image composition process is executed to synthesize the noticeable captured image and the noticeable template image.

  In step S9, it is determined whether a template image update operation has been performed. In step S13, it is determined whether an image recording operation has been performed. In step S17, it is determined whether or not a position correction operation has been performed. In step S23, it is determined whether or not a magnification correction operation has been performed.

  When the template image update operation is performed, the process proceeds from step S9 to step S11, and the template image of interest is updated to the next template image on the list L2. When the update process is completed, the process returns to step S7. When an image recording operation is performed, the process proceeds from step S13 to step S15 to execute a recording process. The composite image created by the image composition processing in step S7 is recorded on the recording medium 30 in the JPEG file format.

  When the position correction operation is performed, the process proceeds from step S17 to step S19, and the multiple position of the template image of interest is corrected. In step S21, the normalized multiple position correction amount is calculated according to the above-described equations 1 and 2. When the magnification correction operation is performed, the process proceeds from step S23 to step S25, and the magnification of the noticeable captured image is corrected. In step S27, the normalized magnification correction amount is calculated according to the above equation 3. When the process of step S21 or S23 is completed, the target template image is multiplexed again with the target captured image in step S29. In this multiplex processing, the multiplex position corrected in step S19 or the magnification corrected in step S25 is referred to. In step S31, an unnecessary partial image (a part of the target captured image that protrudes from the outer edge of the target template image) is cut out. When the process of step S31 is completed, the process returns to step S9.

  The captured image selection process in step S3 shown in FIG. 12 is executed according to a subroutine shown in FIG. First, in step S41, a JPEG file containing the first photographed image is reproduced from the recording medium 30. As a result, the first captured image is displayed on the LCD monitor 24. In step S43, it is determined whether or not a captured image update operation has been performed. In step S45, it is determined whether or not a selection operation has been performed.

  When the photographed image update operation is performed, the process proceeds from step S43 to step S45, where the photographed image to be reproduced is updated to the photographed image stored in the next JPEG file. When the update process is completed, the process returns to step S43. When the selection operation is performed, the process proceeds from step S47 to step S49, and tag data is detected from the JPEG file containing the selected photographed image.

  In step S51, it is determined based on the tag data detected in step S49 whether or not a face exists on the selected photographed image. If NO, a warning is generated in step S53 and the process returns to step S43. If YES, the process proceeds to step S55, and the photographed image is developed in the work area 20f in a state in which the object scene is rotated so as to stand upright. The captured image developed in this manner is set as a noticed captured image. When the process of step S55 is completed, the process returns to the upper-level routine.

  The template image update process in step S5 shown in FIG. 12 is executed according to the subroutine shown in FIGS. First, in step S61, the variable NT is set to “1”, and it is determined whether or not the variable NT exceeds the set value NTmax. The set value NTmax corresponds to the total number of template images stored in the flash memory 36, and the process proceeds to step S65 as long as NT ≦ NTmax.

  In step S65, tag data is detected from the JPEG file containing the NTth template image. In step S67, whether or not a transparent area frame exists is determined based on the detected tag data, and in step S69, whether or not a seasonal setting exists is determined based on the detected tag data.

  If YES in both steps S67 and S69, it is determined in step S71 whether or not the season of the NT-th template image matches the date of the target captured image. If “YES” here as well, the NT-th template image is registered in the list L1 in step S73. If “YES” in the step S67, if “NO” in the step S69, the process proceeds to the step S73 without performing the process of the step S71, and the above-described registration process is executed.

  When the process of step S73 is completed, the variable NT is incremented in step S75, and then the process returns to step S63. If NO in step S67 or step S71, the increment process is executed in step S75 without passing through the process in step S73, and then the process returns to step S63.

  If YES is determined in the step S63, the horizontal / vertical size of the noticeable face frame is detected in a step S77, and the center coordinates of the noticeable face frame are detected in a step S79. The noticeable face frame is the main face frame on the noticeable photographed image, and the tag data detected in step S49 in FIG. 14 is referred to in both steps S77 and S79. When the noticeable photographed image is a photographed image rotated on the work area 20f, the detected center coordinate value is converted in consideration of this rotation.

  In step S81, the variable NL is set to “1”, and it is determined whether or not the variable NL exceeds the set value NLmax. The set value NLmax corresponds to the total number of template images registered in the list L1, and the process proceeds to step S85 as long as NL ≦ NLmax. In step S85, the NL-th template image on the list L1 is reproduced. The reproduced template image is secured in the expanded image area 20d.

  In step S87, the size of the transparent area frame on the reproduced template image is detected. In detecting the size, the tag data detected in step S65 is referred to. When there are a plurality of transparent area frames on the reproduced template image, all sizes of the plurality of transparent area frames are detected. In step S89, the difference between the size of the face frame of interest detected in step S77 and the size of the transparent area frame detected in step S87 is calculated.

  In step S91, a transparent area frame that satisfies the size condition is selected from the transparent area frames on the reproduced template image. As described above, the size condition is a condition that the difference from the size of the target face frame is the smallest. Therefore, the process in step S91 is meaningful when there are a plurality of transparent area frames on the template image, and a transparent area frame having a size closest to the size of the face frame of interest is selected from the plurality of transparent area frames. The

  In step S93, the center coordinates of the selected transparent area frame are detected. In step S95, the difference between the center coordinates of the face frame of interest and the center coordinates detected in step S91 is calculated. When the calculation process is completed, the variable NL is incremented in step S97, and the process returns to step S83.

  When NL> NLmax, the process proceeds from step S83 to step S99, and a template image having a seasonal setting is specified from the list L1. In step S101, the identified template images are moved to the list L2 in order from the smallest center coordinate difference (in descending order of the combined fitness). In step S103, the remaining template images in the list L1, that is, the template images having no seasonal setting, are moved to the list L2 in order from the smallest center coordinate difference (in descending order of the combined fitness). As a result, the template image having the seasonal setting and the smallest difference in the center coordinates is registered at the top of the list L2, and the difference in the center coordinates is the smallest at the bottom of the list L2. A template image is registered.

  When the list L2 is completed, the process proceeds to step S105, and the highest template image in the list L2 is reproduced. The reproduced template image is stored in the expanded image area 20d. When the reproduction process is completed, the process returns to the upper hierarchy routine.

  The image composition process in step S5 shown in FIG. 12 is executed according to a subroutine shown in FIG. First, in steps S111 and S113, processing similar to that in steps S87 and S89 described above is performed on the template image of interest. As a result, the target transparent area frame is specified. In step S115, the magnification of the noticeable photographed image in which the horizontal size of the noticeable face frame matches the horizontal size of the noticeable transparent area frame is calculated. In step S117, the multiple positions of the target template image in which the center coordinates of the target transparent area frame coincide with the center coordinates of the target face frame are calculated.

  In step S119, the magnification calculated in step S115 is corrected with reference to the horizontal size of the target transparent area frame and the normalized magnification correction amount calculated in step S27 shown in FIG. In step S121, the multiple position calculated in step S117 is corrected with reference to the horizontal size of the target transparent area frame and the normalized multiple position correction amount calculated in step S21 shown in FIG.

  In step S123, the target captured image is enlarged / reduced with reference to the magnification corrected in step S119, and the target template image is multiplexed on the enlarged / reduced target captured image with reference to the multiple position corrected in step S121. To do. In step S125, a part of the noticeable photographed image that protrudes from the outer edge of the noticeable template image is cut out. When the composite image is completed in this manner, the process returns to the upper hierarchy routine.

  According to this embodiment, the attributes (attributes: center coordinates and size) of the transparent area frame (transparent area) provided in each of the plurality of template images (M default images) are detected by the CPU 32 (S87). ~ S93). The attributes of the face image (specific object image) or the face frame (specific object image area) of the photographed image are also detected by the CPU 32 (S77, S79). The CPU 32 calculates the combined fitness of each of the plurality of template images based on the detected transparent area frame and face frame attributes (S95), and the calculated combined fitness of each of the plurality of template images is higher. Specify in order (S99 to S105, S11). Further, the CPU 32 executes a process of multiplexing the designated template image on the photographed image in such a manner that the degree of composition suitability is improved (S111 to S123).

  As described above, the composite suitability is calculated based on the attribute of the transparent area frame on the template image and the attribute of the face frame on the photographed image. Template images to be multiplexed with the captured image are designated in order from the highest composite fitness level, and the process of multiplexing the template image with the captured image is executed in a manner that improves the composite fitness level. As a result, a good composite image can be obtained.

  The attribute of the face frame (specific object image area) corresponds to one of the attributes of the face image (specific object image). That is, the face frame attribute exists as a subordinate concept of the face image attribute.

  When this embodiment is viewed from another viewpoint, the attributes of each of the plurality of transparent area frames provided on the template image are detected by the CPU 32 (S87). The attribute of the face frame included in the captured image is also detected by the CPU 32 (S77). The CPU 32 calculates a difference between each of the plurality of attributes detected from the template image and the attribute detected from the captured image (S89), and selects one of the plurality of transparent area frames based on the calculated difference. (S91). The CPU 32 executes the process of multiplexing the template image on the photographed image in such a manner that the difference between the attribute of the selected transparent area frame and the attribute of the face frame is suppressed (S111 to S123).

  As described above, the transparent area frame to be noticed for the multiple processing is selected based on the difference between the attribute of each of the plurality of transparent area frames and the attribute of the face frame. Further, the multiple processing is executed in a manner in which the difference between the transparent area frame and the face frame is suppressed. As a result, a default image having a plurality of transmission areas can be satisfactorily combined with a captured image.

  If this embodiment is viewed from a different viewpoint, the CPU 32 determines the composition processing parameter (composition processing parameter: photographing) based on the face frame attribute of the photographed image and the transparent area frame attribute provided on the template image. Image magnification and template image multiple position) are calculated (S115, S117). The CPU 32 also executes processing for multiplexing the template image on the captured image based on the calculated synthesis processing parameter and the normalized magnification / multiple position correction amount (correction coefficient) (S117 to S123). The normalized magnification / multiple position correction amount is updated by the CPU 32 in response to the magnification / position correction operation after completion of the multiplexing process (S27, S21). The process of multiplexing the template image on the captured image is executed again in a manner according to the magnification / position correction operation (S29). When receiving a template image update operation, the CPU 32 updates the template image and restarts the synthesis process parameter calculation process.

  Therefore, the mode of the multiprocessing depends on the synthesis processing parameter based on the attributes of the face frame and the transparent area frame and the normalization magnification / multiple position correction amount. When the magnification / position correction operation is performed, the mode of multiple processing changes, and the normalized magnification / multiple position correction amount is updated. When the template image update operation is performed, the template image is updated, and the synthesis processing parameter is calculated again. The mode of multiple processing using the updated template image depends on the newly calculated synthesis processing parameter and the updated normalized magnification / multiple position correction amount. As a result, the mode of multiple processing using the template image before update is reflected in the multiple processing using the template image after update, and the template image after update can be satisfactorily combined with the captured image.

  In this embodiment, when there are a plurality of transparent area frames on the template image, the transparent area frame having the smallest difference from the face frame size is selected (S89, S91). Further, when the template image is moved from the list L1 to the list L2, the template image is selected in order from the smallest difference between the center coordinates of the transparent area frame and the center coordinates of the face frame (S95, S101, S103).

  However, a transparent area frame having a center coordinate with the smallest difference from the center coordinate of the face frame may be selected from a plurality of transparent area frames on the template image, or the template image is moved from the list L1 to the list L2. When performing the selection, the transparent area frame size and the face frame size may be selected in ascending order.

  In this embodiment, the season set in the template image is taken into consideration when assigning the rank to the template image (S69, S99 to S103). However, time information such as a monthly event may be set in the template image instead of the season, and the order may be assigned to the template image in consideration of the set time. Here, “time” is considered as a superordinate concept of “season”.

  In this embodiment, the face image is assumed as the specific object image, but a soccer ball, a flower, or the like may be assumed as the specific object image.

  Further, in this embodiment, the magnification adjustment is performed on the captured image during the multiple processing, while the position adjustment is performed on the template image. However, the position adjustment is performed on the captured image. You may do it.

It is a block diagram which shows the structure of one Example of this invention. It is an illustration figure which shows an example of the mapping state of SDRAM applied to FIG. 1 Example. It is an illustration figure which shows an example of the structure of the JPEG file which stored the picked-up image. It is an illustration figure which shows an example of the structure of the JPEG file which accommodated the template image. (A) is an illustration figure which shows an example of a picked-up image, (B) is an illustration figure which shows another example of a picked-up image. (A) is an illustrative view showing an example of a template image, (B) is an illustrative view showing another example of the template image, (C) is an illustrative view showing another example of the template image, ( D) is an illustrative view showing still another example of a template image. (A) is an illustration figure which shows an example of a template image extraction process operation, (B) is an illustration figure which shows another example of a template image extraction process operation. (A) is an illustrative view showing an example of a photographed image selected for image composition processing, (B) is an illustrative view showing an example of a template image selected for image composition processing, (C ) Is an illustrative view showing an example of a transparent area frame and a face frame detected from a template image and a photographed image, respectively. (D) is an illustrative view showing an example of a face frame. It is an illustration figure which shows an example of arrangement | positioning of a face frame, (F) is an illustration figure which shows an example of the multiplexing state of the template image to a picked-up image, and (G) is an illustration figure which shows an example of a synthesized image. . (A) is an illustrative view showing another example of the arrangement of the transparent area frame and the face frame, (B) is an illustrative view showing another example of the synthesized image, and (C) is for image synthesis processing. It is an illustration figure which shows another example of the selected template image, (D) is an illustration figure which shows an example of arrangement | positioning of a transparent area frame and a face frame, (E) is arrangement | positioning of a transmission area frame and a face frame. It is an illustration figure which shows another example, (F) is an illustration figure which shows the other example of a synthesized image, (G) is an illustration figure which shows another example of a synthesized image. (A) is an illustrative view showing an example of a photographed image selected for image composition processing, (B) is an illustrative view showing an example of a template image selected for image composition processing, (C ) Is an illustrative view showing an example of a transparent area frame and a face frame detected from a template image and a photographed image, respectively. (D) is an illustrative view showing an example of a face frame. It is an illustration figure which shows an example of arrangement | positioning of a face frame, (F) is an illustration figure which shows an example of the multiplexing state of the template image to a picked-up image, and (G) is an illustration figure which shows an example of a synthesized image. . (A) is an illustrative view showing another example of the arrangement of the transparent area frame and the enlarged face frame, (B) is an illustrative view showing another example of the synthesized image, and (C) is for image synthesis processing. FIG. 4D is an illustrative view showing another example of the template image selected in FIG. 2D, FIG. 4D is an illustrative view showing an example of the arrangement of the transparent area frame and the face frame, and FIG. FIG. 8F is an illustrative view showing another example of the above, (F) is an illustrative view showing another example of the composite image, and (G) is an illustrative view showing still another example of the composite image. It is a flowchart which shows a part of operation | movement of CPU applied to the FIG. 1 Example. It is a flowchart which shows a part of other operation | movement of CPU applied to the FIG. 1 Example. It is a flowchart which shows a part of other operation | movement of CPU applied to the FIG. 1 Example. FIG. 12 is a flowchart showing yet another portion of behavior of the CPU applied to the embodiment in FIG. 1. It is a flowchart which shows a part of other operation | movement of CPU applied to the FIG. 1 Example. It is a flowchart which shows a part of other operation | movement of CPU applied to the FIG. 1 Example.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Digital camera 14 ... Imaging device 16 ... Camera processing circuit 20 ... SDRAM
24 ... LCD monitor 26 ... JPEG codec 30 ... Recording medium 32 ... CPU
34 ... Tilt sensor 36 ... Flash memory

Claims (18)

First detection means for detecting an attribute of a transparent area provided in each of M (M: an integer of 2 or more) predetermined images;
Second detection means for detecting an attribute of the specific object image included in the captured image;
Calculating means for calculating a combined fitness corresponding to a small difference between each of the M attributes detected by the first detecting means and the attributes detected by the second detecting means;
Designation means for designating one of the M default images in descending order of the composite fitness calculated by the calculation means, and first synthesis for synthesizing the captured image with the default image designated by the designation means An image composition device comprising means. The first detection means includes first position detection means for detecting the position of the transparent area as one of the attributes,
The second detection means includes second position detection means for detecting the position of the specific object image as one of the attributes,
The image synthesis apparatus according to claim 1, wherein the calculation unit increases the combination suitability as a difference between the position detected by the first position detection unit and the position detected by the second position detection unit is small. Rotating means for rotating the photographed image so that a specific object stands upright,
The image synthesizing apparatus according to claim 2, wherein the calculating unit focuses on a position of the specific object image after being rotated by the rotating unit. Each of the M default images has one or more transmissive areas,
The first detection means further includes first size detection means for detecting the size of each of the one or more transparent areas as another one of the attributes,
The image synthesizing apparatus according to claim 2 or 3, wherein the first position detecting unit executes a position detecting process on a transmissive area having a size that satisfies a size condition among the sizes detected by the first size detecting unit. The second detection means further includes second size detection means for detecting the size of the specific object image as another one of the attributes,
The image composition device according to claim 4, wherein the size condition includes a condition that a difference from the size detected by the second detection unit is the smallest. Extraction means for extracting a default image having a time setting from N (N: integer greater than or equal to M) default images, and time setting that does not match the date of the photographed image among the default images extracted by the extraction means An exclusion means for eliminating the default image having
The image composition device according to any one of claims 1 to 5, wherein a predetermined image remaining after the exclusion processing of the exclusion unit among the N predetermined images corresponds to the M predetermined images.   The designation means assigns a rank according to the composite fitness to a default image having a time setting, and a rank assigned to the default image not having a time setting according to the synthesis fitness and by the first allocation means. The image synthesizing apparatus according to claim 6, further comprising a second assigning unit that assigns a lower rank.   The image synthesizing apparatus according to claim 1, wherein the specifying unit includes an updating unit that updates the default image synthesized by the first synthesizing unit every time an image updating operation is received.   The first combining means adjusts the magnification of the captured image so that the size of the specific object image approaches the size of the transparent area, and the position of the transparent area approaches the position of the specific object image. The image synthesizing apparatus according to claim 1, further comprising position adjusting means for adjusting the synthesis position. First magnification correction means for correcting the magnification of the captured image in response to a magnification correction operation;
A first position correction unit that corrects a composite position in response to a position correction operation; and a correction of the first magnification correction unit and / or the first position correction unit by using the predetermined image and the photographed image designated by the designation unit. The image synthesizing apparatus according to claim 9, further comprising second synthesizing means for synthesizing again with reference to the result. A magnification correction amount normalizing unit for normalizing a magnification correction amount by the first magnification correcting unit based on a size of a transparent area on the default image designated by the designation unit;
The first synthesizing unit corrects the magnification of the photographed image based on the size of the transparent area on the default image designated by the designation unit and the magnification correction amount normalized by the magnification correction amount normalizing unit. The image composition apparatus according to claim 10, further comprising a two-magnification correction unit. A position correction amount normalizing unit that normalizes the position correction amount by the first position correcting unit based on the size of the transparent area on the default image specified by the specifying unit;
The first combining unit corrects the combined position based on the size of the transparent area on the default image specified by the specifying unit and the position correction amount normalized by the position correction amount normalizing unit. The image composition device according to claim 10 or 11, further comprising means.   The image synthesizing apparatus according to claim 1, further comprising a cutting unit that cuts out a part of the photographed image protruding from an outer edge of the predetermined image synthesized by the first synthesizing unit.   The image composition device according to claim 1, wherein the specific object image corresponds to a face image. Receiving means for receiving an image selection operation for selecting a photographed image to be noticed by the first detection means; and generating means for generating a warning when the photographed image selected by the image selection operation does not have the specific object image. The image composition device according to claim 1, further comprising:   The image synthesizing apparatus according to claim 1, wherein the first synthesizing unit executes the synthesizing process in such a manner that the synthesis fitness calculated by the calculating unit increases. In the processor of the image synthesizer,
A first detection step of detecting an attribute of a transparent area provided in each of M (M: an integer of 2 or more) predetermined images;
A second detection step of detecting an attribute of the specific object image included in the captured image;
A calculation step of calculating a composite fitness corresponding to a small difference between each of the M attributes detected by the first detection step and the attribute detected by the second detection step;
A designation step for designating one of the M default images in descending order of the synthesis fitness calculated by the calculation step; and a synthesis step for synthesizing the captured image with the default image designated by the designation step. An image composition program to be executed. An image composition method executed by an image composition device,
A first detection step of detecting an attribute of a transmission area provided in each of M (M: an integer of 2 or more) predetermined images;
A second detection step of detecting an attribute of the specific object image included in the captured image;
A calculation step of calculating a composite fitness corresponding to a small difference between each of the M attributes detected by the first detection step and the attribute detected by the second detection step;
A designation step for designating one of the M default images in descending order of the synthesis fitness calculated by the calculation step; and a synthesis step for synthesizing the captured image with the default image designated by the designation step. An image composition method.

Images