JP2011211717A - Three-dimensional image output device and method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform weighting adjustment (variation of intensity) to parallax amounts which are generated in the foreground and the background in accordance with the parallax amounts, and to display a parallax image with more preferable stereoscopic vision particularly on stereoscopic display print.SOLUTION: A plurality of viewpoint images in which the same subject is photographed from a plurality of viewpoints are acquired, and the parallax amounts at a plurality of feature points are acquired from coordinate values of the respectively corresponding feature points of the respective viewpoint images. A conversion table (table for printing) for adjusting the parallax amounts different in weighting corresponding to the parallax amounts converts the parallax amounts so that the absolute value of the maximum value of the parallax amount in a depth direction may be larger than the absolute value of the maximum value of the parallax amount in a protruding direction among the parallax amounts acquired from the respective viewpoint images.

Description

  The present invention relates to a three-dimensional image output apparatus and method, and more particularly to a technique for displaying a good three-dimensional image (stereoscopic image) on a three-dimensional image display device.

  Conventionally, a stereoscopic image generation method that automatically adjusts the viewpoint position of a viewpoint image that constitutes a stereoscopic image according to the stereoscopic display device to be used, and can display a natural stereoscopic image with reduced crosstalk during observation. Has been proposed (Patent Document 1).

  The stereoscopic image generation method described in Patent Document 1 determines a plurality of viewpoints corresponding to a stereoscopic display device from device information regarding the stereoscopic display device, and parallax between the determined plurality of viewpoints and a plurality of first viewpoint images. The first viewpoint image is converted into a plurality of second parallax images corresponding to the plurality of viewpoints based on the parallax information regarding.

  In addition, when displaying a 3D image on various 3D display devices, if the size of the 3D display device is large or the resolution is low, the amount of parallax is increased, so that the 3D image cannot be viewed stereoscopically. Therefore, when the amount of parallax of a stereoscopic image in a certain stereoscopic display device is larger than the amount of parallax in the optimal stereoscopic display device of the stereoscopic image, the amount of parallax is changed by reducing the stereoscopic image to be displayed. A stereoscopic video playback device has been proposed (Patent Document 2).

JP 2006-115198 A Japanese Patent Laying-Open No. 2005-73049 JP 2003-209858 A Japanese Patent Laid-Open No. 10-40420 JP-A-8-331607 International Publication No. 2004/082297 International Publication No. 2004/084560 JP 2004-221700 A

  However, the stereoscopic image generation method described in Patent Document 1 determines a virtual viewpoint based on information such as the size of a stereoscopic display device to be used, and generates a parallax image in which a heel is also captured from the virtual viewpoint. Crosstalk can be reduced and can be displayed as a stereoscopic image, but since the parallax image is uniformly generated regardless of the perspective of the subject, the stereoscopic image may have a poor stereoscopic effect depending on the original viewpoint images. Or a stereoscopic image in which the stereoscopic effect is excessively emphasized, and it is not always possible to generate a stereoscopic image that provides a preferable stereoscopic effect.

  In addition, the stereoscopic video reproduction apparatus described in Patent Document 2 cannot reproduce a stereoscopic video with a more preferable stereoscopic effect, as in the invention described in Patent Document 1, and further, the stereoscopic video in a certain stereoscopic display device. When the parallax amount of the stereoscopic image is larger than the parallax amount of the optimal stereoscopic display device of the stereoscopic video, the stereoscopic video is reduced and displayed, so that the entire display screen of the stereoscopic display device cannot be effectively used. There is.

  That is, the inventions described in Patent Documents 1 and 2 generate a stereoscopic image so that it can be visually recognized as a stereoscopic image regardless of the stereoscopic display device to be used, and generate a stereoscopic image with a more preferable stereoscopic effect. No ingenuity has been made.

  The present invention has been made in view of such circumstances, and the amount of parallax generated in the foreground and background can be adjusted (weighted) according to the amount of parallax, and is particularly preferable on a stereoscopic display print. An object of the present invention is to provide a three-dimensional image output apparatus and method capable of displaying a parallax image having a stereoscopic effect.

  In order to achieve the above object, a three-dimensional image output apparatus according to an aspect of the present invention includes a viewpoint image acquisition unit that acquires a plurality of viewpoint images obtained by photographing the same subject from a plurality of viewpoints, and the acquired plurality of viewpoints. Parallax information acquisition means for acquiring parallax amounts at a plurality of sets of feature points with matching features from an image, and parallax amount adjustment means for adjusting the parallax amounts of the acquired feature points for stereoscopic display printing, The parallax for adjusting the parallax amount with different weighting to the parallax amount so that the absolute value of the maximum value of the parallax amount in the depth direction is larger than the absolute value of the maximum value of the parallax amount in the pop-out direction among the parallax amounts An amount adjusting unit, a parallax image generating unit that generates a parallax image corresponding to the parallax amount of each feature point after the adjustment, and a plurality of parallax images including the generated parallax image are output for the stereoscopic display print. It is characterized by comprising a parallax image output unit.

  According to one aspect of the present invention, the degree of freedom in adjusting the amount of parallax compared to the case where the amount of parallax is uniformly adjusted regardless of the distance of the subject by determining the virtual viewpoint as in the invention described in Patent Document 1. The parallax amount can be adjusted freely for the foreground and background (the parallax intensity can be freely set), and the pop-out direction of the parallax amount of each feature point acquired from multiple viewpoint images The parallax amount is adjusted so that the absolute value of the maximum value of the parallax amount in the depth direction is larger than the absolute value of the maximum value of the parallax amount. It is possible to suppress and generate a parallax image in which the sense of depth of the background is enhanced, and it is possible to display a parallax image having a more preferable stereoscopic effect on the stereoscopic display print.

  In the three-dimensional image output device according to another aspect of the present invention, the parallax amount adjusting means adjusts the parallax amount so that the change rate of the parallax amount decreases as the parallax amount in the pop-out direction increases, It is characterized in that the amount of parallax is adjusted so that the rate of change of the amount of parallax increases as the amount of parallax in the depth direction increases.

  In the three-dimensional image output device according to still another aspect of the present invention, the parallax amount adjusting means is configured to calculate a parallax amount maximum value in the pop-out direction and a parallax amount in the depth direction among the acquired parallax amounts of the feature points. The maximum value is adjusted so as to be a predetermined amount of parallax.

  In the three-dimensional image output device according to still another aspect of the present invention, the parallax amount adjusting means is configured to calculate a parallax amount maximum value in the pop-out direction and a parallax amount in the depth direction among the acquired parallax amounts of the feature points. The maximum value is adjusted to be 3 mm and -8 mm, respectively. Note that the amount of parallax in the pop-out direction is positive, and the amount of parallax in the depth direction is negative.

  In the three-dimensional image output apparatus according to still another aspect of the present invention, the parallax amount adjusting means has a conversion table indicating an input / output relationship between the normalized parallax amount and the parallax amount after the parallax amount is adjusted. Then, the parallax amount of each acquired feature point is normalized, and the adjusted parallax amount corresponding to the normalized parallax amount is read from the conversion table.

  In the three-dimensional image output method according to still another aspect of the present invention, the viewpoint image acquisition step of acquiring a plurality of viewpoint images obtained by photographing the same subject from a plurality of viewpoints matches the characteristics from the acquired plurality of viewpoint images. A parallax information acquisition step of acquiring parallax amounts at a plurality of sets of feature points, and a parallax amount adjustment step of adjusting the parallax amounts of the acquired feature points for stereoscopic display printing, wherein the jumping out of the parallax amounts A parallax amount adjustment step of adjusting a parallax amount having a different weighting with respect to the parallax amount so that an absolute value of a maximum value of the parallax amount in the depth direction is larger than an absolute value of a maximum value of the parallax amount in the direction; A parallax image generating step for generating a parallax image corresponding to the parallax amount of each feature point after adjustment, and outputting a plurality of parallax images including the generated parallax image for the stereoscopic display print It is characterized by comprising a parallax image output step.

  In the three-dimensional image output method according to still another aspect of the present invention, the parallax amount adjusting step adjusts the parallax amount so that the rate of change of the parallax amount decreases as the parallax amount in the pop-out direction increases. The parallax amount is adjusted so that the change rate of the parallax amount increases as the parallax amount in the depth direction increases.

  In the three-dimensional image output method according to still another aspect of the present invention, the parallax adjustment step includes the step of calculating the maximum parallax amount in the pop-out direction and the parallax amount in the depth direction among the parallax amounts of the acquired feature points. The maximum value is adjusted so as to be a predetermined amount of parallax.

  In the three-dimensional image output method according to still another aspect of the present invention, the parallax adjustment step includes the step of calculating the maximum parallax amount in the pop-out direction and the parallax amount in the depth direction among the parallax amounts of the acquired feature points. The maximum value is adjusted to be 3 mm and -8 mm, respectively.

  In the three-dimensional image output method according to still another aspect of the present invention, the method includes a conversion table indicating an input / output relationship between the normalized parallax amount and the parallax amount after the parallax amount is adjusted, and the parallax amount adjusting step Is characterized by normalizing the acquired parallax amount of each feature point and reading out the adjusted parallax amount corresponding to the normalized parallax amount from the conversion table.

  According to the present invention, since the amount of parallax between a plurality of foreground and background parallax images caused by the perspective of a subject is adjusted by weighting according to the amount of parallax, a desired intensity is added to the foreground and background parallax. A parallax image can be generated, and in particular, the absolute value of the maximum value of the parallax amount in the depth direction than the absolute value of the maximum value of the parallax amount in the pop-out direction among the parallax amounts of each feature point acquired from a plurality of viewpoint images Since the amount of parallax with different weighting is adjusted to increase the amount of parallax, it is possible to generate a parallax image that suppresses the amount of projection of the foreground and emphasizes the depth of the background. A parallax image having a more preferable stereoscopic effect can be displayed.

FIG. 1 is an external view showing a three-dimensional image output apparatus according to a first embodiment of the present invention. FIG. 2 is a block diagram showing an internal configuration of the three-dimensional image output apparatus shown in FIG. FIG. 3 is a diagram illustrating an example of feature points in an image. FIG. 4 is a diagram illustrating an example of assigning a plurality of imaging units and viewpoint numbers. FIG. 5 is a diagram illustrating an example of coordinate values of the same feature points l and m of four parallax images photographed from each viewpoint. FIG. 6 is a diagram illustrating the relationship between the virtual viewpoint position and the parallax adjustment parameter Δt. FIG. 7 is a graph showing an example of the parallax adjustment parameter Δt. FIG. 8 is a block diagram showing an internal configuration of a digital camera provided with a 3D image output apparatus according to the present invention. FIG. 9 is a diagram illustrating a data structure of a 3D image file for 3D images. FIG. 10 is a diagram illustrating an example of parallax amount metadata in a 3D image file for 3D images. FIG. 11 is a chart showing an example of the representative parallax amount of each viewpoint. FIG. 12 is a diagram illustrating another example of the parallax amount metadata in the 3D image file for 3D images. FIG. 13 is a graph showing an example of the parallax adjustment parameter Δx. FIG. 14 is a chart showing the relationship between the parallax amount of each feature point and the parallax adjustment parameter Δx. FIG. 15 is a graph showing an example of the parallax adjustment parameter Δp. FIG. 16 is a graph illustrating an example of a parallax amount conversion table. FIG. 17 is a graph showing another example of the parallax amount conversion table. FIG. 18 is a flowchart showing an embodiment in which the parallax amount is adjusted by the display device. FIG. 19 is a graph showing another example of the parallax adjustment parameter Δx. FIG. 20 is a graph showing still another example of the parallax adjustment parameter Δx.

  Embodiments of a 3D image output apparatus and method according to the present invention will be described below with reference to the accompanying drawings.

[First embodiment of three-dimensional image output apparatus]
FIG. 1 is an external view showing a three-dimensional image output apparatus according to a first embodiment of the present invention.

  As shown in FIG. 1, this three-dimensional image output device (3D image output device) 10 is a digital photo frame equipped with a color three-dimensional liquid crystal display (hereinafter referred to as “3D LCD”) 12. An operation unit 14 for selecting a power switch, normal display / slide show, and the like is provided, and a memory card slot 16 is provided on a side surface.

  FIG. 2 is a block diagram showing an internal configuration of the 3D image output apparatus 10.

  As shown in FIG. 2, in addition to the 3D LCD 12 and the operation unit 14, the 3D image output apparatus 10 includes a central processing unit (CPU) 20, a work memory 22, a card interface (card I / F) 24, a display controller 26, A buffer memory 28, an EEPROM 30, and a power supply unit 32 are provided.

  The 3D LCD 12 displays a plurality of parallax images (right-eye image, left-eye image) with directional images having a predetermined directivity by a lenticular lens, a parallax barrier, etc., polarized glasses, liquid crystal shutter glasses, and the like It is possible to apply the one that allows the user to see the right-eye image and the left-eye image individually by wearing the dedicated glasses.

  The CPU 20 functions as a control unit that comprehensively controls the overall operation of the 3D image output apparatus 10 according to a predetermined control program based on an input from the operation unit 14. The contents of control by the CPU 20 will be described later.

  The work memory 22 includes a calculation work area for the CPU 20 and a temporary storage area for image data.

  The card I / F 24 is electrically connected to the memory card 34 when a memory card 34 that is a recording medium of the digital camera is inserted into the memory card slot 16, and data (image data) is exchanged with the memory card 34. It is an apparatus for performing transmission and reception.

  The display controller 26 repeatedly reads out 3D display image data (a plurality of image data) from the buffer memory 28, which is a temporary storage area dedicated to display image data, and converts it into a 3D display signal on the 3D LCD 12. Output to the 3D LCD 12. As a result, a 3D image is displayed on the 3D LCD 12.

  The power supply unit 32 controls power from a battery (not shown) or a commercial power supply, and supplies operating power to each unit of the 3D image output apparatus 10.

[Operation of 3D image output apparatus 10]
When the power switch of the operation unit 14 is turned on and slide show playback is set as the playback mode, the CPU 20 starts from the memory card 34 installed in the memory card slot 16 in the order of file numbers via the card I / F 24. Read the image file at predetermined intervals. This image file is a 3D image file for 3D display in which a plurality of parallax images are stored in one file. Details of the data structure of this 3D image file will be described later.

  The CPU 20 calculates a plurality of parallax images from the read 3D image file, a coordinate value of a plurality of feature points whose features match on the plurality of parallax images, and a parallax that is a difference between the coordinate values of the plurality of feature points. Get the quantity. If the coordinate information of the plurality of sets of feature points is not included in the attached information of the 3D image file, the coordinate values of the feature points are acquired by analyzing the plurality of parallax images.

  Here, the feature points are feature points l and m having features that can be uniquely specified in the parallax image as shown in FIG. 3, and the same feature points l and m can be specified with other parallax images. It is.

  As shown in FIG. 5, the coordinate values of the same feature points l and m of four parallax images taken from four viewpoints take different values. In the example shown in FIG. 5, the feature point l has a larger viewpoint number. As the point becomes (as the viewpoint position moves to the left as shown in FIG. 4), the x coordinate value becomes smaller, and the feature point m becomes larger as the viewpoint number becomes larger.

  This is because the subject including the feature point l is a distant view at a position farther from the position where the optical axes of the imaging units intersect, and the subject including the feature point m is than the position where the optical axes of the imaging units intersect. Is also based on being a foreground in a close position.

  Various methods have been proposed to detect the above feature points. For example, the block matching method, KLT method (Tomasi & Kanade, 1991, Detection and Tracking of Point Features), SIFT (Scale Invariant Feature Transform), etc. In addition, recent face detection techniques can be applied to feature point detection.

  In addition, it is desirable that the feature points in the parallax images are all points at which the features can be uniquely specified between the plurality of parallax images.

  When the block matching method is applied to the detection of feature points whose features match among a plurality of parallax images, the image is cut out based on an arbitrary pixel from one image (for example, the left image) of the plurality of parallax images. The degree of coincidence between a block having a predetermined block size and a block of another parallax image (for example, the right image) among a plurality of parallax images is evaluated, and the right image when the degree of coincidence between the blocks is maximized is evaluated. The reference pixel of the block is a pixel of the right image corresponding to an arbitrary pixel of the left image.

  As a function for evaluating the degree of coincidence between blocks in the block matching method, for example, there is a function that uses a square sum (SSD) of luminance differences between pixels in each block (SSD block matching method).

  In this SSD block matching method, the following equation is calculated for each pixel f (i, j), g (i, j) in the blocks of both images.

  The calculation of the above [Equation 1] is performed while moving the position of the block in the predetermined search area on the right image, and the pixel at the position in the search area when the SSD value is minimum is set as the search target pixel. To do.

  And a parallax indicating the amount and direction of displacement between the pixel position on the left image and the corresponding pixel on the right image searched for (if the left and right images are taken in a horizontal state, the displacement The direction can be expressed as positive or negative).

<First Example of Adjustment of Parallax Amount>
Next, a first embodiment of the parallax adjustment method according to the present invention will be described.

  The CPU 20 adjusts the parallax amount of the acquired feature points of each set, but performs adjustment with different weighting on the parallax amount according to the magnitude of the parallax amount.

  Now, as shown in FIG. 6, assuming that the distance between the viewpoints of the two parallax images (the right-eye image R and the left-eye image L) is S, the virtual viewpoint (the viewpoint at a distance t from the viewpoint of the left-eye image L). The parallax amount d ′ (x, y) of an arbitrary feature point on the image (virtual viewpoint image) indicated by the dotted line captured in () can be expressed by the following equation.

  Here, d (x, y) is the amount of parallax between the image R for the right eye and the image L for the left eye at the arbitrary feature point.

  The invention described in Patent Document 1 determines the position of the virtual viewpoint as described above, and obtains the parallax amount d ′ (x, y) at the virtual viewpoint position, so that the parallax amount d (x, y) is calculated. Adjustments are being made.

  On the other hand, in the embodiment of the present invention, the parallax amount is adjusted by calculating the parallax amount d ′ (x, y) of an arbitrary feature point at the virtual viewpoint position by the following equation.

  Here, Δt is a parallax adjustment parameter for adjusting the position of the virtual viewpoint, and is a function of the parallax amount of the feature point as shown in graphs (1), (2), and (3) of FIG. With the parallax adjustment parameter Δt, the position of the virtual viewpoint is adjusted as indicated by a one-dot chain line.

  In the meaning of the above equation (3), when the position of the virtual viewpoint is further adjusted by the parallax adjustment parameter Δt, for example, the position of the virtual viewpoint is adjusted so that t is increased by the parallax adjustment parameter Δt, the amount of parallax d When '(x, y) is larger than before the adjustment, and when the position of the virtual viewpoint is adjusted so that t becomes smaller by the parallax adjustment parameter Δt, the parallax amount d' (x, y) is smaller than before the adjustment. Become smaller.

  Also, the parallax adjustment parameter Δt shown in the graphs (1) and (2) of FIG. 7 has a larger value in the positive direction and a larger parallax amount in the negative direction as the parallax amount is larger (closer to the foreground). 7), the parallax adjustment parameter Δt shown in the graph (3) of FIG. 7 shows a tendency that the parallax amount increases in the opposite direction to the graphs (1) and (2). A larger value in the direction indicates a larger value in the negative direction, and a larger value in the negative direction indicates a tendency to have a larger value in the positive direction.

  The CPU 20 reads the conversion table corresponding to any one of the graphs (1), (2), and (3) from the EEPROM 30 in accordance with the display unit information including at least the size information of the 3D LCD 12, and displays the parallax image. For each feature point, the parallax adjustment parameter Δt corresponding to the parallax amount of the feature point is read, and the parallax amount d ′ (x, y) is calculated by the above equation [3]. Note that the position of the virtual viewpoint can be determined by information such as the viewpoint image size, the parallax information regarding the parallax between the viewpoint images, and the size of the stereoscopic display device, as in the invention described in Patent Document 1.

  When a parallax image for the right eye is used as a reference image and a parallax image after parallax adjustment is generated from the parallax image for the left eye, the amount of parallax with respect to the reference image of each feature point of the parallax image for the left eye and the parallax The parallax image for the left eye so that the coordinate value of each feature point of the parallax image for the left eye becomes a coordinate value having the parallax amount after the adjustment based on the parallax amount adjusted by the equation [3]. Is geometrically deformed. This geometrical deformation can be performed by projective transformation using projective transformation parameters, affine transformation using affine transformation parameters, Helmart transformation using Helmart transformation parameters, or the like.

  The left-eye parallax image generated by the CPU 20 is output to the buffer memory 28 together with the right-eye parallax image (reference image) and temporarily stored therein.

  The display controller 26 reads out two parallax images (right-eye image and left-eye image) from the buffer memory 28, converts them into 3D display signals, and outputs them to the 3D LCD 12. Thereby, a 3D image (left and right parallax images) is displayed on the 3D LCD 12.

  The 3D image output apparatus 10 of this embodiment is a digital photo frame having a 3D LCD 12 mounted thereon, but the present invention is not limited to this, and a stereoscopic display device having various screen sizes, a 3D plasma display, a 3D organic display, and the like. The present invention can also be applied to 3D image output devices that output parallax images to different types of stereoscopic display devices such as EL displays, printers that generate stereoscopic display prints, and the like. In this case, it is preferable to acquire display unit information including at least the size information of the display unit among the information about the display unit used for stereoscopic display, and adjust the parallax amount based on the display unit information.

  For example, in the case of a stereoscopic display device with a small display size, the parallax adjustment parameter Δt is obtained from the conversion table corresponding to the graphs (1) and (2) in FIG. 7, and the parallax amount adjusted by the parallax adjustment parameter Δt is obtained. By generating the parallax image in this way, it is possible to display a stereoscopic image (a stereoscopic image having a powerful stereoscopic effect) in which the foreground appears to protrude forward while suppressing the depth of the background.

  On the other hand, in the case of a stereoscopic display device having a large display size, the parallax adjustment parameter Δt is obtained from the conversion table corresponding to the graph (3) in FIG. 7, and the parallax image is adjusted to have the parallax amount adjusted by the parallax adjustment parameter Δt. Thus, it is possible to display a stereoscopic image (a stereoscopic image having a soft stereoscopic effect) in which the amount of projection of the foreground is suppressed and the sense of depth of the background is emphasized.

[Second Embodiment of Three-Dimensional Image Output Device]
FIG. 8 is a block diagram showing an internal configuration of the digital camera 100 provided with the 3D image output apparatus according to the present invention.

  As shown in FIG. 8, the digital camera 100 is a compound eye camera including two photographing units 112R and 112L. Note that there may be two or more photographing units 112.

  The digital camera 100 can record a 3D image including a plurality of images captured by the plurality of imaging units 112R and 112L as one 3D display 3D image file.

  A main CPU 114 (hereinafter referred to as “CPU 114”) controls the overall operation of the digital camera 100 according to a predetermined control program based on an input from the operation unit 116.

  A ROM 124, an EEPROM 126, and a work memory 128 are connected to the CPU 114 via a system bus 122. The ROM 124 stores a control program executed by the CPU 114 and various data necessary for control. The EEPROM 126 stores various setting information relating to the operation of the digital camera 100 such as user setting information. The work memory 128 includes a calculation work area for the CPU 114 and a temporary storage area for image data.

  The operation unit 116 is a means for a user to perform various operation inputs, and includes a power / mode switch, a mode dial, a release switch, a cross key, a zoom button, a MENU / OK button, a DISP button, and a BACK button. . The operation display unit 118 is a means for displaying an operation input result from the operation unit 116, and includes, for example, a liquid crystal panel or a light emitting diode (LED).

  The power / mode switch is means for switching on / off the power of the digital camera 100 and switching the operation mode (playback mode and photographing mode) of the digital camera 100. When the power / mode switch is turned on, power supply from the power supply unit 120 to each unit of the digital camera 100 is started, and various operations of the digital camera 100 are started. When the power / mode switch is turned off, the supply of power from the power supply unit 120 to each unit of the digital camera 100 is stopped.

  The mode dial is an operating means for switching the shooting mode of the digital camera 100. The 2D still image shooting mode for shooting a 2D still image according to the setting position of the mode dial, and the 2D movie shooting for shooting a 2D movie. The shooting mode is switched between a mode, a 3D still image shooting mode for shooting a 3D still image, and a 3D moving image shooting mode for shooting a 3D moving image. When the shooting mode is set to the 2D still image shooting mode or the 2D moving image shooting mode, a flag indicating that the 2D mode for shooting a 2D image is set in the shooting mode management flag 130. When the shooting mode is set to the 3D still image shooting mode or the 3D moving image shooting mode, the shooting mode management flag 130 is set to indicate a 3D mode for shooting a 3D image. The CPU 114 refers to the shooting mode management flag 130 to determine the shooting mode setting.

  The release switch is composed of a two-stage stroke type switch composed of so-called “half-press” and “full-press”. When the release switch is pressed halfway in still image shooting mode, shooting preparation processing (for example, AE (Automatic Exposure) processing, AF (Auto Focus) processing, AWB (Automatic White Balance)) (Balance) processing) is performed, and when the release switch is fully pressed, still image shooting / recording processing is performed. In the moving image shooting mode, moving image shooting starts when the release switch is fully pressed, and moving image shooting ends when the release switch is fully pressed again. Note that a release switch for still image shooting and a release switch for moving image shooting may be provided separately.

  The 3D LCD 150 is a 3D image display similar to the 3D LCD 12 of the 3D image output apparatus 10 shown in FIG. 1, and functions as an image display unit for displaying a captured 2D image or 3D image, and at the time of various settings. It functions as a GUI. The 3D LCD 150 functions as an electronic viewfinder for confirming the angle of view in the shooting mode.

  The vertical / horizontal shooting detection circuit 132 includes, for example, a sensor for detecting the orientation of the digital camera 100, and inputs the detection result of the orientation of the digital camera 100 to the CPU 114. The CPU 114 switches between vertical shooting and horizontal shooting when the digital camera 100 is oriented.

  Next, the shooting function of the digital camera 100 will be described. In FIG. 8, each part in each of the photographing units 112R and 112L is distinguished from each other by adding symbols R and L. However, since the function of each unit is substantially the same, in the following description, the symbols R and L are denoted. The description is omitted.

  The photographing lens 160 includes a zoom lens, a focus lens, and a diaphragm. The zoom lens and the focus lens move back and forth along the optical axis (LR and LL in the drawing) of each photographing unit. The CPU 114 performs zooming by controlling the position of the zoom lens by controlling the driving of a zoom actuator (not shown) via the photometry / ranging CPU 180, and controls the driving of the focus actuator via the photometry / ranging CPU 180. Thus, focusing is performed by controlling the position of the focus lens. Further, the CPU 114 controls the aperture amount (aperture value) of the diaphragm by controlling the driving of the diaphragm actuator via the photometry / ranging CPU 180, and controls the amount of light incident on the image sensor 162.

  When shooting a plurality of images in the 3D mode, the CPU 114 drives the shooting lenses 160R and 160L of the shooting units 112R and 112L in synchronization. That is, the photographing lenses 160R and 160L are always set to the same focal length (zoom magnification). In addition, the aperture is adjusted so that the same incident light amount (aperture value) is always obtained. Further, in the 3D mode, focus adjustment is performed so that the same subject is always in focus.

  The flash light emitting unit 176 is constituted by, for example, a discharge tube (xenon tube), and emits light as necessary when photographing a dark subject or in backlight. The charge / light emission control unit 178 includes a main capacitor for supplying a current for causing the flash light emission unit 176 to emit light. The CPU 114 transmits a flash emission command to the photometry / ranging CPU 180 to perform charge control of the main capacitor, discharge (light emission) timing of the flash light emission unit 176, control of the discharge time, and the like. Note that a light emitting diode may be used as the flash light emitting unit 176.

  The imaging unit 112 captures a distance light emitting element 186 (for example, a light emitting diode) for irradiating the subject with light and an image of the subject irradiated with light by the distance light emitting element 186 (ranging image). A distance image sensor 184.

  The photometry / ranging CPU 180 causes the distance light emitting element 186 to emit light at a predetermined timing based on a command from the CPU 114 and controls the distance imaging element 184 to photograph a distance measuring image.

  The distance measurement image captured by the distance image sensor 184 is converted into digital data by the A / D converter 196 and input to the distance information processing circuit 198.

  The distance information processing circuit 198 uses the distance measurement image acquired from the distance image sensor 184 to connect between the subject photographed by the photographing units 112R and 112L and the digital camera 100 based on the principle of so-called triangulation. The distance (subject distance) is calculated. The subject distance calculated by the distance information processing circuit 198 is recorded in the distance information storage circuit 103.

  As a method for calculating the subject distance, the light flight time from when the distance light emitting element 186 emits light until the light emitted by the distance light emitting element 186 is reflected by the subject and reaches the distance imaging element 184 is used. A TOF (Time of Flight) method for calculating the subject distance from the (delay time) and the speed of light may be used.

  The photographing unit 112 includes an interval / convergence angle driving circuit 188 and an interval / convergence angle detection circuit 190.

  The interval / convergence angle driving circuits 188R and 188L drive the imaging units 112R and 112L, respectively. The CPU 114 operates the interval / convergence angle driving circuits 188R and 188L via the interval / convergence angle control circuit 192 to adjust the interval and the convergence angle between the photographing lenses 160R and 160L.

  The interval / convergence angle detection circuits 190R and 190L include means for transmitting and receiving radio waves, for example. The CPU 114 operates the interval / convergence angle detection circuits 190R and 190L via the interval / convergence angle control circuit 92 to transmit and receive radio waves, thereby measuring the interval and the convergence angle between the photographing lenses 160R and 160L. . The measurement results of the interval between the photographing lenses 160R and 160L and the convergence angle are stored in the lens interval / convergence angle storage circuit 102.

  The image sensor 162 is constituted by a color CCD solid-state image sensor, for example. A large number of photodiodes are two-dimensionally arranged on the light receiving surface of the image sensor 162, and color filters of three primary colors (R, G, B) are arranged in a predetermined arrangement in each photodiode. The optical image of the subject formed on the light receiving surface of the image sensor 62 by the photographing lens 160 is converted into signal charges corresponding to the amount of incident light by the photodiode. The signal charge accumulated in each photodiode is sequentially read out from the image sensor 162 as a voltage signal (R, G, B signal) corresponding to the signal charge based on a drive pulse given from the TG 164 according to a command from the CPU 114. The image sensor 162 has an electronic shutter function, and the exposure time (shutter speed) is controlled by controlling the charge accumulation time in the photodiode.

  As the image sensor 162, an image sensor other than a CCD such as a CMOS sensor can be used.

  The analog signal processing unit 166 outputs a correlated double sampling circuit (CDS), R, G, and B signals for removing reset noise (low frequency) included in the R, G, and B signals output from the image sensor 162. An AGS circuit for amplifying and controlling to a certain level is included. Analog R, G, and B signals output from the image sensor 162 are subjected to correlated double sampling processing and amplified by the analog signal processing unit 166. The analog R, G, and B signals output from the analog signal processing unit 166 are converted into digital R, G, and B signals by the A / D converter 168 and input to the image input controller (buffer memory) 170. The

  The digital signal processing unit 172 includes a synchronization circuit (a processing circuit that interpolates a spatial shift of the color signal associated with the color filter arrangement of the single-plate CCD and converts the color signal into a simultaneous expression), a white balance adjustment circuit, a gradation A conversion processing circuit (gamma correction circuit), a contour correction circuit, a luminance / color difference signal generation circuit, and the like are included. Digital R, G, and B signals input to the image input controller 170 are subjected to predetermined processing such as synchronization processing, white balance adjustment, gradation conversion, and contour correction by the digital signal processing unit 172, It is converted into a Y / C signal composed of a luminance signal (Y signal) and a color difference signal (Cr, Cb signal).

  When the live view image (through image) is displayed on the 3D LCD 150, the Y / C signal generated in the digital signal processing unit 172 is sequentially supplied to the buffer memory 144. The display controller 142 reads the Y / C signal supplied to the buffer memory 144 and outputs it to the YC-RGB conversion unit 146. The YC-RGB conversion unit 146 converts the Y / C signal input from the display controller 142 into R, G, and B signals and outputs them to the 3D LCD 150 via the driver 148. Thereby, the through image is displayed on the 3D LCD 150.

  Here, when the camera mode is the shooting mode and in the 2D mode, a recording image is shot by a predetermined one shooting unit (for example, 112R). In the 2D mode, an image photographed by the photographing unit 112R is compressed by the compression / decompression processing unit 174R. The compressed image data is recorded on the memory card 34 as an image file in a predetermined format via the memory controller 134 and the card I / F 138. For example, JPEG (Joint Photographic Experts Group) for still images, MPEG2 or MPEG4 for moving images, and H.264. It is recorded as a compressed image file conforming to the H.264 standard.

  The camera mode is a shooting mode, and in the 3D mode, images are shot in synchronization by the shooting units 112R and 112L. In the 3D mode, AF processing and AE processing are performed based on the image signal acquired by one of the imaging units 112R and 112L. In the 3D mode, the two viewpoint images captured by the imaging units 112R and 112L are respectively compressed by the compression / decompression processing units 174R and 174L, stored in one 3D image file, and recorded on the memory card 34. The 3D image file stores subject distance information, information on the distance between the taking lenses 160R and 160L, information on the convergence angle, and the like, as well as compressed image data of two viewpoints.

  On the other hand, when the operation mode of the camera is the playback mode, the last image file (last recorded image file) recorded on the memory card 34 is read out, and the compression / decompression processing unit 174 performs uncompressed Y / After being expanded to the C signal, it is input to the buffer memory 144. The display controller 142 reads the Y / C signal supplied to the buffer memory 144 and outputs it to the YC-RGB conversion unit 146. The YC-RGB conversion unit 146 converts the Y / C signal input from the display controller 142 into R, G, and B signals and outputs them to the 3D LCD 150 via the driver 148. As a result, the image of the image file recorded on the memory card 34 is displayed on the 3D LCD 150.

  Here, when the image file read from the memory card 34 is a 3D image file, a parallax image in which the parallax amount is adjusted is generated as in the above-described 3D image output apparatus 10, and the adjusted parallax image is displayed. Two parallax images including the image are displayed on the 3D LCD 150.

[Data structure of 3D image file]
FIG. 9 is a diagram showing an example of the data structure of a 3D image file for 3D display recorded on the memory card 34 by the digital camera 100.

  As shown in FIG. 9, the 3D image file F uses the Exif standard file format and adopts a file format in which a plurality of images are integrated and recorded. The viewpoint image (1) (first image) A1, the viewpoint image (2) A2, viewpoint image (2) A2,..., Viewpoint image (n) An are connected to each other.

  Each viewpoint image area is divided into SOI (Start of Image) indicating the start position of each viewpoint image and EOI (End of Image) indicating the end position. Exif attached information of the viewpoint image is next to the SOI. An APP1 marker segment to be recorded and an APP2 marker segment to record multi-viewpoint accessory information are provided, and then a viewpoint image is recorded.

  The APP1 marker segment is provided with Exif identification information, a TIFF header, and an IFD (Image file directory) area (IFD0 area (0th IFD) and IFD1 area (1st IFD)). A thumbnail image generated from the viewpoint image is stored in the IFD1 area (1st IFD). The APP2 marker segment includes individual information IFD.

  The individual information IFD includes the number of viewpoint images, the viewpoint image number, the convergence angle, the baseline length, and the like. In this embodiment, as shown in FIG. 10, the tag value of the reference viewpoint number and the foreground representative parallax amount The tag value of the background representative parallax amount is further recorded.

  The foreground representative parallax amount is a value representative of the foreground parallax between the reference viewpoint (“1” in the example of FIG. 10) and the viewpoint (i) of the viewpoint image. A value representing parallax, parallax at the center of the face recognition position, or the like can be used.

  The background representative parallax amount is a value representing the parallax of the background between the reference viewpoint (“1” in the example of FIG. 10) and the viewpoint (i) of the viewpoint image. Value can be used.

  Note that the numerical examples of the foreground representative parallax amount and the background representative parallax amount shown in FIG. 10 indicate the number of pixels corresponding to the parallax amount in the left-right direction of the parallax image (negative and the parallax direction is reverse). The number of pixels in the left-right direction of the parallax image is 1000 pixels.

  FIG. 11 is a chart showing an example of the representative parallax amount (foreground representative parallax amount, background representative parallax amount) of each viewpoint in the case of the four viewpoints shown in FIG.

  As shown in FIG. 12, the individual information IFD of the 3D image file further records the coordinate values of the feature points 1, 2, 3,..., K in the viewpoint image.

  Note that extraction of feature points between the parallax images, calculation of the foreground representative parallax amount, background representative parallax amount, and the like are performed by the digital camera 100 and are recorded as additional information of the parallax image when the 3D image file is created.

<Second Example of Adjustment of Parallax Amount>
FIG. 13 is a graph showing an example of the parallax adjustment parameter Δx for adjusting the parallax amount.

  The parallax adjustment parameter Δt of the first embodiment shown in FIG. 7 is a parameter for adjusting the virtual viewpoint position, but the parallax adjustment parameter Δx shown in the graphs (1) and (2) of FIG. 13 is added to the parallax amount. This is a parameter for adjusting the parallax amount.

  Now, when n feature points are extracted from the parallax image as shown in FIG. 14 and the parallax amount is obtained for each feature point, the conversion corresponding to the graph (1) in FIG. 13 is performed based on the parallax amount. The corresponding parallax adjustment parameter Δx is read from the table. According to the graph (1) in FIG. 13, the parallax adjustment parameter Δx increases in the positive direction as the parallax amount increases in the positive direction (closer to the foreground) and as the parallax amount increases in the negative direction (closer to the background). It is supposed to take a value.

  The parallax amount can be adjusted by adding the parallax adjustment parameter Δx corresponding to the parallax amount to the parallax amount for each feature point.

  According to the parallax adjustment parameter Δx shown in the graph (1) of FIG. 13, weighting adjustment is performed to increase the parallax amount of a feature point that appears close (a large parallax amount in the positive direction), while the feature that appears far away Weighting adjustment is performed to reduce the amount of parallax at the point (the amount of parallax that increases in the negative direction).

  Then, by generating a parallax image obtained by geometrically deforming the parallax image so as to have the feature point with the parallax amount adjusted as described above, the foreground appears to protrude more forward while suppressing the sense of depth of the background. A viewpoint image can be generated, and a more powerful stereoscopic image can be displayed.

  Note that, according to the parallax adjustment parameter Δx in the graph (2) having the opposite characteristics to the graph (1) in FIG. 13, the foreground pop-out amount is suppressed and the background depth feeling is emphasized, contrary to the above. A parallax image can be generated, and a stereoscopic image with a soft stereoscopic effect can be displayed.

<Third Example of Adjustment of Parallax Amount>
FIG. 15 is a graph showing an example of the parallax adjustment parameter Δp for adjusting the parallax amount.

  The parallax adjustment parameter Δp shown in the graphs (1) and (2) of FIG. 15 is a parameter for adjusting the parallax amount by being multiplied by the parallax amount. According to the graph (1), the parallax adjustment parameter Δp is The larger the amount of parallax (closer to the foreground), the larger the value than 1, and the larger the amount of parallax in the negative direction (closer to the background), the smaller the value.

  By multiplying the parallax amount for each feature point by the parallax adjustment parameter Δp corresponding to the parallax amount, the parallax amount can be adjusted.

  According to the parallax adjustment parameter Δp shown in the graph (1) of FIG. 15, weighting adjustment is performed to increase the parallax amount of a feature point that appears close (a large parallax amount in the positive direction), while the feature that appears far away Weighting adjustment is performed to reduce the amount of parallax of the points (the amount of parallax that increases in the negative direction), and it is possible to generate a viewpoint image in which the foreground appears to protrude forward while suppressing the depth of the background. This makes it possible to display a more powerful stereoscopic image.

  On the other hand, according to the parallax adjustment parameter Δp of the graph (2) having characteristics opposite to those of the graph (1) in FIG. 15, the foreground pop-out amount is suppressed and the background depth feeling is emphasized contrary to the above. A parallax image can be generated, and a stereoscopic image with a soft stereoscopic effect can be displayed.

<Fourth Example of Adjustment of Parallax Amount>
FIG. 15 is a graph showing an example of a conversion table for converting an input parallax amount into a desired parallax amount and outputting the same.

  The conversion tables shown in graphs (1) and (2) in FIG. 16 are conversion tables for converting the parallax amount of each feature point of the input image into an appropriate parallax amount according to the magnitude of the parallax amount. According to the conversion table shown in (1), the conversion is performed to increase the parallax amount of the feature point that appears close (a large amount of parallax in the positive direction), while the parallax amount of the feature point that appears far away (in the negative direction) Conversion for reducing the larger amount of parallax) is performed.

  On the other hand, according to the conversion table shown in the graph (2) in FIG. 16, the conversion is performed so that the disparity amount of the feature point that appears nearby (a large amount of disparity in the positive direction) is further reduced, and the disparity amount of the feature point that appears far away Conversion to increase (the amount of parallax that increases in the negative direction) is performed.

<Fifth Example of Adjustment of Parallax Amount>
FIG. 17 is a graph showing another example of a conversion table for converting an input parallax amount into a desired parallax amount and outputting the same.

  The conversion table shown in FIG. 17 is a conversion table used when a parallax image is printed on a print paper (display unit). The maximum parallax amount (max) on the input image is 3 mm on the print paper. Thus, conversion is performed such that the maximum parallax amount (min) in the negative direction on the input image is −8 mm on the print paper.

  For example, the maximum value (foreground representative value) and the negative maximum value (background representative value) among the parallax amounts of each feature point of the input parallax image are obtained, and each feature point is determined by these foreground representative value and background representative value. 17 is normalized (for example, normalized so that the foreground representative value is 1 and the background representative value is −1), and this normalized amount of parallax is converted into the amount of parallax on the print paper using the conversion table shown in FIG. Convert to

  A lenticular sheet is affixed to the surface of the print paper on which the parallax image with the parallax amount adjusted as described above is printed.

  In a photographic print capable of stereoscopic viewing, a more preferable stereoscopic image is visually recognized when the maximum parallax amount (max) of each parallax image on the print is 3 mm and the maximum parallax amount (min) in the negative direction is −8 mm. Experiments have confirmed that this can be done.

[Example of adjusting the amount of parallax by a display device]
FIG. 18 is a flowchart showing an embodiment in which the parallax amount is adjusted by the display device.

  First, a plurality of parallax images are read (step S10), and the parallax amounts of feature points (corresponding points) in each parallax image are acquired (step S12). As shown in FIG. 12, when the coordinate information of each feature point is included in the individual information of the 3D image file, the amount of parallax can be calculated from the coordinate value. When the coordinate value of each feature point is not included in the individual information, the feature point is extracted from each parallax image, the coordinate value of the viewpoint is acquired, and the parallax amount is calculated.

  Next, display device information is acquired (step S14). The display device information may be acquired by connecting the 3D image output apparatus according to the present invention to the display device, so that the display device information may be automatically acquired from the display device side, or manually. Information may be entered.

  Subsequently, the device type (display, printer) is determined from the acquired display device information (step S16). If it is further determined as a display, the screen size of the display is set to a predetermined size (the length in the horizontal direction). 65mm) or more is discriminated (step S18). Note that 65 mm is the distance between the left and right eyes of a person.

  Here, when the device type is a display and the screen size is determined to be 65 mm or more, the large screen table is selected as the conversion table for converting the parallax amount (step S20), and the screen size is determined to be less than 65 mm. Then, the large screen table is selected as the conversion table (step S22).

  On the other hand, when the device type is determined to be a printer, a printing table is selected as a conversion table (step S24).

  For example, the conversion table shown in the graph (2) in FIG. 16 is applied as the large screen table, and the conversion table shown in the graph (1) in FIG. 16 is applied as the small screen table. Conceivable.

  According to this, in the case of a large-screen display display, it is possible to generate a parallax image that suppresses the amount of foreground projection and emphasizes the depth of the background, and displays a stereoscopic image with a soft stereoscopic effect. It becomes possible, and an observer's feeling of fatigue can be reduced. In addition, in the case of a small-screen display, it is possible to generate a viewpoint image in which the foreground appears to protrude forward while suppressing the depth of the background, and display a more powerful stereoscopic image. Is possible.

  On the other hand, the conversion table shown in FIG. 17 is used as the printing table. According to this, it is possible to print a photographic print capable of more preferable stereoscopic viewing. In this embodiment, when the device type is a printer, the conversion table is not further switched depending on the print size. This is because according to the conversion table shown in FIG. 17, it has been confirmed that a photographic print capable of preferable stereoscopic viewing can be obtained regardless of the print size of the photographic print.

  When the conversion table is selected as described above, the input parallax amount is converted by the selected conversion table, and a parallax image corresponding to the parallax amount of each feature point after the conversion (after parallax adjustment) is generated. (Step S26). Each parallax image generated in this way is output to a 3D display device to enable stereoscopic viewing of the 3D image (step S28).

<Sixth Example of Adjustment of Parallax Amount>
FIG. 19 is a graph showing another example of the parallax adjustment parameter Δx for adjusting the parallax amount.

  A graph (1) -1 in FIG. 19 is a graph obtained by shifting the graph (1) (corresponding to the graph (1) in FIG. 13) in the positive direction of the parallax amount, and the graph (1) -2 in FIG. It is the graph which shifted graph (1) to the negative direction of the amount of parallax.

  Therefore, when the parallax amount of the parallax image is adjusted by the parallax adjustment parameter Δx shown in the graph (1) -1, the foreground is compared with the parallax image in which the parallax amount is adjusted by the parallax adjustment parameter Δx shown in the graph (1). A parallax image in which the amount of popping out is reduced and the depth of the background is also reduced is generated.

  On the other hand, when the parallax amount of the parallax image is adjusted by the parallax adjustment parameter Δx shown in the graph (1) -2, compared to the parallax image in which the parallax amount is adjusted by the parallax adjustment parameter Δx shown in the graph (1), Is generated, and a parallax image in which the depth of the background is also emphasized is generated.

  Further, the parallax adjustment parameter Δx shown in the graphs (1) -1, (1), (1) -2 may be selected according to the screen size of the 3D display. In this case, it is conceivable to apply the parallax adjustment parameter Δx shown in the graphs (1) -1, (1), and (1) -2 according to the large screen, the medium screen, and the small screen, respectively.

  The graphs (1) -1, (1), and (1) -2 may be selected according to the viewing distance, not limited to the screen size of the 3D display. When (1) -1 is far, it is conceivable to select graph (1) -2. Note that the viewing distance may be automatically acquired by a distance measuring unit disposed in the 3D display or in the vicinity thereof, or the viewing distance may be manually input.

<Seventh Example of Adjustment of Parallax Amount>
FIG. 20 is a graph showing still another example of the parallax adjustment parameter Δx for adjusting the parallax amount.

  The parallax adjustment parameters Δx shown in the graphs (1) -3 and (1) -4 in FIG. 20 are limited so as not to enhance the parallax when the parallax becomes larger than a certain value in the positive direction.

  That is, in the conversion table showing the input / output relationship shown in graphs (1) -3 and (1) -4 of FIG. 20, the parallax amount after conversion is limited so as not to exceed the preset maximum parallax amount. Yes. This is because if the parallax is applied too much, it cannot be viewed as 3D.

  Further, the parallax adjustment parameter Δx shown in the graphs (1) -3 and (1) -4 of FIG. 20 may be selected according to the screen size of the 3D display. In this case, it is conceivable to apply the parallax adjustment parameter Δx shown in the graphs (1) -3 and (1) -4 according to the small screen and the large screen, respectively.

  The sixth and seventh embodiments shown in FIGS. 19 and 20 are modifications of the parallax adjustment parameter shown in the graph (1) of FIG. 13, but the parallax shown in the graph (2) of FIG. The present invention can also be applied to a case where the adjustment parameter, the parallax adjustment parameter Δp shown in FIG. 15, and the parallax amount conversion table shown in FIG.

[Others]
The 3D image output apparatus (digital photo frame, digital camera) of this embodiment is an apparatus in which a 3D display is integrated. However, the present invention is not limited to this, and the present invention does not include a display device (for example, It can also be applied to personal computers. In addition, the 3D image output apparatus according to the present invention may be realized by hardware, or may be realized by software installed in a personal computer body.

  The 3D display is not limited to the 3D LCD of the embodiment, and may be another 3D display such as a 3D plasma display or a 3D organic EL display.

  Furthermore, the present invention is not limited to the above-described embodiment, and any modification is possible as long as the amount of parallax can be freely adjusted regardless of the distance of the subject, and various modifications are possible without departing from the spirit of the present invention. It goes without saying that is possible.

  DESCRIPTION OF SYMBOLS 10 ... Three-dimensional image output device (3D image output device), 12, 150 ... Three-dimensional liquid crystal display (3D LCD), 20, 114 ... Central processing unit (CPU), 22, 128 ... Work memory, 26, 142 ... Display Controller, 28, 136 ... Buffer memory, 30, 126 ... EEPROM

Claims (10)

Viewpoint image acquisition means for acquiring a plurality of viewpoint images obtained by photographing the same subject from a plurality of viewpoints;
Parallax information acquisition means for acquiring parallax amounts at a plurality of sets of feature points whose features match from the acquired plurality of viewpoint images;
A parallax amount adjusting unit that adjusts the parallax amount of each acquired feature point for stereoscopic display printing, wherein the parallax amount in the depth direction is larger than the absolute value of the maximum parallax amount in the pop-out direction of the parallax amount. A parallax amount adjusting means for adjusting a parallax amount different in weighting with respect to the parallax amount so that the absolute value of the maximum value becomes large;
Parallax image generation means for generating a parallax image corresponding to the parallax amount of each feature point after the adjustment;
Parallax image output means for outputting a plurality of parallax images including the generated parallax images for the stereoscopic display print;
A three-dimensional image output apparatus comprising:   The parallax amount adjusting means adjusts the parallax amount so that the change rate of the parallax amount decreases as the parallax amount in the pop-out direction increases, and the change rate of the parallax amount as the parallax amount in the depth direction increases. The three-dimensional image output apparatus according to claim 1, wherein the amount of parallax is adjusted so as to increase.   The parallax amount adjusting means adjusts the parallax amount maximum value in the pop-out direction and the maximum parallax amount value in the depth direction among the acquired parallax amounts of the feature points so that the parallax amount is set in advance. The three-dimensional image output apparatus according to claim 1 or 2, wherein   The parallax adjustment means adjusts the maximum parallax amount in the pop-out direction and the maximum parallax amount in the depth direction among the parallax amounts of the acquired feature points to be 3 mm and -8 mm, respectively. The three-dimensional image output apparatus according to claim 1, wherein adjustment is performed.   The parallax amount adjusting means includes a conversion table indicating an input / output relationship between the normalized parallax amount and the parallax amount after the parallax amount is adjusted, normalizes the parallax amount of each acquired feature point, 5. The three-dimensional image output apparatus according to claim 1, wherein an adjusted parallax amount corresponding to the normalized parallax amount is read from the conversion table. 6. A viewpoint image acquisition step of acquiring a plurality of viewpoint images obtained by photographing the same subject from a plurality of viewpoints;
A parallax information acquisition step of acquiring parallax amounts at a plurality of sets of feature points with matching features from the acquired plurality of viewpoint images;
A parallax adjustment step for adjusting the parallax amount of each acquired feature point for stereoscopic display printing, wherein the parallax amount in the depth direction is larger than the absolute value of the maximum parallax amount in the pop-out direction of the parallax amount; A parallax amount adjusting step for adjusting a parallax amount having a different weighting with respect to the parallax amount so that an absolute value of the maximum value becomes large;
A parallax image generation step of generating a parallax image corresponding to the parallax amount of each feature point after the adjustment;
A parallax image output step of outputting a plurality of parallax images including the generated parallax images for the stereoscopic display printing;
A three-dimensional image output method.   The parallax amount adjusting step adjusts the parallax amount so that the change rate of the parallax amount decreases as the parallax amount in the pop-out direction increases, and the change rate of the parallax amount as the parallax amount in the depth direction increases. The three-dimensional image output method according to claim 6, wherein the parallax amount is adjusted so as to increase.   The parallax adjustment step adjusts the parallax amount maximum value in the pop-out direction and the maximum parallax amount value in the depth direction out of the parallax amounts of the acquired feature points so that the parallax amount is set in advance. The three-dimensional image output method according to claim 6 or 7, characterized in that:   In the parallax adjustment step, the maximum parallax amount in the pop-out direction and the maximum parallax amount in the depth direction among the parallax amounts of the acquired feature points are 3 mm and -8 mm, respectively. The three-dimensional image output method according to claim 6 or 7, wherein adjustment is performed.   A conversion table indicating an input / output relationship between the normalized parallax amount and the parallax amount after adjustment of the parallax amount, and the parallax amount adjustment step normalizes the acquired parallax amount of each feature point; The three-dimensional image output method according to any one of claims 6 to 9, wherein an adjusted parallax amount corresponding to the normalized parallax amount is read from the conversion table.

Images