WO2012101916A1 - 立体動画処理装置、立体動画処理プログラム及びその記録媒体、立体撮像装置並びに立体動画処理方法 - Google Patents
立体動画処理装置、立体動画処理プログラム及びその記録媒体、立体撮像装置並びに立体動画処理方法 Download PDFInfo
- Publication number
- WO2012101916A1 WO2012101916A1 PCT/JP2011/078647 JP2011078647W WO2012101916A1 WO 2012101916 A1 WO2012101916 A1 WO 2012101916A1 JP 2011078647 W JP2011078647 W JP 2011078647W WO 2012101916 A1 WO2012101916 A1 WO 2012101916A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- maximum
- stereoscopic
- stereoscopic video
- section
- viewpoint
- Prior art date
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
- H04N13/10—Processing, recording or transmission of stereoscopic or multi-view image signals
- H04N13/189—Recording image signals; Reproducing recorded image signals
-
- G—PHYSICS
- G03—PHOTOGRAPHY; CINEMATOGRAPHY; ANALOGOUS TECHNIQUES USING WAVES OTHER THAN OPTICAL WAVES; ELECTROGRAPHY; HOLOGRAPHY
- G03B—APPARATUS OR ARRANGEMENTS FOR TAKING PHOTOGRAPHS OR FOR PROJECTING OR VIEWING THEM; APPARATUS OR ARRANGEMENTS EMPLOYING ANALOGOUS TECHNIQUES USING WAVES OTHER THAN OPTICAL WAVES; ACCESSORIES THEREFOR
- G03B35/00—Stereoscopic photography
- G03B35/08—Stereoscopic photography by simultaneous recording
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
- H04N13/10—Processing, recording or transmission of stereoscopic or multi-view image signals
- H04N13/106—Processing image signals
- H04N13/128—Adjusting depth or disparity
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
- H04N13/10—Processing, recording or transmission of stereoscopic or multi-view image signals
- H04N13/106—Processing image signals
- H04N13/172—Processing image signals image signals comprising non-image signal components, e.g. headers or format information
- H04N13/178—Metadata, e.g. disparity information
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N2213/00—Details of stereoscopic systems
- H04N2213/007—Aspects relating to detection of stereoscopic image format, e.g. for adaptation to the display format
Definitions
- the present invention relates to a stereoscopic moving image processing apparatus, a stereoscopic moving image processing program and a recording medium thereof, a stereoscopic imaging apparatus, and a stereoscopic moving image processing method, and more particularly, to a stereoscopic moving image in which a stereoscopic image composed of a plurality of viewpoint images is continuous in the time axis direction.
- the present invention relates to a technique for recording in three stereoscopic video files.
- the stereoscopic imaging device captures a subject from left and right viewpoints using two imaging units arranged with parallax on the left and right, acquires a left-eye viewpoint image and a right-eye viewpoint image, and records them on a recording medium. ing.
- the acquired left and right viewpoint images are read from the recording medium and input to a 3D display capable of three-dimensional (3D) display, and the left eye viewpoint image and the right eye viewpoint image are visually recognized separately by the left and right eyes. By being displayed so that it can be made, it can be recognized as a three-dimensional moving image.
- the parallax amount of the recorded left and right viewpoint images changes due to changes in the shooting scene, movement of the subject, changes in the zoom magnification, and the like, while 3D displays have various screen sizes. Therefore, when a stereoscopic video is to be reproduced and displayed, there is a case where the parallax amount of the stereoscopic video is not appropriate for the size of the 3D display. In such a case, the amount of popping out or retracting from the screen becomes excessive, causing a problem that it cannot be recognized as a natural three-dimensional moving image.
- Japanese Patent Laid-Open No. 2004-228561 includes video information that can be viewed stereoscopically, compatible screen size information related to a screen size suitable for playback of stereoscopic video, 3D display screen size information, and distance to a display screen that is suitable for an observer to view during playback. Information on the suitable viewing distance, viewing distance information on the distance from the observer to the display screen of the 3D display, and a shift amount (offset) between the left-eye video and the right-eye video is set based on the information and displayed.
- a technique for adjusting the stereoscopic effect of an image is disclosed.
- the stereoscopic video described in Patent Document 1 is a still image or a moving image.
- the amount of parallax between the left and right viewpoint images changes due to changes in the shooting scene, movement of the subject, changes in zoom magnification, and the like. If the parallax amount of the stereoscopic video is adjusted so that the stereoscopic video has a natural pop-out amount by the technique described in Patent Document 1, it is conceivable to adjust the parallax amount in units of frames of the stereoscopic video.
- the feature as a stereoscopic moving image in which the stereoscopic effect changes is lost or an unnatural stereoscopic moving image is obtained.
- the stereoscopic image is visually recognized as an image having a depth than the display surface, but the screen size of the 3D display is large. Then, this parallax also becomes large, and when the parallax exceeds the distance between both eyes of a human, binocular fusion becomes impossible (stereoscopic viewing becomes impossible).
- Japanese Patent Application Laid-Open No. 2004-228561 describes a method for obtaining a stereoscopic image in which the optimal stereoscopic degree (depth amount) is adjusted according to the display screen size of the 3D display. There is no description to adjust the shift amount with the right-eye image. In addition, since there is no description regarding the maximum disparity amount on the far side (depth side) of the stereoscopic image, the stereoscopic image display device described in Patent Document 1 has a parallax that exceeds the distance between human eyes regardless of the screen size of the 3D display. Therefore, the shift amount between the left eye image and the right eye image cannot be adjusted.
- the present invention has been made in view of such circumstances, and information necessary for adjusting the parallax amount of a stereoscopic moving image so that the stereoscopic moving image playback apparatus can fuse both eyes regardless of the screen size of the stereoscopic display, Create and record a 3D movie file that has pre-attached information of a 3D movie, and create and record a 3D movie file with attached information that can be converted into a natural 3D movie even if the parallax of the 3D movie is adjusted.
- An object of the present invention is to provide a stereoscopic video processing device, a stereoscopic video processing program, a stereoscopic imaging device, and a stereoscopic video processing method.
- the stereoscopic video processing apparatus includes a stereoscopic video acquisition unit that acquires a stereoscopic video in which a stereoscopic image including a plurality of viewpoint images is continuous in the time axis direction, A parallax amount calculating means for calculating a parallax amount indicating a deviation amount between feature points whose features match from a plurality of viewpoint images for each frame of the acquired stereoscopic video, and among the calculated parallax amounts of the feature points for each frame
- Maximum disparity amount acquisition means for acquiring the maximum disparity amount on the distant view side, and the maximum disparity amount within the section that is maximum within the section for each predetermined section of the stereoscopic video among the acquired maximum disparity amount on the distant view side
- the stereoscopic video is displayed on the stereoscopic display for each predetermined section based on the intra-section maximum parallax amount acquiring means and the intra-section maximum parallax amount acquired for each predetermined section of
- Eye fusion An intra-section maximum display size acquisition means for acquiring an internal maximum display size; and a recording means for generating a stereoscopic video file in which the stereoscopic video is recorded and recording the stereoscopic video file on a recording medium, the stereoscopic video being And recording means for recording in the stereoscopic moving image file and recording the maximum display size in the section for each predetermined interval in the stereoscopic moving image file as attached information.
- a disparity amount indicating a deviation amount between feature points having matching features is calculated from a plurality of viewpoint images for each frame of the acquired stereoscopic video, and a distant view of the calculated disparity amounts of the feature points.
- the maximum parallax amount on the side is acquired, and the maximum intra-section parallax amount that is the maximum in the section is acquired for each predetermined section of the stereoscopic video out of the maximum parallax amount for each frame, and based on the intra-section maximum parallax amount
- the maximum display size within the section that can be binocularly fused is acquired, and this maximum display size within the section is recorded in the 3D video file as attached information.
- the maximum display size in the section for each predetermined section of the 3D video file recorded in the attached information is displayed.
- the display size of the body display it is possible to easily determine whether the stereoscopic view possible for each predetermined interval. Further, when it is determined that stereoscopic viewing is impossible, the parallax of the stereoscopic video in the corresponding section can be corrected using the maximum display size in the section, and the parallax correction processing in the stereoscopic video playback device can be reduced. Can do.
- the maximum display size in the section is recorded for each predetermined section of the stereoscopic video, when the parallax is corrected, it can be performed for each section, and even if the parallax of the stereoscopic video is corrected, a natural stereoscopic video is obtained. be able to.
- the recording unit calculates the intra-section maximum parallax amount acquired for each predetermined section of the stereoscopic moving image by the intra-section maximum parallax amount acquiring unit. Further recorded as ancillary information of the video file.
- the stereoscopic video processing device includes a stereoscopic video acquisition unit that acquires a stereoscopic video in which a stereoscopic image consisting of a plurality of viewpoint images is continuous in the time axis direction, and each frame of the acquired stereoscopic video.
- a parallax amount calculating means for calculating a parallax amount indicating a deviation amount between feature points having the same feature from a plurality of viewpoint images, and a disparity maximum parallax amount on a distant side among the calculated parallax amounts of each feature point for each frame
- a maximum parallax amount acquisition means for acquiring the maximum parallax amount within the section for acquiring the maximum intra-section parallax amount within the section for each predetermined section of the stereoscopic video out of the acquired maximum disparity amount on the distant view side
- a volume acquisition means for generating a stereoscopic video file in which the stereoscopic video is recorded, and recording the stereoscopic video file on a recording medium, recording the stereoscopic video in the stereoscopic video file, and A section within the maximum parallax amount for each section provided with a recording means for recording the three-dimensional moving image file as additional information.
- the maximum intra-section parallax amount can be used as information for determining whether or not the stereoscopic video in the corresponding section can be stereoscopically viewed by the stereoscopic video playback device, as with the maximum display size in the section. It can be used as information for correcting the parallax of the 3D moving image.
- the stereoscopic video is converted into a stereoscopic display for each predetermined section based on the intra-section maximum parallax amount acquired for each predetermined section of the stereoscopic video.
- the recording means is a distance between a viewer and a stereoscopic display suitable for observing the stereoscopic video. A certain assumed viewing distance is further recorded as attached information of the stereoscopic moving image file.
- the maximum parallax amount acquisition unit is configured to calculate the maximum near-field side of the calculated parallax amounts of the feature points.
- the parallax amount is acquired, and the intra-section maximum parallax amount acquisition unit acquires, for each predetermined section, the maximum intra-section parallax amount that is the largest in the predetermined section among the acquired maximum near-parallax amount on the foreground. .
- a seventh aspect of the present invention is the stereoscopic video processing apparatus according to the first aspect, the second aspect, or the fourth aspect, out of viewpoint images of N (N is an integer of 3 or more) viewpoints acquired by the stereoscopic video acquisition means.
- Representative image setting means for setting one as a representative image wherein the maximum parallax amount acquiring means excludes a representative image from the representative image and the N viewpoint image (N ⁇ 1) viewpoint image, (N ⁇ 1) maximum disparity amounts on the distant view side are acquired in relation to two viewpoint images each having a different combination, and the intra-section maximum display size acquisition means acquires the acquired (N ⁇ 1) (N-1) maximum display sizes in the section that are maximum in the predetermined section based on the maximum disparity amount on the distant view side of each of the predetermined sections, and the recording unit acquires the acquired ( N-1) maximum in number of sections The I spray size recorded in the three-dimensional moving image file as the attached information.
- the maximum display size in (N-1) sections can be referred to and compared with the display size to be displayed.
- At least one viewpoint image of the plurality of viewpoint images acquired by the stereoscopic video acquisition unit and the parallax Virtual viewpoint image generation means for generating one or a plurality of virtual viewpoint images corresponding to an arbitrary virtual viewpoint based on the parallax amount calculated by the amount calculation means is further provided.
- the stereoscopic video playback device that reads the stereoscopic video file, when the virtual viewpoint image is recorded, it can be displayed using the virtual viewpoint image without performing parallax adjustment.
- a ninth aspect of the present invention is the stereoscopic video processing apparatus according to the first aspect, the second aspect, or the fourth aspect, wherein N (N is an integer of 3 or more) viewpoint images including the plurality of viewpoint images and virtual viewpoint images.
- Representative image setting means for setting one of them as a representative image wherein the maximum parallax amount obtaining means excludes a representative image from the representative image and the N viewpoint image (N-1) viewpoints of viewpoint (N-1) maximum disparity amounts on the far side are acquired in relation to two viewpoint images each having a different combination with the image, and the intra-section maximum display size acquisition means acquires the acquired (N-1 (N-1) maximum display sizes in the section that are maximum in the predetermined section based on the maximum disparity on the far view side are acquired for each predetermined section, and the recording means Within (N-1) sections It is recorded on the three-dimensional moving image file a large display size as the auxiliary information.
- the (N-1) maximum intra-section parallax amount is referred to and a stereoscopic video suitable for stereoscopic viewing is easily selected. be able to.
- the stereoscopic moving image file is an MPEG in which a plurality of viewpoint images are sequentially recorded for each 1 GOP composed of a plurality of frames. It is a file, and the predetermined section of the stereoscopic video is a section corresponding to one or two or more predetermined number of GOPs.
- An eleventh aspect of the present invention is the stereoscopic video processing apparatus according to any one of the first to tenth aspects, further comprising scene change detection means for detecting a scene change of the acquired stereoscopic video,
- the section is a section for each scene divided by the scene change detected by the scene change detecting means.
- the recording unit includes the maximum display size in a section and the maximum disparity amount in a section acquired for each section. At least one of them is recorded at the head of a section preceding the section by a predetermined number of sections or at the head of a stereoscopic video file.
- the stereoscopic video playback device that has read the stereoscopic video file can detect in advance whether or not to adjust the parallax of the stereoscopic video within a predetermined section, and calculate the parallax amount to be adjusted in advance. Can do.
- the recording unit records the viewpoint images of the N viewpoints in the order of viewpoints or in the order closer to the central viewpoint.
- a fourteenth aspect of the present invention is the stereoscopic moving image processing apparatus according to the eighth aspect or the ninth aspect, wherein the recording means records each of the images when the plurality of viewpoint images and the virtual viewpoint image are recorded in the stereoscopic moving image file. Viewpoint information indicating whether the image is a viewpoint image or a virtual viewpoint image is recorded as attached information of the stereoscopic video file.
- the stereoscopic video playback device that reads out the stereoscopic video file, it is possible to distinguish between a real viewpoint video and a virtual viewpoint video.
- a fifteenth aspect of the present invention is the stereoscopic video processing apparatus according to the first aspect, the second aspect, the fourth aspect, the seventh aspect, or the ninth aspect, wherein the intra-section maximum display size acquisition means is the intra-section maximum parallax amount. Binocular fusion is possible for the maximum display size among the display sizes that do not exceed the human binocular interval, and the amount of image shift on the stereoscopic display corresponding to the maximum disparity amount in the far-view side section acquired by the acquisition means Acquired as the maximum display size in the section. Thereby, the maximum display size in a section can be acquired appropriately.
- the intra-section maximum display size acquisition means is based on the far-field maximum intra-section parallax amount acquired by the intra-section maximum parallax amount acquisition means.
- the maximum display size corresponding to the maximum amount of parallax is read in advance from the storage means in which the maximum display size is recorded, or the maximum display size in the section is calculated using a calculation formula including a predetermined value indicating the human binocular interval. calculate. Thereby, the maximum display size can be acquired appropriately.
- the stereoscopic moving image processing program according to the seventeenth aspect of the present invention causes a stereoscopic moving image processing apparatus according to any one of the first to sixteenth aspects to be realized by a computer.
- a programmable record in which a three-dimensional moving image processing apparatus according to any one of the first to sixteenth aspects is tangentially stored with a three-dimensional moving image processing program of machine-readable instructions that is realized by a computer. Consists of medium.
- a stereoscopic imaging apparatus includes the imaging unit that functions as the stereoscopic moving image acquisition unit, and the stereoscopic moving image processing apparatus according to any one of the first to sixteenth aspects.
- the stereoscopic video processing method includes a step of acquiring a stereoscopic video in which a stereoscopic image consisting of a plurality of viewpoint images is continuous in the time axis direction, and a plurality of frames for each frame of the acquired stereoscopic video.
- a stereoscopic video processing method includes a step of acquiring a stereoscopic video in which a stereoscopic image composed of a plurality of viewpoint images is continuous in the time axis direction, and a plurality of frames for each frame of the acquired stereoscopic video.
- the maximum display size within the section or the maximum intra-section parallax amount that is maximum within the section is obtained for each predetermined section of the stereoscopic video, and is recorded in the stereoscopic video file as attached information.
- the stereoscopic video playback device that has read the stereoscopic video file displays the stereoscopic video displayed on the stereoscopic display based on the maximum display size in the section or the maximum amount of parallax in the section for each predetermined section of the stereoscopic video file recorded in the attached information. Can be easily determined for each predetermined section.
- the parallax of the stereoscopic video in the corresponding section can be corrected using the maximum display size in the section or the maximum amount of parallax in the section, and the parallax in the stereoscopic video playback device Correction processing can be reduced. Furthermore, since the maximum display size in the section is recorded for each predetermined section of the stereoscopic video, when the parallax is corrected, it can be performed for each section, and even if the parallax of the stereoscopic video is corrected, a natural stereoscopic video can be obtained. can do.
- a figure showing how to capture two viewpoint images The figure which shows the 1st example which showed the data structure of the 3D moving image file typically Flow chart showing shooting and recording processing The figure which shows an example of the viewpoint image of 2 viewpoints The figure which shows the 2nd example which showed the data structure of 3D moving image file typically The figure which shows the 3rd example which showed the data structure of 3D moving image file typically The figure which shows the 4th example which showed the data structure of 3D moving image file typically A diagram showing how to capture four viewpoint images The figure which showed typically the data structure of the 3D moving image file in which the viewpoint image of 4 viewpoints was recorded Illustration for explaining the virtual viewpoint The figure which shows the external appearance of a stereoscopic imaging device The figure which shows the external appearance of a stereoscopic imaging device Block diagram showing the internal configuration of the stereoscopic imaging device Flow chart showing a first embodiment of 3D video playback Diagram for explaining the principle of parallax shifting Diagram showing left and right viewpoint images and parallax shift Flow chart showing a second embodiment of 3D video playback
- FIG. 1 is a diagram showing a state in which two viewpoints (left viewpoint and right viewpoint) are taken from different viewpoints of a subject 100 by two imaging apparatuses 101-2 and 101-3.
- 2 is a diagram schematically showing a data structure of a 3D moving image file in which a 3D moving image composed of moving images of the left viewpoint and the right viewpoint captured by the imaging devices 101-2 and 101-3 shown in FIG. 1 is recorded. .
- the 3D video file shown in FIG. 2 is an MPEG file conforming to the MPEG (moving picture picture expert group) format, and in 1 GOP (Group of picture: MPEG compression, the number of video signals as a unit at the time of compression and playback / editing
- the left-viewpoint viewpoint image and the right-viewpoint viewpoint image are alternately connected and combined into one file in units of frames).
- a header area is provided at the head of the area where the viewpoint image of each GOP is recorded, and additional information such as an image size, an aspect ratio, and a frame rate is recorded in this header area.
- additional information such as an image size, an aspect ratio, and a frame rate is recorded in this header area.
- all or part of the following attached information is further recorded as attached information.
- GOP maximum display size width, height, unit: mm
- GOP maximum parallax amount near view: parallax amount (%) with respect to image width
- GOP maximum parallax amount disant view: parallax amount with respect to image width (%)
- Convergence angle, baseline length, and imaging unit layout viewpoint number of the device that captured each viewpoint image Shooting and recording processing for recording such a 3D moving image file will be described with reference to the flowchart of FIG.
- left and right viewpoint images for one frame among a plurality of frames in one GOP corresponding to the left viewpoint and the right viewpoint are acquired (step S11).
- the left and right viewpoint images for one frame in the 3D moving image obtained by capturing the 3D moving image of the subject 100 by the two imaging devices 101-2 and 101-3 are acquired.
- the imaging apparatus 101-2 is referred to as viewpoint number 2
- the imaging apparatus 101-3 is referred to as viewpoint number 3.
- a plurality of feature points are extracted from the left and right viewpoint images (step S12), and the amount of parallax for each feature point is calculated (step S13).
- the amount of parallax refers to a difference in distance between corresponding feature points between viewpoint images from the left end of each viewpoint image, and the unit is a pixel.
- the maximum parallax amount on the near view side and the maximum parallax amount on the far view side are acquired from the parallax amount for each feature point thus calculated (step S14).
- FIG. 4 shows an example of the left and right viewpoint images
- FIG. 4 (a) shows the left viewpoint image 201-2
- FIG. 4 (b) shows the right viewpoint image 201-3.
- the maximum parallax amount on the foreground side is 213N, and feature points having this parallax amount (maximum parallax amount position (near view)) are 211N and 212N, respectively.
- the maximum disparity amount on the distant view side is 213F, and feature points having the disparity amount (maximum disparity amount position (distant view)) are 211F and 212F, respectively. That is, when the stereoscopic image based on the left and right viewpoint images is stereoscopically viewed, the maximum parallax amount position (near view) appears closest and the maximum parallax amount position (distant view) appears closest.
- the ratio (%) of the maximum parallax amount 213N on the near view side and the maximum parallax amount 213F on the near view side to the image width is the maximum parallax amount (near view) (%) and the maximum parallax amount (far view) (%).
- step S15 it is determined whether or not the acquisition of the maximum amount of parallax of all the frames in one GOP has been completed. If the acquisition has not been completed (in the case of “NO”), the process returns to step S11. When the processing from S11 to step S14 is executed and completed (“YES”), the process proceeds to step S16.
- step S16 the maximum parallax amount of all the frames in one GOP is acquired as the GOP maximum parallax amount. Further, the GOP maximum display size is acquired based on the GOP maximum parallax amount.
- the maximum display size (GOP maximum display size) among the display sizes in which the parallax does not exceed the human binocular interval is obtained from the ratio (%) of the GOP maximum parallax amount (distant view) to the image width.
- the maximum display width allowed for stereoscopic viewing is 500 mm. That is, if the display has a width of 500 mm or less, the left and right images corresponding to the GOP maximum parallax amount (distant view) are displayed without exceeding the binocular interval of 50 mm, and as a result, the viewer can view stereoscopically.
- the maximum display height is 281.25 mm.
- interval according to the target viewer. For example, when only an adult is targeted, a wide value such as 65 mm may be set.
- the maximum GOP display size allowed for stereoscopic viewing is about 433 mm.
- the height of the GOP maximum display size is about 244 mm.
- a table in which the maximum display size corresponding to the maximum amount of parallax is recorded is prepared in advance, and the GOP maximum display size is obtained by referring to this table. May be.
- the left viewpoint image and the right viewpoint image for 1 GOP are respectively MPEG-compressed, and alternately connected as shown in FIG. 2 and recorded as a 3D moving image file, and the beginning of the area where the viewpoint image of each GOP is recorded
- the GOP maximum display size, GOP maximum parallax amount (distant view) (%), GOP maximum parallax amount (near view) (%), etc. acquired in step S16 are recorded as attached information in the header area (step S17).
- the 3D moving image file shown in FIG. 2 can be recorded by performing the above processing on all GOPs of the 3D moving image.
- the 3D moving image file recorded in this way is read by the stereoscopic moving image playback device when displayed on the 3D display.
- it is easy to determine whether or not stereoscopic viewing is possible by comparing the GOP maximum display size recorded in the attached information of the 3D video file with the display size of the 3D display to be displayed. Can be determined. If it is determined that stereoscopic viewing is possible, the recorded 3D video may be displayed as it is. If it is determined that stereoscopic viewing is impossible, the corresponding GOP maximum display size and GOP maximum parallax amount are used. The parallax between the left and right viewpoint images in the GOP can be corrected, and the parallax correction processing in the stereoscopic video playback device can be reduced.
- the GOP maximum display size is determined based only on the far-field GOP maximum parallax amount, but may be determined in consideration of the near-field GOP maximum parallax amount.
- a display size where the amount of parallax in the foreground is 50 mm or less may be set as the GOP maximum display size. This is because when the amount of parallax on the near view side increases, the viewer feels fatigue when viewing stereoscopically, so it is preferable to set the amount to a predetermined amount or less.
- the GOP maximum parallax amount and the GOP maximum display size are recorded as attached information in the header area at the head of the GOP from which they are acquired.
- the present invention is not limited to this, as shown in FIG. May be recorded in the header area at the beginning of the GOP two GOPs before (one GOP time before), and as shown in FIG. 6, the header area at the beginning of the 3D video file is obtained from all GOPs.
- the GOP maximum parallax amount and the GOP maximum display size may be recorded together. According to this, the stereoscopic video playback device that has read the 3D video file can detect in advance whether or not the parallax of the stereoscopic video within 1 GOP should be adjusted, and calculate the parallax amount to be adjusted in advance. Can do.
- the maximum GOP parallax amount and the GOP maximum display size that are maximum within one GOP are acquired.
- the present invention is not limited to this, and the maximum within the section is set for each predetermined number of GOPs (sections).
- the amount of parallax and the maximum display size in the section may be acquired and recorded as attached information for the 3D moving image in the section.
- -Flag indicating the beginning of the scene-Scene maximum display size (width, height, unit: mm) -Assumed viewing distance (distance between the viewer and the 3D display when viewing a stereoscopic video) (unit: mm) -Scene maximum parallax amount (near view): Parallax amount relative to image width (%) ⁇ Scene maximum parallax amount (distant view): parallax amount with respect to image width (%) ⁇ Convergence angle, baseline length, and imaging unit layout (viewpoint number) of the device that captured each viewpoint image ⁇ Convergence angle, baseline length, and imaging unit layout (viewpoint number) of the device that captured each viewpoint image Further, a flag indicating that zoom in / out and a flag indicating that camera shake correction has been turned on / off / reset may be recorded as attached information.
- Scene change can be performed by scene change detection means for detecting a scene change from the acquired 3D video. For example, when the time stamp indicating the recording time is separated by the preceding and following 3D moving image frames, the correlation between the preceding and following 3D moving image frames is obtained, and when there is no correlation, the GOP maximum parallax amount of the adjacent GOP is a predetermined threshold value When the above changes, it can be detected from the case where the camera shake correction state changes (ON ⁇ OFF, OFF ⁇ ON, Reset).
- the scene change detection method is not limited to the above example, and various methods are conceivable.
- the scene head flag is a flag indicating the head of each scene in the 3D video file in which the scenes of the 3D video are connected as shown in FIG. 7, and the scene maximum parallax amount is the maximum among the 3D video in the scene.
- the maximum display size of the scene is the maximum of the display sizes acquired based on the maximum amount of parallax in the scene and capable of binocular fusion when the scene is displayed on the stereoscopic display. The display size.
- a 3D video file in which scenes of a 3D video are connected is attached with a scene head flag at the head of the scene, a maximum parallax amount in the scene, a maximum scene display size, etc. for each scene.
- Information is recorded.
- scene length information of the scene is also recorded in the attached information at the beginning of the scene.
- the scene length information can be represented by the number of GOPs constituting the scene.
- the left and right viewpoint images taken from two viewpoints have been described as an example.
- the number of viewpoints of the 3D moving image according to the present invention is not limited to two viewpoints, and the viewpoint images include three or more viewpoints. There may be.
- FIG. 8 is a diagram illustrating a state in which four viewpoint images are photographed from different viewpoints with respect to the subject 100 by the four imaging devices 101-1 to 101-4.
- the viewpoint numbers of the imaging devices 101-1 to 101-4 are sequentially set to 1 to 4.
- FIG. 9 is a diagram schematically showing the data structure of a 3D moving image file in which four viewpoint images captured by the four imaging devices 101-1 to 101-4 are recorded.
- This 3D moving image file is an MPEG file similar to the file shown in FIG. 2, and the viewpoint images of four viewpoints are in units of 1 GOP, and the viewpoint images (1) to (4) corresponding to the viewpoint numbers 1 to 4 are displayed. , In order and repeatedly concatenated into one file.
- the GOP maximum display size, the GOP maximum parallax amount, the assumed viewing distance, and each viewpoint image are captured in the header area at the head of the recording area in which each viewpoint image for each GOP is recorded. Attached information such as the convergence angle, base line length, and imaging unit arrangement (viewpoint number) of the device is recorded.
- the GOP maximum display size, the GOP maximum parallax amount, and the like recorded at the head of each viewpoint image (1) to (4) are values calculated between the viewpoint image and other viewpoint images. .
- the GOP maximum display size and the GOP maximum parallax amount for the viewpoint image (1) have the largest parallax amount with reference to the viewpoint image (1) taken from the left-most reference viewpoint as shown in FIG.
- the GOP maximum parallax amount calculated from the viewpoint image (here, the viewpoint image (4)), the GOP maximum display size acquired from the maximum parallax amount, the assumed viewing distance, and the like are recorded.
- the convergence angle, the base line length, etc. the convergence angle and the base line length with the apparatus that captured the same viewpoint image (here, the viewpoint image (4)) are recorded.
- the 3D moving image file in which the viewpoint images of three or more viewpoints are recorded in this way is read by the stereoscopic moving image playback device when displayed on the 3D display.
- the stereoscopic video playback device is the stereoscopic display possible by comparing the GOP maximum display size for each viewpoint image recorded in the attached information of the 3D video file with the display size of the 3D display to be displayed? It is possible to easily determine whether or not.
- the recording order of the multi-viewpoint images is not limited to the order shown in FIG. 9, and the recommended images (two viewpoint images) to be displayed on the 3D display may be recorded first.
- the viewpoint images (2) and (3) are recorded first, followed by the viewpoint image (1 ), (4), and if stereoscopic display with two viewpoint images (1), (4) at both ends is recommended, the viewpoint images (1), (4) are recorded first. Subsequently, viewpoint images (2) and (3) may be recorded.
- the multi-viewpoint image as in the third embodiment does not need to be an image that is actually actually captured, and may include a virtual viewpoint image corresponding to the virtual viewpoint.
- two viewpoint images are taken from different viewpoints (viewpoint number 1 and viewpoint number 4) with respect to the subject 100 by the two imaging devices 101-1 and 101-4.
- viewpoint number 1 and viewpoint number 4 may be different viewpoints, and viewpoint image 2 and viewpoint image 3 with viewpoint number 2 and viewpoint number 3 in a virtual viewpoint that does not actually exist may be generated.
- viewpoint image 2 and viewpoint image 3 with viewpoint number 2 and viewpoint number 3 in a virtual viewpoint that does not actually exist may be generated.
- a virtual viewpoint image there are a method of internally dividing each pixel of a plurality of captured images, a method of generating using a parallax map generated from a plurality of captured images and a single captured image, and the like. There is no particular limitation.
- the viewpoint images of each viewpoint are attached to the first viewpoint image (1) of the multi-viewpoint images.
- the information indicating whether the viewpoint image is a virtual viewpoint image or a virtual viewpoint image is recorded together (see FIG. 9).
- the attached information of the first viewpoint image (1) not only when the information of the real viewpoint and the virtual viewpoint is recorded together, but as the attached information of each viewpoint image (1) to (4), the actual viewpoint is individually added.
- Information indicating whether a viewpoint image or a virtual viewpoint viewpoint image may be recorded.
- Information indicating whether the viewpoint image is a real viewpoint or a virtual viewpoint can be used when a 3D moving image is displayed as a 2D moving image.
- FIG. 11A and 11B are views showing an external appearance of the stereoscopic imaging apparatus according to the present invention
- FIG. 11A is a perspective view of the stereoscopic imaging apparatus viewed from the front side
- FIG. 11B is a rear view.
- This stereoscopic imaging device (compound-eye camera) 10 is a digital camera capable of recording and reproducing 2D / 3D still images and 2D / 3D moving images. As shown in FIGS. 11A and 11B, a thin rectangular parallelepiped camera is used. A shutter button 11 and a zoom button 12 are disposed on the upper surface of the main body.
- a lens barrier 13 having a width substantially equal to the width in the left-right direction of the camera body is disposed so as to be movable in the up-down direction of the camera body.
- the front surfaces of the pair of left and right photographic optical systems 14-1 and 14-2 can be opened and closed simultaneously by moving in the vertical direction between the position indicated by and the position indicated by the solid line. Note that as the photographing optical systems 14-1 and 14-2, a zoom lens of a bending optical system is used.
- the camera power supply can be turned on / off in conjunction with the opening / closing operation of the lens front surface by the lens barrier 13.
- a 3D liquid crystal monitor 16 is disposed at the center of the back of the camera body.
- the liquid crystal monitor 16 can display a plurality of parallax images (right viewpoint image and left viewpoint image) as directional images each having a predetermined directivity by a parallax barrier.
- the 3D liquid crystal monitor 16 uses a lenticular lens, or can display a right viewpoint image and a left viewpoint image individually by wearing dedicated glasses such as polarized glasses or liquid crystal shutter glasses. Is applicable.
- the operation switch 18A is a changeover switch for switching between still image shooting and moving image shooting
- the operation switch 18B is a parallax adjustment switch for adjusting the amount of parallax between the right viewpoint image and the left viewpoint image
- the operation switch 18C is 2D imaging. This is a changeover switch for switching between 3D imaging.
- the operation switch 18D is a seesaw key that functions as both a MENU / OK button and a playback button
- the operation switch 18E is a multifunction cross key
- the operation switch 18F is a DISP / BACK key.
- the MENU / OK button is an operation switch having both a function as a menu button for instructing to display a menu on the screen of the liquid crystal monitor 16 and a function as an OK button for instructing confirmation and execution of selection contents. It is.
- the playback button is a button for switching from the shooting mode to the playback mode.
- the cross key is an operation switch for inputting instructions in four directions, up, down, left, and right.
- a macro button, a flash button, a self-timer button, or the like is assigned to the menu key. When a menu is selected, the menu screen is displayed. Function as a switch (cursor moving operation means) for selecting an item from the menu or instructing selection of various setting items from each menu.
- the left / right key of the cross key functions as a frame advance (forward / reverse feed) button in the playback mode.
- the DISP / BACK key is used for switching the display form of the liquid crystal monitor 16, canceling the instruction content on the menu screen, or returning to the previous operation state.
- 15 is a stereo microphone.
- FIG. 12 is a block diagram showing an internal configuration of the stereoscopic imaging apparatus 10.
- the stereoscopic imaging apparatus 10 mainly includes a plurality of imaging units 20-1, 20-2, a central processing unit (CPU) 32, the shutter button 11, the zoom button 12, and various operation switches described above. Including an operation unit 34, a display control unit 36, a liquid crystal monitor 16, a recording control unit 38, a compression / expansion processing unit 42, a digital signal processing unit 44, an AE (Automatic Exposure) detection unit 46, and an AF (Auto Focus): An automatic focus detection unit 48, an AWB (Automatic White Balance) detection unit 50, a VRAM 52, a RAM 54, a ROM 56, an EEPROM 58, and the like. Note that the imaging units 20-1 and 20-2 capture two parallax images, a left-eye image and a right-eye image, that have parallax with each other, but there may be three or more imaging units 20.
- the imaging units 20-1 and 20-2 capture two parallax images, a left-eye image and a right-eye image, that have par
- the imaging unit 20-1 that captures an image for the left eye includes an imaging optical system 14-1 (FIG. 14) including a prism (not shown), a focus lens and a zoom lens 21, an optical unit including a diaphragm 22 and a mechanical shutter 23.
- the imaging unit 20-2 that captures the image for the right eye has the same configuration as the imaging unit 20-1 that captures the image for the left eye, and thus the description of the specific configuration is omitted.
- the CPU 32 controls the overall operation of the camera according to a predetermined control program based on the input from the operation unit 34. Calculation of parallax amount by feature point extraction, acquisition of maximum parallax amount, calculation of GOP maximum display size, calculation of scene maximum display size, calculation of GOP maximum parallax amount, calculation of scene maximum parallax amount, detection of scene change, parallax shift The CPU 32 also generates a virtual viewpoint image.
- the ROM 56 stores a control program executed by the CPU 32 and various data necessary for the control, a 3D moving image processing program, a 3D moving image reproduction program, and the like.
- the EEPROM 58 stores various adjustment results at the time of adjustment before product shipment. For example, pixel defect information of the CCD 24, correction parameters used for image processing, a correspondence table between the maximum amount of parallax and the maximum display size, and the like are stored.
- the VRAM 52 is a memory for temporarily storing image data for display displayed on the liquid crystal monitor 16, and the RAM 54 includes a calculation work area for the CPU 32 and a temporary storage area for image data.
- the focus lens and zoom lens 21 included in the photographing optical system are driven by the lens driving unit 28 and moved back and forth along the optical axis.
- the CPU 32 controls the driving of the lens driving unit 28 to control the position of the focus lens so as to adjust the focus so that the subject is in focus, and in response to a zoom command from the zoom button 12 in the operation unit 34. Control the zoom position of the zoom lens to change the zoom magnification.
- the diaphragm 22 is configured by an iris diaphragm, for example, and is driven by the diaphragm driving unit 29 to operate.
- the CPU 32 controls the aperture amount (aperture value) of the aperture 22 via the aperture drive unit 29 and controls the amount of light incident on the CCD 24.
- the mechanical shutter 23 determines the exposure time in the CCD 24 by opening and closing the optical path, and prevents unwanted light from entering the CCD 24 when the image signal is read from the CCD 24, thereby preventing smear.
- the CPU 32 outputs a shutter close signal synchronized with the exposure end time corresponding to the shutter speed to the shutter control unit 30 to control the mechanical shutter 23.
- the CCD 24 is composed of a two-dimensional color CCD solid-state imaging device. A large number of photodiodes are two-dimensionally arranged on the light receiving surface of the CCD 24, and color filters are arranged in a predetermined arrangement on each photodiode.
- the optical image of the subject imaged on the CCD light receiving surface via the optical unit having the above configuration is converted into a signal charge corresponding to the amount of incident light by the photodiode.
- the signal charge accumulated in each photodiode is sequentially read out from the CCD 24 as a voltage signal (image signal) corresponding to the signal charge based on a drive pulse given from the CCD control unit 31 according to a command from the CPU 32.
- the CCD 24 has an electronic shutter function, and the exposure time (shutter speed) is controlled by controlling the charge accumulation time in the photodiode.
- the electronic shutter controls the charge accumulation start time corresponding to the shutter speed, and the exposure end time (charge accumulation end time) is controlled by closing the mechanical shutter 23.
- the CCD 24 is used as the image pickup device, but an image pickup device having another configuration such as a CMOS sensor may be used.
- the analog signals R, G, and B read from the CCD 24 are subjected to correlated double sampling (CDS) and amplification by the analog signal processing unit 25, and then the R, G, and B analog signals are output by the A / D converter 26. Converted to a digital signal.
- CDS correlated double sampling
- the image input controller 27 has a built-in line buffer having a predetermined capacity, and temporarily stores R, G, B image signals (CCDRAW data) A / D converted by the A / D converter 26 and then a bus 60. And stored in the RAM 54.
- the CPU 32 controls the imaging unit 20-2 that captures the right viewpoint image in the same manner as the imaging unit 20-1 that captures the left viewpoint image in the 3D imaging mode.
- the AE detection unit 46 calculates subject brightness necessary for AE control based on an image signal captured when the shutter button 11 is half-pressed, and outputs a signal indicating the subject brightness (shooting EV value) to the CPU 32.
- the CPU 32 sets the shutter speed (exposure time), aperture value, and imaging sensitivity in the plurality of imaging units 20-1 and 20-2 according to a predetermined program diagram based on the input imaging EV value.
- the AF detection unit 48 integrates the absolute value of the high frequency component of the image signal in the AF area captured when the shutter button 11 is half-pressed, and outputs this integrated value (AF evaluation value) to the CPU 32.
- the CPU 32 moves the focus lens from the closest position to the infinity side, searches for a focus position where the AF evaluation value detected by the AF detection unit 48 is maximum, and moves the focus lens to the focus position. Adjust the focus on the subject (main subject).
- so-called hill climbing control is performed in which the focus lens is moved so that the AF evaluation value always takes the maximum value.
- the AWB detection unit 50 automatically obtains the light source type (the color temperature of the object scene) based on the R, G, and B image signals acquired at the time of the main imaging, and R, G, The corresponding white balance gain is read out from the table storing the B white balance gain (white balance correction value).
- the digital signal processing unit 44 interpolates a spatial shift of color signals such as R, G, and B accompanying a white balance correction circuit, a gradation conversion processing circuit (for example, a gamma correction circuit), and a color filter array of a single-plate CCD.
- the image processing is performed on the R, G, and B image signals (CCDRAW data) stored in the RAM 54, including a synchronization circuit for aligning the position of each color signal, a contour correction circuit, a luminance / color difference signal generation circuit, and the like. .
- the R, G, and B CCDRAW data are multiplied by the white balance gain detected by the AWB detection unit 50 in the digital signal processing unit 44 to be subjected to white balance correction, and thereafter, gradation conversion processing (for example, After predetermined processing such as gamma correction is performed, the signal is converted into a YC signal including a luminance signal (Y signal) and a color difference signal (Cr, Cb signal).
- Y signal a luminance signal
- Cr, Cb signal color difference signal
- the YC signal processed by the digital signal processing unit 44 is stored in the RAM 54.
- the digital signal processing unit 44 cuts out an image of a predetermined cutout area from each of the distortion correction circuit and the left and right viewpoint images for correcting the lens distortion correction of the imaging optical systems of the plurality of imaging units 20-1 and 20-2.
- the image pickup processing circuit includes an image cutout processing circuit that corrects an optical axis shift of the imaging optical system of the plurality of imaging units 20-1 and 20-2.
- the compression / decompression processing unit 42 compresses the YC signal stored in the RAM 54 in accordance with a command from the CPU 32 during recording on the memory card 40, and decompresses the compressed compressed data recorded on the memory card 40. To YC signal.
- the recording control unit 38 converts the compressed data compressed by the compression / decompression processing unit 42 into a predetermined format image file (for example, 3D still image is MP file, 3D moving image is MPEG4, MPEG4-MVC, motion JPEG, H.264, etc.). H.264 moving image file) is recorded on the memory card 40, or the moving image file is read from the memory card 40.
- a predetermined format image file for example, 3D still image is MP file, 3D moving image is MPEG4, MPEG4-MVC, motion JPEG, H.264, etc.
- the recording control unit 38 when recording the 3D moving image file according to the present invention, the recording control unit 38, as described in the first to fourth embodiments, attached information such as an image size, an aspect ratio, and a frame rate. Besides, GOP maximum display size, scene maximum display size, assumed viewing distance, GOP maximum parallax amount (near view) (%), GOP maximum parallax amount (far view) (%), scene maximum parallax amount (near view) (%), The scene maximum parallax amount (far view) (%) is recorded in the memory card 40 as attached information.
- the 3D moving image file in addition to the case of creating and recording the 3D moving image file described in the first to fourth embodiments, it is once recorded on the memory card 40 as a normal 3D moving image file, Thereafter, the 3D moving image file may be read from the memory card 40, the 3D moving image file described in the first to fourth embodiments may be created, and recorded on the memory card 40 again.
- the liquid crystal monitor 16 is used as an image display unit for displaying captured images, and is used as a GUI (graphical user interface) at various settings.
- the liquid crystal monitor 16 is used as an electronic viewfinder that displays a live view image (hereinafter referred to as “through image”) for confirming the angle of view in the shooting mode.
- the display control unit 36 alternately displays the left viewpoint image and the right viewpoint image held in the VRAM 52 pixel by pixel.
- the left and right images alternately arranged pixel by pixel are visually recognized separately by the left and right eyes of the user observing from a predetermined distance. This enables stereoscopic viewing.
- the above-described embodiment can be realized by the stereoscopic imaging device 10 configured as described above.
- the stereoscopic imaging device 10 described here includes two imaging units for capturing images of two left and right viewpoints, but includes three or more imaging units and captures three or more viewpoint images. You may comprise as follows. For example, as shown in FIG. 8, it may be configured to capture four viewpoint images by providing four imaging units such as the four imaging devices 101-1 to 101-4.
- the output display size of the output 3D display for displaying the 3D moving image is acquired (step S21).
- This output display size can be obtained from the HDMI terminal of the 3D display.
- the left-view GOP_L and the right-view GOP_R are read from the 3D video file, and these are held in the cache (step S22), and the left-view GOP_L and the right view are kept until the cached amount reaches a predetermined maximum cache amount.
- GOP_R is added to the cache (step S23).
- step S24 it is determined whether or not reading of all left-view GOP_L and right-view GOP_R of the 3D video file has ended (step S24), and reading has not ended.
- the GOP maximum display size is determined from the attached information of the left-view GOP_L and the right-view GOP_R to be output to the 3D display of the cached left-view GOP_L and right-view GOP_R.
- Read and temporarily store step S25). Subsequently, a comparison is made between the width of the output display size acquired in step S21 and the width of the GOP maximum display size stored in step S25 (step S26).
- the left-view GOP_L and the right-view GOP_R may be displayed as they are.
- the data is output as it is to the 3D display (step S27).
- the width of the output display size is larger than the width of the GOP maximum display size
- the GOP_L for the left viewpoint and the GOP_R for the right viewpoint having the GOP maximum display size as attached information are displayed.
- the amount of parallax at the parallax position exceeds the distance between human eyes, and that portion cannot be stereoscopically viewed. Therefore, it is necessary to correct the amount of parallax between the left and right viewpoint images.
- the correction of the amount of parallax is performed by shifting the parallax (step S28).
- FIG. 14 is a diagram for explaining the principle of parallax shifting.
- FIG. 15A shows a left viewpoint image
- FIG. 11B shows a right viewpoint image.
- the viewer's left eye is at coordinates (0, D)
- the viewer's right eye is at coordinates (X B , D).
- the object displayed at the coordinates (X L , 0) of the left viewpoint image and the coordinates (X R , 0) of the right viewpoint image is the coordinates (X P , Y P ).
- the amount of parallax can be adjusted by shifting the parallax. Therefore, when the amount of parallax on the distant view side exceeds the human binocular interval, the parallax shift can be performed to fit within the human binocular interval, and appropriate stereoscopic viewing can be achieved.
- the GOP maximum parallax amount P ′ after the parallax correction is expressed by the following equation [Equation 1].
- P ′ ⁇ W2 / W1 ⁇ * P To be.
- the parallax on the 3D display of the 3D video in 1 GOP is within the human binocular interval by shifting the pixels in the direction in which one or both of the left and right viewpoint images approach by the amount of parallax (%) shown in the following equation Can fit.
- the parallax correction is not limited to the above-described parallax shift, and may be performed by parallax compression.
- the parallax compression can be performed by generating a virtual viewpoint image having a smaller amount of parallax than the left and right viewpoint images and displaying the generated virtual viewpoint image.
- Which of the parallax amounts is to be corrected may be determined in advance, or may be configured to be selectable by the user. By adjusting either amount of parallax, at least the disparity on the far side can be reduced.
- the left-view GOP_L and the right-view GOP_R corrected for parallax in step S28 are output to the 3D display (step S27).
- the GOP maximum display size recorded as the attached information is read out, and compared with the output display size. If the GOP maximum display size is larger, it is determined that there is no problem in stereoscopic view.
- GOP_R is output to the 3D display as it is, and if the output display size is larger, it is determined that there is a region where stereoscopic viewing is not possible, and parallax correction is performed for all frames of GOP_L for the left viewpoint and GOP_R for the right viewpoint.
- the above processing is performed by reading the new left-view GOP_L and the right-view GOP_R and adding them to the cache. 3D video can be played.
- step S24 when reading of all left-view GOP_L and right-view GOP_R of the 3D video is finished (“YES”), the left-view GOP_L and right-view GOP_R in the cache are output to the 3D display ( Step S29) The playback of the 3D moving image is terminated. Note that the processing in steps S25 to S28 is also performed when the left-view GOP_L and the right-view GOP_R in the cache are output to the 3D display.
- whether or not the stereoscopic video for each GOP displayed on the output 3D display is stereoscopic is determined using the GOP maximum display size recorded in the attached information of the 3D video file.
- the present invention is not limited to this, and the same operation can be performed using the GOP maximum parallax amount.
- the allowable amount of parallax is calculated based on the display size of the 3D display at the output destination and a predetermined value (for example, 50 mm) indicating the distance between both eyes of the human.
- the allowable parallax amount is (human binocular interval) / (display size) (%).
- the human binocular interval is 50 mm and the display size (width) is 1300 mm
- the allowable parallax amount is 3.85 ( %)become.
- whether or not binocular fusion is possible can be determined based on whether or not the GOP maximum parallax amount (%) is equal to or less than the allowable parallax amount.
- an image shift amount on a 3D display corresponding to the GOP maximum parallax amount is calculated based on the GOP maximum parallax amount and the display size of the output 3D display, and the calculated image shift is calculated. Whether or not binocular fusion is possible may be determined based on whether or not the amount exceeds a predetermined value indicating a human binocular interval.
- the parallax correction can be performed for each GOP.
- the present invention is not limited to this, and the maximum display size and the maximum parallax within the predetermined number of GOPs are not limited thereto.
- the parallax may be corrected based on the amount or the like.
- FIG. 16 is a flowchart showing a second embodiment of 3D moving image playback processing.
- the same step number is attached
- the first embodiment is directed to a 3D video file in which the GOP maximum display size and the GOP maximum parallax amount are recorded as attached information for each GOP, but the second embodiment is as shown in FIG. The difference is that a 3D moving image file in which the maximum scene display size and the maximum scene parallax amount are recorded as attached information is targeted for each scene.
- step S30 determines whether there is a scene head flag.
- the scene head flag is detected (“YES”)
- the maximum scene display size is read from the header at the head of the scene and temporarily stored (step S31). Subsequently, the width of the output display size acquired in step S21 is compared with the width of the maximum scene display size stored in step S31 (step S32).
- the width of the maximum scene display size is equal to or larger than the width of the output display size, even if a 3D moving image in the scene is displayed, there is no problem with stereoscopic viewing, and thus the image is directly output to the 3D display (step S27).
- the parallax of the 3D video in the scene is corrected, and the 3D video in the scene is corrected.
- the parallax on the 3D display is set to fall within the distance between human eyes.
- the scene maximum parallax amount may be used to determine whether or not the 3D video of the scene is stereoscopically viewable.
- FIG. 17 is a flowchart showing a third embodiment of 3D moving image playback processing.
- the same step number is attached
- the first embodiment is directed to a 3D moving image file in which the GOP maximum display size and the GOP maximum parallax amount are recorded as attached information for each GOP, but the third embodiment has a GOP maximum display for each GOP.
- the size and the GOP maximum parallax amount are recorded as attached information, and as shown in FIG. 7, a 3D moving image file in which the scene maximum display size and the scene maximum parallax amount are recorded as attached information is targeted. Is different.
- step S40 reads and temporarily stores the GOP maximum parallax amount from the cached left viewpoint GOP_L and right viewpoint GOP_R attached information.
- a step S41 determines whether or not there is a scene head flag. When the scene head flag is detected (“YES”), the scene length, the maximum scene display size, and the maximum scene parallax amount are read from the header at the head of the scene and temporarily stored (steps S42, S43, S45).
- step S45 it is determined whether or not the scene length stored in step S42 is sufficiently longer than 1 GOP (step S45). If it is determined to be short (in the case of “NO”), the parallax correction is performed for each GOP as in the first embodiment (step S46), and the GOP_L of the left viewpoint and the GOP_R of the right viewpoint subjected to the parallax correction are determined. It outputs to a 3D display (step S47). If the maximum display size of the GOP is equal to or larger than the output display size, the parallax correction is not performed and the data is output to the 3D display as it is.
- step S45 determines whether the length is long (“YES”)
- the process returns to step S22, and then passes through the processing in steps S41 and S48 until the predetermined maximum cache amount is reached from the scene head. , Left-view GOP_L and right-view GOP_R are cached.
- parallax correction is performed on the 3D video in the scene using the parallax correction table shown below (Ste S49).
- FIG. 18 is a graph showing an example of the parallax correction table.
- the horizontal axis indicates the GOP maximum parallax amount / scene maximum parallax amount
- the vertical axis indicates the parallax correction coefficient (0 to 1).
- the parallax correction coefficient is GOP maximum parallax amount / scene maximum.
- the parallax amount is set to a value that changes from 1 to X as the parallax amount approaches 0 to 1.
- the parallax correction coefficient is multiplied for each GOP maximum parallax amount in the scene.
- the parallax correction is performed so that each GOP maximum parallax amount in the scene becomes the multiplied value (that is, the corrected GOP maximum parallax amount).
- the parallax is shifted so that the GOP maximum parallax amount P ′ is obtained (that is, the parallax is shifted by the parallax amount expressed by the formula 2), and thereby the parallax on the 3D display of the 3D video in the GOP is changed by the human. It is possible to fit within the distance between both eyes.
- the parallax correction coefficient is determined.
- step S49 by performing parallax correction using the parallax correction table, the parallax amount is gradually decreased as the maximum parallax amount of each GOP approaches the maximum parallax amount in the scene (scene maximum parallax amount).
- the parallax correction is performed to reduce the parallax reduction amount as the distance from the maximum parallax amount in the scene decreases.
- the left-view GOP_L and the right-view GOP_R corrected for parallax in step S49 are output to the 3D display (step S47).
- the maximum scene display size is equal to or larger than the output display size (when there is no problem in 3D display)
- the above-described parallax correction is not performed, and the left viewpoint GOP_L and the right viewpoint GOP_R that are not parallax corrected are output to the 3D display. Is done.
- parallax correction even if some GOPs in the scene have parallax that causes a problem in 3D display, parallax correction of other GOPs is not uniformly corrected, while suppressing excessive parallax. The stereoscopic effect of the entire scene can be ensured.
- the parallax shifting process may be performed in consideration of the maximum amount of parallax (near view).
- the required shift amount is added to the product of the display size (width) of the output 3D display and the maximum parallax amount (near view), and whether or not the added value is equal to or smaller than the binocular interval (for example, 50 mm). Determine whether.
- the binocular interval for example, 50 mm.
- the parallax shift is performed, and the viewpoint images of the two viewpoints subjected to the parallax shift are displayed in 3D on the 3D display.
- the distance is larger than 50 mm, there is a part that cannot be stereoscopically viewed on the far side unless the parallax is shifted, and the near side cannot be stereoscopically viewed properly if the parallax is shifted so that the maximum parallax amount on the far side is appropriate. Therefore, 3D display is not performed and 2D display is performed on the 3D display.
- the 2D display is performed by outputting one viewpoint image from the viewpoint images recorded in the 3D moving image file to the 3D display.
- a warning may be displayed to the effect that the 3D video is not displayed because the display size is too large.
- an appropriate 3D moving image can be displayed.
- step S51 the display size (width) of a 3D display that displays a 3D moving image is acquired.
- the acquired horizontal width of the display is 600 mm.
- GOPs for four viewpoints shown in FIG. 9 are read (step S52).
- the GOP maximum display size is acquired from each header area of the read GOP in the order of viewpoint numbers (step S53).
- the GOP maximum display size, the assumed viewing distance, and the GOP maximum parallax amount (distant view) of a GOP for four viewpoints are as shown in Table 1 below.
- the GOP maximum display size 500 mm of the viewpoint image (1) having the viewpoint number 1 is acquired.
- step S54 it is determined whether or not the acquired GOP maximum display size is equal to or larger than the display size acquired in step S51 (step S54).
- the process proceeds to step S56.
- step S56 it is determined whether or not acquisition of the GOP maximum display size has been completed for all viewpoint images.
- step S53 the viewpoint number is incremented by 1, and then the GOP maximum display size 1200 mm of the viewpoint image (2) that is the viewpoint number 2 is acquired.
- step S54 it is determined whether or not the acquired GOP maximum display size is equal to or larger than the display size. Since the GOP maximum display size is 1200 mm and the display size is 600 mm or more this time, the process proceeds to step S55.
- step S55 the current viewpoint number is acquired. Here, viewpoint number 2 is acquired.
- step S53 to step S55 is repeated for all viewpoint images.
- viewpoint number of the viewpoint image whose GOP maximum display size is equal to or larger than the display size is acquired.
- viewpoint number 2 and viewpoint number 3 are acquired.
- a viewpoint image of two viewpoints to be output to the 3D display is selected from the viewpoint images of the viewpoint number acquired in step S55 (step S57).
- the viewpoint image (1) which is the image of the reference viewpoint, or the image of viewpoint number 2 (viewpoint image (2)) or the image of viewpoint number 3 (viewpoint image (3)) is selected. .
- a viewpoint image with the largest amount of parallax a viewpoint image with the smallest amount of parallax, a viewpoint image close to the central viewpoint, and the like can be considered.
- the actual viewpoint image actually taken may be selected with priority over the virtual viewpoint image.
- This reference may be determined in advance or may be configured so that the viewer can freely set it.
- the viewpoint image (1) and the viewpoint image (3) are selected.
- the viewpoint images of the two viewpoints selected in step S57 are displayed in 3D on the 3D display (step S58). That is, here, 3D display is performed based on the viewpoint image (1) and the viewpoint image (3).
- step S59 it is determined whether or not all GOPs have been read from the 3D moving image file (step S59). If not completed (“NO”), the process proceeds to step S52, and the above process is repeated. If completed (“YES”), 3D video playback is terminated.
- step S55 If there is no viewpoint number acquired in step S55, that is, if the GOP maximum display size of all viewpoint images is smaller than the display size, 2D display is performed.
- the reference for selecting the viewpoint image for 2D display may be determined as appropriate, but the viewpoint image for the reference viewpoint is preferable as the viewpoint image for the actual viewpoint.
- the GOP maximum display size recorded as ancillary information of each viewpoint image is read, compared with the display size of the output destination 3D display, and a viewpoint image pair having a larger GOP maximum display size is selected.
- a viewpoint image pair having a larger GOP maximum display size is selected.
- processing can also be performed based on the GOP maximum parallax amount recorded in the attached information of each viewpoint image of the 3D moving image file.
- FIG. 20 is a flowchart when selecting a pair of viewpoint images to be 3D reproduced and displayed based on the GOP maximum parallax amount.
- symbol is attached
- the GOP maximum display size is not recorded in the attached information of each viewpoint image of the 3D moving image file to be read, and other attached information is the same as the file shown in Table 1.
- the horizontal width of the output 3D display is 600 mm.
- the display size is acquired (step S51), and GOPs for four viewpoints are read from the 3D moving image file shown in FIG. 9 (step S52).
- two viewpoint images are selected from each header region of the read GOP (step S61). Since the combination of all viewpoint images is finally selected, the selection order may be determined as appropriate.
- the viewpoint image (1) and the viewpoint image (2) are first selected.
- the GOP maximum parallax amount (distant view) of the selected two viewpoint images is acquired (step S62).
- the GOP maximum parallax amount (distant view) recorded in the attached information of each viewpoint image is a parallax amount with respect to the reference viewpoint image. Therefore, when the reference viewpoint image is not included in the selected two viewpoint images, it is necessary to recalculate the GOP maximum parallax amount.
- the GOP maximum parallax amount (distant view) recorded in the attached information of the viewpoint image (2) is the same. This is the GOP maximum parallax amount (distant view) of these two viewpoints.
- Step S63 it is determined whether or not the product of the display width acquired in step S51 and the GOP maximum parallax amount (distant view) of the two viewpoint images acquired in step S62 is equal to or less than a human binocular interval of 50 mm. It should be noted that the distance between human eyes is not limited to 50 mm, and a numerical value such as 65 mm may be used.
- step S65 If this product is larger than 50 mm, the GOP maximum parallax position on the far side cannot be properly stereoscopically viewed, so the combination of the viewpoint images of the two viewpoints is determined to be a combination that is not suitable for stereoscopic viewing. Control goes to step S65.
- step S64 the viewpoint numbers of the two viewpoint images are acquired. That is, the combination of viewpoint numbers 1 and 2 is acquired here.
- step S65 it is determined whether or not acquisition of the GOP maximum parallax amount has been completed for all combinations. Here, since all combinations have not been completed yet, the process returns to step S61.
- two different viewpoint (one set) viewpoint images are selected.
- two viewpoint images, viewpoint image (2) and viewpoint image (3) are selected.
- step S62 the GOP maximum parallax amount (distant view) of these two viewpoints is acquired.
- the maximum parallax amount (distant view) recorded in the auxiliary information of the viewpoint image (2) and the viewpoint image (3) is the maximum parallax amount with respect to the viewpoint image (1) that is the reference viewpoint image. Therefore, the GOP maximum parallax amount at the two viewpoints of the viewpoint image (2) and the viewpoint image (3) needs to be recalculated.
- step S63 it is determined whether the product of the calculated GOP maximum parallax amount (distant view) of the two viewpoint images and the display width is 50 mm or less.
- step S64 the viewpoint numbers of the two viewpoint images, that is, the combination of viewpoint numbers 2 and 3 are acquired.
- step S62 to step S64 is repeated for all the viewpoint image combinations.
- step S64 only the viewpoint number of the combination of viewpoint images whose product of the GOP maximum parallax amount (distant view) and the display size is 50 mm or less is acquired.
- viewpoint numbers 1 and 2, 2 and 3, 2 and 4, 3 and 4 are acquired.
- a viewpoint image of two viewpoints to be output to the 3D display is selected from the combination of viewpoint numbers acquired in step S64 (step S57).
- a combination having the largest GOP parallax amount, a combination having the smallest GOP parallax amount, a combination including a viewpoint image close to the central viewpoint, and giving priority to the actual viewpoint image may be considered.
- This reference may be determined in advance or may be configured so that the viewer can freely set it.
- viewpoint number 2 and viewpoint number 3 which is the combination with the smallest amount of parallax, that is, viewpoint image (2) and viewpoint image (3) is selected.
- the viewpoint images of the two viewpoints selected in step S57 are displayed in 3D on the 3D display (step S58). That is, here, 3D display is performed based on the viewpoint image (2) and the viewpoint image (3).
- step S59 it is determined whether or not all GOPs have been read from the 3D moving image file (step S59). If not completed (“NO”), the process proceeds to step S52, and the above process is repeated. If completed (“YES”), 3D video playback is terminated.
- step S64 If there is no viewpoint number acquired in step S64, that is, if all the combinations do not meet the condition of step S63, 2D display is performed.
- the criteria for selecting an image for 2D display may be determined as appropriate.
- the GOP maximum parallax amount (distant view) recorded as the attached information of each viewpoint image is read, and the product with the display size of the output 3D display is calculated, and the product is smaller than the human binocular interval.
- the viewpoint image is selected using both information.
- FIG. 21 is a flowchart showing a case where a viewpoint image is selected based on the GOP maximum display size and the GOP maximum parallax amount (distant view).
- symbol is attached
- a viewpoint image capable of 3D display is selected based on the GOP maximum display size (steps S53 to S56).
- a viewpoint image that can be displayed in 3D is selected based on the GOP maximum parallax amount (distant view) (steps S61 to S65). In this case, it is not necessary to perform processing for the combination with the reference viewpoint image.
- the two-viewpoint viewpoint image to be output to the 3D display is selected from the combination with the reference viewpoint image acquired in step S55 and the combination of the two viewpoint images acquired in step S64 (step S57).
- the criteria to be selected may be appropriately determined as before.
- the viewpoint image of two viewpoints to be displayed is selected based only on the GOP maximum parallax amount on the far side, but may be determined in consideration of the GOP maximum parallax amount on the near side.
- the GOP maximum parallax amount on the near view side not only the distant view side but also the close view side can appropriately determine a combination of images that can be stereoscopically viewed.
- a combination in which the amount of parallax in the foreground is equal to or less than a predetermined value may be selected from the combinations of the viewpoint images of the two viewpoints acquired in step S64. This is because, when the amount of parallax on the near view side increases, the viewer feels tired when viewing stereoscopically, so it is preferable to set the value to 50 mm or less.
- the value can be used. Further, as described with reference to FIG. 10, feature points may be extracted from each viewpoint image and calculated from the maximum parallax amount position on the foreground side.
- FIG. 22 is a flowchart illustrating image reproduction processing according to the seventh embodiment.
- the read 3D moving image file is the same as the file shown in Table 1.
- the horizontal width of the display is assumed to be 1300 mm.
- the display size of the output 3D display is acquired (step S51).
- the display size can be acquired from a 3D display connected by an HDMI cable.
- GOPs for four viewpoints are read from the 3D moving image file shown in FIG. 9 (step S52), and the value of the reference viewpoint number tag is read from each header area of the read GOP (step S71).
- the reference viewpoint number 1 is read.
- the GOP maximum display size is acquired from the attached information of the viewpoint image of the reference viewpoint number 1, that is, the viewpoint image (1) (step S72).
- the value recorded here is the GOP maximum display size in the combination of viewpoints that has the maximum amount of parallax in combination with the viewpoint image (1).
- the viewpoint image (4) GOP maximum display size in combination is the GOP maximum display size in the combination of viewpoints that has the maximum amount of parallax in combination with the viewpoint image (1).
- the GOP maximum display size is compared with the display size acquired in step S71 (step S73). If the GOP maximum display size is larger than the display size (in the case of “YES”), there is no problem even if the viewpoint image (1) and the viewpoint image (4) are displayed in 3D as they are. Is displayed in 3D on the 3D display (step S74).
- step S73 if the comparison result in step S73 is NO, the process proceeds to step S75.
- Step S75 acquires the viewpoint numbers of all the read viewpoint images. Furthermore, the value of the GOP maximum parallax amount of each viewpoint image is acquired from the attached information of each viewpoint image (step S76). As shown in Table 1, the GOP maximum parallax amount 4% is acquired from the viewpoint image (2), the GOP maximum parallax amount 7% is acquired from the viewpoint image (3), and the GOP maximum parallax amount 10% is acquired from the viewpoint image (4).
- the allowable parallax amount in the 3D display that performs 3D display is calculated (step S77).
- the allowable parallax amount is calculated by the following [Equation 3].
- the display size refers to the horizontal width of the display.
- the human binocular interval is 50 mm here, other numerical values may be used.
- Allowable parallax amount [%] 50 mm / display size [mm] ⁇ 100
- the allowable parallax amount is about 3.85%.
- step S78 the arrangement of the reference viewpoint is confirmed (step S78). If the reference viewpoint is not near the center of all viewpoints, the viewpoint position near the center is set as a new reference viewpoint (step S79), and the maximum parallax amount of each viewpoint image with respect to the new reference viewpoint is calculated (step S80). .
- viewpoint number 2 is set as a new reference viewpoint as the viewpoint near the center.
- the GOP maximum parallax amount of each viewpoint image with respect to the viewpoint number 2 which is the new reference viewpoint is calculated.
- the absolute value of the difference between the GOP maximum parallax amount described in the attached information of each viewpoint image and the GOP maximum parallax amount described in the attached information of the viewpoint image (2) is a new value. This is the GOP maximum parallax amount at the reference viewpoint. Therefore, the GOP maximum parallax amount of each viewpoint image is as follows.
- GOP maximum parallax of viewpoint image (1)
- 6% Note that such processing is performed because viewpoint images arranged closer to the center are more suitable for 3D display.
- step S80 When the process of step S80 ends, or when it is determined in step S78 that the reference viewpoint is near the center of all viewpoints, the maximum value that is equal to or smaller than the allowable parallax amount among these GOP maximum parallax amounts Is selected (step S81).
- the allowable parallax amount is 3.85% or less, the viewpoint satisfying this condition is the viewpoint number 3.
- step S82 when a viewpoint satisfying the condition can be selected (step S82), 3D display is performed using the reference viewpoint image and the selected viewpoint image (step S83). If it cannot be selected (“NO” in step S82), 2D display is performed (step S84).
- the allowable parallax amount is 2.94%, and a viewpoint having the maximum maximum parallax amount value that is equal to or smaller than the allowable parallax amount cannot be selected. In this case, 2D display is performed.
- the pair of viewpoint images is selected based on the allowable amount of parallax, it is possible to appropriately select a viewpoint image suitable for stereoscopic viewing.
- step S85 it is determined whether or not all GOPs have been read from the 3D moving image file. If not completed ("NO"), the process proceeds to step S52, and the above process is repeated. If completed (“YES”), 3D video playback is terminated.
- step S82 2D display may be performed after performing parallax shifting or parallax compression to adjust to an appropriate amount of parallax, instead of performing 2D display.
- parallax correction is performed for each GOP.
- the present invention is not limited to this. For each scene using the maximum scene display size and the maximum scene parallax amount. You may make it correct parallax.
- the first to seventh embodiments of the 3D moving image reproduction may be realized by the 3D moving image reproduction function of the stereoscopic imaging apparatus 10 shown in FIGS. 11 and 12, or the stereoscopic moving image reproduction without the imaging unit. You may make it implement
- FIG. 23 is a diagram showing the overall configuration of the stereoscopic video playback device 300 and the 3D display 320. As shown in the figure, the stereoscopic video playback device 300 and the 3D display 320 are separately configured devices, and are communicably connected by a communication cable 310.
- the 3D display 320 is a parallax barrier type or lenticular type display, and alternately displays the left viewpoint image and the right viewpoint image input from the stereoscopic video playback device 300 to the 3D display 320 for each line.
- the 3D display 320 may be one that displays the left viewpoint image and the right viewpoint image by alternately switching in time. In this case, the viewer visually recognizes the 3D display 320 using special glasses.
- FIG. 24 is a block diagram showing an internal configuration of the stereoscopic video playback device 300.
- the stereoscopic moving image playback apparatus 300 includes a CPU 301, a recording control unit 305, a memory card 306, a display control unit 307, a communication interface 308, and the like.
- the CPU 301 controls the overall operation of the stereoscopic moving image playback apparatus 300 based on a control program such as a stereoscopic moving image playback program recorded in the ROM 302.
- the RAM 303 is used as a calculation work area for the CPU 301.
- the recording control unit 305 and the display control unit 307 are connected to the CPU 301 via the bus 304.
- the recording control unit 305 controls data reading / writing of the 3D moving image file with respect to the memory card 306.
- the memory card 306 is, for example, the same as the memory card 40 of the stereoscopic imaging device 10 shown in FIG. 12, and stores a 3D moving image file including each viewpoint image captured by the stereoscopic imaging device 10 and attached information. .
- the communication interface 308 is a connector unit to which the communication cable 310 is connected, and the display control unit 307 displays a 3D moving image on the 3D display 320 via these.
- the communication interface 308 and the communication cable 310 it is conceivable to adopt an HDMI standard.
- the stereoscopic video playback device 300 can acquire the display size of the 3D display 320 connected via the communication cable 310.
- the stereoscopic video playback device 300 may be provided with a compound eye imaging unit for capturing each viewpoint image, and the captured viewpoint image may be recorded in the memory card 306.
- the stereoscopic video playback device 300 and the 3D display 320 may be configured as an integrated device.
- a stereoscopic video playback program may be installed in a personal computer so that the personal computer functions as the stereoscopic video playback device 300.
- SYMBOLS 10 Stereoscopic imaging device, 14-1, 14-2 ... Shooting optical system, 16 ... Liquid crystal monitor, 20-1, 20-2 ... Imaging part, 21 ... Focus lens and zoom lens, 24 ... CCD, 25 ... Analog signal Processing unit 32 ... Central processing unit (CPU) 34 ... Operation unit 40 ... Memory card 44 ... Digital signal processing unit 100 ... Subject, 101-1 to 101-4 ... Imaging device 211N, 212N ... Maximum parallax Amount position (near view), 211F, 212F ... Maximum parallax amount position (far view), 213N ... Maximum parallax amount on the near view side, 213F ... Maximum parallax amount on the far view side, 213F ... Maximum parallax amount on the far view side, 300 ... Stereoscopic video playback Device, 320 ... 3D display
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Library & Information Science (AREA)
- Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)
- Stereoscopic And Panoramic Photography (AREA)
Abstract
Description
第1の実施形態に係る立体動画ファイル(3D動画ファイル)について説明する。
・想定視距離(立体動画を観察する際の視認者と3Dディスプレイとの距離)(単位:mm)
・GOP最大視差量(近景):画像幅に対する視差量(%)
・GOP最大視差量(遠景):画像幅に対する視差量(%)
・各視点画像を撮影した装置の輻輳角、基線長、撮像部配置(視点番号)
このような3D動画ファイルを記録するための撮影、記録処理について、図3のフローチャートを用いて説明する。
第2の実施形態では、図7に示すように3D動画のシーンが変わる場合、そのシーンの先頭の付属情報として、下記の付属情報の全部又は一部を記録する。
・シーン最大ディスプレイサイズ(幅、高さ、単位:mm)
・想定視距離(立体動画を観察する際の視認者と3Dディスプレイとの距離)(単位:mm)
・シーン最大視差量(近景):画像幅に対する視差量(%)
・シーン最大視差量(遠景):画像幅に対する視差量(%)
・各視点画像を撮影した装置の輻輳角、基線長、撮像部配置(視点番号)
・各視点画像を撮影した装置の輻輳角、基線長、撮像部配置(視点番号)
更に、ズームIN/OUTされたことを示すフラグ、手ブレ補正がON/OFF/Resetされたことを示すフラグを付属情報として記録するようにしてよい。
第1の実施形態では、2視点から撮影した左右の視点画像を例に説明したが、本発明に係る3D動画の視点の数は2視点に限らず、3視点以上の多視点の視点画像であってもよい。
第3の実施形態のような多視点画像は、全てが実際に撮影された画像である必要はなく、仮想視点に対応した仮想視点画像を含んでいてもよい。
図11A及び図11Bは本発明に係る立体撮像装置の外観を示す図であり、図11Aは立体撮像装置を前面側から見た斜視図であり、図11Bは背面図である。
図12は上記立体撮像装置10の内部構成を示すブロック図である。
次に、上記のように記録された3D動画ファイルを読み出して3D動画再生する処理について、図13のフローチャートを用いて説明する。
[数1]
P’={W2/W1}*P
になるようにする。したがって、次式に示す視差量(%)だけ、左右視点画像の一方又は両方を接近する方向に画素ずらしすることにより、1GOP内の3D動画の3Dディスプレイ上の視差を人間の両眼間隔内に収めることができる。
視差量(%)=P-P’
また、視差の補正は、上記視差ずらしに限らず、視差圧縮により行うようにしてもよい。視差圧縮は、図10を用いて説明したように、左右の視点画像よりも視差量の小さい仮想視点の画像を生成し、生成した仮想視点の画像を表示することにより行うことができる。どちらによって視差量を補正するのかは、予め決めておいてもよいし、ユーザが選択可能に構成してもよい。どちらの視差量調整によっても、少なくとも遠景側の視差を小さくすることができる。
図16は、3D動画再生処理の第2の実施形態を示すフローチャートである。尚、図13に示した第1の実施形態と共通する部分には同一のステップ番号を付し、その詳細な説明は省略する。
図17は、3D動画再生処理の第3の実施形態を示すフローチャートである。尚、図13に示した第1の実施形態と共通する部分には同一のステップ番号を付し、その詳細な説明は省略する。
次に、図9に示したように3視点以上の視点画像が記録された3D動画ファイルを読み出して再生表示する処理について、図19のフローチャートを用いて説明する。
図20は、GOP最大視差量に基づいて3D再生表示する視点画像のペアを選択する場合のフローチャートである。尚、図19に示すフローチャートと共通する部分には同一の符号を付し、その詳細な説明は省略する。また、読み出される3D動画ファイルの各視点画像の付属情報には、GOP最大ディスプレイサイズが記録されていないものとし、その他の付属情報は表1に示すファイルと同様であるものとする。また、出力先の3Dディスプレイの横幅も同様に600mmであるとする。
第4の実施形態で説明したように、各視点画像の付属情報に記録されているGOP最大ディスプレイサイズに基づいて表示画像を選択する場合には、基準視点画像との組み合わせでしか画像の選択ができない。これに対し、第5の実施形態で説明したように、各視点画像の付属情報に記録されているGOP最大視差量(遠景)に基づいて視点画像を選択する場合には、全視点画像の組み合わせについて判断することができるが、処理が煩雑となる。
図22は、第7の実施形態の画像再生処理を示すフローチャートである。
許容視差量[%]=50mm÷ディスプレイサイズ[mm]×100
本実施形態では、ディスプレイサイズは1300mmであるので、許容視差量は約3.85%となる。
視点画像(2)のGOP最大視差量=|4%-4%|=0%
視点画像(3)のGOP最大視差量=|7%-4%|=3%
視点画像(4)のGOP最大視差量=|10%-4%|=6%
尚、このような処理を行うのは、より中央に近く配置されている視点画像が、3D表示により好適であるためである。
上記3D動画再生の第1から第7の実施形態は、図11及び図12に示した立体撮像装置10の3D動画再生機能により実現するようにしてもよいし、撮像部を有しない立体動画再生装置により実現するようにしてもよい。
Claims (21)
- 複数の視点画像からなる立体画像が時間軸方向に連続している立体動画を取得する立体動画取得手段と、
前記取得した立体動画のフレーム毎に複数の視点画像から特徴が一致する特徴点間のずれ量を示す視差量を算出する視差量算出手段と、
前記算出したフレーム毎の各特徴点の視差量のうちの遠景側の最大視差量を取得する最大視差量取得手段と、
前記取得した遠景側の最大視差量のうちの前記立体動画の所定の区間毎に該区間内で最大となる区間内最大視差量を取得する区間内最大視差量取得手段と、
前記立体動画の所定の区間毎に取得した区間内最大視差量に基づいて、前記所定の区間毎に前記立体動画を立体ディスプレイに表示させる際に両眼融合可能な区間内最大ディスプレイサイズを取得する区間内最大ディスプレイサイズ取得手段と、
前記立体動画が記録される立体動画ファイルを生成し、該立体動画ファイルを記録媒体に記録する記録手段であって、前記立体動画を前記立体動画ファイルに記録するとともに、前記所定の区間毎の区間内最大ディスプレイサイズを付属情報として前記立体動画ファイルに記録する記録手段と、
を備えた立体動画処理装置。 - 前記記録手段は、前記区間内最大視差量取得手段により前記立体動画の所定の区間毎に取得した区間内最大視差量を前記立体動画ファイルの付属情報として更に記録する請求項1に記載の立体動画処理装置。
- 複数の視点画像からなる立体画像が時間軸方向に連続している立体動画を取得する立体動画取得手段と、
前記取得した立体動画のフレーム毎に複数の視点画像から特徴が一致する特徴点間のずれ量を示す視差量を算出する視差量算出手段と、
前記算出したフレーム毎の各特徴点の視差量のうちの遠景側の最大視差量を取得する最大視差量取得手段と、
前記取得した遠景側の最大視差量のうちの前記立体動画の所定の区間毎に該区間内で最大となる区間内最大視差量を取得する区間内最大視差量取得手段と、
前記立体動画が記録される立体動画ファイルを生成し、該立体動画ファイルを記録媒体に記録する記録手段であって、前記立体動画を前記立体動画ファイルに記録するとともに、前記所定の区間毎の区間内最大視差量を付属情報として前記立体動画ファイルに記録する記録手段と、
を備えた立体動画処理装置。 - 前記立体動画の所定の区間毎に取得した区間内最大視差量に基づいて、前記所定の区間毎に前記立体動画を立体ディスプレイに表示させる際に両眼融合可能な区間内最大ディスプレイサイズを取得する区間内最大ディスプレイサイズ取得手段を備え、
前記記録手段は、前記所定の区間毎の区間内最大ディスプレイサイズを付属情報として更に前記立体動画ファイルに記録する請求項3に記載の立体動画処理装置。 - 前記記録手段は、前記立体動画を観察するのに適した視認者と立体ディスプレイとの距離である想定視距離を前記立体動画ファイルの付属情報として更に記録する請求項1から4のいずれか1項に記載の立体動画処理装置。
- 前記最大視差量取得手段は、前記算出した各特徴点の視差量のうちの近景側の最大視差量を取得し、
前記区間内最大視差量取得手段は、前記取得した近景側の最大視差量のうち前記所定の区間内で最大となる区間内最大視差量を該所定の区間毎に取得する請求項1から5のいずれか1項に記載の立体動画処理装置。 - 前記立体動画取得手段により取得されたN(Nは3以上の整数)視点の視点画像のうちの1つを代表画像として設定する代表画像設定手段を備え、
前記最大視差量取得手段は、前記代表画像と前記N視点の視点画像のうちの代表画像を除く(N-1)視点の視点画像とのそれぞれ異なる組み合わせからなる2つの視点画像に関連して(N-1)個の遠景側の最大視差量を取得し、
前記区間内最大ディスプレイサイズ取得手段は、前記取得した(N-1)個の遠景側の最大視差量に基づいて前記所定の区間内で最大となる(N-1)個の区間内最大ディスプレイサイズを前記所定の区間毎に取得し、
前記記録手段は、前記取得した(N-1)個の区間内最大ディスプレイサイズを前記付属情報として前記立体動画ファイルに記録する請求項1、2又は4に記載の立体動画処理装置。 - 前記立体動画取得手段により取得された複数の視点画像のうちの少なくとも1つの視点画像と前記視差量算出手段により算出された視差量とに基づいて任意の仮想視点に対応する1又は複数の仮想視点画像を生成する仮想視点画像生成手段を更に備えた請求項1から7のいずれか1項に記載の立体動画処理装置。
- 前記複数の視点画像及び仮想視点画像を含むN(Nは3以上の整数)視点の視点画像のうちの1つを代表画像として設定する代表画像設定手段を備え、
前記最大視差量取得手段は、前記代表画像と前記N視点の視点画像のうちの代表画像を除く(N-1)視点の視点画像とのそれぞれ異なる組み合わせからなる2つの視点画像に関連して(N-1)個の遠景側の最大視差量を取得し、
前記区間内最大ディスプレイサイズ取得手段は、前記取得した(N-1)個の遠景側の最大視差量に基づいて前記所定の区間内で最大となる(N-1)個の区間内最大ディスプレイサイズを前記所定の区間毎に取得し、
前記記録手段は、前記取得した(N-1)個の区間内最大ディスプレイサイズを前記付属情報として前記立体動画ファイルに記録する請求項1、2又は4に記載の立体動画処理装置。 - 前記立体動画ファイルは、複数のフレームからなる1GOP毎に複数の視点画像が順次記録されたMPEGファイルであり、
前記立体動画の所定の区間は、1又は2以上の所定数のGOPに対応する区間である請求項1から9のいずれか1項に記載の立体動画処理装置。 - 前記取得した立体動画のシーンチェンジを検知するシーンチェンジ検知手段を備え、
前記立体動画の所定の区間は、前記シーンチェンジ検知手段により検知されたシーンチェンジにより区分されるシーン毎の区間である請求項1から10のいずれか1項に記載の立体動画処理装置。 - 前記記録手段は、前記区間毎に取得される区間内最大ディスプレイサイズ及び区間内最大視差量のうちの少なくとも一方を、該区間よりも所定の区間数だけ前の区間の先頭、又は立体動画ファイルの先頭に記録する請求項1から11のいずれか1項に記載の立体動画処理装置。
- 前記記録手段は、前記N視点の視点画像を視点の並び順に、又は中央の視点に近い順に記録する請求項7又は9に記載の立体動画処理装置。
- 前記記録手段は、前記複数の視点画像及び仮想視点画像を前記立体動画ファイルに記録する際に、各画像が実視点の画像か仮想視点の画像かを示す視点情報を前記立体動画ファイルの付属情報として記録する請求項8又は9に記載の立体動画処理装置。
- 前記区間内最大ディスプレイサイズ取得手段は、前記区間内最大視差量取得手段により取得した遠景側の区間内最大視差量に対応する立体ディスプレイ上での画像ずれ量が、人間の両眼間隔を越えないディスプレイサイズのうちの最大のディスプレイサイズを、両眼融合可能な区間内最大ディスプレイサイズとして取得する請求項1、2、4、7、又は9に記載の立体動画処理装置。
- 前記区間内最大ディスプレイサイズ取得手段は、前記区間内最大視差量取得手段により取得した遠景側の区間内最大視差量に基づいて、予め最大視差量に対応する最大ディスプレイサイズが記録された記憶手段から対応する最大ディスプレイサイズを読み出し、又は人間の両眼間隔を示す所定値を含む計算式を使用して前記区間内最大ディスプレイサイズを算出する請求項15に記載の立体動画処理装置。
- 請求項1から16のいずれか1項に記載の立体動画処理装置を、コンピュータにより実現させる立体動画処理プログラム。
- 請求項1から16のいずれか1項に記載の立体動画処理装置を、コンピュータにより実現させる機械読み取り可能な指示の立体動画処理プログラムを有形的に格納したプログラム可能な記録媒体。
- 前記立体動画取得手段として機能する撮像手段と、
請求項1から16のいずれか1項に記載の立体動画処理装置と、
を備えた立体撮像装置。 - 複数の視点画像からなる立体画像が時間軸方向に連続している立体動画を取得するステップと、
前記取得した立体動画のフレーム毎に複数の視点画像から特徴が一致する特徴点間のずれ量を示す視差量を算出するステップと、
前記算出したフレーム毎の各特徴点の視差量のうちの遠景側の最大視差量を取得するステップと、
前記取得した遠景側の最大視差量のうちの前記立体動画の所定の区間毎に該区間内で最大となる区間内最大視差量を取得するステップと、
前記立体動画の所定の区間毎に取得した区間内最大視差量に基づいて、前記所定の区間毎に前記立体動画を立体ディスプレイに表示させる際に両眼融合可能な区間内最大ディスプレイサイズを取得するステップと、
前記立体動画が記録される立体動画ファイルを生成し、該立体動画ファイルを記録媒体に記録するステップであって、前記立体動画を前記立体動画ファイルに記録するとともに、前記所定の区間毎の区間内最大ディスプレイサイズを付属情報として前記立体動画ファイルに記録するステップと、
を含む立体動画処理方法。 - 複数の視点画像からなる立体画像が時間軸方向に連続している立体動画を取得するステップと、
前記取得した立体動画のフレーム毎に複数の視点画像から特徴が一致する特徴点間のずれ量を示す視差量を算出するステップと、
前記算出したフレーム毎の各特徴点の視差量のうちの遠景側の最大視差量を取得するステップと、
前記取得した遠景側の最大視差量のうちの前記立体動画の所定の区間毎に該区間内で最大となる区間内最大視差量を取得するステップと、
前記立体動画が記録される立体動画ファイルを生成し、該立体動画ファイルを記録媒体に記録するステップであって、前記立体動画を前記立体動画ファイルに記録するとともに、前記所定の区間毎の区間内最大視差量を付属情報として前記立体動画ファイルに記録するステップと、
を含む立体動画処理方法。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2012554636A JP5750457B2 (ja) | 2011-01-25 | 2011-12-12 | 立体動画処理装置、立体動画処理プログラム及びその記録媒体、立体撮像装置並びに立体動画処理方法 |
CN201180066035.XA CN103329549B (zh) | 2011-01-25 | 2011-12-12 | 立体视频处理器、立体成像装置和立体视频处理方法 |
US13/949,707 US9380290B2 (en) | 2011-01-25 | 2013-07-24 | Stereoscopic video processor, recording medium for stereoscopic video processing program, stereoscopic imaging device and stereoscopic video processing method |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2011-013331 | 2011-01-25 | ||
JP2011013331 | 2011-01-25 |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US13/949,707 Continuation US9380290B2 (en) | 2011-01-25 | 2013-07-24 | Stereoscopic video processor, recording medium for stereoscopic video processing program, stereoscopic imaging device and stereoscopic video processing method |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2012101916A1 true WO2012101916A1 (ja) | 2012-08-02 |
Family
ID=46580499
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2011/078647 WO2012101916A1 (ja) | 2011-01-25 | 2011-12-12 | 立体動画処理装置、立体動画処理プログラム及びその記録媒体、立体撮像装置並びに立体動画処理方法 |
Country Status (4)
Country | Link |
---|---|
US (1) | US9380290B2 (ja) |
JP (1) | JP5750457B2 (ja) |
CN (1) | CN103329549B (ja) |
WO (1) | WO2012101916A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104361627A (zh) * | 2014-11-07 | 2015-02-18 | 武汉科技大学 | 基于sift的双目视觉沥青路面微观纹理三维图像重构方法 |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP5449535B2 (ja) * | 2010-04-28 | 2014-03-19 | 富士フイルム株式会社 | 立体撮像装置及びその制御方法 |
CN103329549B (zh) | 2011-01-25 | 2016-03-09 | 富士胶片株式会社 | 立体视频处理器、立体成像装置和立体视频处理方法 |
CN106464856B (zh) * | 2014-06-24 | 2019-01-15 | 谷歌有限责任公司 | 用于分析和确定虚拟环境的特性的计算机化系统和方法 |
CN114650405B (zh) * | 2022-03-21 | 2023-01-24 | 嘉兴智瞳科技有限公司 | 一种立体视频最佳融像视差成像方法及装置 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH09121370A (ja) * | 1995-08-24 | 1997-05-06 | Matsushita Electric Ind Co Ltd | 立体tv装置 |
JP2005217874A (ja) * | 2004-01-30 | 2005-08-11 | Konica Minolta Photo Imaging Inc | ファイル管理方法 |
JP2008103820A (ja) * | 2006-10-17 | 2008-05-01 | Sharp Corp | 立体画像処理装置 |
WO2011136137A1 (ja) * | 2010-04-28 | 2011-11-03 | 富士フイルム株式会社 | 立体撮像装置及びその制御方法 |
Family Cites Families (33)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0888017A2 (en) * | 1993-08-26 | 1998-12-30 | Matsushita Electric Industrial Co., Ltd. | Stereoscopic image display apparatus and related system |
JP3157384B2 (ja) * | 1994-06-20 | 2001-04-16 | 三洋電機株式会社 | 立体映像装置 |
US6005607A (en) | 1995-06-29 | 1999-12-21 | Matsushita Electric Industrial Co., Ltd. | Stereoscopic computer graphics image generating apparatus and stereoscopic TV apparatus |
US6163337A (en) * | 1996-04-05 | 2000-12-19 | Matsushita Electric Industrial Co., Ltd. | Multi-view point image transmission method and multi-view point image display method |
EP0830034B1 (en) | 1996-09-11 | 2005-05-11 | Canon Kabushiki Kaisha | Image processing for three dimensional display of image data on the display of an image sensing apparatus |
US6864910B1 (en) | 1999-06-30 | 2005-03-08 | Canon Kabushiki Kaisha | Optical apparatus |
WO2004051577A1 (en) | 2002-11-27 | 2004-06-17 | Vision Iii Imaging, Inc. | Parallax scanning through scene object position manipulation |
JP3978392B2 (ja) | 2002-11-28 | 2007-09-19 | 誠次郎 富田 | 立体映像信号生成回路及び立体映像表示装置 |
EP1617684A4 (en) | 2003-04-17 | 2009-06-03 | Sharp Kk | THREE-DIMENSIONAL IMAGE CREATION DEVICE, THREE-DIMENSIONAL IMAGE REPRODUCING DEVICE, THREE-DIMENSIONAL IMAGE PROCESSING DEVICE, THREE-DIMENSIONAL IMAGE PROCESSING PROGRAM, AND RECORDING MEDIUM CONTAINING THE SAME |
JP4490074B2 (ja) * | 2003-04-17 | 2010-06-23 | ソニー株式会社 | 立体視画像処理装置、立体視画像表示装置、立体視画像提供方法、および立体視画像処理システム |
JP4679804B2 (ja) * | 2003-04-17 | 2011-05-11 | シャープ株式会社 | 画像ファイル作成装置 |
JP4469159B2 (ja) * | 2003-11-06 | 2010-05-26 | 学校法人早稲田大学 | 立体映像評価装置および立体映像チューナ |
JP2005167310A (ja) | 2003-11-28 | 2005-06-23 | Sharp Corp | 撮影装置 |
JP4698961B2 (ja) * | 2004-03-12 | 2011-06-08 | カシオ計算機株式会社 | 電子カメラ、及び電子カメラの制御プログラム |
JP4328311B2 (ja) | 2005-04-14 | 2009-09-09 | 株式会社東芝 | 三次元画像表示用多視点画像の作成方法およびプログラム |
JP4763571B2 (ja) * | 2006-10-24 | 2011-08-31 | シャープ株式会社 | 立体画像生成装置及び立体画像復号装置 |
JP4932660B2 (ja) | 2007-10-05 | 2012-05-16 | 富士フイルム株式会社 | 画像記録装置及び画像記録方法 |
KR101313740B1 (ko) | 2007-10-08 | 2013-10-15 | 주식회사 스테레오피아 | 원소스 멀티유즈 스테레오 카메라 및 스테레오 영상 컨텐츠제작방법 |
US8390674B2 (en) | 2007-10-10 | 2013-03-05 | Samsung Electronics Co., Ltd. | Method and apparatus for reducing fatigue resulting from viewing three-dimensional image display, and method and apparatus for generating data stream of low visual fatigue three-dimensional image |
MX2010004275A (es) * | 2007-10-19 | 2010-07-29 | Samsung Electronics Co Ltd | Metodo para grabar datos de imagenes tridimensionales. |
JP4940111B2 (ja) * | 2007-11-26 | 2012-05-30 | 富士フイルム株式会社 | 撮影装置及び画像記録方法 |
JP4657313B2 (ja) | 2008-03-05 | 2011-03-23 | 富士フイルム株式会社 | 立体画像表示装置および方法並びにプログラム |
US9013551B2 (en) * | 2008-12-01 | 2015-04-21 | Imax Corporation | Methods and systems for presenting three-dimensional motion pictures with content adaptive information |
JP5369952B2 (ja) * | 2009-07-10 | 2013-12-18 | ソニー株式会社 | 情報処理装置および情報処理方法 |
JP5299214B2 (ja) * | 2009-10-20 | 2013-09-25 | ソニー株式会社 | 画像処理装置、および画像処理方法、並びにプログラム |
CN102860019B (zh) | 2010-04-28 | 2015-07-29 | 富士胶片株式会社 | 立体图像再生装置及方法、立体摄像装置、立体显示器装置 |
JP5449537B2 (ja) | 2010-04-28 | 2014-03-19 | 富士フイルム株式会社 | 立体画像再生装置及び方法、立体撮像装置、立体ディスプレイ装置 |
EP2624571A4 (en) * | 2010-10-01 | 2014-06-04 | Samsung Electronics Co Ltd | DISPLAY DEVICE, SIGNAL PROCESSING DEVICE AND CORRESPONDING METHODS |
JP4956658B2 (ja) | 2010-10-12 | 2012-06-20 | シャープ株式会社 | 立体映像変換装置及び立体映像表示装置 |
KR20120062589A (ko) * | 2010-12-06 | 2012-06-14 | 삼성전자주식회사 | 3d 디스플레이 장치 및 그 방법 |
CN103329549B (zh) | 2011-01-25 | 2016-03-09 | 富士胶片株式会社 | 立体视频处理器、立体成像装置和立体视频处理方法 |
JP5466773B2 (ja) | 2011-02-03 | 2014-04-09 | 富士フイルム株式会社 | 立体動画再生装置、立体動画再生プログラムならびにその記録媒体、立体ディスプレイ装置、立体撮像装置及び立体動画再生方法 |
JP5486697B2 (ja) | 2011-02-03 | 2014-05-07 | 富士フイルム株式会社 | 立体動画再生装置、立体動画再生プログラムならびにその記録媒体、立体ディスプレイ装置、立体撮像装置及び立体動画再生方法 |
-
2011
- 2011-12-12 CN CN201180066035.XA patent/CN103329549B/zh not_active Expired - Fee Related
- 2011-12-12 JP JP2012554636A patent/JP5750457B2/ja not_active Expired - Fee Related
- 2011-12-12 WO PCT/JP2011/078647 patent/WO2012101916A1/ja active Application Filing
-
2013
- 2013-07-24 US US13/949,707 patent/US9380290B2/en not_active Expired - Fee Related
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH09121370A (ja) * | 1995-08-24 | 1997-05-06 | Matsushita Electric Ind Co Ltd | 立体tv装置 |
JP2005217874A (ja) * | 2004-01-30 | 2005-08-11 | Konica Minolta Photo Imaging Inc | ファイル管理方法 |
JP2008103820A (ja) * | 2006-10-17 | 2008-05-01 | Sharp Corp | 立体画像処理装置 |
WO2011136137A1 (ja) * | 2010-04-28 | 2011-11-03 | 富士フイルム株式会社 | 立体撮像装置及びその制御方法 |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104361627A (zh) * | 2014-11-07 | 2015-02-18 | 武汉科技大学 | 基于sift的双目视觉沥青路面微观纹理三维图像重构方法 |
Also Published As
Publication number | Publication date |
---|---|
US20130308917A1 (en) | 2013-11-21 |
JP5750457B2 (ja) | 2015-07-22 |
CN103329549A (zh) | 2013-09-25 |
CN103329549B (zh) | 2016-03-09 |
JPWO2012101916A1 (ja) | 2014-06-30 |
US9380290B2 (en) | 2016-06-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5449537B2 (ja) | 立体画像再生装置及び方法、立体撮像装置、立体ディスプレイ装置 | |
JP5449536B2 (ja) | 立体画像再生装置及び方法、立体撮像装置、立体ディスプレイ装置 | |
JP4692770B2 (ja) | 複眼デジタルカメラ | |
JP4897940B2 (ja) | 立体撮像装置 | |
JP5166650B2 (ja) | 立体撮像装置、画像再生装置及び編集ソフトウエア | |
JP5486697B2 (ja) | 立体動画再生装置、立体動画再生プログラムならびにその記録媒体、立体ディスプレイ装置、立体撮像装置及び立体動画再生方法 | |
JPWO2011121840A1 (ja) | 立体撮像装置 | |
JP5449535B2 (ja) | 立体撮像装置及びその制御方法 | |
JP5750457B2 (ja) | 立体動画処理装置、立体動画処理プログラム及びその記録媒体、立体撮像装置並びに立体動画処理方法 | |
JP5466773B2 (ja) | 立体動画再生装置、立体動画再生プログラムならびにその記録媒体、立体ディスプレイ装置、立体撮像装置及び立体動画再生方法 | |
JP5580486B2 (ja) | 画像出力装置、方法およびプログラム |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 201180066035.X Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 11857212 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2012554636 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 11857212 Country of ref document: EP Kind code of ref document: A1 |