JP5455876B2 - 3D image processing device, 3D display device, and 3D image processing method - Google Patents

Description

The present invention relates to a color correction technique for stereoscopic video.

  When creating a stereoscopic (3D) image, prepare two cameras or lenses to capture the image viewed from the right eye and the image viewed from the left eye, and simultaneously capture each image captured (polarization or There is a method of expressing a stereoscopic image on a display device in a parallax barrier, a lenticular system, or the like) or time sequential (shutter glasses system, etc.). By showing the left and right images with parallax to the right and left eyes, the viewer can make the left and right parallax three-dimensional with the brain so that the image can be seen in three dimensions. Even when shooting the same image due to individual differences of multiple cameras and lenses, the amount of incident light and the angle of incidence, etc., in such a stereoscopic image creation and display method, the color and brightness of each image May be different.

  In this way, when there is a difference in the color and brightness of the left and right images, and the image is shown to the left and right eyes, if the left and right images are misaligned, a visual field struggle may occur, making it difficult to fuse There is a possibility that color misalignment between the left and right gender causes eye strain.

  As a method for correcting the left and right images for stereoscopic display, Patent Literature 1 discloses a method for correcting the vertical positional deviation between the left and right images. Patent Document 2 discloses a correction method for matching the average luminance level and the dynamic range of the left-eye video signal and the right-eye video signal.

JP 2003-61116 A JP-A-2-58993

  However, the left and right image correction techniques disclosed in Patent Document 1 and Patent Document 2 do not describe a correction technique related to the color shift of the left and right images. When shooting with a plurality of cameras, color misregistration may occur due to differences in characteristics of individual cameras, differences in focus, focus shift, and white balance adjustment of the respective cameras. In that case, the viewer sees a video with a color shift between the right eye and the left eye.

  The present invention has been made in view of the above problems, and an object of the present invention is to more suitably correct the hue of each image entering the left and right eyes in a stereoscopic image.

  In order to solve the above-described problem, an embodiment of the present invention provides an input unit to which a video is input, and a histogram of a color space for each of the input video including a left-eye video and a right-eye video. And analyzing the color space to reduce the difference between the feature parameter of the left-eye video and the right-eye video feature parameter detected by the analysis unit. And a correction unit to be performed.

  In a stereoscopic image, the hue of each image entering the left and right eyes is corrected more preferably.

It is a block diagram which shows an example of a structure of the stereo image processing apparatus which concerns on Example 1 of this invention. It is a block diagram which shows an example of the left-right image color correction process part which concerns on Example 1 of this invention. It is the example which showed the left-right color correction process sequence of the left-right image color correction process part which concerns on Example 1 of this invention. It is the example which showed description of HSV color space. It is the example which showed the hue histogram of the left-right color correction process part which concerns on Example 1 of this invention. It is the example which showed the characteristic of the hue histogram of the left-right color correction process part which concerns on Example 1 of this invention. It is a block diagram which shows an example of a structure of the stereo image processing apparatus which concerns on Example 2 of this invention. It is the figure which showed the example of the interlocking | linkage of the user operation of the stereoscopic video processing apparatus concerning Example 2 of this invention, and a left-right color correction process. It is a block diagram which shows an example of a structure of the stereo image processing apparatus which concerns on Example 3 of this invention. It is the example which showed the right-and-left video image at the time of the occlusion detection of the three-dimensional video processing apparatus concerning Example 3 of this invention. It is the example which showed the operation | movement sequence at the time of the occlusion detection of the stereo image processing apparatus which concerns on Example 3 of this invention.

  Embodiments of the present invention will be described below with reference to the drawings.

  Hereinafter, Example 1 of the present invention will be described with reference to FIGS.

  FIG. 1 is a block diagram illustrating an example of the configuration of the stereoscopic video processing apparatus according to the first embodiment of the present invention.

  In FIG. 1, 100 is a video input I / F, 101 is a video decoder circuit, 102 is a stereoscopic video color correction processing unit, 103 is a stereoscopic video processing unit, 104 is a left / right color correction processing unit, 105 is a stereoscopic display processing unit, 106 Is a display part.

  The video input I / F 100 is an interface for inputting content such as video, audio, and characters from a player device such as a broadcast or an optical disc, a game machine, or the like. The video decoder circuit 101 performs video decoding processing and outputs a video signal. The stereoscopic image color correction processing unit 102 receives a stereoscopic image such as a Side-by-Side (SBS) method, a Top-and-Bottom (TAB) method, a Frame-Packing (FP) method, or a multi-view method. Then, the stereoscopic video processing unit 103 performs video signal processing for separating video corresponding to the left and right eyes. The left and right color correction processing unit 104 performs color correction (correction in the color space) so that the difference between the colors of the images corresponding to the left and right eyes becomes small. The stereoscopic correction processing unit 105 switches the left-eye video L and the right-eye video R to the display unit 106 and outputs the switching timing to the shutter glasses. Perform infrared signal processing. In the case of a polarized glasses type 3D display unit, a process of adjusting display pixels so that L and R images are displayed in accordance with the parallax barrier installed in the display unit 106 is also performed. In the case of the naked eye method, video signal processing is performed in accordance with a lenticular lens or the like installed in the display unit. The display unit 106 is a display unit that uses a liquid crystal device, an OLED (Organic light-emitting diode), plasma, or the like. The L and R images output from the stereoscopic display processing unit 105 are switched at the adjusted timing for display or display. 3D display is performed by performing display with pixels according to the installed parallax barrier or display with pixels according to the installed lenticular lens.

  Next, a mechanism for reducing the difference in color between the right-eye video and the left-eye video during stereoscopic image display will be described.

  FIG. 2 is a block diagram showing an example of a detailed configuration of the left and right video color correction processing device section according to the first embodiment of the present invention. Video for stereoscopic display such as SBS, TAB, and FP is separated into a left-eye video 201 and a right-eye video 202 by the stereoscopic video signal processing unit 103 in FIG. Input to the unit 104. For example, in the SBS system, the left half of the video decoded by the video decoder circuit 101 is separated as the left-eye video and the right half is separated as the right-eye video. Thereafter, the left and right color correction processing unit 104 performs color analysis of the left and right images.

  In the left and right color correction processing unit 104 of FIG. 2, an L video hue histogram analysis unit 203 and an R video hue histogram analysis unit 204 perform color space analysis of the left and right videos, respectively. The hue correction processing unit 205 extracts feature parameters (hereinafter simply referred to as “features”) such as the peak and hue width of each of the left and right histograms, and the center of gravity of the histogram, and obtains the difference between the features. Further, correction is performed to change the hues of the left and right images so as to reduce the difference. As a result, a left-eye image 206 and a right-eye image 207 with a small difference in hue are output.

  Here, the hue histogram used for processing in the L video hue histogram analysis unit 203, the R video hue histogram analysis unit 204, and the hue correction processing unit 205 will be described.

  In this embodiment, the L video hue histogram analysis unit 203 and the R video hue histogram analysis unit 204 calculate the histograms of the hues of the left-eye video and the right-eye video, respectively. First, the definition of the numerical value of the hue is shown in FIG. In the HSV color space, the circular circle of the cylinder represents the hue, with red being 0 degrees, changing 360 degrees from yellow, green, blue, and purple. The hue histogram shown in FIG. 5 is an example in which the hue is plotted on the horizontal axis and the number of pixels for each hue included in one screen for the left (or right) video is counted and graphed. Using the hue histogram distribution, the difference in hue between the left and right images is calculated. Details of the calculation of the histogram and the hue correction will be described later with reference to FIG.

  In this example, an example of histogram analysis in the hue of the HSV color space is shown. As the color analysis, for example, hue (Hue), saturation (Saturation / Chroma), and brightness (Brightness / Lightness / Value) are used. Separated into the used HSV / HSB color space, each histogram distribution or the like may be obtained. This is the HLS / HSI color space using Hue, Saturation, Luminance (Lightness / Luminance or Intensity), CIE color system (RGB, XYZ, xyY, L * u * v *, L * a * b *) and sRGB color triangles may be used for color analysis to calculate a histogram distribution on each axis representing various color spaces.

  In this example, the hue histogram of the histogram expressed by the hue, saturation, and lightness axes of the HSV color space is analyzed, and the difference between the color spaces is reduced by bringing the histogram features of the left and right images closer to each other. . Not only the hue but also the saturation and lightness may be analyzed by analyzing the histogram distribution to calculate the difference amount of the feature, and this difference amount may be reduced. This processing may be performed simultaneously with the histogram analysis of the hue, or may be performed on temporally different frames.

  Next, FIG. 3 describes an example of the left and right color correction processing sequence of the left and right video color correction processing unit 104 according to the first embodiment of the present invention with reference to FIGS. 3, 5, and 6.

  In S300, the left and right images are input to the left and right image color correction processing unit 104, and the process is started. In S301, the L video hue histogram analysis unit 203 counts the number of pixels of each hue included in the left-eye video. Similarly, in S303, the R video hue histogram analysis unit 204 adds up the total number of images of each hue included in the right eye video. In S <b> 302, the L video hue histogram analysis unit 203 detects the hue characteristics of the left-eye video hue histogram. Here, the characteristics of the hue are the peak hue position, the frequency of the peak hue, the distribution width around the peak hue, the position of the center of gravity of the distribution around the peak hue, or a combination thereof. FIG. 5 is an example of a hue histogram compiled by the L video hue histogram analysis unit 203 or the R video hue histogram analysis unit 204 of the left and right color correction processing unit. For example, if the image shows a blue sky, a histogram appears around the blue hue, and the peak hue, the distribution width around the peak hue, and the hue at the center of gravity around the peak are detected. Hue characteristics can be determined. Similarly, in S304, the video hue histogram analysis unit 204 detects the feature of the hue histogram of the right-eye video. In step S <b> 305, the hue correction processing unit 205 compares the hue characteristics of the left-eye image with the hue characteristics of the right-eye image. For example, how much the peak hue position and the peak frequency, the width of the peak hue peripheral distribution, the center of gravity position of the peak peripheral hue distribution, and the like are calculated, and the difference between the features is digitized. In S306, the hue correction processing unit 205 determines not to perform color correction when the difference amount of the feature of the histogram distribution calculated in S305 is equal to or greater than a predetermined threshold (S307). In addition to the determination in S307, if the difference between the features is large as a result of the color analysis of the left and right images, it may be determined that the input is not a 3D image signal. In this case, an instruction not to separate the input video into left and right videos may be given to the stereoscopic video signal processing unit 103 (S308). At this time, the stereoscopic video processing apparatus of FIG. 1 performs 2D video signal processing.

  The advantage of performing the process of S308 will be described below. First, when the feature difference amount is large in S307, the input video signal is not a 3D video image, and the video image is halved as in 3D such as SBS or TAB by setting the stereoscopic display device. There is a possibility that the left eye image and the right eye image are created separately to be displayed in 3D. In this case, since the video has broken down, it is not necessary to match the colors of the left and right images (S307), and it is not necessary to separate the left and right images. Therefore, it instructs the stereoscopic video signal processing unit 103 not to perform the left and right video separation processing (S308).

  When the stereoscopic video signal processing unit 103 receives the instruction in S308, the stereoscopic video signal processing unit 103 uses the video signal output from the video decoder circuit as it is as the left-eye video output and the right-eye video output. The video may be output, and the stereoscopic display process 105 may perform the same display as the 3D display (display the left and right images alternately / display the left and right images in a line-by-line manner) to realize 2D display.

  As another example of processing when the stereoscopic video signal processing unit 103 receives the instruction of S308, the stereoscopic video color correction processing unit 102 does nothing and the 2D output video signal of the video decoder circuit 101 is stereoscopically processed as it is. The stereoscopic display unit 105 may transmit the output video signal to the display unit 105 and display the display processing setting by switching the display processing setting from 3D display to 2D display.

  In S305, even if the difference between the hue histogram features is large, if the difference is a certain difference, it is determined that the video is input in an anaglyph method using red-blue glasses, and color correction is performed. Alternatively, the process of outputting the left and right images may be performed as is.

  In addition, when the difference amount of the histogram histogram feature is large, there is a possibility that the image is shot using two cameras that are not in focus on the left and right or white balance. In this case, the stereoscopic effect disappears in consideration of eye strain, etc., but only the left-eye or right-eye video is selected, the same left and right video is output, and the left and right visible video colors You may perform the process which eliminates deviation.

  Further, the threshold value of the difference between the left and right images of the color distribution determined as the anaglyph method and the threshold value of the difference between the left and right images of the color distribution due to the difference of the camera focus and white balance may have different threshold values. That is, by setting a plurality of threshold values, the plurality of determinations may be switched according to the magnitude of the difference.

  FIG. 6 shows an example showing the characteristics of the hue histogram of the left and right devices in the left and right color correction processing calculated in S305. The hue position of the hue peak, the frequency of the peak hue, the distribution width around the peak hue (histogram concentration range), and the like represent the characteristics of the histogram. Here, using the histogram frequency and the histogram concentration range, the histogram centroid of the portion where the histogram is concentrated can be calculated. FIG. 6 shows an example of numerical values calculated to represent the characteristics of the hue histogram. For example, first, the hue positions (PL0, PL1, PR0, PR1) indicating the hue peaks of the left-eye video and the right-eye video, and the respective peak frequencies (ΔPL0, ΔPR0, ΔPL1, ΔPR1) of the histogram are calculated. Further, ranges ΔWL0, ΔWL1, ΔWR0, ΔWR1 in which histogram frequencies equal to or higher than the threshold set in the histogram are concentrated are calculated. Using these as the features of the histogram, the difference amount between the features of the left and right images is calculated. As described above, when determining the magnitude of the feature difference amount between the hues of the left and right images in S306 of FIG. 3, | PL0−PR0 | and | PL1−PR1 | You may compare. Alternatively, | ΔPL0−ΔPR0 | or | ΔPL1−ΔPR1 |, which is a difference in peak frequency, may be compared with a predetermined threshold. Further, the difference in the histogram concentration range may be used for the determination.

  In step S309, as described above, the feature difference amount calculated in step S305 is determined as the hue correction amount. For example, the hue shift of the peak of the hue of the left-eye video and the right-eye video calculated in S305 (PL0-PR0, PL1-PR1; PL0, PR0, PL1, PR1 are hue positions of the respective peak frequencies in the histogram), the peak frequency difference (ΔPL0−ΔPR0, ΔPL1−ΔPR1; ΔPL0, ΔPR0, ΔPL1, and ΔPR1 are the respective peak frequencies of the histogram) is set as the hue correction amount. Furthermore, in S309, a range ΔWn (n: integer) in which histogram frequencies equal to or higher than the threshold set in the histogram are concentrated is calculated as a hue correction range.

  In S310, the hue correction processing for the left-eye video and the right-eye video is performed according to the correction amount determined in S309. The hue correction of the video may be performed on at least one video. For example, a hue that shifts the hue around the peak hue PR0 of the right-eye image PR0 to the right by | PL0−PR0 | and a hue that shifts the hue around the ΔΔ1 around PR1 by | PL1-PR1 | to the left. Make corrections. In this way, by limiting the correction range to ΔWL0 and ΔWL1 and correcting the pixels of the hue in the range, it is possible to increase the accuracy of the correction and reduce the adverse effects of the correction.

  Furthermore, it is possible to correct the pixels within the respective ranges of ΔWL0 and ΔWL1 and perform processing for adjusting the peak frequency and the frequency of surrounding hues.

  When the characteristics of the histogram distribution of the left and right images are approximated by such hue correction, the characteristics of the histogram distribution of the right eye image may be changed to approximate the characteristics of the histogram distribution of the left eye image. Further, the feature of the left-eye video may be changed to be close to the feature of the right-eye video. Alternatively, the histogram distribution feature of the left-eye video and the histogram distribution feature of the right-eye video may be averaged and approximated to the average histogram distribution feature obtained by calculating both the left and right videos.

  Further, this process may be performed every frame or every several frames. When it is performed every several frames, the processing load can be reduced by performing the same hue correction for several frames.

  Through the processing described above, the amount of hue shift between the left and right images can be quantified, and the hue histogram of the left and right images can be approximated according to the amount of shift.

  Although the example of correcting the hue has been described in the above-described example, the color correction of the left and right images may be performed according to the amount of difference in the color characteristics of the left and right images in another color space.

  Note that regarding which of the left-eye video and the right-eye video is to be corrected, a video that is input later to the processing unit may be the correction target. In this case, the pre-input video is stored in the line memory, the characteristics of the histogram distribution of the pre-input video are analyzed first, and the hue of the post-input video is determined according to the analysis result of the pre-input video using the analysis result. The hue correction is advantageous from the viewpoint of speeding up the processing. For example, in the case of the Frame Packing method, when the left-eye video is transmitted first, the analysis result of the histogram distribution feature of the left-eye video is used to change the hue of the right-eye video to the histogram distribution of the left-eye video. It is desirable to match the characteristics of The same applies to the case where the received video is stored in the line memory from the left-eye video in the SBS system and the TAB system.

  In FIG. 1, the video input I / F 100 to the stereoscopic display processing unit 105 may be integrated to form a stereoscopic video processing device and connected to a separate display device having the display unit 106. A stereoscopic display device may be configured by integrating F100 to the display unit 106.

  A second embodiment of the present invention will be described below with reference to FIGS. FIG. 7 is a block diagram illustrating an example of the configuration of the stereoscopic video processing apparatus according to the second embodiment of the present invention.

  The configuration shown in FIG. 7 is a configuration in which a user I / F (interface) 700 that is an interface for inputting a user operation is added to the configuration of FIG. By installing this user I / F, it is possible to perform the left and right color correction processing after detecting that the user has selected 3D viewing.

  FIG. 8 shows an example of the linkage between the user operation and the left and right color correction processing. First, the display device 106 displays a 3D video signal received from the broadcast or the network by the SBS method as in a display example 800. Further, the 3D video signal received by the TAB method is displayed as in a display example 801. Next, when a switching instruction signal from 2D to 3D is transmitted from the remote controller 802, the user I / F 700 receives the switching instruction signal. The user I / F 700 transmits the reception of the switching instruction signal to the stereoscopic video color correction processing unit 102, and the stereoscopic video color correction processing unit 102 starts the right and left color correction processing at that timing. You can see the images made. The reverse operation is the same, and right and left color correction is performed at the time of 3D display, but control to end the left and right color correction processing of the stereoscopic video color correction processing unit 102 at the timing of switching to 2D display by an instruction signal from the remote controller 802. I do.

  Accordingly, it is possible to easily determine whether or not the user is in a state of performing 3D viewing, and the determination process as in S306 of FIG. 3 in the embodiment is not necessary.

  During 2D display such as display example 8900 and display example 8901, when viewing images taken using two cameras whose focus, focus, white balance, etc. are not adjusted on the display device, left and right color misalignment, etc. You can also check if there is any. At the timing when the user switches to 3D display, right and left color correction is performed, and good display characteristics can be provided.

  Also, (1) whether the color distribution of the left-eye video is corrected to match the color distribution of the right-eye video, or (2) whether the color distribution of the right-eye video is corrected to match the color distribution of the left-eye video, Alternatively, (3) the user may be able to select via the remote controller 802 and the user I / F 700 whether to match the color distribution of the left and right images to the average of the color distribution characteristics of the left and right images.

  A human sensor may be mounted instead of the user I / F. For example, when a sensor that detects the presence of a person is installed in the viewing range by detecting a person using an infrared sensor that generates heat from the human body, detecting a person using a camera, or detecting a motion, It is also possible to perform left / right color correction only when it is determined, and not to perform left / right color correction when there is no person. At this time, in addition to the condition that there is a person, it may be added to the condition that 3D display setting is made. When the viewer is absent by such processing, the left and right color correction processing can be omitted, and the processing load can be reduced.

  Hereinafter, Example 3 of the present invention will be described with reference to FIGS.

  FIG. 9 is a block diagram illustrating an example of the configuration of the stereoscopic display control apparatus according to the third embodiment of the present invention.

  The configuration shown in FIG. 9 is a configuration in which an occlusion detection unit is added to the configuration of FIG. The occlusion detection unit detects an object that cannot be seen by one eye (viewpoint) but can be seen by the other eye (viewpoint) as shown in FIG. This occlusion area exists only in the image of one eye and does not exist in the image of the other eye. Therefore, if it is calculated as the color difference between the left and right images, it becomes an error component of color complementarity and there is a risk of adverse effects. Therefore, in this embodiment, an occlusion area is detected, and the area is excluded from the correction range, and is not used for histogram analysis so as not to cause adverse effects on other portions. The occlusion detection is realized by detecting an object in the video or detecting a left / right color difference area.

  Here, FIG. 10 is an example showing left and right video images during occlusion detection processing in the occlusion detection unit according to the third embodiment of the present invention. When the cube 1000 and the sphere 1001 having different colors are slightly shifted from each other, only the cube is visible with the left eye, and both the cube and the sphere are visible with the right eye. When this is shot using two cameras corresponding to both eyes, a left-eye video 1002 and a right-eye video 1003 are shot.

  When this is input to the stereoscopic image processing apparatus and histogram analysis is performed on the hues of the left and right images, a sphere color region that does not exist in the left eye image exists in the right eye image. It will be detected greatly. However, this is not a difference in hue but an original image pickup target is different, and it can be seen that this is not a left-right color difference due to a difference in camera focus or white balance.

  FIG. 11 shows an example of an occlusion detection operation sequence of the stereoscopic video processing apparatus according to the third embodiment of the present invention.

  The process starts from S1100, and in S1101, an occlusion area in the left and right video is detected. In step S1102, the occlusion area is excluded from histogram analysis. Next, in S1103, the occlusion area is removed from the left and right color correction areas. The operation of FIG. 11 may be performed in conjunction with the processing of the left and right color correction processing unit 104 in the occlusion detection unit 900, for example. At this time, in order to exclude the occlusion area from the histogram analysis target, S1101 in FIG. 11 needs to be performed before S301 and S302 in FIG. Further, in S1102 of FIG. 11, in conjunction with S301 and S302 of FIG. 3, the number of occlusion areas is excluded from the histogram aggregation processing of S301 and S302. In S1103 in FIG. 11, the occlusion area is excluded from the correction range in S309 in conjunction with S309 in FIG.

  Through the above processing, it is possible to perform the left and right video color corrections outside the occlusion area without being affected by the color of the occlusion area. Furthermore, there is a merit that correction is not necessary for colors in the occlusion area that are relevant only to one eye.

  In each of the embodiments described above, the same processing is possible regardless of whether the 3D display method is a liquid crystal shutter method, a polarized glasses method, or a naked eye method.

  In addition, the program that operates on the stereoscopic display control apparatus may be installed in the stereoscopic display control apparatus, may be recorded on a recording medium and provided, or may be downloaded and provided via a network. You may make it do.

  In each of the embodiments described above, an example of hue analysis and an example of correction have been described. However, application of these analyzes and corrections is not limited to hue. For other color spaces such as saturation, lightness, and luminance, color correction of the left and right images may be performed according to the amount of difference in color characteristics between the left and right images. Alternatively, analysis and correction for a plurality of color spaces among a plurality of color spaces such as hue, saturation, brightness, and luminance may be performed together.

100 video input I / F unit 101 video decoder circuit unit 102 stereoscopic video color correction processing unit 103 stereoscopic video signal processing unit 104 left and right color correction processing unit 105 stereoscopic display processing unit 106 display device unit 201 left-eye video (input)
202 Right-eye video (input)
203 Left-eye image hue histogram analysis 204 Right-eye image hue histogram analysis 205 Color correction process 206 Left-eye image (output)
207 Right-eye video (output)
400 HSV color space 500 Hue histogram 600 Hue histogram of video for left eye
601 Hue histogram of right eye video
700 User I / F section 800 SBS display image (2D display)
801 TAB display image (2D display)
802 Remote control 803 3D display image 900 Occlusion detection unit 1000 Cube (subject)
1001 Sphere (subject)
1002 Left eye image 1003 Right eye image

Claims (14)

An input unit for inputting video;
An analysis unit that generates a color space histogram for each of the input video when the left-eye video and the right-eye video are included, and detects a feature parameter for each video;
A correction unit that performs correction on a color space that reduces a difference between the feature parameter of the left-eye video and the right-eye video feature parameter detected by the analysis unit ;
Wherein the correction processing by the correcting unit, a stereoscopic image difference between the feature parameters of the characteristic parameter and the right-eye video of the left-eye image detected by said analyzing unit characterized in that it is carried out in the case of less than a predetermined threshold value Processing equipment. The stereoscopic image processing apparatus according to claim 1 ,
The correction unit does not perform the correction when the difference between the feature parameter of the left-eye image detected by the analysis unit and the feature parameter of the right-eye image is larger than a predetermined threshold value. Processing equipment. The stereoscopic image processing apparatus according to claim 2 ,
A stereoscopic video processing unit that selectively performs signal processing for 3D video and signal processing for 2D video on the input video;
When the difference between the feature parameter of the left-eye video and the feature parameter of the right-eye detected by the analysis unit is larger than the predetermined threshold, the signal processing of the stereoscopic video processing unit is changed to signal processing for 2D video. A stereoscopic image processing apparatus characterized by switching. The stereoscopic image processing apparatus according to claim 1,
With user interface,
The stereoscopic image processing apparatus, wherein the correction unit starts correction on the color space when an instruction to switch from 2D display to 3D display is input to the user interface. The stereoscopic video processing apparatus according to claim 4 ,
The stereoscopic image processing apparatus, wherein when the switching instruction from 3D display to 2D display is input to the user interface, the correction unit ends the correction on the color space. The stereoscopic image processing apparatus according to claim 1,
Equipped with human sensor,
When the human sensor determines that there is a person, the correction unit performs correction on the color space,
The stereoscopic image processing apparatus, wherein when the human sensor determines that there is no person, the correction unit does not perform correction on the color space. The stereoscopic image processing apparatus according to claim 1,
With an occlusion detector,
The occlusion detection unit detects an occlusion region in which an object that is not visible at one viewpoint and visible at the other viewpoint exists in the left-eye video and the right-eye video included in the input video,
The analysis unit excludes the occlusion area in the histogram generation process of the color space,
The stereoscopic image processing apparatus, wherein the correction unit excludes the occlusion area from correction targets in a color space. A stereoscopic video processing apparatus according to any one of claims 1 to 7,
The feature parameter detected by the analysis unit is a peak position of a histogram of the color space of the image, a frequency at the peak position, a distribution width around the peak position, or a center of gravity of the distribution around the peak position. apparatus. The stereoscopic image processing apparatus according to claim 8 ,
The correction on the color space in the correction unit is correction that brings the peak position of the left-eye image and the right-eye image, the frequency at the peak position, the width of the distribution around the peak position, or the center of gravity of the distribution around the peak position. A 3D image processing apparatus characterized by that. Stereoscopic image display device, characterized in that it comprises a further display section on the three-dimensional image processing apparatus according to claim 1 to 9. An input step for video input;
An analysis step of generating a color space histogram for each of the input video when the left-eye video and the right-eye video are included, and detecting feature parameters for each video;
A correction step for performing correction on a color space to reduce the difference between the feature parameter of the left-eye video feature detected in the analysis step and the right-eye video feature parameter;
Whether or not to perform the correction process in the correction step is switched according to a comparison result between a difference between the feature parameter of the left-eye video and the feature parameter of the right-eye video detected in the analysis step and a predetermined threshold value. 3D image processing method. The stereoscopic image processing method according to claim 11 ,
In conjunction with the 2D display and 3D display switching instruction from the user, a stereoscopic image processing method characterized by you switch whether to perform the correction process by the correction step. The stereoscopic image processing method according to claim 11 ,
Depending on the detection result by the motion sensor, a stereoscopic image processing method characterized by you switch whether to perform the correction process by the correction step. The stereoscopic image processing method according to claim 11 ,
An occlusion detection step of detecting an occlusion area for the left-eye video and the right-eye video,
The stereoscopic image processing method, wherein the occlusion region detected in the occlusion detection step is excluded from generation of a histogram in the analysis step and excluded from a correction target in the correction step.

Images