JP2003319417A - Stereoscopic image generating device, stereoscopic image display device, and their methods - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a stereoscopic image generating device that can generate high-accuracy data for stereoscopic images, and to provide a stereoscopic image display device that displays the data for stereoscopic images. <P>SOLUTION: The image generating sections 102 and 103 for left and right eyes of the stereoscopic image generating device generate images of a prescribed format by sampling picture elements, by thinning out images photographed by means of cameras 101a and 101b in the horizontal direction, and calculating the color information on the picture elements by changing the calculating method of the color information in accordance with the positions of the sampled picture elements. Therefore, the image generating device can generate the high- accuracy data for stereoscopic images. In addition, since the multiplexing section 105 of the image generating device multiplexes decoded images for left and right eyes with the information on a color information calculating method, the user of this device can easily judge which calculating method can generate color information resulting in the most legible stereoscopic image, by displaying stereoscopic images in a data displaying section 112 for stereoscopic image. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for creating and displaying a left-eye image and a right-eye image for a stereoscopic image, and particularly, by performing format conversion of the left-eye image and the right-eye image. The present invention relates to a stereoscopic image creating apparatus that performs encoding, a stereoscopic image display apparatus that displays stereoscopic images after decoding encoded data and performing format conversion, and methods thereof.

[0002]

2. Description of the Related Art In recent years, techniques for displaying stereoscopic images have been actively studied. One of them is a stereoscopic image device that creates and displays a stereoscopic image using the parallax between the left and right eyes of the user.

FIG. 6 is a diagram for explaining a method of creating image data in a conventional stereoscopic image device. This stereoscopic image device is a stereoscopic image creating device for creating a stereoscopic image,
It is divided into a stereoscopic image display device that displays a stereoscopic image created by the stereoscopic image creation device.

The stereoscopic image creating apparatus takes images for the left eye and the right eye respectively with two cameras when creating image data. Then, as shown in FIG. 6A, pixels are thinned out for each pixel to reduce the horizontal size of the left-eye image and the right-eye image to ½. Next, the left-eye image and the right-eye image that are half the size in the horizontal direction are encoded, the encoded data are multiplexed, and the multiplexed data is transmitted or recorded in a recording medium.

The stereoscopic image display device receives or reads the multiplexed data created by the stereoscopic image creation device from the recording medium, and reproduces the left-eye image and the right-eye image, respectively.
Then, as shown in FIG. 6B, the left-eye image (L) and the right-eye image (R) are alternately arranged in the horizontal direction one pixel at a time, or as shown in FIG. An image for stereoscopic display is created by alternately arranging each line in the vertical direction.
The stereoscopic image display device uses a lenticular method, a parallax barrier method, a shutter (time division) method, or the like, and only the left-eye image can be seen by the user's left eye and the right-eye image by the user's right eye. By making only the image visible, the user can see the image as a stereoscopic image.

As an example of such a stereoscopic image device,
An example is the stereoscopic video transmission device disclosed in Japanese Patent Laid-Open No. 11-18111. This Japanese Patent Laid-Open No. 11-181
In the stereoscopic video transmission device disclosed in Japanese Patent No. 11,
The compression circuit thins out the image for the left eye and the image for the right eye pixel by pixel, and compresses the image in the horizontal scanning line direction.

For example, as shown in FIG. 6 (d), only the even-numbered RGB pixel values from the left of the photographed image are extracted to create a left-eye image and a right-eye image. In addition, FIG.
As shown in (e), RG of the image for the left eye and the image for the right eye
Among B, the G value of the pixel is shifted by 1 pixel with respect to the RB value and thinned out to create the image for the left eye and the image for the right eye. Here, the numbers added after R, G, and B shown in FIGS. 6D and 6E indicate the coordinates of R, G, and B on one line.

[0008]

However, when transmitting an image, particularly a moving image, or recording it on a recording medium, it is common to perform an encoding process because the data amount of the image is very large. MPE which is a general coding algorithm
G (Moving Picture Experts Group) -1, MPEG-
2, DV-4 compression used in MPEG-4 and digital video cameras are all YU as input image data.
V data is used.

Further, the image taken by the camera is not always RGB data, but is often YUV data. Furthermore, the format of YUV data is 4:
Since there is not one type such as 4: 4, 4: 2: 2, 4: 2: 0, 4: 1: 1, etc., the format is different even if the input image is YUV data, and the conventional stereoscopic image display device is used. However, there is a problem that an appropriate format conversion may be necessary to display a stereoscopic image correctly.

The present invention has been made to solve the above problems, and a first object thereof is to provide a stereoscopic image creating apparatus and method capable of creating highly accurate stereoscopic image data. It is to be.

A second object is to provide a stereoscopic image display apparatus and method capable of interpolating and displaying stereoscopic image data so that color information becomes optimum.

[0012]

According to one aspect of the present invention, there is provided a stereoscopic image creating apparatus for creating stereoscopic image data from an input image in which luminance information and color information are separated, and a left eye image. Image creating means for creating a left eye image, a right eye image creating means for creating a right eye image, a left eye image and a right eye image created by the left eye image creating means Encoding means for encoding the image for the right eye created by means, including a multiplexing means for multiplexing the image for the left eye and the image for the right eye encoded by the encoding means, The image forming means for the left eye and the image forming means for the right eye sample pixels while thinning out the input image in the horizontal direction, change the calculation method of color information according to the positions of the sampled pixels, and change the color of the pixel. Calculate the information and To create the image of Tsu door.

Since the color information calculation method is changed by changing the color information calculation method according to the position of the sampled pixel, it is possible to create highly accurate stereoscopic image data.

Preferably, the multiplexing means multiplexes the left-eye image and the right-eye image encoded by the encoding means and the information regarding the predetermined format.

Therefore, by displaying the stereoscopic image on the stereoscopic image display device, the user can easily determine in which format the stereoscopic image created is likely to be stereoscopic.

Preferably, the multiplexing means multiplexes the left-eye image and the right-eye image coded by the coding means and the information regarding the color information calculation method.

Therefore, by displaying the stereoscopic image on the stereoscopic image display device, the user can easily determine by which calculation method the stereoscopic image for which the color information is created is likely to be stereoscopic.

Preferably, the multiplexing means multiplexes the left-eye image and the right-eye image coded by the coding means and the information on the interpolation method of the color information of the left-eye image and the right-eye image. Turn into.

Therefore, the stereoscopic image display device extracts information about the interpolation method of the color information to interpolate the stereoscopic image data so that the color information becomes optimum, and display an image more visible as a stereoscopic image. Is possible.

Preferably, the multiplexing means multiplexes the image for the left eye and the image for the right eye encoded by the encoding means and the information on the positions of the sampled pixels.

Therefore, by extracting the information on the pixels sampled by the stereoscopic image display device, the stereoscopic image data is interpolated and displayed so that the color information is optimized, and which left-eye image and right-eye image is displayed. It is possible to improve the convenience when the user decides whether or not the combination with the use image easily looks three-dimensional.

According to another aspect of the present invention, there is provided a stereoscopic image display device for combining and displaying an image for the left eye and an image for the right eye in which luminance information and color information are separated, and which is multiplexed. Demultiplexing for demultiplexing data to extract the encoded left-eye image, the encoded right-eye image, and the interpolation method of the color information of the left-eye image and the right-eye image Demultiplexing means, decoding means for decoding the encoded left-eye image and the encoded right-eye image extracted by the demultiplexing means, and the left-eye image extracted by the demultiplexing means And a format conversion means for interpolating the color information of the left-eye image and the right-eye image decoded by the decoding means to convert the color information into a predetermined format image And converted by the format conversion means Comprising synthesizing means for synthesizing the eye image and the right eye image, in accordance with the synthesized image by synthesizing means and display means for displaying a stereoscopic image.

The format converting means interpolates the color information of the left-eye image and the right-eye image and converts it into an image of a predetermined format with reference to the color information interpolation method, so that the color information is optimized. It is possible to interpolate and display the stereoscopic image data.

According to still another aspect of the present invention, the stereoscopic image display device is a stereoscopic image display device for synthesizing and displaying a left-eye image and a right-eye image in which luminance information and color information are separated. Then, by demultiplexing the multiplexed data, the encoded image for the left eye, the encoded image for the right eye,
Demultiplexing means for extracting information regarding the positions of sampled pixels of the left-eye image and the right-eye image,
Decoding means for decoding the encoded left-eye image and the encoded right-eye image extracted by the demultiplexing means, and the left-eye image and the right-eye image extracted by the demultiplexing means With reference to the information on the position of the sampled pixel of the image, a format conversion unit for horizontally expanding the image for the left eye and the image for the right eye decoded by the decoding unit and converting the image into a predetermined format image,
It includes a synthesizing unit for synthesizing the left-eye image and the right-eye image converted by the format converting unit, and a display unit for displaying a stereoscopic image according to the image combined by the synthesizing unit.

The format conversion means refers to the information on the positions of the sampled pixels, horizontally expands the left-eye image and the right-eye image, converts them into images of a predetermined format, and arranges pixel data. It is possible to prevent a phase shift between the stereoscopic image creating device and the stereoscopic image display device.

According to still another aspect of the present invention, there is provided a stereoscopic image creating method for creating stereoscopic image data from an input image in which luminance information and color information are separated, the step of creating a left eye image. A step of creating a right-eye image, a step of encoding the created left-eye image and right-eye image, and a step of multiplexing the encoded left-eye image and right-eye image Including the step of creating the image for the left eye and the step of creating the image for the right eye, sampling pixels while thinning out the input image in the horizontal direction, and calculating the color information according to the positions of the sampled pixels. Changing and calculating the color information of the pixel to create an image in a predetermined format.

Since the color information calculation method is changed by changing the color information calculation method in accordance with the sampled pixel position, it is possible to create highly accurate stereoscopic image data.

According to still another aspect of the present invention, there is provided a stereoscopic image display method for synthesizing and displaying a left-eye image and a right-eye image in which luminance information and color information are separated, which are multiplexed. Demultiplexing the data, the encoded left-eye image, the encoded right-eye image, a step of extracting the interpolation method of the color information of the left-eye image and the right-eye image, Decoding with reference to the step of decoding the extracted encoded left-eye image and the encoded right-eye image, and the interpolation method of the color information of the extracted left-eye image and right-eye image, A step of interpolating the color information of the left-eye image and the right-eye image that have been converted into an image of a predetermined format, and a step of combining the left-eye image and the right-eye image converted into the predetermined format, Step to display a stereoscopic image according to the captured image. Including the door.

Since the color information of the left-eye image and the right-eye image is interpolated and converted into an image of a predetermined format by referring to the color information interpolation method, the stereoscopic image data is optimized so that the color information is optimized. Can be interpolated and displayed.

According to still another aspect of the present invention, there is provided a stereoscopic image display method for combining and displaying a left-eye image and a right-eye image in which luminance information and color information are separated, which are multiplexed. Demultiplexing the data and extracting the encoded left-eye image, the encoded right-eye image, and information regarding the positions of the sampled pixels of the left-eye image and the right-eye image And a step of decoding the extracted encoded left-eye image and the encoded right-eye image, and information regarding the positions of the sampled pixels of the extracted left-eye image and right-eye image Then, the step of horizontally enlarging the decoded left-eye image and right-eye image to convert into a predetermined format image, and combining the left-eye image and right-eye image converted into the predetermined format Steps, And displaying a stereoscopic image according to made the image.

By referring to the information on the positions of the sampled pixels, the left-eye image and the right-eye image are enlarged in the horizontal direction, converted into images of a predetermined format, and the pixel data is arranged. It is possible to prevent a phase shift that occurs in the stereoscopic image display device.

[0032]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS (First Embodiment) FIG. 1 is a block diagram showing a schematic configuration of a stereoscopic image device according to a first embodiment of the present invention. The stereoscopic image device includes a stereoscopic image creating device (hereinafter, referred to as a stereoscopic image data creating unit) 111 and a stereoscopic image display device (hereinafter, referred to as a stereoscopic image data display unit) 112.

The stereoscopic image data creation unit 111 includes the cameras 101a and 101b and the left-eye image creation unit 102.
And a right-eye image creating unit 103, encoders 104a and 104b, and a multiplexing unit 105. Camera 10
1a captures an image for the left eye. Also, the camera 101
b captures an image for the right eye.

The image forming unit 102 for the left eye is connected to the camera 101.
The image data for the left eye captured by a is converted into a format that can be encoded by the encoder 104a. Similarly, the right-eye image creating unit 103 uses the camera 1
The image data for the right eye captured by 01b is converted into a format that can be encoded by the encoder 104b.

The encoder 104a uses MPEG-1, M
An image captured by the camera 101a is encoded using an algorithm such as PEG-2, MPEG-4, or DV compression method. The encoder 104b uses the same algorithm as the encoder 104a, and uses the camera 1
The image taken by 01b is encoded.

The multiplexing unit 105 multiplexes the coded data of the left-eye image encoded by the encoder 104a, the coded data of the right-eye image encoded by the encoder 104b, and various control information. The data multiplexed by the multiplexing unit 105 is transmitted to the stereoscopic image data display unit 112 or recorded on a recording medium.

Further, the stereoscopic image data display section 112 is
Demultiplexer 106 and decoders 107a and 107b
And format conversion units 108a and 108b, a combining unit 109, and a display unit 110.

The demultiplexing unit 106 demultiplexes the multiplexed data received from the stereoscopic image data creating unit 111 and the multiplexed data read from the recording medium, and encodes the left-eye image data and the right-eye image data. The encoded data of the image and various control information are extracted. The encoded data of the extracted image for the left eye is transferred to the decoder 107a, the encoded data of the image for the right eye is transferred to the decoder 107b, and the control information is transferred to the format conversion units 108a and 108b.

The decoder 107a is the encoder 104a.
It supports MPEG-1, MPEG-2, MPE
The encoded data of the image for the left eye received from the demultiplexing unit 106 is decoded by using an algorithm such as G-4 or DV compression method. The decoder 107b corresponds to the encoder 104b, and uses the same algorithm as the decoder 107a to decode the encoded data of the right-eye image received from the demultiplexing unit 106.

The format conversion unit 108a refers to the control information received from the demultiplexing unit 106 and refers to the decoder 10
The image for the left eye decoded by 7a is converted into a format that can be displayed by the display unit 110. Similarly, the format conversion unit 108b refers to the control information received from the demultiplexing unit 106, and converts the right-eye image decoded by the decoder 107b into a format that can be displayed by the display unit 110.

The synthesizing section 109 includes a format converting section 10
The image for the left eye whose format has been converted by 8a and the image for the right eye whose format has been converted by the format conversion unit 108b are combined to create stereoscopic image data.

The display unit 110 is an LCD (Liquid Crystal).
Display) module, etc., lenticular system, parallax barrier system, shutter (time division)
A stereoscopic image is displayed using a method or the like.

2 and 3 are views showing examples of various image formats. The numerical value above each pixel represents the coordinate from the left end of the pixel, and in this specification, the position of the pixel is represented by the coordinate from the left end of the image. Also, unless otherwise specified, the U component and the V component are collectively expressed as C (chroma component). In FIG. 2, only one line is shown because the relationship between Y and C is the same for all lines.

FIG. 2A is a diagram showing the YUV 4: 2: 2 format, in which C is ½ of Y in the horizontal direction.
It has been reduced to. FIG. 2D shows YUV4: 1 :.
1 is a diagram showing one format, in which C is reduced to 1/4 in the horizontal direction with respect to Y. FIG. Figure 3 (a) shows YUV
FIG. 3 is a diagram showing a 4: 2: 0 (for MPEG-2) format, in which C is ½ in the horizontal direction and 1 in the vertical direction with respect to Y.
It was reduced to / 2.

As shown in FIGS. 2A and 2D, the YUV 4: 2: 2 format and the YUV 4: 1: 1 format are used.
In the format, Y and C are sampled at the same position on each line. However, as shown in FIG. 3A, in the YUV 4: 2: 0 (for MPEG-2) format, in order to thin out C in the vertical direction,
The vertical sampling positions of Y and C are different. Here, the numerical value at the left end of FIG. 3A is a numerical value indicating the coordinates of the pixel from the upper end of the image.

As can be seen by comparing FIG. 2A and FIG. 3A, the relationship between Y and C in the horizontal direction is YUV.
4: 2: 2 format and YUV 4: 2: 0 (MPEG
(For -2) format is the same. Since only horizontal processing is involved in the present invention, in the following description, YUV4: 2: 0 is included in the YUV4: 2: 2 format.
We will include the format (for MPEG-2).

In the present embodiment, the camera 101a
And 101b are for outputting data in YUV 4: 2: 2 format. Left eye image creation unit 1
02 and the right-eye image creating unit 103, when the image data shown in FIG. 2A is input, thin the image data by 1/2 in the horizontal direction. When the image data after thinning is also in the YUV 4: 2: 2 format, the data is as shown in FIG. 2B or 2C. Also, the image data after thinning out is YUV.
In the case of the 4: 1: 1 format, the format is as shown in FIG. 2 (e) or FIG. 2 (f).

Here, Y of the pixel at the coordinate i before the thinning process is represented as Y (i), and C of the pixel at the coordinate i is represented as C (i). In FIG. 2B, Y (0), Y
(2), Y (4), Y (6), Y (8) ... Are present, as well as C (0), C (4), C (8). Further, in FIG. 2C, Y (1), Y (3), Y
(5), Y (7), Y (9) ... Are present and C
(1), C (5), C (9), ... Are present.

The difference between FIG. 2B and FIG. 2C is the difference between sampling even-numbered pixels and odd-numbered pixels. In the combination of the image for the left eye and the image for the right eye, FIG.
2C may be used, and the image shown in FIG. 2C may be used as the image for the right eye. In addition, the image shown in FIG. 2C is used as the image for the left eye, and the image for the right eye is shown in FIG.
The image shown in (b) may be used.

Further, both the left-eye image and the right-eye image may use the image shown in FIG. 2 (b), or both may use the image shown in FIG. 2 (c). The combination of the left-eye image and the right-eye image can be freely selected. The same applies to the second and subsequent embodiments of the present invention.

Further, as an image after thinning, FIG.
When using the image shown in (b), the data at the same position before sampling may be used as it is for both Y and C.
When the image shown in FIG. 2C is used as the image after thinning, the data at the same position before sampling may be used as Y as it is. However, as an image after thinning out,
When the image shown in FIG. 2C is used, the position of C differs before and after sampling, and there is no data at the same position before sampling. Therefore, C of the image shown in FIG. 2C is C (0), C before sampling.
From (2), C (4), C (6), C (8) ... to C
(1), C (5), C (9) ... need to be calculated.

For example, C in the image shown in FIG.
2a and C of the image shown in FIG. 2C is represented by Cc,
The relationship between Ca and Cc is one of the following expressions (1) to (3).

Cc (i * 4 + 1) = Ca (i * 4) (1) Cc (i * 4 + 1) = Ca (i * 4 + 2) (2) Cc (i * 4 + 1) = (Ca (i * 4) ) + Ca (i * 4 + 2)) / 2 (3) Here, Ca (i) is C of the pixel at the coordinate i of the image shown in FIG. 2A, and Cc (i) is shown in FIG. 2C. Image coordinate i
Represents the C of the pixel. As in Expression (1), Expression (2), or Expression (3), calculation is performed using pixel values in the vicinity of the pixel corresponding to the position C of the image shown in FIG. 2C before sampling. Become. That is, the formula (1) uses the pixel value of the C on the left side, the formula (2) uses the pixel value of the C on the right side, and the formula (3) uses the average pixel value of the C on both sides. Shows.

Similarly, when the left-eye image creating unit 102 and the right-eye image creating unit 103 create images in the YUV 4: 1: 1 format, selection of a combination of the left-eye image and the right-eye image Is free. When the image shown in FIG. 2E is used as the image after thinning, Y and C are used.
In both cases, the data at the same position before sampling may be used as it is. When the image shown in FIG. 2 (f) is used as the image after thinning, the data at the same position before sampling may be used as Y as it is. However, when the image shown in FIG. 2F is used as the image after thinning, the position of C of the pixel is different before and after sampling, and therefore, data at the same position before sampling does not exist. Therefore, for C of the image shown in FIG. 2F, C (0), C (2), C (4), C (6), C before sampling
It is necessary to calculate C (1), C (9) ... From (8).

For example, C in the image shown in FIG.
If a is expressed as a and C of the image shown in FIG. 2 (f) is expressed as Cf,
The relationship between Ca and Cf is one of the expressions (4) to (6) shown below.

Cf (i * 8 + 1) = Ca (i * 8) (4) Cf (i * 8 + 1) = Ca (i * 8 + 2) (5) Cf (i * 8 + 1) = (Ca (i * 8) ) + Ca (i * 8 + 2)) / 2 (6) Here, Ca (i) is C of the pixel at the coordinate i of the image shown in FIG. 2 (a), and Cf (i) is shown in FIG. 2 (f). Image coordinate i
Represents the C of the pixel. As shown in Expression (4), Expression (5), or Expression (6), the pixel value in the vicinity of the pixel corresponding to the position of C in FIG. That is, the equation (4) uses the pixel value of the C on the left side, the equation (5) uses the pixel value of the C on the right side, and the equation (6) uses the average pixel value of the C on both sides. Shows.

In order to convert the images taken by the two cameras 101a and 101b into a left-eye image and a right-eye image, respectively, the images (Y) shown in FIGS. 2B and 2C are displayed.
In the case of the UV4: 1: 1 format, the user selects which of the images shown in FIG. 2 (e) and FIG. 2 (f) is to be selected according to the situation. Then, a flag indicating which is selected is output to multiplexing section 105. The multiplexing unit 105 multiplexes this flag into the coded data of the left-eye image and the coded data of the right-eye image as control information.

When the optical axes are taken in parallel, the image (YUV4: 1: 1) shown in FIG. 2B is used as the image for the left eye.
2 (e) in the case of the format) and the image shown in FIG. 2 (c) as the image for the right eye (the image shown in FIG. 2 (f) in the case of YUV format 4: 1: 1). ) Is selected, a highly accurate image is obtained.

Further, both the image for the left eye and the image for the right eye,
If the image shown in FIG. 2B (the image shown in FIG. 2E in the case of YUV4: 1: 1 format) is selected, the process is simple. In this way, the user may select the image suitable for the situation each time.

As described above, in the stereoscopic image device according to the present embodiment, the cameras 101a and 10a are provided.
When the image captured by 1b is in the YUV 4: 2: 2 format, thinning processing is performed when the left-eye image creating unit 102 and the right-eye image creating unit 103 create the left-eye image and the right-eye image. Since the color information is calculated according to the position of the color information of the image that is created in-line, it is possible to create a highly accurate stereoscopic image.

Further, the control information including the information indicating which YUV format the image corresponds to and the information indicating at which position the pixel has been encoded are the encoded data of the image for the left eye and the image for the right eye. Therefore, when the stereoscopic image data display unit 112 displays a stereoscopic image, it is convenient for the user to determine which combination of images is more likely to be stereoscopically viewed. It has become possible to improve. In addition, by feeding back to the stereoscopic image data creation unit 111 information indicating which combination is likely to look stereoscopic, the left-eye image and the right-eye image in the left-eye image creation unit 102 and the right-eye image creation unit 103 are displayed. It becomes possible to easily select an image.

(Second Embodiment) The schematic configuration of the stereoscopic image device according to the second embodiment of the present invention is the same as the schematic configuration of the stereoscopic image device according to the first embodiment shown in FIG. . Therefore, detailed description of overlapping configurations and functions will not be repeated.

In the present embodiment, the camera 101a
And 101b are for outputting image data in YUV 4: 1: 1 format. The left-eye image creating unit 102 and the right-eye image creating unit 103 are shown in FIG.
When the image data shown in (1) is input, 1/2 of the image data is thinned out in the horizontal direction. When the image data after thinning is also in the YUV 4: 1: 1 format, the data is as shown in FIG. 2 (e) or FIG. 2 (f).

When the image shown in FIG. 2 (e) is used as the image after thinning, the data at the same position before sampling may be used as it is for both Y and C. When using the image shown in FIG. 2 (f) as the image after thinning,
For Y, the data at the same position before sampling may be used as it is. However, the image after thinning out is shown in FIG.
In the case of using the image shown in, the position of C differs before and after sampling, and there is no data at the same position before sampling. Therefore, C of the image shown in FIG. 2F is C (0), C (4), C (8) before sampling.
It is necessary to calculate C (1), C (9), ...

For example, C in the image shown in FIG.
2d, and C of the image shown in FIG. 2F is represented by Cf,
The relationship between Cd and Cf is one of the following expressions (7) to (9).

Cf (i * 8 + 1) = Cd (i * 8) (7) Cf (i * 8 + 1) = Cd (i * 8 + 4) (8) Cf (i * 8 + 1) = Cd (i * 8) * W1 + Cd (i * 8 + 4) * W2 (9) Here, Cd (i) is C of the pixel at the coordinate i of the image shown in FIG. 2 (d), and Cf (i) is the image shown in FIG. 2 (f). Coordinate i
Represents the C of the pixel. Further, W1 and W2 are weighting factors, and W1 + W2 = 1. As in Expression (7), Expression (8), or Expression (9), calculation is performed using pixel values in the vicinity of the pixel corresponding to the position C of the image shown in FIG. Become. That is, the formula (7) is for the case of using the pixel value of the C on the left side, the formula (8) is for using the pixel value of the C of the right side, and the formula (9) is for using the average pixel value of the C on both sides. is there.

In the case of the YUV 4: 1: 1 format, the amount of information of C is originally small, and the amount of information is further reduced by thinning out to 1/2 in the horizontal direction. Therefore, the image after thinning is the image shown in FIG. 2 (e). Even in the case of C, the image quality is better when the weighted sum of neighboring pixels before sampling is calculated for C.

In order to convert the images photographed by the two cameras 101a and 101b into a left-eye image and a right-eye image, respectively, one of the images shown in FIGS. 2 (e) and 2 (f) is used. The user makes a selection depending on the situation. Then, a flag indicating which is selected is output to multiplexing section 105. The multiplexing unit 105 multiplexes this flag into the coded data of the left-eye image and the coded data of the right-eye image as control information.

When the optical axes are taken in parallel, if the image shown in FIG. 2 (e) is selected as the image for the left eye and the image shown in FIG. 2 (f) is selected as the image for the right eye, a highly accurate image is obtained. Is obtained. Further, for both the left-eye image and the right-eye image, selecting the image shown in FIG. 2 (e) simplifies the processing.
In this way, the user may select the image suitable for the situation each time.

As described above, in the stereoscopic image device according to this embodiment, the cameras 101a and 10a are provided.
When the image captured by 1b is in the YUV 4: 1: 1 format, the thinning process is performed when the left-eye image creating unit 102 and the right-eye image creating unit 103 create the left-eye image and the right-eye image. Since the color information is calculated according to the position of the color information of the image that is created in-line, it is possible to create a highly accurate stereoscopic image.

Further, the control information including the information indicating which YUV format the image corresponds to and the information indicating at which position the pixel has been encoded are the encoded data of the image for the left eye and the image for the right eye. Therefore, when the stereoscopic image data display unit 112 displays a stereoscopic image, it is convenient for the user to determine which combination of images is more likely to be stereoscopically viewed. It has become possible to improve. In addition, by feeding back to the stereoscopic image data creation unit 111 information indicating which combination is likely to look stereoscopic, the left-eye image and the right-eye image in the left-eye image creation unit 102 and the right-eye image creation unit 103 are displayed. It becomes easy to select images, etc.
It is possible to obtain the same effect as the effect described in the first embodiment.

(Third Embodiment) The schematic configuration of the stereoscopic image device according to the third embodiment of the present invention is the same as the schematic configuration of the stereoscopic image device according to the first embodiment shown in FIG. . Therefore, detailed description of overlapping configurations and functions will not be repeated.

In this embodiment, the cameras 101a and 101b are YUV 4: 2: 0 (for MPEG-1).
This is for outputting format data. Left eye image creation unit 102 and right eye image creation unit 103
When the image data shown in FIG. 3B is input, the image data is thinned out to 1/2 in the horizontal direction. The image data after thinning is also YUV 4: 2:
In case of 0 format, the format shown in FIG.
It becomes what is shown in (d).

As an image after thinning, FIG.
When using the image shown in either (c) or FIG. 3 (d), the data at the same position before sampling may be used as Y as it is. However, the position of C differs before and after sampling, and there is no data at the same position before sampling.

Therefore, C of the image shown in FIG.
Is C (0_1), C (2_3), C before sampling
From (4_5), C (6_7), C (8_9) ... to C
(1), C (5), C (9) ... need to be calculated.
Here, C (i_ (i + 1)) is the coordinate i and the coordinate (i +
It means the intermediate position of 1).

For example, C in the image shown in FIG.
When represented by b and C of the image shown in FIG. 3C is represented by Cc,
The relationship between Cb and Cc is one of the following expressions (10) to (12).

Cc (i * 4 + 1) = Cb ((i * 4) _ (i * 4 + 1)) ... (10) Cc (i * 4 + 1) = Cb ((i * 4 + 2) _ (i * 4 + 3)) ... (11) Cc (i * 4 + 1) = Cb ((i * 4) _ (i * 4 + 1)) * W1 + Cb ((i * 4 + 2) _ (i * 4 + 3)) * W2 (12) where Cc ( i) represents C at the coordinate i in FIG. W1 and W2 are weighting factors, and W1 +
W2 = 1.

C of the image shown in FIG. 3D is C (0_1), C (2_3), C (4_) before sampling.
5), C (6_7), C (8_9) ... to C (2), C
(6), C (10) ... need to be calculated.

For example, C in the image shown in FIG.
3b and C of the image shown in FIG. 3D is represented as Cd,
The relationship between Cb and Cd is one of the following expressions (13) to (15).

Cd (i * 4 + 2) = Cb ((i * 4) _ (i * 4 + 1)) (13) Cd (i * 4 + 2) = Cb ((i * 4 + 2) _ (i * 4 + 3)) ... (14) Cd (i * 4 + 2) = Cb ((i * 4) _ (i * 4 + 1)) * W1 + Cb ((i * 4 + 2) _ (i * 4 + 3)) * W2 (15) where Cd ( i) represents C at the coordinate i in FIG. W1 and W2 are weighting factors, and W1 + W2 =
It is 1.

As shown in equations (10) to (15), FIG.
The calculation is performed by using the pixel value of the pixel corresponding to the position C in (b) before sampling. That is,
Equations (10) and (13) use the pixel value of C on the left side, and equations (11) and (14) use the pixel value of C on the right side of equations (12) and (15). ) Is a case where the average pixel value of C on both sides is used.

In order to convert the images captured by the two cameras 101a and 101b into a left-eye image and a right-eye image, respectively, one of the images shown in FIGS. 3 (c) and 3 (d) is used. The user makes a selection depending on the situation. Then, a flag indicating which is selected is output to multiplexing section 105. The multiplexing unit 105 multiplexes this flag into the coded data of the left-eye image and the coded data of the right-eye image as control information.

In the case where the optical axes are taken in parallel, if the image shown in FIG. 3 (c) is selected as the image for the left eye and the image shown in FIG. 3 (d) is selected as the image for the right eye, a highly accurate image is obtained. Is obtained. Further, for both the left-eye image and the right-eye image, the process is easy if FIG. 3C is selected. In this way, the user may select the image suitable for the situation each time.

As described above, in the stereoscopic image device according to the present embodiment, the cameras 101a and 10a are provided.
The image captured by 1b is YUV 4: 2: 0 (MPEG-1
(For use) format, when the left-eye image creating unit 102 and the right-eye image creating unit 103 create the left-eye image and the right-eye image, the color information of the image created by performing the thinning-out process. Since the color information is calculated according to the position of, it becomes possible to create a highly accurate stereoscopic image.

Further, the control information including the information indicating which YUV format the image corresponds to and the information indicating at which position the pixel is coded are the coded data of the left eye image and the right eye image. Therefore, when the stereoscopic image data display unit 112 displays a stereoscopic image, it is convenient for the user to determine which combination of images is more likely to be stereoscopically viewed. It has become possible to improve. In addition, by feeding back to the stereoscopic image data creation unit 111 information indicating which combination is likely to look stereoscopic, the left-eye image and the right-eye image in the left-eye image creation unit 102 and the right-eye image creation unit 103 are displayed. It becomes easy to select images, etc.
It is possible to obtain the same effect as the effect described in the first embodiment.

Incidentally, the first to third embodiments of the present invention
Through the embodiment of FIG.
When communicating between the stereoscopic image data display unit 112 and the stereoscopic image data display unit 112, the combination of the left-eye image and the right-eye image is changed by the stereoscopic image data creation unit 111 so that the stereoscopic image data display unit 112 selects the most. When a combination that makes it easy to see a stereoscopic image is found, the combination may be fed back from the stereoscopic image data display unit 112, and a flag indicating which one has been selected may be multiplexed by the multiplexing unit 105.

(Fourth Embodiment) The schematic configuration of the stereoscopic image device according to the fourth embodiment of the present invention is the same as the schematic configuration of the stereoscopic image device according to the first embodiment shown in FIG. . Therefore, detailed description of overlapping configurations and functions will not be repeated.

In the present embodiment, the display section 110 is compatible with the lenticular system and the parallax barrier system. Decoders 107a and 107b
2B output image data in YUV 4: 2: 2 format, and the format conversion units 108a and 108b expand the image data to image data in YUV 4: 4: 4 format.

FIG. 4 is a diagram for explaining the interpolation processing of the format conversion units 108a and 108b in the fourth embodiment. As shown in FIG. 4A, the output data from the decoders 107a and 107b does not include the C of the pixel at the coordinates 2, 6 ... Figure 4 (c)
Shows an example in which C of this pixel is interpolated.

For example, C in the image shown in FIG.
c and C of the image shown in FIG. 4A is represented by Ca,
The relationship between Cc and Ca is one of the expressions (16) to (18) shown below. Equations (16) and (17) show the case where neighboring pixel values are copied as they are,
Expression (18) shows a case where a weighted sum of a plurality of pixel values is calculated.

Cc (4 * i + 2) = Ca (4 * i) (16) Cc (4 * i + 2) = Ca (4 * i + 4) (17) Cc (4 * i + 2) = W1 * Ca (4 * i) + W2 * Ca (4 * i + 4) (18) Here, Cc (i) is C of the pixel at the coordinate i of the image shown in FIG. 4 (c), and Ca (i) is shown in FIG. 4 (a). Image coordinate i
Represents the C of the pixel. Further, W1 and W2 are weighting factors, and W1 + W2 = 1.

When the YUV 4: 2: 2 format image data shown in FIG. 2C is output from the decoders 107a and 107b, equations (16) to (18) are as follows.

Cd (4 * i + 3) = Cc (4 * i + 1) (19) Cd (4 * i + 3) = Cc (4 * i + 5) (20) Cd (4 * i + 3) = W1 * Cc (4 * i + 1) + W2 * Cc (4 * i + 5) (21) where Cd (i) is C of the pixel at the coordinate i of the image shown in FIG. 4 (d), and Cc (i) is shown in FIG. 2 (c). Image coordinate i
Represents the C of the pixel. Further, W1 and W2 are weighting factors, and W1 + W2 = 1. When using the pixel value of C on the left side of the equations (16) and (19), the equation (1)
7) and equation (20) show the case where the pixel value of C on the right is used, and equations (18) and (21) show the case where the average pixel value of C on both sides is used.

Next, the image data in the YUV 4: 1: 1 format shown in FIG. 2E is output from the decoders 107a and 107b, and the format conversion units 108a and 108b expand the image data in the YUV 4: 4: 4 format. The case will be described.

As shown in FIG. 4B, the decoder 107
The output data of a and 107b do not include the C of the pixel at coordinates 2, coordinates 4, coordinates 6 ... FIG. 4C shows an example in which C of this pixel is interpolated.

For example, C in the image shown in FIG.
c and C of the image shown in FIG. 4B is represented as Cb,
The relationship between Cc and Cb is expressed by the following equations (22) to (24).
Will be either. Note that equations (22) and (23)
Shows the case where the neighboring pixel values are copied as they are, and Expression (24) shows the case where the weighted sum of a plurality of pixel values is calculated.

Cc (8 * i + 2 * k) = Cb (8 * i) (22) Cc (8 * i + 2 * k) = Cb (8 * i + 8) (23) Cc (8 * i + 2 * k) = W1 * Cb (8 * i) + W2 * Cb (8 * i + 8) (24) Here, Cc (i) is the pixel C at the coordinate i of the image shown in FIG. 4C, and Cb (i) is the figure. Coordinate i of the image shown in FIG.
Represents the C of the pixel. Further, W1 and W2 are weighting coefficients, and W1 + W2 = 1 and k = 1, 2, and 3.

When the YUV 4: 1: 1 format image data shown in FIG. 2 (f) is output from the decoders 107a and 107b, equations (22) to (24) are as follows.

Cd (8 * i + 2 * k + 1) = Cf (8 * i + 1) (25) Cd (8 * i + 2 * k + 1) = Cf (8 * i + 9) (26) Cd (8 * i + 2 * k + 1) = W1 * Cf (8 * i + 1) + W2 * Cf (8 * i + 9) (27) where Cd (i) is the pixel C at the coordinate i of the image shown in FIG. 4D, and Cf (i) is the figure. Coordinate i of the image shown in 2 (f)
Represents the C of the pixel. Further, W1 and W2 are weighting coefficients, and W1 + W2 = 1 and k = 1, 2, and 3.

The method for selecting the interpolation pixel calculation method is based on which calculation method is selected to make it easier to see a stereoscopic image. If the characteristics of the display unit 110 in FIG. 1 are largely dependent, the stereoscopic image data display unit 112 may be uniquely determined. Further, when the user's individual difference greatly depends, the user may select the interpolation method in the stereoscopic image data display unit 112.

When the optimum interpolation method depends on the data created by the stereoscopic image data creation unit 111,
Information indicating the optimum interpolation method is multiplexed in advance by the multiplexing unit 105. Then, the stereoscopic image data display unit 1
The demultiplexing unit 106 in 12 may take out this information, and the format conversion units 108a and 108b may select the interpolation method accordingly.

As shown in FIG. 4 (e), the synthesizing section 109 alternates the image data converted into the YUV 4: 4: 4 format in the horizontal direction with the image for the left eye and the image for the right eye one pixel row at a time. Create an image for display by arranging. By displaying the display image on the display unit 110, the user can see the stereoscopic image.

As described above, in the stereoscopic image device according to this embodiment, the format conversion unit 108
Since a and 108b are adapted to interpolate YUV 4: 2: 2 format image data or YUV 4: 1: 1 format image data to generate YUV 4: 4: 4 format image data, the display unit 110 displays the lenticular. It became possible to display a highly accurate stereoscopic image when the system or the parallax barrier system was supported.

Further, since the control information indicating how to interpolate the color information is multiplexed with the coded data of the left-eye image and the coded data of the right-eye image, the stereoscopic image data is obtained. When the display unit 112 displays a stereoscopic image, the image can be interpolated so that the color information becomes optimum.

(Fifth Embodiment) A schematic configuration of a stereoscopic image device according to a fifth embodiment of the present invention is similar to that of the stereoscopic image device according to the first embodiment shown in FIG. . Therefore, detailed description of overlapping configurations and functions will not be repeated.

In the present embodiment, the display unit 110 corresponds to the shutter (time division) system.

FIG. 5 is a diagram for explaining the interpolation processing of the format conversion units 108a and 108b in the fifth embodiment. As shown in FIG. 5D, since the left-eye image and the right-eye image are arranged alternately line by line in the vertical direction, the output data of the decoders 107a and 107b are vertically thinned to 1/2 and horizontally output. It is necessary to double the size in the direction (return to the size before the left-eye image and the right-eye image are created).

Therefore, the format converter 108a
And 108b, in the same manner as described in the fourth embodiment, except that C is interpolated so as to obtain image data of YUV 4: 4: 4 format, and YUV is thinned out to 1/2 in the vertical direction. In addition, a process to double the size in the horizontal direction is newly required.

FIG. 5 (a) corresponds to FIG. 2 (b) or FIG.
The case where the image data shown in (e) is converted into the image data in the YUV 4: 4: 4 format by the same method as in the fourth embodiment is shown, and corresponds to FIG. 4 (c). Further, in FIG. 5B, the image data shown in FIG. 2C or 2F is subjected to YUV in the same manner as in the fourth embodiment.
It shows a case where the image data is converted into a 4: 4: 4 format image data, and corresponds to FIG.

Here, FIG. 2B and FIG.
Left eye image creation unit 102 and right eye image creation unit 103
Is created by sampling even-numbered pixels. 2C and 2F are created by the left-eye image creating unit 102 and the right-eye image creating unit 103 sampling odd-numbered pixels.
A flag indicating which of the sampling methods has been selected is multiplexed in the encoded data by the multiplexing unit 105.

Therefore, the pixels in which C is interpolated in the fourth embodiment are arranged by the format conversion units 108a and 108b according to this flag. That is, in the case of FIG. 4C, the pixels are arranged as shown in FIG. 5A, and in the case of FIG. 4D, the pixels are arranged as shown in FIG. 5B. Then, the pixels for which Y and C are interpolated in the horizontal direction are calculated. Accordingly, the stereoscopic image data creation unit 111 and the stereoscopic image data display unit 11
It is possible to prevent the phase difference between 2 and.

The above description is for the YUV 4: 2: 2 format or the YUV 4: 1: 1 format, but the YU as shown in FIGS. 3C and 3D is used.
The same applies to the case of the V4: 2: 0 (for MPEG-1) format.

In the above description, the image data is first converted to the YUV 4: 4: 4 format, then thinned out in the vertical direction and enlarged in the horizontal direction to create the image for the left eye and the image for the right eye. The order of processing is not limited to this, and for example, the order may be such that horizontal expansion is performed and then vertical thinning is performed.

As shown in FIG. 5D, the synthesizing unit 110 makes the left-eye image (L) and the right-eye image (R) line by line in the vertical direction, as shown in FIG. 5D. Images are arranged alternately to create a stereoscopic display image.

As described above, in the stereoscopic image device according to this embodiment, the format conversion section 108 is used.
Since a and 108b are interpolated so as to be image data of YUV 4: 4: 4 format, YUV is decimated to 1/2 in the vertical direction and further doubled in the horizontal direction. It becomes possible to display a highly accurate stereoscopic image when the shutter (time division) system is supported.

Since the control information indicating the positions of the sampled pixels is multiplexed with the coded data of the left-eye image and the coded data of the right-eye image, the stereoscopic image data display unit 112 is When displaying a stereoscopic image,
It is possible to interpolate the image so that the color information is optimum and prevent the stereoscopic image data creation unit 111 and the stereoscopic image data display unit 112 from being out of phase.

In the above embodiments, the case of binocular vision has been described, but the present invention is not limited to this, and the N-eye vision (N
The same applies to the case of ≧ 2).

The embodiments disclosed this time must be considered as illustrative in all points and not restrictive. The scope of the present invention is shown not by the above description but by the claims, and is intended to include meanings equivalent to the claims and all modifications within the scope.

[0119]

According to one aspect of the present invention, since the color information calculation method is changed according to the position of the sampled pixel to calculate the pixel color information, highly accurate stereoscopic image data can be obtained. It is now possible to create.

Further, since the encoded left-eye image and right-eye image are multiplexed with the information regarding the predetermined format, the user can select which format by displaying the stereoscopic image on the stereoscopic image display device. It has become possible to easily judge whether the created stereoscopic image looks stereoscopic.

Since the encoded left-eye image and right-eye image and the information regarding the color information calculation method are multiplexed, the user can display the stereoscopic image on the stereoscopic image display device. It has become possible to easily judge by which calculation method the stereoscopic image in which the color information is created is likely to be stereoscopic.

Since the coded left-eye image and right-eye image and the information regarding the interpolation method of the color information of the left-eye image and the right-eye image are multiplexed, the stereoscopic image display device By extracting the information about the interpolation method of the color information, it becomes possible to interpolate the stereoscopic image data so as to optimize the color information and display an image that is more stereoscopically visible.

Further, since the encoded left-eye image and right-eye image and information regarding the positions of the sampled pixels are multiplexed, the three-dimensional image display device displays the pixels according to the information indicating the sampled positions of the pixels. Can be arranged, and it has become possible to prevent a phase shift between the stereoscopic image creating apparatus and the stereoscopic image display apparatus.

According to another aspect of the present invention, the format conversion means refers to the color information interpolation method and interpolates the color information of the left-eye image and the right-eye image to convert it into an image of a predetermined format. Therefore, it becomes possible to interpolate the stereoscopic image data so that the color information becomes optimal and display an image that is more stereoscopically visible.

According to still another aspect of the present invention, the format conversion means refers to the information on the positions of the sampled pixels to horizontally expand the image for the left eye and the image for the right eye to obtain a predetermined format. Since the image is converted into an image, it is possible to prevent the phase shift between the stereoscopic image creating device and the stereoscopic image display device.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of a stereoscopic image device according to a first embodiment of the present invention.

FIG. 2: YUV 4: 2: 2 format and YUV
It is a figure which shows an example of a 4: 1: 1 format.

FIG. 3 is a diagram showing an example of a YUV 4: 2: 0 format.

FIG. 4 is a diagram for explaining an interpolation process of format conversion units 108a and 108b in the fourth embodiment.

FIG. 5 is a diagram for explaining an interpolation process of format conversion units 108a and 108b in the fifth embodiment.

FIG. 6 is a diagram for explaining a method of creating image data in a conventional stereoscopic image device.

[Explanation of symbols]

101a, 101b camera, 102 left-eye image creating unit, 103 right-eye image creating unit, 104a, 104b
Encoder, 105 multiplexer, 106 demultiplexer,
107a, 107b Decoders, 108a, 108b
Format conversion unit, 109 composition unit, 110 display unit, 111 stereoscopic image data creation unit, 112 stereoscopic image data display unit.

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Hiroyuki Katata             22-22 Nagaikecho, Abeno-ku, Osaka-shi, Osaka             Inside the company F term (reference) 2H059 AA07 AA18 AA35 AA38                 5C061 AB04 AB08 AB12 AB17

Claims (10)

[Claims] 1. A stereoscopic image creating apparatus for creating stereoscopic image data from an input image in which luminance information and color information are separated, and left-eye image creating means for creating a left-eye image, A right-eye image creating means for creating a right-eye image, a left-eye image created by the left-eye image creating means, and a right-eye image created by the right-eye image creating means Encoding means for encoding, including a multiplexing means for multiplexing the left-eye image and the right-eye image encoded by the encoding means, the left-eye image creating means and the right The eye image creating unit samples pixels while thinning out the input image in the horizontal direction, changes the color information calculation method according to the positions of the sampled pixels, calculates the color information of the pixels, and then calculates a predetermined format. Image of To, the stereoscopic image generating apparatus. 2. The stereoscopic image creation according to claim 1, wherein the multiplexing unit multiplexes the left-eye image and the right-eye image encoded by the encoding unit and the information regarding the predetermined format. apparatus. 3. The multiplexing unit multiplexes the left-eye image and the right-eye image encoded by the encoding unit and information regarding a method of calculating the color information. The three-dimensional image creation device described. 4. The information relating to the interpolation method of the color information of the left-eye image and the right-eye image, the left-eye image and the right-eye image encoded by the encoding means, and the multiplexing means. The stereoscopic image creating apparatus according to claim 1, wherein the and are multiplexed. 5. The method according to claim 1, wherein the multiplexing unit multiplexes the left-eye image and the right-eye image encoded by the encoding unit and the information regarding the positions of the sampled pixels. 3D image creation device. 6. A stereoscopic image display device for synthesizing and displaying an image for the left eye and an image for the right eye, in which luminance information and color information are separated, wherein the multiplexed data is demultiplexed to obtain a code. Demultiplexing means for extracting the encoded left-eye image, the encoded right-eye image, the interpolation method of the color information of the left-eye image and the right-eye image, and the demultiplexing Decoding means for decoding the encoded left-eye image and the encoded right-eye image extracted by the means, the left-eye image and the right-eye for the demultiplexing means extracted A format conversion unit for interpolating the color information of the left-eye image and the right-eye image decoded by the decoding unit to convert the image into a predetermined format image by referring to an interpolation method of image color information, and Converted by format conversion means Comprising synthesizing means for synthesizing an image and a right-eye image for the left eye, and display means for displaying a stereoscopic image according to the synthesized image by said synthesizing means, three-dimensional image display device. 7. A stereoscopic image display device for synthesizing and displaying a left-eye image and a right-eye image in which luminance information and color information are separated, wherein the multiplexed data is demultiplexed to obtain a code. Demultiplexing means for extracting the encoded left-eye image, the encoded right-eye image, and information regarding the positions of the sampled pixels of the left-eye image and the right-eye image, the inverse Decoding means for decoding the encoded left-eye image and the encoded right-eye image extracted by the multiplexing means, the left-eye image and the right extracted by the demultiplexing means A format conversion for horizontally enlarging the left-eye image and the right-eye image decoded by the decoding means and converting them into an image of a predetermined format with reference to information on the positions of sampled pixels of the eye image means A combining unit for combining the left-eye image and the right-eye image converted by the format converting unit, and a display unit for displaying a stereoscopic image according to the image combined by the combining unit. Including a stereoscopic image display device. 8. A stereoscopic image creating method for creating stereoscopic image data from an input image in which luminance information and color information are separated, the step of creating an image for the left eye and an image for the right eye. Step, including a step of encoding the created left-eye image and right-eye image, including a step of multiplexing the encoded left-eye image and right-eye image, the left eye In the step of creating an image and the step of creating the image for the right eye, pixels are sampled while the input image is thinned out in the horizontal direction, and the calculation method of color information is changed according to the position of the sampled pixel. A stereoscopic image creating method, comprising: calculating color information of pixels to create an image in a predetermined format. 9. A stereoscopic image display method for synthesizing and displaying an image for the left eye and an image for the right eye, in which luminance information and color information are separated, wherein the multiplexed data is demultiplexed to obtain a code. Image for the left eye, the image for the right eye that has been encoded, a step of extracting the interpolation method of the color information of the image for the left eye and the image for the right eye, the extracted encoded And a step of decoding the left eye image and the encoded right eye image, with reference to the interpolation method of the color information of the extracted left eye image and right eye image, the decoded left eye For interpolating color information of the image for right eye and the image for right eye to convert the image into an image of a predetermined format, a step of combining the left eye image and the right eye image converted to the predetermined format, A step to display a stereoscopic image according to the image. Including the door, three-dimensional image display method. 10. A stereoscopic image display method for synthesizing and displaying an image for the left eye and an image for the right eye in which luminance information and color information are separated, wherein the multiplexed data is demultiplexed to obtain a code. The image for the left eye that has been encoded, the image for the right eye that has been encoded, a step of extracting information regarding the positions of the sampled pixels of the image for the left eye and the image for the right eye, and the extracted encoded A step of decoding the left-eye image and the encoded right-eye image, and referring to the information regarding the positions of the sampled pixels of the extracted left-eye image and right-eye image, the decoded A step of horizontally expanding the image for the left eye and the image for the right eye to convert the image into a predetermined format image, a step of combining the left eye image and the right eye image converted into the predetermined format, and the combining It It was and displaying a stereoscopic image according to the image, the stereoscopic image display method.