JP2001298754A - Crosstalk cancellation method and system for stereoscopic image - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize an image processing method or an image processing system with which a crosstalk caused in a stereoscopic vision by both eyes eliminated or reduced y applying gradation processing to images for left and right eyes and using images resulting from a weighted difference of the images for left and right eyes for respective stereoscopic images. SOLUTION: The image processing method or the image processing system employs a linear image conversion and weight difference processing sections that display images R', L' on a stereoscopic image display screen, which images are obtained by substituting images R, L to an expression R'=(R-aL)÷(1-a2) as a right eye image and an expression L'=(L-aR)÷(1-a2) as a left eye image, wherein R, L are images subjected to linear gradation conversion under the conditions of a<=R<=1 and a<=L<=1 and 'a' is the minimum luminance of the images R, L.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing system and an image processing apparatus for reducing or eliminating crosstalk in binocular stereoscopic vision.

[0002]

2. Description of the Related Art Conventionally, an attempt to extract a pure stereo image from mixed stereo images by using respective difference images is described in Japanese Patent Application Laid-Open Nos.
090241, JP-A-10-048538 and JP-A-1
Although the present inventor has shown in Japanese Patent Application Laid-Open No. 0-216082, in the method shown here, eliminating the crosstalk which indicates the coexistence occurring on the three-dimensional display is not possible because a negative luminance display cannot be performed on the display. It was impossible.

[0003]

SUMMARY OF THE INVENTION The present invention has been made in view of the above points, and provides an image processing method and an image processing apparatus for reducing or eliminating crosstalk in binocular stereoscopic vision. .

[0004]

Accordingly, the present invention firstly performs a linear gradation conversion of a stereoscopic video signal so that a difference between image components corresponding to crosstalk can be performed in advance, and performs the weighting difference. According to another aspect of the present invention, there is provided an image processing method and an image processing apparatus for displaying a stereoscopic image, wherein crosstalk during stereoscopic viewing is eliminated or reduced. Second, when the display of the brightness of each pixel of each parallax image of the stereoscopic video signal is standardized and expressed in a range of 0 to 1, the darkest brightness 0 of the parallax image brightness is set to be larger than 0 and larger than 1 By changing the value of the coefficient to a small coefficient, the gradation relating to the luminance of the parallax image is linearly converted, and each pixel luminance of the left and right eye images of a pair of stereo images of the parallax images after the linear conversion, A pixel difference obtained by multiplying the luminance of each pixel of the image for the right and left eyes and the image for the eye opposite to the coefficient by the coefficient is taken for each pixel, and a value obtained by subtracting the square of the coefficient from 1 is further multiplied for each pixel. 2. The stereoscopic apparatus according to claim 1, wherein when the image reconstructed by the groups is used as a stereoscopic image for the left and right eyes, the coefficient is optimized and determined to reduce or eliminate crosstalk generated during stereoscopic vision. Image processing method and image for image display Due to the fact that the management apparatus.

[0005]

According to the present invention, the crosstalk in the binocular stereopsis is subjected to gradation processing of the images for the left and right eyes, and the images obtained by taking the weighted difference between the images for the left and right eyes are converted into stereo images. Can be eliminated or reduced.

[0006]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing an embodiment of a crosstalk canceling method for stereoscopic images according to the present invention. FIG. 1 is an explanatory view showing gradation conversion of a crosstalk elimination method for a stereoscopic image in this embodiment, and FIG. 2 is a block diagram of a crosstalk elimination method including the gradation conversion. The pixel group observed by the observer includes a crosstalk state in which a left-eye image L is mixed with a right-eye image R indicated by R + aL for each pixel, and L + a
This is a crosstalk state in which the right-eye image R is mixed with the left-eye image L shown in R. However, 0 <a <1. Here, in order to eliminate this crosstalk, R ′ + aL ′ =
I want to create R and L by inputting special R 'and L' on the stereoscopic video display screen such as R, L '+ aR' = L. Theoretically, R ′ = (R−aL) as an image for the right eye
{(1-a ² ), L ′ = (L−aR) as the left-eye image}
It is sufficient that (1-a ² ) can be displayed on the stereoscopic image display surface.
= 0, R '<0 when the image is black and L ≠ 0, and such negative luminance cannot be physically displayed.
Therefore, as shown in FIG. 1, the minimum luminance of R and L is a.
R obtained by performing a linear gradation conversion satisfying ≦ R ≦ 1, a ≦ L ≦ 1,
According to the algorithm shown in FIG.
And L ′, 0 ≦ R ′ ≦ 1, 0 ≦ L ′ ≦ 1, and display is possible. Therefore, when the R ′ and the L ′ are displayed as a right-eye image and a left-eye image on the stereoscopic image display surface, crosstalk disappears. In this embodiment, the crosstalk elimination method for observing a stereoscopic image is described, but this method may be used as an image conversion device. Needless to say, any algorithm having the same effect may be used.

[0007]

Since the present invention is configured as described above, the crosstalk in the binocular stereopsis is subjected to gradation processing of the left and right eye images, and the weighted difference between the left and right eye images is calculated. This is a unique effect that can be eliminated or reduced by converting each image into a stereo image.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram showing gradation conversion by a crosstalk elimination method in a stereoscopic image according to the invention based on claims 1 and 2;

FIG. 2 is a block diagram showing an algorithm of a crosstalk elimination method in a stereoscopic image according to the invention based on claims 1 and 2;

[Explanation of symbols]

A Normalized luminance signal before gradation conversion B Normalized luminance signal after gradation conversion a Parameter indicating crosstalk C Linear gradation conversion using intercept C R ′ = (R−aL ) 重 み (1-a ² ) weighted difference processing E L ′ = (L−aR) ÷ (1-a ² ) weighted difference processing R image signal for right eye before processing L for left eye before processing Image signal R ′ Image signal for right eye after processing L ′ Image signal for left eye after processing

Claims (2)

[Claims] 1. A stereoscopic video signal is subjected to linear gradation conversion so that a difference between image components corresponding to crosstalk can be performed in advance, and the weighting difference is performed, thereby eliminating or reducing crosstalk during stereoscopic viewing. An image processing method and an image processing apparatus for displaying a stereoscopic image, which are characterized. 2. A standardized representation of brightness of each pixel of each parallax image of a stereoscopic video signal, which is expressed in a range from 0 to 1,
The gradation regarding the luminance of the parallax image is linearly converted by changing the darkest luminance 0 of the parallax image luminance to a value of a coefficient larger than 0 and smaller than 1, and among the parallax images after the linear conversion, The difference between each pixel luminance of the left and right eye images of the pair of stereo images and the luminance obtained by multiplying each pixel luminance of the opposite eye image and each pixel luminance of the ophthalmic image by the coefficient is taken for each pixel. When an image reconstructed from a pixel luminance group obtained by further multiplying each pixel by a value obtained by subtracting the square is used as a stereo image for the left and right eyes, by optimizing and determining the coefficient, crosstalk generated in stereoscopic vision 2. An image processing method and an image processing apparatus for displaying a three-dimensional image according to claim 1, wherein the image processing method is configured to reduce or eliminate the difference.