KR101080988B1 - A 3D Picture Producing Method Using Optimum Parallax, And A Storage Medium - Google Patents

Abstract

The present invention relates to a method for producing optimal parallax 3D images. According to an aspect of the present invention, there is provided a method for producing an optimal parallax 3D image, comprising: an environment setting step of setting a screen size, a viewing distance, and a resolution of a device for outputting a 3D stereoscopic image; Determining a pixel reference value corresponding to a binocular parallax that is optimal for formation of a three-dimensional effect in response to the screening condition set in the environment setting step; Determining a pixel reference value corresponding to a maximum limit binocular parallax in which a person can feel a three-dimensional feeling in response to the screening condition set in the environment setting step; The original image is shifted by the pixel value corresponding to the optimal binocular disparity to form a left eye image and a right eye image, and a pixel change occurring in the original image during the image conversion process is within a range of pixel reference values corresponding to the maximum limiting binocular disparity. It characterized in that it comprises a converting step of forming a three-dimensional image to be.

Description

A 3D Picture Producing Method Using Optimum Parallax, And A Storage Medium}

The present invention relates to a 3D video production method, and more particularly, to determine the pixel reference value corresponding to the optimal binocular disparity and maximum binocular disparity according to the size, resolution, and viewing distance of the screen to use for converting to a 3D video object. The present invention relates to a method for producing optimal parallax 3D images.

There are two ways to create 3D stereoscopic content. There are two methods of capturing images by using two cameras imitating human left and right eyes, and converting existing 2D contents into 3D stereoscopic images.

Among them, the method of converting 2D contents into 3D stereoscopic image contents is in the spotlight in terms of creating new profits by recycling previously produced contents.

The method of converting 2D content into 3D stereoscopic content is a method of assuming an original image as a left image and generating a virtual parallax right image, and then replaying it to an exclusive display.

However, this method itself involves a process that is difficult to machine, so most of the conversion work now relies on human hand using retouching graphic tools.

In addition, since the cost of producing content is increased due to the manual labor, most of the converted contents are mostly high-definition screen-only contents requiring HD or higher quality.

The production method of 3D stereoscopic video contents must be different depending on the media to be reproduced, so you can feel the stereoscopic effect of the content that was originally intended. It becomes difficult and the advantages of three-dimensional content disappear.

In particular, the optimal value (13.5 minutes) or maximum value (1 degree or 60 minutes) of human binocular disparity for stereoscopic images is determined as an experimental value, but specific pixel values applicable to PCs, HDTVs, and mobile devices are not determined. It is applied to other stereoscopic video screening devices such as mobile or PC monitor only by converting the image or by changing the size of high-definition screen-only contents by the sense of the operator. There is a problem of overlapping visual confusion or eye strain.

An object of the present invention is to determine an optimal binocular parallax and maximum limit binocular parallax according to the size, resolution, and viewing distance of the screen, and to use it for converting into a 3D image to form the best stereoscopic effect. A video production method, an optimal parallax video image, and a recording medium are implemented.

Optimum parallax 3D image production method according to the present invention for achieving the above object is a method for producing an optimal parallax 3D image to produce a 3D image of the optimal parallax by shifting the 2D original image, the screen of the device for outputting a 3D stereoscopic image Setting a size, a viewing distance, and a resolution; In response to the screening condition set in the environment setting step, determining a optimal pixel reference value, which is a pixel reference value corresponding to binocular parallax, which is optimal for forming a stereoscopic feeling; Determining a threshold pixel reference value that is a pixel reference value corresponding to a maximum marginal binocular parallax in which a person can feel a three-dimensional feeling in response to the screening condition set in the environment setting step; Converting the original image according to the optimal pixel reference value to form a left eye image and a right eye image, and converting the stereoscopic image to form a stereoscopic image such that a pixel change occurring in the original image is within a range of the limit pixel reference value during an image conversion process. In the converting step, for some scenes where the degree of stereoscopic feeling may be relatively low, a frame corresponding to the optimal pixel reference value is selected to replace the left eye image and the right eye image.

The limit pixel reference value used in the converting step may select a minimum value among pixel reference values corresponding to each maximum binocular disparity determined according to positions of a viewer watching the 3D image.

The limit pixel reference value used in the converting step may select a minimum value among pixel values of each maximum binocular parallax corresponding to the size of screens on which the 3D image is to be displayed.

The limit pixel reference value used in the converting step may select a minimum value among pixel reference values of each maximum binocular parallax corresponding to the resolution of screens on which the 3D image is to be displayed.

The converting step includes a shift for the object and a skew for the background, and the maximum pixel displacement of all the conversion processes is less than the threshold pixel reference value.

The method for producing an optimal parallax 3D image according to the present invention for achieving the above object is a method for producing an optimal parallax 3D image for producing a 3D video by binocular disparity with respect to a delayed frame of a 2D video, and an apparatus for outputting a 3D stereoscopic image. A preference step of setting the size, viewing distance and resolution of the screen; Determining an optimal pixel reference value that is a pixel reference value corresponding to a binocular parallax that is optimal for formation of a stereoscopic sense, in response to the screening condition set in the environment setting step; In converting the 2D image into a 3D image, converting the left eye image and the right eye image by selecting a frame corresponding to the optimal pixel reference value for some scenes where the degree of stereoscopic feeling may be relatively low. Characterized in that.

A computer-readable recording medium recording a computer program according to the present invention for achieving the above object is characterized by recording a computer program for performing each step of the method for producing an optimal parallax 3D image on any one of the above described on a computer system. .

The recording medium according to the present invention for achieving the above object is characterized in that the 3D video material produced by each step of the method for producing the optimal parallax 3D video object described above is readable on a computer system.

The present invention provides the optimal parallax and the maximum parallax for forming a stereoscopic image as the pixel information according to the size and resolution of the screen on which the image is to be screened according to the above configuration, while giving the maximum stereoscopic effect for the image, and various distance and angle positions. It is possible to provide an appropriate stereoscopic image for the viewer of.

In addition, since the parallax range is converted to the pixel reference value according to the size and resolution of the screen so that the image can be produced by the optimal binocular disparity that gives the maximum stereoscopic effect even for the image screened on a smartphone having a small screen size, It is possible to give an optimized stereoscopic effect even on a small screen.

1 is a view illustrating a screening apparatus and a viewing distance for outputting a stereoscopic image.
2 is a schematic diagram illustrating a process of calculating a pixel reference value corresponding to an optimal binocular disparity and a maximum binocular disparity according to the size and viewing distance of a screening apparatus.
3 is a diagram illustrating a pixel reference value corresponding to an optimal value of binocular disparity and a pixel reference value corresponding to a maximum value of binocular disparity according to the size of the screening apparatus.
4 is a diagram illustrating a pixel reference value corresponding to an optimal value of binocular disparity and a pixel reference value corresponding to a maximum value of binocular disparity according to the size of the screening apparatus.
5 is a schematic diagram illustrating a process of calculating a maximum and optimal pixel reference value corresponding to a viewing position;
FIG. 6 is a diagram illustrating a process of manufacturing a stereoscopic image by applying a pixel reference value of optimal binocular disparity and a pixel reference value of maximum binocular disparity according to the present embodiment; FIG.
7 is a diagram illustrating a process of applying a pixel reference value of maximum and optimal binocular parallax to video converting.

Hereinafter, a configuration of a preferred embodiment of an optimal parallax 3D image production method according to the present invention having the above configuration will be described in detail with reference to the accompanying drawings.

3D video production method according to an embodiment of the present invention converts the 2D video to 3D video by converting the maximum / optimal binocular disparity according to the conditions of the screening device and the viewer to the pixel reference value to input the screening device and viewer conditions of the 3D video The process of calculating the optimal / maximum parallax to the pixel reference value is preceded.

Left and right images produced for 3D stereoscopic images are best perceived when the binocular disparity of the two images is about 13 minutes. In addition, when the time difference exceeds 60 minutes, the stereoscopic images do not begin to form well. Most mobile devices have screen sizes ranging from 2.5 inches to 7 inches, the average viewing distance is about 30cm, and the horizontal viewing angle of a person covers about 6 to 10 degrees. There is a significant difference in the pixel reference values corresponding to the optimal / maximum binocular parallax that can be felt.

For example, a movie theater has a screen size of 20m or more, a viewing distance of about 40m, a resolution of about 4K (4096 * 2048), a home TV or monitor of about 80cm, a viewing distance of about 2m, and a resolution of Full HD 1920 * 1080. For mobile, 3.2 inch viewing distance 30 cm, resolution 320 * 240, the maximum horizontal binocular disparity and the optimal horizontal binocular disparity as shown in the diagram shown in Figures 3 and 4.

In an embodiment of the present invention, a process of determining a pixel reference value corresponding to a maximum and optimal binocular disparity in consideration of the size and resolution of a screen as a condition of a screening device and a viewing distance and viewing angle as a condition of a viewer is performed.

For example, in FIG. 2, the process of determining the pixel reference value corresponding to the maximum and optimal binocular disparity for the conditions of the screening apparatus and the viewer is as follows.

Viewing distance (distance from where the person's view begins to the screen) = D

Screen size (width) = W

Hypotenuse length of right triangle with D as height and W / 2 as base = B

In this case, the range that the screen occupies in the human viewing angle is calculated as follows using the trigonometric function.

On the other hand, when the viewer's position is not the position corresponding to the center of the screen and is biased to one side of the screen, as shown in FIG. 5, the optimal binocular disparity and the maximum binocular disparity are the same, but because the viewer's position is different with respect to the same condition screening device, the pixel reference value Is different.

That is, in the exemplary embodiment of the present invention, the pixel reference value corresponding to the maximum binocular disparity calculated according to the viewer's position is calculated to form a stereoscopic image having an appropriate stereoscopic effect within the range of the viewer's position change. In other words,

The pixel reference value is calculated.

for example,

Becomes

For example, since the pixel reference value corresponding to the maximum binocular parallax of the viewer at the position centered closest to the screen is the smallest, in the embodiment of the present invention, the maximum number of pixels shifted to form the left eye image and the right eye image is the pixel reference value. It is done.

6 illustrates a process of forming a stereoscopic image using pixel reference values corresponding to the maximum marginal binocular disparity and optimal binocular disparity calculated by the above process.

For example, in the left original image, the top of the background is the deepest point on the screen, and the bottom is the most protruding part of the screen, and skew (distortion) that distorts the background to generate a stereoscopic effect. Shifts are made on objects 1 and 2.

That is, the object 1 and the object 2 may be shifted by pixel reference values corresponding to optimal and maximum binocular disparity to form a left eye image and a right eye image to form a three-dimensional effect. However, in this case, the three-dimensional effect of the object itself may not be reproduced, which may result in a sense of disconnection with the background, thereby allowing the background to match naturally with the background skew.

In this case, if the difference with the original pixel exceeds the maximum binocular disparity (1 °), the binocular fusion is not achieved, so the pixel shift due to skew should be performed below the pixel reference value corresponding to the maximum binocular disparity. . .

In this case, the object may be shifted by a pixel reference value (optimum pixel reference value) corresponding to an optimal binocular disparity for forming a virtual parallax between the left eye image and the right eye image, and maximum in background skew for a stereoscopic effect. The pixel displacement value of is based on the pixel reference value (limit pixel reference value) corresponding to the virtual maximum limit binocular parallax.

On the other hand, the pixel displacement value for the background skew and the object shift in this embodiment is smaller than the smallest value (the limit pixel reference value) among the pixel reference values corresponding to the maximum binocular disparity for the viewer (ie, expected viewing position) located in a certain range. It is desirable to set it to be small.

On the other hand, the pixel displacement value that may occur due to the background skew and the object shift of the present embodiment may be complemented by the range of the resolution of the screen on which the image is to be displayed.

That is, an optimal pixel reference value for the resolution of the most common screen can be calculated as the pixel reference value of image converting, and pixel reference values corresponding to the maximum binocular disparity for the screen having the lowest resolution of the screen on which the stereoscopic image is to be displayed. The pixel shift due to the background skew and the object shift is performed within a range smaller than the smallest value (limit pixel reference value), so that the stereoscopic image may be formed.

As shown in FIG. 6, the background (excluding the object 1 and the object 2) is skewed (the black part in the second background drawing from the left). As a result, the upper left and lower right parts of the background drawing have holes formed by screen distortion ( HOLE, black part).

In addition, as shown in the third from the left of FIG. 6, the object 1 is shifted to the left so that a hole, which is a black part, is formed in the right edge of the object 1. On the other hand, since the object 2 is an object in which the horizontal and vertical planes are clearly distinguished, the vertical plane is shifted and the horizontal plane is converted by skew to form a screen with no background.

In this case, the maximum pixel shift due to background skew should be set below the threshold pixel reference value, and the pixel change due to the shift of object 1 should also be below the threshold pixel reference value. As a result of the shift and skew of the entire pixel due to object 2, it is also necessary to be less than the threshold pixel reference value to form an appropriate stereo effect.

In FIG. 6, since the lowermost part of the background part is the closest distance where both eyes can be fused, the lowermost part of the background picture has a maximum amount of parallax and a sense of depth is about several tens of cm in front. In addition, the longest distance that a person can feel the depth is about 15 ~ 20m, and the top part of the background picture is the maximum sound parallax that a person can feel and can feel a depth of about 20m. It is possible to set the criteria for the pixel movement by.

Meanwhile, in another embodiment of the present invention, the pixel displacement occurring in processes for generating various stereoscopic effects other than the shift skew for the background object should be performed below the threshold pixel reference value, and on the final result screen, the threshold pixel reference is also performed. It is preferable to become less than a value.

Another embodiment of the present invention illustrated in FIG. 7 illustrates a process of giving a 3D stereoscopic effect from a 2D video.

When converting a 2D video into a 3D video, the left image and the right image, which becomes the difference image, are found in a frame close to time in order to form a virtual binocular disparity. The optimal binocular disparity is the vertical parallax of half the horizontal parallax. It can be usefully used in that it can cover even.

That is, even though the right image, which is the difference image of the left image, is found in a frame close in time, the vertical value is often incorrect. In this case, the complexity of the converting operation can be reduced by selecting a frame corresponding to the optimal binocular disparity.

For example, if the frame from the T-2 frame to the T + 2 frame is close to the left image as shown in the drawing, the pixel movement of the object with respect to the left image corresponds to selecting a frame having pixel movement corresponding to the optimal parallax. This reduces the complexity of the conversion.

In addition, in scenes where stereoscopic effect is not important, converting the optimal parallax instead of the maximum parallax into the transform enables the conversion by simply finding an approximate right image in a nearby frame without retouching.

The scope of the present invention is not limited to the above embodiments, but is determined by the matters described in the claims, and it is obvious that the present invention includes various modifications and adaptations made by those skilled in the art in the claims and equivalents.

Claims (8)

Optimal parallax 3D image production method that produces 2D original image by shifting the 2D original image,
An environment setting step of setting a size, viewing distance, and resolution of a screen of a device for outputting a 3D stereoscopic image;
In response to the screening condition set in the environment setting step, determining a optimal pixel reference value, which is a pixel reference value corresponding to binocular parallax, which is optimal for forming a stereoscopic feeling;
Determining a threshold pixel reference value that is a pixel reference value corresponding to a maximum marginal binocular parallax in which a person can feel a three-dimensional feeling in response to the screening condition set in the environment setting step;
Converting the original image by the optimal pixel reference value to form a left eye image and a right eye image, and converting the stereoscopic image to form a stereoscopic image such that a pixel change occurring in the original image within the image conversion process is within a range of the limit pixel reference value. Including;
And selecting a frame corresponding to the optimal pixel reference value to replace the left eye image and the right eye image for some scenes where the degree of stereoscopic feeling may be relatively low in the converting step.
The method of claim 1, wherein the threshold pixel reference value used in the converting step,
And selecting a minimum value among pixel reference values corresponding to each maximum binocular parallax determined according to positions of viewers viewing the 3D image.
The method of claim 1, wherein the threshold pixel reference value used in the converting step,
A method for producing an optimal parallax 3D image, characterized in that for selecting a minimum value among the pixel values of each maximum binocular parallax corresponding to the size of the screen on which the 3D image will be displayed.
The method of claim 1, wherein the threshold pixel reference value used in the converting step,
A method for producing an optimal parallax 3D image, characterized in that for selecting a minimum value among the pixel reference value of each maximum binocular parallax corresponding to the resolution of the screen on which the 3D image will be displayed.
The method of claim 1, wherein in the converting step,
And a shift of an object and a skew of a background, wherein the maximum pixel displacement of all the conversion processes is less than the threshold pixel reference value.
Optimal parallax 3D video production method that produces 3D video by binocular parallax for delayed frame of 2D video
An environment setting step of setting a size, viewing distance, and resolution of a screen of a device for outputting a 3D stereoscopic image;
Determining an optimal pixel reference value that is a pixel reference value corresponding to a binocular parallax that is optimal for formation of a stereoscopic sense, in response to the screening condition set in the environment setting step;
In converting the 2D image into a 3D image, converting the left eye image and the right eye image by selecting a frame corresponding to the optimal pixel reference value for some scenes where the degree of stereoscopic feeling may be relatively low. Optimal parallax 3D video production method characterized in that.
A computer-readable recording medium having recorded thereon a computer program for performing each step of the method for producing an optimal parallax 3D image object according to any one of claims 1 to 6.
A recording medium, characterized in that the 3D video material produced by each step of the method for producing optimal parallax 3D video material according to any one of claims 1 to 6 is recorded.

Images