KR20120095059A - Method of converting 2d images to 3d images - Google Patents

Abstract

PURPOSE: A method for converting a two-dimensional image into a three-dimensional image is provided to maximize distance the sense of distance and the sense of depth by giving three-dimensional effect. CONSTITUTION: A lighting layer is generated. A light is generated by a brush. The light is generated by a mask. A depth map is generated. A three-dimensional effect is generated through the depth map. Right and left images are generated. A lens blur is applied to the left image and the right image.

Description

How to convert 2D image to 3D stereoscopic image {METHOD OF CONVERTING 2D IMAGES TO 3D IMAGES}

The present invention relates to a method of converting a 2D animation into a 3D image by a 3D stereoscopic conversion method, and a method of stereoscopically working a 2D animation into a 3D animation.

A person feels three-dimensional by looking at an object from different angles in his left and right eyes. In order to implement and provide 3D stereoscopic images based on the above principle, stereoscopic images are artificially implemented by displaying two images respectively captured at the left and right eye visual locations, respectively, to the left and right eyes. do.

However, in order for the conventional technology to be effective, the image to be seen in the left eye and the image to be seen in the right eye must exist respectively as described above. Therefore, in order to create a stereoscopic image, two images must be newly produced. There was a difficulty. This means that many existing two-dimensional images (including animation) cannot be used in a stereoscopic image implementation system.

In particular, when the 2D animation is converted to a 3D image, due to the characteristics of the picture, it does not create a sense of volume (three-dimensional feeling), only the effect of a plate like a card. As such, two-dimensional images such as animations are more difficult to convert to 3D images than general images. This is because conventional 2D animation does not generally include not only lighting effects but also stereoscopic effects. In addition, it is because it is not easy to predict lighting effects and stereoscopic effects.

Accordingly, the present invention has been invented to solve the above problems, and an object thereof is related to a method for converting general 2D image data, preferably 2D animation into 3D animation.

As used herein, a depth map refers to a contrast image showing a depth value of an object on a 2D image. It is a map of the reference image to represent the part far from the viewer. For example, in the present invention, the portions near the viewing side are white, the far portions are black, and each position is expressed in a gradient manner of white and black according to the perspective from the viewing side. 1 and 9 are exemplarily referred to. For example, in FIG. 1, a depth map of a cylinder is shown. The convex portion of the cylinder corresponds to the closest portion when the cylinder is viewed, and is shown in white, and when viewed based on the most convex portion of the cylinder. It is getting farther to the left and right, so it is gradually darkened in a gradient manner. This is used as a basis for the movement of the pixels of the original image, for example the application of the displacement map effect, which will be described in detail below.

As used herein, “in white and black gradient” means the difference in height (perspective difference on the viewing side) of each pixel in an image, the color between white and black, that is, the distance away from the camera. It means that the difference is expressed.

According to the present invention, a method of converting a 2D image into a 3D stereoscopic image, in which a highlight portion that receives the most light is white, and a portion that does not receive light is displayed in white and a black gradient method in the 2D image. Forming an illumination layer; Forming a depth map layer in which the portion closest to the viewing side is displayed in white, and the portion furthest from the viewing side is in black and displayed in a gradient manner of white and black; Based on the white and black gradation values of the depth map, the pixels of the image for the right eye image are moved to the left, and the pixels of the image for the left eye image are to the right, where the white portion is previously It relates to a method of converting a 2D image into a 3D stereoscopic image, which is moved to the largest range of the predetermined range, and the black part is moved to the smallest range of the predetermined range.

In addition, in order to further increase the stereoscopic effect, the present invention may provide a minimum blur effect to a portion closest to a focal length to the left and right eye images, and include a maximum blur effect to a portion farthest from the focal length. .

1 is a depth map image corresponding to a cylindrical 2D image.
FIG. 2 is an image in which a light effect is applied to a cylindrical 2D image and is divided into pixels.
FIG. 3 is an image divided before the lighting effect is applied to a cylindrical 2D image, and is divided into pixels.
4 is an image in which the image of FIG. 2 is arranged side by side.
FIG. 5 is an image in which pixels of an image are shifted by applying a displacement map effect based on the depth map of FIG. 1.
6 to 9 are images showing a depth map generation process by a mask.
10 is a view for explaining the principle that a stereoscopic image is projected onto a flat film to form a 2D image.
11 is a view showing a state of seeing an object with both eyes.
12 is a view for explaining the principle of making the right eye and left eye images.
13 is a flowchart illustrating a method of converting a 2D animation of the present invention into a 3D animation.

As referred to in Figure 13, the method of converting the schematic 2D animation of the present invention to 3D animation is as follows. Extract target images from original image. Keep the size of the old film (e.g., 640 x 480) larger than HD (e.g., 3840 x 2160) while keeping the resolution (e.g., 640 x 480). The extracted and resized image is printed directly on the drawing paper, and the pencil lines and color portions on the printed image are corrected. The corrected image is scanned at high resolution using a scanner. Since 3D TVs are based on FULL HD, they scan at the appropriate resolution, such as 1920 x 1080. The scanned image file is loaded into a computer image editing program such as Photoshop or After Effects.Nuke.Cinema 4D to insert image lighting effects using a separate layer. This will be explained in detail below.

Meanwhile, a depth script and a depth map are generated based on the image file scanned at the high resolution. The depth map will be described in detail below. Based on the created depth map, a displacement map effect is generated to generate left and right eye images, which will be described in detail below. The generated left and right eye images are corrected, and a stereoscopic image is subsequently generated and stored as a 3D animation image.

Below is an illumination layer of the present invention; Depth map layer; And a representative example of implementing pixel movement by way of example.

1. Insert lighting effect (lighting) Layer  produce)

A feature of the present invention is that it adds an illumination effect using a gradation to implement a three-dimensional effect on a 2D image. In the conversion of 3D animations from conventional 2D animations, we did not think about inserting lighting effects, but simply shadowed them. The inventors have found that the insertion of the lighting effect can be properly implemented three-dimensional. Thus, the present inventors applied the effect of lighting to further increase the three-dimensional effect on the 2D animation. The method of applying this lighting effect is implemented by the method introduced in detail below. In this way, the light effect was applied to the solid surface, which is the characteristic of 2D animation, to change the color. Such addition of the lighting effect of the method using the gradation of the present invention allows the three-dimensional (volume) feeling to be properly alive. However, if the effect of the lighting is too strong there is a possibility that the original feeling may change, preferably to give a lighting effect of 50% or less to take full advantage of the original feeling. 2 is an exemplary diagram to add a lighting effect of the method using a gradient.

1.1. How to put lighting effects (lighting by brush)

Referring to FIG. 2, reference may be made to an example in which a lighting effect is added. For image sources that have finished layer separation or finished resolution enhancement, create one or two layers in Photoshop and highlight them with a brush tool to apply white overlay effects and darken shadows. The part (the “shadow part” in this specification does not mean the shadow of the actual image, but is a dark part of the part opposite to the direction of light coming in), and the black overlay effect is applied and the highlight part The part of the shadow that becomes darker and darker is selected by considering the direction of illumination from the target image. The boundary between the highlight and the shadow part has a feather effect, making the gradation smooth and natural. Wherein the feather value is preferably 4-6. Transparency was applied to the layer so that this effect was applied quietly. Preferably the transparency value is 50%. Referring to FIG. 2, it can be seen that the left side represents a bright highlight portion and the shadow portion gradually darkens toward the right side. In other words, the original image of the cylinder is illuminated from the left, so the left part of the cylinder corresponds to the bright part (highlight part) and the right part corresponds to the dark part (shadow part), and Fig. 2 is a cylindrical 2D The lighting effect is expressed in the image.

1.2. How to add lighting effects (creating lights by masks)

The brush effect described above was created using the mask and pen tool which is the basic function of the video effects program (After effects. Nuke. Cinema 4D, etc.). This method works even if the layers of the original source are not separated. I created a solid layer over the original source image and masked the highlights or shadows with the Pen tool. At this point, the mask line is precisely cut with a knife, and the edge outline is smoothed and natural as described above. For example, it was a feather value of 4-6. 6-9 may be referred to.

2. Depth map  to make

6 to 9 are referred to for the depth map operation. In the present invention, the portion near the viewing side is expressed in white and the portion farther in black. To implement this, a separate black layer was created (FIG. 6). A white layer was made on the near part of the object on it (Fig. 7). After the white layer was made, contours representing depth were made in a frame unit using a pen tool. As described above, by adjusting the feather value of the mask (feather value 4-6) to create a natural and smooth gradation (Figs. 8 and 9). FIG. 1 is an example of a depth map made in this manner, and FIG. 1 shows a depth map for a cylinder as already described.

3. Depth map  How to create 3D effect

Normally, a 2D picture is not shown in three dimensions because it reflects a three-dimensional plane on a two-dimensional plane (film in FIG. 10) as shown in FIG. The reason why a person feels things three-dimensionally is because they observe things with the eyes of two people (Fig. 11). Therefore, to express a picture in three dimensions requires two pictures viewed from two viewpoints, such as the human eye. If there is an original image, as shown in FIG. 12, creating a left eye image is an image of turning the original to the right, and a right eye image of an image of turning the original to the left. But because you can't actually turn it, you have to move the pixels left and right. At this time, since the protruding part of the object closest to both eyes is close to the eye from the side of looking at the object, the moving distance should be the largest when viewed only with the left eye and the right eye, respectively. will be. Therefore, when making a right eye image viewed from the right eye and a left eye image viewed from the left eye, the part closest to the eye should be moved more and the part farthest is less. As mentioned earlier, the white part of the depth map is close to the eye and the black part is far from the eye, so when creating the right and left eye images using the depth map, the degree of movement of pixels is determined by using the created depth map. By moving a lot and moving the dark part less, a three-dimensional effect can be realized (applied with the display map effect). Generating such left and right eye images is described further below.

To move the pixels of the original image in this way, we used Photoshop or a video effects program. The generated depth map was used to give a displacement map effect which is an effect tool. As described above, the position of the pixels is appropriately shifted using the depth map by using white to black gradation color data by using the displacement map effect, thereby giving a stereoscopic effect.

4. Create left and right video

Referring to FIG. 4, the original images to which the lighting effect is applied are arranged side by side. The left eye image was placed on the right side and the right eye image on the left side. The depth map (depth map of FIG. 1) of the image is placed under the image. Thereafter, displacement maps were applied to the original images arranged left and right. The displacement map affected by the depth map moved the original pixels. The bright pixels in the depth map were shifted to the left by giving a positive value to the left eye image. Bright pixels in the depth map were shifted to the right by giving a negative value to the original to be the left eye image on the right. That is, as shown in FIG. 5, pixels of the right eye image move to the left and pixels of the left eye image move to the right. As described above, in the right eye image and the left eye image, the direction in which the light color moves and the direction in which the dark color moves is not different.

5. Lens Blur ( Lenz Blur )

As a method for maximizing the sense of distance (depth), the present invention considered providing additional lens blur. The closer to the focal length, the sharper and farther away from the focal point, the greater the blur value, and the greater the blurring of the three-dimensional effect generated by the pixel movement due to the depth map.

When the lens blur is applied to the left and right eye images, the bright part receives the most blur and the dark part receives the blur due to the depth map. If you check Invert in the option of lens blur, on the contrary, the blur value of the bright part is the smallest and the blur value of the dark part is increased.

In addition to the effects for displaying the three-dimensional effect described above, it is also possible to maximize the sense of distance (depth) by the air (air) expression. In general, the distance between objects is expressed by simply spreading them to the left and right. One of the methods used to draw an animated background is to express an organism affected by light in the air. For example, there is sunlight that leaks through a window. You can maximize the sense of distance by adding a layer of organisms affected by light. The expression of the layer is invisible, but the difference in depth with and without the layer is visible.

Although described above by way of example only, it will be understood by those skilled in the art that the present invention may be variously modified and changed without departing from the spirit and spirit of the present invention as set forth in the claims below. will be.

Claims (2)

As a method of converting a 2D image into a 3D stereoscopic image,
In a 2D image, the highlights that receive the most light are white, the areas that don't receive light are black, forming a light layer that is displayed in a gradient between white and black,
The depth closest to the viewing side is shown in white and the farthest away from the viewing side is in black, forming a depth map layer that is displayed in a gradient between white and black.
Based on the white and black gradation values of the depth map, the image for the right eye image is shifted to the left, and the image for the left eye image is shifted to the right, wherein in the movement, the white portion is the most out of a predetermined range. Is moved to a larger range, and the black part is moved to the smallest of the predefined ranges,
How to convert 2D image into 3D stereoscopic image.
The method of claim 1,
In the left and right eye images, a minimum blur effect is given to a portion closest to a focal length, and a maximum blur effect is given to a portion farthest from a focal length.
How to convert 2D image into 3D stereoscopic image.

Images