JP2008090750A - Method and device for generating three-dimensional image - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To more easily generate a more natural three-dimensional image consisting of a number of object images. <P>SOLUTION: A method for generating a three-dimensional image constituted by a plurality of object images included in a basic image is employed to generate a grayscale image the depth direction of which is indicated by grayscale a position of the respective object images, to form a respective image layer by a pixel group corresponding to the respective grayscale level indicated by the basic image, and to move the respective image layers so as to get parallax according to the grayscale level, thereby forming a three-dimensional image. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a method and apparatus for creating a three-dimensional image having a plurality of object images as constituent elements.

  Conventionally, three-dimensional images (stereoscopic images) are created by various methods. Traditionally, a subject such as a landscape or a person is photographed a plurality of times while moving the camera by a certain distance in the horizontal direction, thereby creating a series of images. Using these images, a three-dimensional photograph can be obtained by exposing and developing each image and the printing material with different positions in the parallel direction by a certain distance.

  In addition, a special camera is also used in which a plurality of cameras are arranged linearly and a plurality of images can be obtained simultaneously by one shooting. However, when a special camera is used, since the pitch between the cameras cannot be varied, the range in which the stereoscopic effect can be obtained is narrow. In addition, since it is necessary to shoot in a place where the subject exists, a lot of time and labor are required.

Therefore, using object images that are a plurality of planar art materials, moving each of these object images to create different composite images with their mutual positions, and using the created composite image and lenticular screen, three-dimensional It has been proposed to create an image (Patent Document 1). As a result, a three-dimensional image can be arbitrarily created from a planar image regardless of the type of subject. Further, a method for improving this and creating a more natural three-dimensional image has been proposed (Patent Document 2).
JP-A-2-293733 JP-A-8-147495

  However, in the case of the methods of Patent Documents 1 and 2 described above, the extent to which the object image can be moved when moving the object image can be obtained on the spot according to the intuition and experience of the creator. It was decided.

  Therefore, there is not much problem if the object image is a simple three-dimensional image of about 2 to 3, but when the number of object images increases, it takes a great deal of time and labor, and it is natural satisfaction. It was difficult to obtain a smooth 3D image.

  The present invention has been made in view of the above-described problems, and provides a method and apparatus that can more easily create a three-dimensional image including a plurality of object images and can create a more natural three-dimensional image. For the purpose.

  A method according to the present invention is a method of creating a three-dimensional image having a plurality of object images included in a basic image as constituent elements, and a grayscale image indicating a position in the depth direction of each of the plurality of object images by grayscale. Create and form each image layer by pixel groups corresponding to each shade level shown in the basic image, and move each image layer so that the parallax according to the shade level is obtained. A three-dimensional image is formed.

  Further, a first step of creating a gray image indicating the position of each of the plurality of object images in the depth direction by shading and using this as a depth designation image, and corresponding to the respective levels of shading shown in the depth designation image A second step of forming an image layer configured by the pixel group and displaying an original image of the pixel group, and moving each image layer so as to obtain a parallax corresponding to each gray level A third step of acquiring a plurality of parallax images by combining and a fourth step of combining the plurality of parallax images to form a three-dimensional image.

  Preferably, in the third step, when one object image extends over a plurality of image layers, the parallax image is obtained by performing interpolation correction so that the shape of the object image over the plurality of image layers is maintained. To do.

  In the fourth step, a plurality of parallax images are combined as a compressed image observed through a lenticular lens.

  Further, in the fourth step, a plurality of parallax images are synthesized by drawing as images observed by the right eye or the left eye through stereo glasses.

  In the fourth step, a plurality of parallax images are synthesized by drawing as images observed by the right eye or the left eye through the grid barrier.

  In addition, for a basic image including a plurality of object images, a grayscale image that indicates the position in the depth direction of each pixel that constitutes the basic image by grayscale is created, and pixels corresponding to the respective grayscale levels shown in the basic image Are extracted to form an image layer composed of a pixel group corresponding to each level, and each image layer is moved so as to obtain a parallax corresponding to the level of light and shade to form a three-dimensional image.

  An apparatus according to the present invention is a three-dimensional image creation apparatus including a plurality of object images included in a basic image as constituent elements, and a grayscale image indicating a position in the depth direction of each of the plurality of object images by grayscale. A means for creating, a means for forming each image layer by a pixel group corresponding to each shade level shown in the basic image, and a parallax corresponding to each shade level can be obtained for each image layer. And a means for forming a three-dimensional image.

  1 is a diagram showing an example of the basic image KG1, FIG. 2 is a diagram showing positions in the depth direction of the object images JG1 to JG1-3 included in the basic image KG1, and FIG. 3 is a diagram showing positions in the depth direction, gray levels NL, and movement amounts. FIG. 4 is a diagram illustrating an example of a grayscale image NG1 for the object images JG1 to JG1, FIG. 5 is a diagram illustrating the image layers GL1 to 3, and FIG. 6 is a diagram illustrating the spatial relationship between the image layers GL1 to GL3. FIG. 7 is a diagram showing the movement amount ML of each of the image layers GL1 to GL3.

  In FIG. 1, the basic image KG1 includes three object images JG1 to JG1. That is, the basic image KG1 is composed of three objects: a square object image JG1, a circular object image JG2, and a triangular object image JG3.

  Hereinafter, a method for creating the three-dimensional image DG from the basic image KG1 will be described.

  First, the positions in the depth direction of these three object images JG1 to JG1 are determined. The position in the depth direction is a position that can be seen in the depth direction when the basic image KG is viewed as the three-dimensional image DG. Normally, the position in the depth direction is between an infinite distance and the closest position. However, it can also be between any finite two distances.

  In any case, the distance range from the farthest position in the depth direction to the closest position is expressed by the gradation level KL. For example, it is expressed by a level of 256 gradations, and the position in the depth direction is assigned somewhere between 0 and 255 gradation levels KL. In this case, the gradation level KL of “0” is the farthest position, and the gradation level KL of “255” is the closest position.

  Further, in order to visually represent the position in the depth direction, the gradation level KL of 0 to 255 is assumed to be the light and shade level (grayscale) NL, and the position in the depth direction assigned to the object images JG1 to JG1 is represented by the light and shade. It is expressed as a grayscale image by level NL. That is, in creating the three-dimensional image DG, the depth of each object image JG is expressed in gray scale.

  In the example shown in FIG. 2, the gradation levels KL of “51”, “153”, and “230” are assigned to the object images JG1 to JG1. That is, among the three object images JG, the object image JG1 is the farthest, the object image JG3 is the closest, and the object image JG2 is in the middle.

  Moreover, each of the object images JG1 to JG1 has the same gradation level KL as a whole, and therefore has no depth in itself.

  As shown in FIG. 3, the gradation level KL and the light and shade level NL (gray scale GS) correspond to each other, and the movement amount ML is determined corresponding to the gradation level KL from 0 to 255. In FIG. 3, the movement amount ML is indicated by the width of the triangular figure at the horizontal position corresponding to each gradation level KL.

  FIG. 4 shows a grayscale image NG1 represented by a grayscale level NL corresponding to the gradation level KL shown in FIG. In this grayscale image NG1, the object image JG1 is represented by a density close to black, the object image JG3 is represented by a density close to pure white, and the object image JG2 is represented by an intermediate density. This grayscale image NG1 is also a depth designation image YG1.

  That is, in the depth designation image YG1, the shade level NL of each object image JG1 to 3 indicates the position of the object image JG in the depth direction. Therefore, for the user, the position in the depth direction of each object image JG can be intuitively recognized by the shading. That is, by assigning shades to the object images JG1 to JG1 to the basic image KG1 shown in FIG. 1, the positions in the depth direction of the object images JG1 to JG1 can be designated.

  Further, the position in the depth direction can be easily corrected by correcting the shading of the object image JG.

  In creating the grayscale image NG1 or the depth designation image YG1, the actual object is photographed with a camera and photographed from a plurality of different positions by changing the photographing position, and the amount of movement of an object reflected from the obtained plurality of images is determined. The grayscale image NG1 may be created based on the measured movement amount. In this way, it is possible to specify the position in the depth direction more accurately.

  Next, based on the depth designation image YG1, image layers GL1 to GL3 shown in FIG. 5 are formed. That is, the image layers GL1 to GL3 are configured by pixel groups corresponding to the respective gray levels NL shown in the depth designation image YG1, and the original images of the pixel groups are displayed.

  That is, in the depth designation image YG1 shown in FIG. 4, since the object images JG1, JG2, and JG3 have uniform gray levels NL, the pixel groups that constitute the object images JG1, 2, and 3, respectively, Layers GL1 to GL3 are configured.

  In each of the image layers GL1 to GL3, portions other than the object image JG are transparent. Therefore, the image layers GL1 to GL3 do not necessarily have to be in the form of an image frame of a certain size. For example, the object image JG may be designated in the depth direction.

  Now, the image layer GL1 shown in FIG. 5A includes only the object image JG1 in the basic image KG1. The image layer GL2 shown in FIG. 5B includes only the object image JG2 in the basic image KG1. The image layer GL3 shown in FIG. 5C includes only the object image JG3 in the basic image KG1.

  Thus, the image layers GL1 to GL3 can be created based on the depth designation image YG1. When each object image JG has a uniform gray level NL, a number of image layers GL equal to the number of object images JG are created.

  Since these image layers GL1 to GL3 are each composed of a pixel group of one gray level NL, they are associated with the gradation level KL according to the gray level NL of the pixel group included in each of the image layers GL1 to GL3. Can do.

  Note that the basic image KG, the grayscale image NG, the depth designation image YG, the image layer GL, the object image JG, and the like can be generated and stored by digital image processing, and each image is obtained from a pixel or a pixel group. Can be configured.

  FIG. 6 shows a state in which the three image layers GL1 to GL3 are arranged at the gradation level KL corresponding to the respective light and shade levels NL. The positions in the depth direction of the object images JG1 to JG3 in this state are based on the positions designated by the depth designation image YG1.

  Then, such image layers GL1 to GL3 are moved and combined so as to obtain parallax according to the respective light and shade levels NL, thereby obtaining a plurality of parallax images UG. The image layer GL can be moved by shifting the arrangement position by the movement amount ML.

  Now, since the relationship between the light and shade level NL and the movement amount ML is shown in FIG. 3, the movement amount ML can be derived from the light and shade level NL. In FIG. 3, the movement amount ML is in the range of 0 to MLmax, and is determined in proportion to the light and shade level NL within the range. In the present embodiment, the movement amount ML is proportional to the gray level NL. However, the movement amount ML is not simply proportional, but may be expressed by a predetermined linear function, quadratic function, or other function. Good.

  The parallax image UG1 shown in FIG. 7A is the same as the basic image KG1 shown in FIG. In the parallax image UG1, the positions of the object images JG1 to JG1 are indicated by distances L1 to L3 from the left end, respectively.

  In the parallax image UG2 shown in FIG. 7B, the positions of the object images JG1 to JG1 to the right of the parallax image UG1 shown in FIG. Has moved. These a, b, and c are the movement amounts ML of the object images JG1 to JG1. The ratio of “a”, “b”, and “c” that is the movement amount ML is equal to the ratio of the light and shade levels NL of the object images JG1 to JG1.

  As described above, the movement amount ML is a horizontal distance difference between the positions of the object image JG between the two parallax images UG when the parallax image UG is created. Or it is a parameter equivalent to a distance difference. Depending on the magnitude of the movement amount ML, parallax is generated when the user views the two parallax images UG with the left and right eyes, and a stereoscopic effect can be obtained. When viewed with both eyes, the object image JG with the movement amount ML of 0 is viewed in the front-rear direction as the actual position of the object image JG, and the object image JG with the positive movement amount ML is more than the actual position. The object image JG having a negative movement amount ML moves backward from the actual position.

  Therefore, as for the relationship between the position in the depth direction and the movement amount ML, the closer the position in the depth direction, the larger the movement amount ML in the positive direction. As described above, if the movement amount ML of the object image JG existing in the middle in the depth direction is set to 0, the object image JG existing behind the object image JG has a negative movement amount ML, and the object image JG existing before the object image JG. The movement amount ML becomes positive. If the movement amount ML of the innermost object image JG is set to 0, the movement amount ML of the other object image JG is positive, and the movement amount ML is larger as the object image JG is closer.

  Usually, the most important object image JG in the basic image KG is set as a key image, and the movement amount ML is set to “0” in order to make the key image appear clear. Then, the movement amount ML of the object image JG arranged behind the key image is negative, and the movement amount ML of the object image JG arranged in front of the key image is positive.

  In creating the parallax image UG, the object image JG, that is, the image layer GL is moved in this example. However, the image layers GL1 to GL3 are arranged corresponding to the positions in the depth direction as shown in FIG. 6 and are photographed by the camera from the front, and the positions of the cameras are moved and made different. May be. That is, a plurality of parallax images UG can be acquired by photographing the image layers GL1 to GL3 from a plurality of different positions, and the movement amount ML is used as the displacement amount of the photographing position of each camera in that case. be able to.

  In the parallax image UG3 shown in FIG. 7C, the positions of the respective object images JG1 to JG1 to the parallax image UG1 shown in FIG. Has moved. That is, the positions of the object images JG1 to JG1 are further moved to the right by a, b, and c with respect to the parallax image UG2 illustrated in FIG.

In order to actually obtain the three-dimensional image DG from these parallax images UG1 to UG3, the following various methods can be employed.
(1) The plurality of parallax images UG <b> 1 to UG <b> 3 are combined as a compressed image that is observed through a lenticular lens.
(2) The plurality of parallax images UG <b> 1 to UG <b> 3 are synthesized by drawing as images observed with the right eye or the left eye through stereo glasses.
(3) The plurality of parallax images UG <b> 1 to UG <b> 3 are combined by drawing as images that are observed by the right eye or the left eye through the grid barrier.

  It is also possible to use methods other than these methods. The methods (1) to (3) are known per se and can be applied as necessary.

  For example, with respect to (1) above, using the parallax images UG <b> 1 to UG <b> 3, exposure is performed on a photographic film for stereoscopic photography having a lenticular lens to burn the image to create a stereoscopic photograph. When the object image JG which is a key when creating the parallax images UG1 to UG3 is moved in the printing, the amount corresponding to the moving distance of the object image JG is obtained when the printing film is exposed. Only move in the reverse direction on the photographic film. By doing so, the object image JG which is a key in the stereoscopic photograph is fixed, and the stereoscopic image appears clearly. For details, reference can be made to Patent Documents 1 and 2 described above.

  In FIG. 7, the positions of the object images JG1 to JG1 are indicated by the distances L1 to L3 from the left end of the image layers GL1 to GL1, but at the distance from the right end or the center of the image layers GL1 to GL3. You may show by the distance from the centerline located.

  Also, although three parallax images UG are shown, only two or four or more may be used depending on the type of method for obtaining the three-dimensional image DG. In the above example, the basic image KG1 includes three object images JG, but may include two object images JG or four or more object images JG.

  In the above example, each object image JG is contained in one image layer GL, but one object image JG may extend over a plurality of image layers GL. In that case, interpolation correction is performed so that the shape of the object image JG across the plurality of image layers GL is maintained, and a parallax image UG is acquired.

  That is, for example, instead of taking one object image JG in a plane and arranging it at a position in one depth direction, one object image JG itself is arranged obliquely or along the depth direction so that it has a depth. May be. In that case, although the image layer GL is not continuous but discrete even though the object image JG itself is a continuum, the portion between the image layers GL of the object image JG that spans multiple image layers GL Since there is no image, the object image JG cannot be accurately reproduced because the portion is cut or thinned when viewed from the front. In order to avoid this, interpolation correction is performed so that the shape of the object image JG over the image layer GL is maintained, and a parallax image UG subjected to the interpolation correction is acquired.

  In the example described above, a simple image composed of three object images JG1 to JG1 to 3 is shown as the basic image KG, but any image may be used as the basic image KG. Various objects such as photographs, paintings, printed materials, illustrations, graphics, and characters can be used as the object image JG.

  For example, when the basic image is an image in which water splash is applied around a can for beverages and polka dots or splashes are splashed, the object image JG is a can, polka dots, water spray, or the like. In this case, the can is a key image, for example, arranged at the center position in the depth direction, and the can is arranged obliquely along the depth direction in order to give a stereoscopic effect to the can itself.

  In the embodiment shown above, the position in the depth direction of each of the object images JG1 to JG1 is designated by the grayscale image NG1 as shown in FIG. The position in the depth direction of the object image JG shown in FIG. 4 is a position corresponding to the gradation level KL as shown in FIG. In this case, if the object image JG2 is a sphere, for example, the object image JG2 in FIG. 2 may be represented by a circle instead of a horizontal line. In that case, the circle may be provided with shading according to the gradation level KL. Further, when the background image is arranged at the innermost position, the gradation level KL of the background image is 0, so that it is represented in black. Although the gradation level KL of “0” is the farthest distance, it may be the closest position. The gradation level KL of “0” is black, but conversely, the gradation level KL of “0” may be white.

  In the above embodiment, the gradation level KL is represented by 256 gradations, that is, 8 bits, but other values may be used. For example, there may be 8 gradations, 16 gradations, 32 gradations, 64 gradations, 128 gradations, 512 gradations, or more.

  Next, a method for creating a three-dimensional image DG from the basic image KG will be described with reference to a flowchart.

  FIG. 8 is a flowchart showing a flow of creating the three-dimensional image DG from the basic image KG.

  In FIG. 8, first, a basic image KG is prepared (# 11). The basic image KG is determined by a plurality of object images JG and their arrangement positions. Based on the basic image KG, a grayscale image NG indicating the depth direction positions of the plurality of object images is created (# 12). The created grayscale image NG is set as a depth designation image YG.

  An image layer GL configured by pixel groups corresponding to the respective shade levels shown in the depth designation image YG and displaying the original image of the pixel group is formed (# 13).

  A plurality of parallax images UG are acquired by moving and synthesizing each image layer GL so as to obtain parallax according to the respective shade levels (# 14). A plurality of parallax images UG are combined to form a three-dimensional image DG (# 15).

  FIG. 9 is a block diagram showing a configuration of the creating apparatus 1 that creates a three-dimensional image DG from a basic image KG.

  In FIG. 9, the creation device 1 includes a grayscale image creation unit 21, an image layer creation unit 22, a parallax image creation unit 23, and a 3D image creation unit 24.

  The grayscale image creation unit 21 creates a grayscale image that indicates the position in the depth direction of each of the plurality of object images JG included in the basic image KG by the grayscale.

  The image layer creation unit 22 forms each image layer GL with a pixel group corresponding to each level of shading shown in the basic image KG.

  The parallax image creation unit 23 creates the parallax image UG by moving each image layer GL so that the parallax according to the level of the shade is obtained.

  Based on the parallax image UG, the three-dimensional image creation unit 24 combines the three-dimensional image DG as an image that can be observed through a lenticular lens, stereo glasses, a grid barrier, or the like as described above. Create

  Such a creation apparatus 1 can use a computer system for part or all of it. An example of such a creation apparatus 1 will be described.

  FIG. 10 is a block diagram showing a hardware configuration of the creation apparatus 1.

  In FIG. 10, the creation apparatus 1 includes a processing device 11, an image scanner 12, a magnetic disk device 13, a magneto-optical disk device 14, a display device 15, a printer device 16, a keyboard 17, a mouse 18, and the like.

  The magnetic disk device 13 and the magneto-optical disk device 14 store various object images JG, basic images KG, grayscale images NG, image layers GL, parallax images UG, three-dimensional images DG, and the like. An image read from the image scanner 12, an image obtained by photographing with a video camera or a still camera, an image created by computer graphics, or the like can be used as the object image JG.

  The display device 15 displays an image stored in the magnetic disk device 13 or the magneto-optical disk device 14 and other images and data.

  The printer device 16 prints an image or the like on a film. The film includes paper and photographic paper. The printing by the printer device 16 includes a process of printing an image on a film or the like.

  The keyboard 17 and the mouse 18 are used for inputting commands to the processing device 11, inputting messages, specifying the position of an image on the frame memory or the screen, and the like. In creating the grayscale image NG, the user instructs the grayscale level by operating the input device.

  The processing device 11 performs various processes by executing a program stored in a ROM or the like. The processing device 11 is provided with a plurality of frame memories, a memory for storing image data, and the like.

  With such a hardware configuration, each unit shown in FIG. 9 is functionally formed in terms of software or hardware.

  According to the three-dimensional image creation method using the creation apparatus 1 described above, the position in the depth direction of the various object images JG in the basic image KG is represented by the grayscale image NG, and the position in the depth direction is indicated by the shade. A three-dimensional image DG composed of a large number of object images JG can be created more easily, and a more natural three-dimensional image DG can be created.

  In the above-described embodiment, the entire creation device 1 or the configuration of each part, the content, configuration, shape, positional relationship, dimensions, number, material, and the like of each image can be appropriately changed in accordance with the spirit of the present invention. it can.

It is a figure which shows the example of a basic image. It is a figure which shows the position of the depth direction of each object image contained in a basic image. It is a figure which shows the relationship between the position of a depth direction, a shading level, and movement amount. It is a figure which shows the example of the grayscale image about an object image. It is a figure which shows an image layer. It is a figure which shows the state which has arrange | positioned the image layer spatially. It is a figure which shows the moving amount | distance of each image layer. It is a flowchart which shows the flow of preparation of a three-dimensional image from a basic image. It is a block diagram which shows the structure of the production apparatus which produces a three-dimensional image from a basic image. It is a block diagram which shows the hardware constitutions of a production apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Creation apparatus 11 Processing apparatus 21 Gray image creation part 22 Image layer creation part 23 Parallax image creation part 24 Three-dimensional image creation part JG Object image KL Grain level NL Gray level ML Movement amount KG Basic image NG Gray image YG Depth designation Image GL Image layer UG Parallax image DG 3D image

Claims (8)

A method for creating a three-dimensional image having a plurality of object images included in a basic image as constituent elements,
Creating a grayscale image that indicates the position in the depth direction of each of the plurality of object images by grayscale;
Each image layer is formed by a pixel group corresponding to each level of shading shown in the basic image,
Each image layer is moved so as to obtain a parallax according to the level of the shading to form a three-dimensional image.
A method of creating a three-dimensional image characterized by the above. A method for creating a three-dimensional image having a plurality of object images as constituent elements,
A first step of creating a grayscale image indicating the position in the depth direction of each of the plurality of object images by grayscale and using this as a depth designation image;
A second step of forming an image layer configured by pixel groups corresponding to respective levels of shading shown in the depth designation image and displaying an original image of the pixel group;
A third step of acquiring a plurality of parallax images by moving and synthesizing each image layer so as to obtain a parallax corresponding to each gray level;
A fourth step of combining the plurality of parallax images to form a three-dimensional image;
A method for creating a three-dimensional image, comprising: In the third step, when one object image extends over a plurality of image layers, the parallax image is obtained by performing interpolation correction so that the shape of the object image over the plurality of image layers is maintained.
The method for creating a three-dimensional image according to claim 2. In the fourth step, a plurality of parallax images are synthesized by arranging them as compressed images observed through a lenticular lens.
The method for creating a three-dimensional image according to claim 2 or 3. In the fourth step, a plurality of parallax images are synthesized by drawing as images observed by the right eye or the left eye through stereo glasses.
The method for creating a three-dimensional image according to claim 2 or 3. In the fourth step, the plurality of parallax images are synthesized by drawing as images observed by the right eye or the left eye through the grid barrier,
The method for creating a three-dimensional image according to claim 2 or 3. For a basic image including a plurality of object images, create a grayscale image indicating the position in the depth direction of each pixel constituting the basic image by grayscale,
Extracting pixels corresponding to each level of shading shown in the basic image, forming an image layer consisting of a pixel group corresponding to each level,
Each image layer is moved so as to obtain a parallax according to the level of the shading to form a three-dimensional image.
A method of creating a three-dimensional image characterized by the above. A three-dimensional image creation device having a plurality of object images included in a basic image as constituent elements,
Means for creating a grayscale image indicating the position in the depth direction of each of the plurality of object images by grayscale;
Means for forming each image layer by a pixel group corresponding to each level of shading shown in the basic image;
Means for moving the respective image layers so as to obtain a parallax according to the level of the light and shade to form a three-dimensional image;
An apparatus for creating a three-dimensional image, comprising: