JPH0421187A - Stereoscopic image display device - Google Patents

Abstract

PURPOSE:To attain stereoscopic observation from an image by forming right and left images by texture mapping processing using three-dimensional shape information. CONSTITUTION:Two projection converting parts 3, 4 in a stereoscopic image forming part 1 are controlled by right and left visual points E, D obtained by a visual point calculating part 2 and two two-dimensional projection models F, F are formed from the three-dimensional structure information such as a polygon by respective projection converting parts 3, 4. Respective texture mapping parts 5, 6 execute the texture mapping processing of the models F and corresponding information to an original image B and form stereoscopic images H, I and a synthetic image output J through a signal switch 7 is displayed on an image display and the displayed contents are observed through shutter spectacles 9 controlled by a switch signal K outputted from the switch 7. In the above constitution, a stereoscopic image based upon a plane image can be effectively observed by a TV camera without using two TV cameras or the like.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a stereoscopic image display device for image processing.

BACKGROUND OF THE INVENTION In recent years, three-dimensional image display devices have become more important than TV telephones, computer display devices, etc. as display devices that provide a more realistic feeling.

An example of the stereoscopic image display device described above will be described below with reference to the drawings.

FIG. 5 shows the overall configuration of a conventional stereoscopic image display device. The entire system is composed of an imaging system 1 and a display system 2. In the imaging system 1, the subject 3 is imaged by cameras 4.5 arranged on the left and right sides so as to produce stereoscopic parallax. Video signals A and B from these cameras 4.5 are supplied to a signal switch 6 of the display system 2. A signal switch 6 switches between video signals A and B using an appropriate synchronizing signal and supplies the signal as a composite video signal C to an image display 7. Further, the signal switch 6 supplies the shutter glasses 8 with a left/right switching signal synchronized with the switching of the video signals A and B. The shutter glasses 8 close the left and right shutters in synchronization with the left and right switching signal. This allows the video signals A and B to be viewed separately by the left and right eyes, and stereoscopic viewing is possible due to the parallax between the left and right eyes.

Note that the video signal A is the output from the imaging system 1.

The imaging system 1 and the display system 2 can be separated by temporarily recording the video signals A and B in the image recording device and inputting the video signals A and B reproduced from the image recording device to the display system 2.

Furthermore, by transmitting the video signals A and B through a communication line, stereoscopic image communication is performed at a remote location.

Problems to be Solved by the Invention However, the conventional three-dimensional image display device as described above has the following problems.

(b) In the above configuration, two video signals A,
Since B is required, there is a problem in that the recording capacity of the image recording device and the transmission amount of image communication are twice as large when performing stereoscopic viewing as compared to ζ in the case of a two-dimensional image.

(b) In the above configuration, the stereoscopic effect of the displayed image is determined only by the position of the camera 4.5, so the camera 4.
．． If you don't keep the position 5 accurate, you won't get a three-dimensional effect. Further, there is a problem in that only a three-dimensional effect can be obtained that is determined in advance by the position of the camera 4.5, and operations such as emphasizing the visual sensation using the display system 2 are not possible.

In view of the above problems, the present invention provides a stereoscopic image display device that can generate a stereoscopic image from a single original image and three-dimensional structure information.

Means for Solving the Problems The present invention includes an input unit that receives an original image, three-dimensional structure information of a display target, and correspondence information that associates the two, and an input unit that receives arbitrary reinforcement from the three-dimensional structure information input by the input unit. a projection conversion unit that generates two two-dimensional projection models viewed from an interval that produces a stereoscopic parallax; and a projection conversion unit that maps the original image to a two-dimensional projection model using the correspondence information to generate input images for the left and right eyes. This is a three-dimensional image display device characterized by comprising a texture mapping section that performs the image processing, and a display section that displays two images obtained by the texture mapping section to the left and right eyes, respectively.

According to the above-described configuration, the present invention generates a three-dimensional shape model from one original image and three-dimensional structural information and correspondence information, which have a smaller amount of data than the image, without using two images.
The positions of the left and right viewpoints are determined to produce a three-dimensional effect of arbitrary strength, a projection model as seen from the left and right viewpoints of the 3D shape model is generated, and each coordinate point of the left and right projection model is From here, the projection model is converted into two left and right images, and by giving the two images separately to the left and right eyes, a three-dimensional effect of arbitrary strength can be obtained. Display a certain stereoscopic image.

EXAMPLE An example of the present invention will be described below with reference to the drawings. FIG. 1 shows the overall configuration of a three-dimensional image display device according to an embodiment of the present invention.

In the figure, reference numeral 1 denotes a stereoscopic image generation system that generates left and right images G and H for stereoscopic viewing from three-dimensional structure information A, original image B, and correspondence information C that associates the two. The stereoscopic image generation system 1 includes a viewpoint calculation unit 2 that calculates the left and right positions of the viewpoint that produces stereoscopic vision, E, and 3D structure information A.
A projection conversion unit 3.4 converts the two-dimensional projection model F and G into two-dimensional projection models F and G when viewed from the viewpoint position E#), and the two-dimensional projection model F.

A texture mapping unit 5.6 inputs G, correspondence information C, and input original image B, and performs texture mapping on the two-dimensional projection models F, G to generate left and right stereoscopic images ti, I. . Note that 7 is a signal switch, J is a composite video output, 8 is an image display, K is a left/right switching signal, and 9 is shutter glasses, which are the same as those in the conventional example.

The operation of the stereoscopic image display device configured as described above will be described below.

First, the input three-dimensional structure information A describes the three-dimensional shape of the object to be displayed three-dimensionally in the original image B using a surface model. A surface model is a well-known three-dimensional shape description method in the field of computer graphics, and is an approximate representation of the surface of a three-dimensional shape using polygonal planes (hereinafter referred to as polygons). Next, 3
An example of the dimensional structure information A will be explained using a diagram. FIG. 2 is an example of a three-dimensional structure of a hexahedron. First, from Vl to the top■
8 and number it. The three-dimensional structure information of this hexahedron is
It has three-dimensional coordinate values of vertices as shown in Table 1.

Table 1 Next, number the faces of the hexahedron from Fl to F6. The three-dimensional structure information has a list of each face and its composing vertices. At this time, to distinguish between the front and back of a surface, the vertices are described in counterclockwise order when the surface is viewed from outside the object. This list is shown in Table 2 for the hexahedron shown in FIG.

Table 2 The coordinate data of the above vertices and the list of surface vertices are shown in Figure 1.
This is the content to dimensional structure information.

Next, in the viewpoint calculation unit 2, when the object of the three-dimensional structure information is viewed from a certain position in the three-dimensional space, the positions E of the left and right viewpoints that cause stereoscopic parallax are determined. At this time, the three-dimensional effect can be emphasized by making the position of the viewpoint, the distance between the left and right sides of E, larger than the distance between human eyes.

A projection conversion unit 3.4 performs perspective projection to convert the three-dimensional shape information A of the object into a two-dimensional planar image on a display. Perspective transformation will be explained using diagrams. Third
In the figure, l is a three-dimensional polygon, 2 is a vertex therein having three-dimensional coordinate values, 3 is a projection plane, and 4 is a viewpoint.

The perspective projection image of vertex l seen from viewpoint 4 becomes point 5.

A three-dimensional polygon l is perspectively projected onto a two-dimensional figure 6.

At this time, the order of the vertices of polygon 1 and two-dimensional polygon 6 does not change. In FIG. 1, the projection transformation unit 3.4 perspectively transforms the coordinate values of the vertices of the three-dimensional shape information A, and generates the two-dimensional coordinate values of the two-dimensional projection models F and G. As the information on the vertices forming the two-dimensional polygons of the two-dimensional projection models F and G, the information on the vertices forming the polygon of the three-dimensional shape information A is used as is. These two-dimensional coordinate values and information on the vertices constituting the two-dimensional polygon are input as two-dimensional projection models F and G to the texture mapping section 5.6.

The texture mapping unit 5.6 pastes the original image B onto the two-dimensional projection models F and G through texture mapping processing using the correspondence information C to generate stereoscopic images H and [. The original image B is expressed in the form of a two-dimensional sampled array, and the correspondence information C indicates, for each vertex, the two-dimensional coordinates on the original image B plane to which each vertex of the three-dimensional image information A corresponds. This is what I described. Texture mapping processing will be explained using figures. In FIG. 4, l is an array of stereoscopic images, and 2 is a two-dimensional polygon forming a two-dimensional projection model. Two-dimensional polygon 2 is from vertex Vl ■
Consists of 5. 3 is the original image B in FIG.
4 is a point Wl to w5 on the original image 3 corresponding to the vertices Vl to ■5 of the two-dimensional polygon 2, which are described in the correspondence information in FIG. 1C. Now, as shown in FIG. 3A, in order to generate a stereoscopic image l, the value of the pixel P is determined.

Point M where a straight line passing through pixel P intersects two-dimensional polygon 2,
Find N, and find the ratio s, t that P internally divides the line segment (M, N)
seek. Also, the ratio m that points M and N internally divide the 20 sides of the two-dimensional polygon (VLV2) and (V4, V5), respectively.
1. Calculate m2 and nLn2. Next, the pixel Q of the original image 3 corresponding to the point P is determined from the above internal division ratio.

First, as shown in the same figure (b), the sides (VLV2), (V
4. V5) d corresponding sides (WLW2), (W4, W5)
Find points U and V that internally divide into internal division ratios ml, m2, and nLn2, respectively. Considering a straight line I passing through points U and V, the coordinates of the point that internally divides the line segment (U, V) with internal division ratios s and t are the pixels in original image 3 that correspond to pixel P in stereoscopic image 1. These are the coordinates of Q. The value of pixel Q can be obtained from the original image 3 as the value of pixel P.

However, normally, the calculated coordinates of pixel Q do not completely match the coordinates of the sampled pixel of original image 3, so
The coordinate value of the pixel Q is determined by linear interpolation from several pixels around the determined coordinate value of the pixel Q. In this way, by calculating values for all points inside the two-dimensional polygon 2 through this texture mapping process, a stereoscopic image (first
One two-dimensional polygon in Figures H, I) is generated. 1st
Stereoscopic images H and I are generated by performing texture mapping processing on all the two-dimensional polygons described in the two-dimensional projection models in Figures F and G.

In FIG. 1, the generated left and right stereoscopic images H=I are supplied to a signal switch 7. The signal switching device 7 switches and outputs a stereoscopic image H when the interlaced and NTSC composite video output J is an odd field, and a stereoscopic image I of an even field, and outputs a synchronization signal for switching between the odd and even fields. It is supplied to the shutter glasses 9 as a left/right switching signal. The image display 8 uses the interlace method to display a stereoscopic image H for the left eye when displaying odd fields.
However, during even field display, the stereoscopic images I for the right eye are alternately displayed. The shutter glasses 9 close the shutter for the right eye when the stereoscopic image H for the left eye is displayed, and close the shutter for the left eye when the stereoscopic image I for the right eye is displayed, according to the left-right switching signal K. Close the shutter. As a result, the left-eye and right-eye stereoscopic images H1I are input independently to the left and right eyes, resulting in stereoscopic vision.

As described above, according to this embodiment, a texture mapping method is used from three-dimensional structure information A and correspondence information C, which are numerical data smaller than image data such as one original image B, coordinate data, and vertex number data. By generating a stereoscopic image, it is possible to display a stereoscopic image with less input data compared to a conventional stereoscopic display device that uses two images. Further, according to this embodiment, by arbitrarily setting the positions of the left and right viewpoints in the viewpoint calculation unit 2, the stereoscopic effect of the generated stereoscopic image can be freely changed.

In the embodiment shown in FIG. 1, it is assumed that the three-dimensional structure information A is input from the outside together with the original image B, but if the target of the input original image B is a specific object such as a face image in a video phone, In the case where the three-dimensional structure information A is limited to a general model, the three-dimensional image generation system 1 has information for a general model, and even if the three-dimensional structure information A is associated with the original image B, the correspondence information C is generated. good. In this case, only the original image B is input, and a three-dimensional stereoscopic image can be generated from a two-dimensional image.

Effects of the Invention As described above, the present invention utilizes three-dimensional shape information to achieve stereoscopic viewing, which was conventionally only possible by combining images taken by two cameras installed specifically for stereoscopic viewing. By generating left and right images through texture mapping processing, stereoscopic viewing can be achieved from a single image. This allows us to reduce the amount of data required for stereoscopic display from two images.
It can be reduced to three-dimensional structure information and corresponding information, which is much smaller than a single image and image data, reducing the transmission capacity when communicating with stereoscopic TV phones, and storing data for stereoscopic viewing. The capacity can be reduced, and stereoscopic viewing can be easily performed using existing communication systems and storage devices.

In addition, the strength of the stereoscopic effect, which was conventionally determined by the left and right positions of the camera at the time of image capture, can be changed by arbitrarily determining the position of the viewpoint at the time of image generation. It becomes possible to perform stereoscopic display with a three-dimensional effect, and more effective stereoscopic display becomes possible.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing the overall configuration of an embodiment of the present invention, Fig. 2 is a perspective view showing an example of a three-dimensional structure in the embodiment, and Fig. 3 is an explanation of perspective projection transformation in this embodiment. FIG. 4 is a configuration diagram for explaining the texture mapping process in the same embodiment, and FIG. 5 is a block diagram showing the configuration of a conventional stereoscopic display device. 1... Three-dimensional image generation system, 2... Viewpoint calculation unit, 3.4
. . . Projection conversion section, 5. 6. Texture mapping section, 7. Signal switch, 8. Image display, 9. Shutter glasses. Name of agent Patent attorney Matsu 1) Tadashi Figure 2 Figure 3 (a) Figure (b)

Claims (2)

[Claims] (1) An input unit that receives the original image, three-dimensional structure information of the display target, and correspondence information that associates the two, and an interval that produces a stereoscopic parallax of arbitrary strength from the three-dimensional structure information input by the input unit. a projection conversion unit that generates two viewed two-dimensional projection models; a texture mapping unit that maps the original image to a two-dimensional projection model using the correspondence information to generate input images for the left and right eyes; A stereoscopic image display device comprising: a display section that displays two images obtained by a mapping section to left and right eyes, respectively. (2) A receiving unit is provided to receive one original image, three-dimensional structure information, and correspondence information from the communication system, and a three-dimensional image is created by transmitting one original image, three-dimensional information, and correspondence information from the transmission system. The stereoscopic image display device according to claim 1, wherein the stereoscopic image display device performs communication.