JP2000149062A - Method and device for generating three-dimensional image - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a more natural and effective three-dimensional(3D) image, in addition to conventional perspective concerning 3D images of a computer. SOLUTION: This device is provided with a transparency designating means 14 for setting transparency, which is improved with the reduction in the distance from the viewpoint of each object, to that an object and an image compositing means 15 for compositing the images of respective objects, to which transparency is added by an image generating means and the transparency designating means, into one image by performing hidden surface processing while taking the transparency into consideration, so that more natural and effective 3D images can be provided.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for generating a three-dimensional image, and more particularly, to a method for generating a three-dimensional image based on three-dimensional shape data more naturally. Regarding the device.

[0002]

2. Description of the Related Art A method of constructing a three-dimensional model based on numerical data in a computer and displaying the model in three dimensions is known as V
It is important in R, CAD, and many other application software. As a method of such a stereoscopic display, an image to be presented to each of the left and right eyes is generated by using the positions of the left and right eyes as viewpoints, and the respective images are provided to the left and right eyes, respectively, so that the images are stereoscopically displayed. A stereoscopic method and a virtual one viewpoint, for example, the center point of the left and right eye position is determined instead of the two left and right viewpoints, and a single image viewed from this single viewpoint is generated and provided. And how to do it. Among them, the former method based on stereoscopic vision includes an anaglyph method, a polaroid method, an alternate method, a lenticular screen method, and the like. (For example, see "CG Introduction Seminar" written by Toshihiro Imama, Nikkei BP)

[0003] The most common of the latter single image methods is perspective projection. That is,
This is based on the fact that objects that are near are large and objects that are far away are small. However, in this state, when there are a plurality of objects in the field of view and one object is blocked by another object, the latter object is not displayed. One of methods for preventing such a problem is a method called front face culling. In this method, a clipping plane is placed at a fixed distance from the viewpoint, objects closer to the viewpoint than the clipping plane are excluded from the calculation, and objects immediately before the viewpoint are excluded from the calculation. This can prevent many other objects from being blocked by an object close to the viewpoint.

As a method using a single image, there is a method of displaying an object with different transparency. For example, Japanese Patent Application Laid-Open No. 4-362794 discloses a method in which a front object is displayed translucently and a rear object is seen through. In Japanese Patent Application Laid-Open No. 4-71082, a method of changing the transparency of each portion to display the shape of the cut surface when the target object is cut in a cross section and the shape before and after the cut surface when the cut object is cut is used. Is disclosed.

[0005]

In any of the above-described methods using stereoscopic vision, special hardware such as stereoscopic glasses and a display device with a lenticular lens is required. Also, when a single image is displayed by front face culling, objects close to the clipping plane become completely invisible, and when the clipping plane and the object intersect, the part of the object before the clipping plane is missing. The displayed image is unnatural. Furthermore, in a mobile driving simulator, it is desirable to display a moving object with a sense of realism, but when the object moves beyond the clipping plane, the image of the object suddenly disappears or appears due to front face culling. There is also a problem that.

In the method disclosed in Japanese Patent Application Laid-Open No. Hei 4-362794, the settable transparency values are discrete, and the number of pairs of the frame buffer and the Z buffer is required. For this reason, the number of transparency levels that can be set practically is limited, and an object approaching from near or far from near cannot be expressed with a sense of realism. Also, the transparency is the first visible from the viewpoint.
An object plane, a second object plane visible when the object plane is removed,
.. Are determined in the order seen from the viewpoint, and are given regardless of the distance. Therefore, the perspective cannot be sufficiently expressed from this point. JP 4-7
The method of No. 1082 gives a cross section and gives different transparency to the top and front and rear shapes by software processing, and there is no problem of heart, but the perspective, the presence and the like are described in JP-A-4-362794. There is a problem similar to the issue.

An object of the present invention is to provide a method and apparatus for generating a three-dimensional image which can display a three-dimensional image, especially a moving object with a sense of realism, without complicating the hardware structure. is there.

[0008]

SUMMARY OF THE INVENTION Accordingly, the present invention provides a shape data storage means for storing three-dimensional shape data of a plurality of objects and their position data, and a viewpoint specification for designating a viewpoint position and orientation in a three-dimensional space. Means, image generation means for generating an image when each object stored in the shape data storage means is viewed from the viewpoint specified by the viewpoint specification means, and a distance from the viewpoint to each of the individual objects. A transparency setting unit that sets a larger value of transparency as the size is smaller, and an image to be subjected to hidden surface processing while taking transparency into account for the image of each object to which transparency has been added by the transparency setting unit, and to be combined into one image And a synthesizing unit.

Further, according to the present invention, the transparency designating means includes:
Distance specifying means for specifying a first distance and a second distance smaller than the first distance;
The transparency is set to 0 for objects larger than the distance, and the transparency is set to 1 for objects whose distance from the viewpoint is smaller than the second distance, and the distance from the viewpoint is smaller than the first distance and the second distance. The three-dimensional object further comprises a transparency calculating means for the larger object to have a transparency varying between 1 and 0 in proportion to a distance obtained by subtracting the distance from the viewpoint from the first distance. An image generation device is disclosed.

Further, according to the present invention, when three-dimensional shape data of a plurality of objects and their positions are given, an image of each object is generated from the position and orientation of a designated viewpoint, and the individual objects are generated. The greater the distance from the viewpoint, the greater the transparency is set, and the images of the individual objects with these transparency settings are combined into a single image by performing hidden surface processing while taking transparency into account. A featured three-dimensional image generation method is disclosed.

[0011]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a functional block diagram showing a configuration example of a three-dimensional image generation apparatus according to the present invention. The input device 10, a shape data storage unit 11, and a viewpoint designation unit 12
, An image generation unit 13, a transparency setting unit 14, an image synthesis unit 15, a frame buffer 16, and a display 17.

The shape data storage means 11 stores the N to be displayed.
It stores the three-dimensional shape data of the individual objects (objects) and sends this information to the image generating means 13. The viewpoint designating unit 12 receives a signal from the input device 10 such as a keyboard or a mouse, determines the position and orientation of the viewpoint in a three-dimensional space, and sends this information to the image generating unit 13. The image generating means 13 generates, by perspective projection, an image viewed from the viewpoint determined by the viewpoint specifying means 12 for each of the three-dimensional shape data of the N objects sent from the shape data storage means 11. For example, as shown in FIG. 2, the shapes transmitted from the shape data storage means are a sphere A, a cube B, and a triangular prism C, and the viewpoint determined by the viewpoint specifying means 12 is located as shown in FIG. Suppose you have At this time, the three images as shown in FIG. 3 are generated for the respective objects by the image generating means 13.

The transparency setting means 14 is a means which is a feature of the present invention.
For each of the N images as in (a), the distance between the object corresponding to the image and the viewpoint is obtained from the shape data storage unit 11, and the transparency is increased so that the smaller the distance between the viewpoint and the object, the greater the transparency. Set. For example, a certain distance D1>D2> 0 is set, and the distance of the object from the viewpoint is d (i) (i = 1, 2,... N are numbers assigned to the object). At this time, the transparency α (i), that is, the transparency of the i-th object is

(Equation 1) If the object is deeper than the distance D1, the object is completely opaque, if it is between the distances D1 and D2, it is translucent, and the transparency increases in proportion to approaching the viewpoint. When it is closer, it is completely transparent. FIG. 4 illustrates the relationship between the distance and the transparency expressed by (Equation 1).

FIG. 5 is a flowchart of the processing performed by the transparency setting means 14 for setting the transparency as shown in (Equation 1). First, D1 and D2 satisfying D1>D2> 0 are set from the input device 10 such as a keyboard (step 101), and it is determined whether the transparency setting for the images of the N objects has been completed (step 102). If you have not finished setting the transparency for the images of the N objects,
The distance d (i) between the position of the i-th object obtained from the shape data storage unit 11 and the position of the viewpoint obtained from the viewpoint specification unit 12 is calculated (step 103). Here, as the position of the object, the position of the center of gravity of the object, a point closest to the viewpoint, or an intermediate point between the closest point and a point far from the viewpoint is used. Next, it is checked whether D1 ≦ d (i) (step 104), and if it is satisfied, the transparency of the image of the i-th object is set to 0 (step 10).
5) Return to step 102. If the condition in step 104 is not satisfied, it is checked whether d (i) ≦ D2 (step 106). If the condition is satisfied, the transparency of the image of the i-th object is set to 1 (step 107), and step 102 Return to If the condition of step 106 is not satisfied, i
Let the transparency of the image of the th object be (D1-d (i))
/ (D1-D2) (step 108), and returns to step 102. When the above processing is performed on the images of the N objects, the processing of the transparency setting means 14 ends.

As a specific display example, consider the three objects A, B, and C shown in FIG. Now, the distance d (a) between the sphere A and the viewpoint is 80, and the distance d (b) between the sphere B and the viewpoint is =
400, distance d (c) between triangular prism C and viewpoint = 140, D1
= 200, D2 = 100 (distance is an appropriate unit). Then, since d (a) <D2, the transparency α (a) of the image of the sphere A is set to 1, that is, completely transparent,
Since D1 ≦ d (b), the transparency α (b) of the cube B becomes 0,
That is, it is set completely opaque, and D2 ≦ d (c) <D1
Therefore, the transparency α (c) of the triangular prism C is (D1-d (c))
/(D1-D2)=0.4, that is, translucent, so that the images of the three objects in FIG.
(B). However, in FIG. 3B, the hatching of the vertical solid line has a transparency of 0, and the hatching of the horizontal solid line has a transparency of 0.4.
In a semi-transparent state.

The image synthesizing means 15 performs a hidden surface process on the projected image of each object to which the transparency has been added by the transparency setting means 14 in consideration of the transparency, and synthesizes the image into one image. Write to the frame buffer 16. Here, it is assumed that the Z-sort method is used as a method of the hidden surface processing. This sorts all N objects by the value of Z (distance from the viewpoint), and superimposes the images of the individual objects based on the result. In other words, this is a method in which the image of the object located farther from the viewpoint is written into the frame buffer 16 so that the image of the object located further in front is left.

At this time, when transparency is set for the image to be written, pixel information before writing is left by alpha blending according to the transparency, for example, by the following formula. That is, when the value of RGB before writing an image to a certain pixel is rgb0, the value of RGB to be written is rgb1, and the transparency is α, it is the value of RGB after writing of the pixel.
rgb2 is

Rgb2 = (1−α) · rgb1 + α · rgb0 For example, when α = 0, it is completely opaque, and rgb2 = rgb1, which is the RGB value to be written. When α = 1/2, the transparency is 50%, and rg
b2 = (rgb1 + rgb0) / 2, and RGB before writing
It is an intermediate value between the value and the RGB value to be written. Furthermore α
When = 1, it is completely transparent, and rgb2 = rgb0, which is the RGB value before writing.

In the examples shown in FIGS. 2 and 3, the Z-sorting method and the above-described transparency calculating method are used to obtain a cube B, a triangular prism C, and a sphere A.
6 are stored in the frame buffer 16 and displayed on the display 17. The RGD value of each image in FIG.
Assuming that b (a), rgb (b), and rgb (c), the RBG value of the image of the body B to be scratched first is rgb (b), which corresponds to the vertical hatch in FIG. . At the time of the next image writing of the triangular prism C, the RGB value of the portion that does not overlap with the body B, that is, the horizontal hatch portion in FIG. 6 is rgb (c). On the other hand, the RGB value of the hatched portion in FIG. 6 that overlaps the rectangular solid B is 0.6 rgb (c) +0.4 rgb (b) from (Equation 2). Thus, a natural image can be displayed by continuously changing the transparency in accordance with the distance from the viewpoint and the overlap.

In the transparency setting method of (Equation 1) and FIG. 4, α = 1 (transparency) when the distance is equal to or less than D2, and α = when the distance D1 or more.
In the meantime, the transparency α changes linearly with distance in the meantime. However, this portion may be a curve, and in general, any setting may be used as long as the transparency α monotonically decreases with distance.

[0020]

As described above, according to the method of continuously changing the transparency according to the distance according to the present invention, even if the object is located immediately before the viewpoint, the image of the object is transparent. Since the image is displayed translucently, an image of the object located behind the object can be displayed, and more information can be provided to the observer. In addition, the appearance of unnatural images due to front face culling and the unnatural display that suddenly causes the image of the object to disappear or appear when the object moves beyond the clipping plane disappears and disappears gradually. Therefore, it is possible to obtain an image with a sense of realism with less discomfort. Further, in the present invention, since a technique based on stereoscopic vision is not used, special hardware such as a display device with stereoscopic glasses or a lenticular lens is not required, and an inexpensive three-dimensional image generating device is provided. There is also an effect that it can be provided.

[Brief description of the drawings]

FIG. 1 is a functional block diagram illustrating a configuration example of a three-dimensional image generation device according to the present invention.

FIG. 2 is an example of setting three objects and a viewpoint.

FIG. 3 is an example in which transparency is set for each object in FIG. 2;

FIG. 4 is an example of a transparency setting function according to a distance.

FIG. 5 is a flowchart of a process for executing the setting method of FIG. 4;

FIG. 6 is a diagram showing an image obtained by combining the objects of FIG. 3 in consideration of transparency.

[Explanation of symbols]

 Reference Signs List 10 input device 11 shape data storage means 12 viewpoint designation means 13 image generation means 14 transparency setting means 15 image synthesis means 16 frame buffer 17 display

Claims (3)

[Claims] 1. Shape data storage means for storing three-dimensional shape data of a plurality of objects and their position data; viewpoint designating means for designating the position and orientation of a viewpoint in a three-dimensional space; Image generating means for generating an image when the stored individual object is viewed from the viewpoint specified by the viewpoint specifying means; and setting the transparency of the individual object to a larger value as the distance from the viewpoint is smaller. Transparency setting means, and image combining means for performing hidden surface processing on the image of each object to which transparency has been added by the transparency setting means while considering transparency, and combining the images into one image. Characteristic three-dimensional image generation device. 2. The object according to claim 1, wherein said transparency designating means designates a first distance and a second distance smaller than said first distance, and an object whose distance from a viewpoint is larger than said first distance. Sets transparency to 0, sets transparency to 1 for objects whose distance from the viewpoint is smaller than the second distance, and assigns transparency to objects whose distance from the viewpoint is smaller than the first distance and larger than the second distance. 1 to 0 in proportion to a distance obtained by subtracting the distance from the viewpoint from the first distance.
The three-dimensional image generation apparatus according to claim 1, further comprising: a transparency calculation unit that sets the transparency to change between the two. 3. When three-dimensional shape data of a plurality of objects and their positions are given, an image of each object is generated from the position and orientation of a specified viewpoint, and each object is generated from that viewpoint. The tertiary method is characterized in that a larger value of transparency is set as the distance is smaller, and images of the individual objects set with these transparency are combined into one image by performing hidden surface processing while taking transparency into account. Original image generation method.