JP2878678B1 - Tree 3D shape generator - Google Patents

Abstract

An object of the present invention is to provide a three-dimensional shape generation device that generates a three-dimensional shape of a model that approximates a real tree shape. SOLUTION: A target tree 1 is photographed by a plurality of cameras 2 from its surroundings, a background image is deleted from the obtained image, color conversion is performed, and the image is input to an image generation computer 3.
In addition to creating volume data, a tree model is created by creating branch segments, and leaves are rendered on each branch segment to display the tree model on the display 4.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for generating a three-dimensional shape of a tree, and more particularly to a three-dimensional shape generating apparatus which actually photographs a target tree and generates a three-dimensional tree model shape from the captured image. It relates to a generating device.

[0002]

2. Description of the Related Art In practice, in order to realize a communication environment through virtual scenes between persons at a distance, it is necessary to establish a method of generating a virtual scene for communication. As a method for generating a virtual scene, use of computer graphics (CG) technology can be considered. Artifact C that is relatively easy to model
Research on G expression has been relatively advanced, and its texture has been improved. On the other hand, it is not easy to model natural terrain, objects such as grass and people, and natural phenomena.

[0003] As a first method for generating a tree model using CG, for example, Norio Shigeba, Shunichi Okawa et al., "CG
Model of Tree Growth for Fostering-Natural Tree Form Generation by Fictitious Plant Hormone, IEICE Transactions DI
I, Vol. J76-D-II, No. 8, pp. 17
22-1734, 1993 have been proposed.

As a second method, Hiroshi Nagahashi, Kenji Komine et al., "Modeling of potted plant by silhouette method", Journal of the Institute of Television Engineers of Japan, Vol. 50, no. 1
0, pp. 1467-1473, 1996 have been proposed.

[0005]

SUMMARY OF THE INVENTION The above-mentioned first proposed method
Although the method of (1) simulates plant growth and can generate a natural shape, trial and error processing is required to create a shape intended by the maker.

In the second method, the shape of a naturally occurring plant can be approximated, but it is difficult to apply it to a tree or the like.
It is not possible to reproduce the branch structure of a tree.

Therefore, a main object of the present invention is to create a model that approximates an existing tree shape by creating volume data, and to easily create a shape intended by a maker by modifying volume data. It is an object of the present invention to provide a three-dimensional shape generating apparatus for trees that can be used.

[0008]

The invention according to claim 1 is
A three-dimensional shape generation device that generates a three-dimensional tree model shape from an image of a target tree, wherein the volume data generation means generates volume data enclosing the tree based on images from a plurality of viewpoints; Tree model creation means for creating a tree model consisting of branches and trunk segments based on the created volume data, and a leaf model added to the created tree model and rendered, and the shape of the target tree Rendering means for forming an image of a tree having

According to a second aspect of the present invention, there is provided a photographing means for photographing a target tree from a plurality of viewpoints, and a background image is removed from a photographed image before volume data is created. The conversion is performed, and the tree size and branch positions are normalized.

In the invention according to claim 3, the normalized volume data is further shaped into a tree shape.

[0011] In the invention according to claim 4, the tree model creating means according to claim 1 is configured to, based on the created volume data, create a branch having the same length and the same thickness, and a branch between the tip and the middle thereof. A tree model is created from the branch segments having the tree model.

[0012] In the invention according to claim 5, the tree model creating means is provided such that the branch is formed within a predetermined angle in the x direction and the z direction with respect to the y direction in which the branch or the segment extends around the root of the parent branch. Set branches at any angle where they will not bend unnaturally.

[0013]

FIG. 1 is a block diagram showing an embodiment of the present invention. In FIG. 1, a plurality of cameras 2 for photographing the target tree 1 from a plurality of viewpoints are installed around the target tree 1. The camera 2 may be a camera using a CCD element or an optical camera. The image data captured by the camera 2 is stored in an image generation computer 3.
Given to.

The image generating computer 3 mainly has a volume data creating function, a tree model creating function, and a rendering function in order to create a three-dimensional tree model from the image data input from the camera 2. I have.

The volume data creation function creates volume data that wraps a tree based on images from a plurality of viewpoints captured by the camera 2. The tree model creation function creates a tree model composed of branch and trunk segments based on the created volume data. The rendering function performs rendering by adding a leaf model to the created tree model, and creates a tree image having the shape of the target tree 1. Image generation computer 3
Is displayed on the display 4.

FIGS. 2 and 3 are flowcharts for explaining the operation of the embodiment of the present invention.

Next, a specific operation of the embodiment of the present invention will be described with reference to FIGS. The target tree 1 is photographed from the surroundings by the plurality of cameras 2 shown in FIG. Subsequently, the background is removed from the captured image, color conversion is performed, and normalization such as tree size and trunk position is performed.

The removal of the background image is performed manually,
The color conversion is performed by gray scale conversion. After the conversion, the dynamic range is normalized, a rectangle including the target tree 1 is obtained, and the size of the target tree 1 is obtained by normalizing the size. .

Next, the steps shown in FIG.
In SP1, the image generating computer 3 creates volume data.

FIGS. 4 and 5 are diagrams for explaining a method of creating volume data. In order to create volume data from live-action video, the concept of silhouette back projection is introduced. When the silhouette obtained from the image is anti-parallel projected into a three-dimensional space, an infinitely long columnar body having a cross section of the silhouette at this time is obtained as shown in FIG.
When the same processing is repeated using images from a plurality of viewpoints, the overlapping area of the pillars, that is, the area where the target shape exists as a logical product is limited. The volume data obtained in this way is not the shape itself of the target tree 1, but can be defined as a shape including the shape.

In this embodiment, unlike the simple silhouette back projection method, the volume data has 256 gradation values as the volume data.

Next, the volume data is shaped. If a large number of images of the object are prepared around the entire circumference, the shape approximation ability of the required volume data also increases, but if the number of given images is small,
Unintended shapes are generated. For example, if an image captured by two cameras as shown in FIG. 4 is read, and the two cameras are arranged at right angles, the volume data to be obtained is a rectangular parallelepiped as shown in FIG. The shape is close to, and is obviously different from the original tree shape.

To avoid this, volume data obtained by back projection is shaped. Generally, a tree has a shape close to a circle when viewed from above vertically. Therefore, the outermost point when the trunk is centered for each scan line of the captured image after processing is determined, and the distance from the trunk is defined as the outermost point of the volume data. Shaping can be performed by deleting the voxels existing outside the outermost point.

Next, at steps SP2 to SP11, modeling of the shape is performed. That is, a root stem segment is generated in step SP2. Figure 6 shows the trunk segment
As shown in (1), they are all represented as a set of segments having the same length and the same thickness. In step SP3, branches of the same generation are sorted. Here, the same generation is a generation when the root segment is set to 1, and all generations are the same until a branch model is generated.

Next, a branch / branch search is performed in steps SP4 to SP6. Here, as shown in FIG. 7, in addition to two branches at the tip, a total of three branches can be obtained by one branch at the middle point of the branch. In step SP4, the search for the leading branch 1 is performed, and in step SP5, the leading branch 2 is searched.
Is performed, and a search for an intermediate branch is performed in step SP6. The search for each branch is executed according to a subroutine shown in FIG.

FIGS. 8 and 9 are diagrams for explaining the branching of the branch. In particular, FIG. 8 shows the range of generation of the temporary branch.
Indicates the existence range of the branch segment and the leaf.

Branch branches are generated vertically upward in the y-direction from the center of the lower base of all branch segments.
That is, the maximum value MAX of the volume data is set to 0 in step SP21 of FIG. In step SP22, as shown in FIG.
Rotate to 0 °, and -6 around the Z axis in step SP23.
By rotating from 0 ° to 60 ° to define a temporary branch generation range, a temporary branch is generated in step SP24. Step SP
At 25, the total value T of the volume data in the small ball set at the tip of each temporary branch is calculated. Here, the small ball shown in FIG.
As shown in FIG. 9, the length of the branch segment is a diameter, and is represented by a sphere centered on the tip of the branch.

At step SP26, it is determined whether or not the total value T of the volume data is larger than the maximum value MAX.
If the total value T is larger than the maximum value MAX, step SP
At 27, the total value T is made the maximum value MAX.

Thereafter, or if the total value T has not reached the maximum value MAX in step SP26, the process proceeds to step S26.
The processing around the Z axis is performed at P28 and the processing around the X axis is performed at step SP29. If the total value is larger than the threshold value at step SP30, a branch of the next generation is generated at step SP31. The area is deleted from the volume data. That is, when one branch is generated, the region of the branch (the branch itself) and the region where the leaf is present (the large sphere in the figure) shown in FIG. 9 are deleted from the volume data.

Returning to FIG. 2 again, it is determined in step SP7 whether segment generation has been completed for all branches, and if not completed, generation of segments for all branches of the same generation has been completed in step SP8. It is determined whether or not. If not completed, a segment of the next branch is generated in step SP10. If completed, the process proceeds to the branch of the next generation in step SP19, and the branch of the next generation is sorted in step SP3. Then, in steps SP4 to SP6, a branch search for each branch is performed. In step SP7, it is determined whether or not all branches have been completed. If the branch has been completed, the thickness of the branches is controlled in step SP11.

FIG. 10 is a diagram showing a method of controlling the thickness of a branch segment, FIG. 11 is a diagram showing a leaf model, FIG. 12 is a diagram showing an example of tree modeling, and FIG.
3 is a diagram showing leaf texture.

In the above description, the calculation is performed while the thickness of the branch segments is all constant. However, in the case of finally forming a tree shape, the thickness is controlled based on the parent-child relationship. In this embodiment, as shown in FIG. 10, in the case of the fore-end branch, the thickness is controlled to be 0.77 times the parent branch. Regardless of the middle branch, if no branch is found ahead of the twig, the value is multiplied by 0.50, the thickness of the branch segment is set to a lower limit, and the thickness is not further reduced.

In this embodiment, since a general hardwood is targeted, a leaf model as shown in FIG. 11 is created, and the twigs are randomly added to each branch segment at a ratio of 2 to 3 branches. Attach. In this case, the branch segment to which the twig is added is required to be smaller than 0.3 times the thickness of the trunk. FIG. 12 shows a result example in which the shape of the tree generated in this manner is represented by a hexagonal prism model.

The leaf texture shown in FIG. 13 is mapped on each tree branch. FIGS. 14 and 15 show the original images used for the actual modeling.
Is taken from the front, and FIG. 15 is also taken from the side. FIGS. 16 and 17 show tree models generated from these images. Each shows an image when viewed from the same direction as the original image. Comparing the two, there is a sense of incongruity due to the difference in the degree of leaf growth, but it seems that the approximate shape of the tree can be approximated.

FIG. 18 shows a case where a plurality of different models are arranged in a virtual environment to simulate the transmission of light rays. FIG. 18 shows a pattern of transmission of sunlight, which is difficult with the technique of treating as volume data. It can also be seen that a field-of-view image from below the tree that is not originally included in the given image is also generated.

In order to improve the likelihood of a tree, the following rules may be applied at the time of branch calculation. That is, the angle of the branch segment with respect to the Y axis is ± 120 °.
Do not exceed. This is to suppress branches that are going to extend vertically downward, that is, in the direction of the ground. In addition, the relative angle with the branch segment several generations before should not exceed ± 90 °. This is a necessary condition for avoiding branch wraparound. If you do not add this condition,
In particular, many branches wrapping around the edge are generated in the outer part of the tree.

Further, branch calculation is performed for branches of the same generation in order from the branch closest to the trunk center. Under this condition, it is possible to express the spread of branches like a tree. This is because, when the branch calculation is performed in the order of generation of the branch segments, the branch first branched tends to extend while meandering in the volume data, and the likelihood of a tree may be reduced.

In the above-described embodiment, in order to create volume data, the camera 2 shoots the target tree 1 from different directions. However, the present invention is not limited to this. The volume data may be created manually, or the volume data may be created from something like a picture written on paper.

[0039]

As described above, according to the present invention, volume data enclosing a tree is created on the basis of images from a plurality of different viewpoints of the tree, and branches and trunks are created based on the created volume data. A tree model consisting of segments was created, a leaf model was added to the created tree model and rendered, and an image of a tree having the shape of the target tree was created. Applications can be expected. Further, since the created model has a tree-like branch structure, it is natural to generate an image viewed from a viewpoint other than the photographed point. Further, it is possible to express light transmission and the like, and it is possible to generate a video with high reality. Further, since the branch structure is expressed as a branch model, it is useful for expressing dynamics.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a flowchart for explaining the operation of the embodiment of the present invention;

FIG. 3 is a specific flowchart of a branch routine shown in FIG. 2;

FIG. 4 is a diagram for explaining a method of creating volume data according to an embodiment of the present invention.

FIG. 5 is a diagram for explaining a method of shaping volume data.

FIG. 6 is a diagram for explaining a method of determining the length of a branch segment.

FIG. 7 is a diagram showing a branch model of a branch.

FIG. 8 is a diagram for explaining a temporary branch generation range;

FIG. 9 is a diagram showing an existing range of a branch segment and a leaf.

FIG. 10 is a diagram for explaining control of the thickness of a branch segment.

FIG. 11 is a diagram showing a leaf model.

FIG. 12 is a diagram illustrating a modeling example of a tree.

FIG. 13 is a diagram showing a leaf texture.

FIG. 14 is a diagram showing an original image for actual modeling.

FIG. 15 is an image showing the modeling image shown in FIG. 14 from the side.

FIG. 16 is a diagram showing a tree model generated from FIG. 14;

FIG. 17 is a diagram showing a tree model generated from FIG. 15;

FIG. 18 is a diagram showing a rendered tree image.

[Explanation of symbols]

 1 target tree 2 camera 3 computer for image generation 4 display

Continuation of the front page (56) References Kuwahara et al., "Automatic estimation method of fractal model of three-dimensional tree shape using tree image as input", Journal of the Institute of Image Electronics Engineers of Japan, 1995, Vol. 24, No. 5, p. 541-549 Chiba, "Simulation of Natural Objects and Natural Phenomena by CG", Measurement and Control, 1995, Vol. 34, No. 7, p. 16-18 Suzuki, "Basic Study on Landscape Management in Forests Modeling Tree Form by Computer Graphics", 1996 Hokkaido Forestry Research Institute Research Report, No. 33, p. 1-43 (58) Field surveyed (Int. Cl. ⁶ , DB name) G06T 17/00 BEILSTEIN (STN) JICST file (JOIS)

Claims (5)

(57) [Claims] 1. A three-dimensional shape generating apparatus for generating a three-dimensional tree model shape from an image of a target tree, wherein volume data enclosing the tree is created based on images from a plurality of viewpoints. Volume data creating means, based on the volume data created by the volume data creating means, a tree model creating means for creating a tree model composed of branches and trunk segments, and a tree model created by the tree model creating means A three-dimensional shape generating apparatus for a tree, comprising rendering means for adding a leaf model and rendering the image to create an image of the tree having the shape of the target tree. 2. A photographing means for photographing the target tree from a plurality of viewpoints, wherein a background image is removed from the photographed image and color conversion is performed before creating the volume data. The three-dimensional shape generating apparatus for trees according to claim 1, characterized in that the tree size and the position of the trunk are normalized by performing the following. 3. The method according to claim 1, further comprising shaping the normalized volume data into a tree shape.
A three-dimensional shape generation device for a tree according to claim 2. 4. The tree model creating means, based on the created volume data, creates a tree model with branches having the same length and the same thickness, and a branch segment having a branch at the tip and middle thereof. The three-dimensional shape generating apparatus for trees according to claim 1, wherein the apparatus generates the three-dimensional shape. 5. The tree model creating means unnaturally bends the branch within a predetermined angle in the x direction and the z direction with respect to the y direction in which the branch segment extends around the root of the parent branch. The three-dimensional shape generating apparatus according to claim 4, wherein the branch is set at an arbitrary angle that cannot be seen.