JPH08293041A - High speed rendering method and device therefor - Google Patents

Abstract

PURPOSE: To shorten the time needed for selection of a drawn polygon by designating the spherical body of the highest hierarchy of a hierarchical spherical model as a rendering ball to assign a rendering model and then clipping the corresponding rendering ball when this ball is excluded out of a clipping area. CONSTITUTION: A rendering ball setting part 2 extracts a spherical body of the highest hierarchy that has its radius smaller than the prescribed value and defines this spherical body as a rendering ball. A rendering model correspondence part 3 prepares a rendering model in response to every set rendering ball, and a clipping area holding part 4 holds the data to prescribe a clipping area. An inside/outside decision part 5 decides whether every rendering ball exists inside or outside the clipping area. Then a clip processing part 6 clips the rendering model corresponding to the rendering ball existing outside the clipping area.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a high-speed rendering method and an apparatus thereof, and more particularly to a high-speed rendering method and an apparatus thereof for reducing the drawing time of a three-dimensional model without reducing the visual effect. This method and its apparatus can be suitably applied to fields such as landscape simulation and artificial reality.

[0002]

2. Description of the Related Art In recent years, researches such as landscape simulation and scientific visualization applying computer graphics (hereinafter abbreviated as CG) technology have been actively conducted, and walkthroughs, flight simulators, etc. Research on reality is also being actively conducted. Here, the former mainly targets a still image, whereas the latter mainly targets a moving image. Therefore, as a CG technique applied to the latter, generally, high-speed drawing is strongly required. However, even in the former case, there is a case where there is a huge amount of data, especially in a landscape simulation, so that high speed drawing is also required. In any case, it is required to achieve high speed without deteriorating the visual effect.

Conventionally, the following three methods have been proposed as methods for achieving this requirement. (1) Solids are grouped by space division,
A method for accelerating the drawing of the corresponding 3D model by performing inside / outside determination on the visual field for each group and setting only the group determined to be within the visual field as the drawing target (“wide area 3D object management system”, Takashi Tamada, Yasuaki Nakamura, TECHNICAL REPORT O
F IEICE HC93-11 (1993-05),
(Institute of Electronics, Information and Communication Engineers) (2) A method for performing stepwise simplified representation of a shape, for example,
When the distance from the viewpoint is small, the shape is displayed as it is, and when the distance from the viewpoint is large, the shape is simplified and displayed to speed up the drawing of the entire 3D model (“ Adaptive Dis
play Algorithm forInterac
five Frame Rate During Vi
suialization of Complex Vi
rtual Environments ", Thomas
s A. Funkhauser and Carlo
H. Outer 1, COMPUTER GRAPHICS
Proceedings, Annual Conf
rence series, 1993), and

[0004]

[Outside 1]

(3) By grouping faces according to the direction of the line of sight, determining whether each group is visible from the viewpoint, and drawing only the group determined to be visible, the corresponding three-dimensional model To speed up the drawing of images (“Drawing time reduction based on shape grouping and multiple expressions for real-time landscape simulator”, Katsuhiro Kitajima, Yoko Yusa, IEICE Transactions D
-II Vol. J77-D-II No. 2 pp.
311-320 February 1994)

[0006]

However, the above-mentioned conventional methods are all methods for selecting polygons to be drawn according to the positional relationship between the viewpoint and the drawing object, and furthermore, using a polyhedral expression such as a cube area or octotree Since the relationship is calculated, there is an inconvenience that the calculation cost is high. Then, due to the high calculation cost, the time required for selecting the drawing polygons becomes long, and the drawing time becomes long.

[0007]

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and an object of the present invention is to provide a high-speed rendering method and an apparatus therefor capable of significantly shortening the time required for selecting drawing polygons.

[0008]

According to a high speed rendering method of claim 1, among the spheres having a predetermined radius or less of a hierarchical sphere model, a sphere of the highest hierarchy is designated as a rendering sphere, and each rendering sphere is designated. This is a method of correspondingly assigning a rendering model, determining whether the rendering sphere is inside or outside the clip area, and clipping the corresponding rendering model when the rendering sphere is outside the clip area.

According to a second aspect of the present invention, in the high-speed rendering method, the surface of an object to be modeled is mesh-divided, and the lengths of all sides of each mesh-divided mesh are set in advance in the lowest layer sphere. Sphere of the highest hierarchy is generated by generating a minimum sphere including the cubic area determined based on all the nodes by setting the diameter equal to or less than the diameter, and centering each node generated by mesh division. Then, all nodes are divided into two in a plane that passes through the center of the sphere of the immediately preceding layer, and the smallest sphere including the cubic area determined based on all the nodes belonging to each divided section is generated,
A sphere of the layer next to the layer before the division, and then a two-part division process of all nodes and a process of generating the smallest sphere including a cubic area determined based on all the nodes belonging to each segment and making it the sphere of the next layer Is repeated until the number of nodes after division becomes 1, and a hierarchical sphere model is created.

According to a third aspect of the present invention, in the high-speed rendering method, in the state where the visual field is defined by the line-of-sight direction vector and the viewing angle with respect to the viewpoint position, the vector from the viewpoint toward the center of the rendering sphere, and this vector as the center. This is a method of calculating the angle defining the spread of the rendering sphere and removing the corresponding rendering sphere on condition that the angle formed by both vectors is larger than the sum of the viewing angle and the angle defining the spread of the rendering sphere. .

A high-speed rendering method according to a fourth aspect is such that, for a rendering sphere of which at least a part is included in the field of view, the polygons included in this rendering sphere are divided into a plurality of groups based on the directions of their normal vectors. A set of polygon normals belonging to each section is approximated by a cone, and a first spread angle that defines the spread of the rendering sphere centered on the vector from the viewpoint to the center of the rendering sphere is calculated. The center axis vector that defines the center axis of the cone that approximates the aggregate and the second spread angle of this cone are calculated, and the angle formed by the vector from the viewpoint to the center of the rendering sphere and the center axis vector is the sum of both spread angles. This is a method of selecting a polygon belonging to the corresponding section as a drawing target on condition that it is the following.

According to a fifth aspect of the present invention, in the high-speed rendering method, a plurality of polygons having different drawing precisions are generated in advance for the polygons corresponding to the respective rendering models in correspondence with the magnitude of the vector from the viewpoint to the center of the rendering sphere. In other words, it is a method of drawing a polygon with a corresponding drawing accuracy based on the actual size of the vector. The high-speed rendering device according to claim 6 is a rendering sphere designating means for designating a sphere of a highest hierarchy among spheres having a radius equal to or less than a predetermined radius of a hierarchical sphere model,
Rendering model assigning means for assigning a rendering model to each rendering sphere, inside / outside determining means for determining whether the rendering sphere is inside or outside the clip area, and inside / outside determining means for indicating that the rendering sphere is outside the clip area. And clipping means for clipping the corresponding rendering model in response to the determination result.

According to a seventh aspect of the present invention, in a high-speed rendering device, the surface of an object to be modeled is divided into meshes, and the lengths of all sides of each mesh-divided mesh are set in advance in the lowest-level sphere. Mesh dividing means to be set to less than or equal to the diameter, the lowest layer sphere generating means for generating the sphere of the lowest layer centering on each node generated by mesh division, and the smallest sphere including a cubic area determined based on all nodes To generate a top-level sphere, a top-level sphere generating means, a dividing means for dividing all nodes into two in a plane passing through the center of the sphere of the immediately previous level, and all the nodes belonging to each divided section. Intermediate sphere generating means for generating a minimum sphere including a cubic area determined based on the sphere, and making it the sphere of the next hierarchy of the hierarchy before division, and processing by the dividing means and intermediate hierarchy sphere generation The process according to stage number of nodes after the division is further and a repeat control means for repeating until 1.

According to another aspect of the present invention, there is provided a high-speed rendering device, which holds a line-of-sight direction vector and a viewing angle set with reference to a viewpoint position to define a field of view, a vector from the viewpoint to a center of the rendering sphere, A calculation unit that calculates an angle that defines the spread of the rendering sphere centered on this vector, and determines whether the angle formed by both vectors is larger than the sum of the viewing angle and the angle that defines the spread of the rendering sphere. In response to the determination result of the first magnitude determining means and the first magnitude determining means indicating that the angle formed by both vectors is larger than the sum of the viewing angle and the angle defining the spread of the rendering sphere, the corresponding rendering sphere is selected. It further has a rendering sphere removing means for removing.

According to a ninth aspect of the present invention, in a high-speed rendering device, a rendering sphere, at least a part of which is included in the field of view, divides the polygons included in the rendering sphere into a plurality of groups based on the directions of their normal vectors. Means, an approximation means for approximating a set of polygon normals belonging to each section by a cone, and a first spread angle defining a spread of the rendering sphere centered on a vector from the viewpoint to the center of the rendering sphere. 1 calculation means, a central axis vector defining a central axis of a cone approximating a collection of normals of a polygon, and a second calculation means for calculating a second spread angle of the cone, and from the viewpoint to the center of the rendering sphere Second determination of whether or not the angle formed by the vector and the central axis vector is less than or equal to the sum of both spread angles
In response to the determination result of the size determination means and the second size determination means indicating that the angle formed by the vector from the viewpoint toward the center of the rendering sphere and the central axis vector is less than or equal to the sum of both spread angles, the corresponding classification is performed. It further has selection means for selecting the polygon to which it belongs as a drawing target.

According to a tenth aspect of the present invention, the high-speed rendering device generates a plurality of polygons having different drawing precisions in advance, corresponding to the size of the vector from the viewpoint to the center of the rendering sphere, for the polygons corresponding to the respective rendering models. It further has a polygon generating means to be placed and a drawing means for drawing a polygon of a corresponding drawing accuracy based on the actual size of the vector.

[0017]

According to the high-speed rendering method of claim 1, of the spheres having a radius less than or equal to a predetermined radius of the hierarchical sphere model, the sphere of the highest hierarchy is designated as the rendering sphere, and rendering is performed corresponding to each rendering sphere. A model is assigned, it is determined whether the rendering sphere is inside or outside the clipping area, and when the rendering sphere is outside the clipping area, the corresponding rendering model is clipped. Therefore, the positional relationship using the polyhedral representation is used. The calculation of is unnecessary, and by performing the interference calculation using the rendering sphere, the clip of the render length model can be achieved at high speed, and the rendering can be speeded up.

According to the high-speed rendering method of claim 2, the surface of the object to be modeled is mesh-divided, and the lengths of all sides of each mesh-divided mesh are preset. Set to less than or equal to the diameter of the sphere, generate the lowest level sphere centered on each node generated by mesh division, and then generate the smallest level sphere including the cube area determined based on all nodes , And then divide all nodes into two in the plane that passes through the center of the sphere of the previous layer, and generate the smallest sphere including the cubic area that is determined based on all the nodes belonging to each divided segment, and divide it. As a sphere of the next level of the previous level, then
The process of dividing all nodes into two and the process of generating the smallest sphere including a cubic area determined based on all the nodes belonging to each segment and making it the sphere of the next layer is repeated until the number of nodes after division becomes 1. Since the hierarchical sphere model is created by this, the sphere included in the object is not originally generated,
The number of generated spheres can be reduced without complicating the processing, and the same operation as in claim 1 can be achieved based on the hierarchical sphere model thus generated.

According to the high-speed rendering method of claim 3, in the state where the visual field is defined by the line-of-sight direction vector and the viewing angle with respect to the viewpoint position, the vector from the viewpoint to the center of the rendering sphere and this vector are The angle that defines the spread of the rendering sphere is calculated as the center, and the corresponding rendering sphere is removed under the condition that the angle formed by both vectors is larger than the sum of the viewing angle and the angle that defines the spread of the rendering sphere. Therefore, it is not necessary to calculate the positional relationship using the polyhedron representation, and it is possible to easily remove the rendering sphere that is not necessary for drawing by the high-speed cone-to-cone interference check, which in turn speeds up the detection of the drawing polygon. Can be achieved.

According to the high-speed rendering method of claim 4, with respect to the rendering sphere of which at least a part is included in the visual field, the polygons included in the rendering sphere are divided into a plurality of groups based on the directions of the respective normal vectors. Then, a set of polygon normals belonging to each section is approximated by a cone, and the first spread angle that defines the spread of the rendering sphere centered on the vector from the viewpoint to the center of the rendering sphere is calculated. The central axis vector that defines the central axis of the cone that approximates the aggregate of lines and the second spreading angle of this cone are calculated, and the angle between the vector from the viewpoint to the center of the rendering sphere and the central axis vector is the spreading angle. The polygons that belong to the corresponding category are selected for drawing, provided that they are less than or equal to the sum of From eliminates the need for calculation of the positional relationship with polyhedral representation of the polygons in the rendering sphere, the selection of polygon to be drawn can be achieved at high speed, it is possible to speed up the turn rendering.

According to the high-speed rendering method of claim 5, a plurality of polygons having different drawing precisions are generated in advance in correspondence with the size of the vector from the viewpoint to the center of the rendering sphere for the polygons corresponding to the respective rendering models. Since polygons with the corresponding drawing accuracy are drawn based on the actual size of the above vector, some polygons with low drawing accuracy can be used to speed up the drawing, and Can be speeded up.

According to the high-speed rendering apparatus of claim 6, among the spheres having a predetermined radius or less in the hierarchical sphere model,
The sphere of the highest hierarchy is designated as the rendering sphere by the rendering sphere designating means, and the rendering model assigning means assigns the rendering model corresponding to each rendering sphere. Then, the inside / outside determination means determines whether the rendering sphere is inside or outside the clip area, and responds to the determination result of the inside / outside determination means indicating that the rendering sphere is outside the clip area. To clip. Therefore, it is not necessary to calculate the positional relationship using the polyhedron representation, and by performing the interference calculation using the rendering sphere, the clip of the render length model can be achieved at high speed, and the rendering can be speeded up.

According to the high-speed rendering apparatus of claim 7, the surface of the object to be modeled is mesh-divided, and the lengths of all sides of each mesh-divided mesh are preset. Mesh dividing means to be set to less than or equal to the diameter of the sphere, the lowest layer sphere generating means to generate the sphere of the lowest layer centering on each node generated by mesh division, and the minimum including the cubic area determined based on all nodes Top-level sphere generation means for generating the sphere of the top level sphere, a dividing means for dividing all nodes into two on a plane that passes through the center of the sphere of the immediately preceding level, and all of the divided divisions. Intermediate layer sphere generation means for generating a minimum sphere including a cubic area determined based on nodes and making it the sphere of the layer next to the layer before division, processing by the dividing means, and intermediate layer sphere Since it further has an iterative control means for repeating the processing by the generation means until the number of nodes after division becomes 1, the sphere included in the object is not originally generated, and the processing becomes complicated. It is possible to reduce the number of spheres generated without undue effect, and it is possible to achieve the same effect as that of claim 6 based on the hierarchical sphere model thus generated.

According to another aspect of the high-speed rendering device of the present invention, in order to define the visual field, a holding means for holding the line-of-sight direction vector and the viewing angle set with the viewpoint position as a reference, and a vector from the viewpoint toward the center of the rendering sphere. And a calculation means for calculating an angle that defines the spread of the rendering sphere centered on this vector, and whether the angle formed by both vectors is greater than the sum of the viewing angle and the angle that defines the spread of the rendering sphere. Responsive to the determination result of the first magnitude determining means and the first magnitude determining means indicating that the angle formed by both vectors is larger than the sum of the viewing angle and the angle defining the spread of the rendering sphere. Since it further has a rendering sphere removing means for removing the sphere, the calculation of the positional relationship using the polyhedron representation becomes unnecessary, and the drawing is performed. Can be the removal of unwanted rendering sphere simpler fast interference checking cone pair cone can thus achieve the detection of the drawing polygons at high speed.

According to the ninth aspect of the high-speed rendering device, with respect to the rendering sphere that includes at least a part in the visual field, the polygons included in the rendering sphere are divided into a plurality of groups based on the directions of the respective normal vectors. A dividing means, an approximating means for approximating a set of normals of polygons belonging to each division by a cone, and a first spread angle defining the spread of the rendering sphere centered on a vector from the viewpoint to the center of the rendering sphere. First calculating means, a second calculating means for calculating a central axis vector defining a central axis of a cone approximating a set of normals of the polygon and a second spreading angle of the cone, and a center of the rendering sphere from the viewpoint. Second magnitude determining means for determining whether or not the angle formed by the vector toward the center and the central axis vector is less than or equal to the sum of both spread angles. , Drawing a polygon belonging to the corresponding section in response to the determination result of the second magnitude determining means indicating that the angle formed by the vector from the viewpoint toward the center of the rendering sphere and the central axis vector is less than or equal to the sum of both spread angles. Since it further has a selection means for selecting as a target of, it is not necessary to calculate the positional relationship using the polyhedron representation, and the polygon to be drawn can be selected at high speed from the polygons in the rendering sphere. This can be achieved, and rendering can be speeded up.

According to the high-speed rendering device of the tenth aspect, a plurality of polygons having different drawing precisions are generated in advance for the polygons corresponding to the respective rendering models in correspondence with the magnitude of the vector from the viewpoint to the center of the rendering sphere. Since it further has a polygon generating means for preserving and a drawing means for drawing a polygon having a corresponding drawing accuracy based on the actual size of the vector, a polygon having a low drawing accuracy for some polygons. By adopting, it is possible to speed up the drawing, and thus speed up the rendering.

[0027]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to the accompanying drawings showing embodiments. FIG. 1 is a schematic diagram showing the relationship between a hierarchical sphere model and a rendering model. In the figure,
The spheres of each layer that make up the hierarchical sphere model are indicated by ellipses of different sizes. Further, this hierarchical sphere model is created by sequentially performing division processing from the sphere of the highest hierarchy. In this case, spheres S1, S2, S3 having radii equal to or less than the value designated in advance are generated. In this case, the spheres S1, S2, S3 are set as rendering spheres and are associated with the rendering models L1, L2, L3, respectively. However, among spheres with a radius less than or equal to the value specified in advance,
Only the sphere at the highest level is set as the rendering sphere.
Therefore, the sphere in the hierarchy lower than the rendering sphere is not the rendering sphere. Also, the above value is specified so that the radius of the rendering sphere is sufficiently larger than the radius of the sphere in the lowest hierarchy. Furthermore, the rendering model L
1, L2 and L3 are models including shape information such as polygons included in the corresponding rendering spheres.

When the rendering sphere and the rendering model are associated with each other as described above, the rendering sphere and the rendering model may not be in one-to-one correspondence. In such a case, as shown in FIG. 2, the rendering model is separated into the one-to-one correspondence portions L1 ′ and L2 ′ and the many-to-one correspondence portion L12 ′, and a new rendering model is generated. Original rendering models L1 and L2 and new rendering models L1 ′ and L2 corresponding to the case of FIG.
The relation with the ‘′, L12’ is as shown in FIG. 3.

Therefore, if the rendering sphere is set in this way and a rendering model is associated with each rendering sphere, it is determined whether each rendering sphere is inside or outside the clip area, and outside the clip area. By clipping only the rendering sphere determined to be, the clipping process of the rendering model can be easily achieved.

In the clipping process of the rendering model, for example, as shown in the flowchart of FIG. 4, the rendering sphere is set in step SP1, the rendering model is associated with each rendering sphere in step SP2, and in step SP3, the rendering model is associated with each rendering model. It is determined whether the rendering sphere is inside or outside the set clip area, and if it is determined that the rendering sphere is outside the clip area, in step SP4
Clip the corresponding rendering sphere, and step SP5
In step S3, it is determined whether or not all rendering spheres have been processed. If there is a rendering sphere that has not been processed, the process of step SP3 is performed again. On the contrary, if it is determined in step SP5 that all rendering spheres have been processed, the series of processes is ended. When it is determined in step SP3 that the rendering sphere is inside the clip area, the process of step SP3 is performed again.

Therefore, in the hierarchical sphere model, the inside / outside determination for the clip area is performed based on only the sphere set as the rendering sphere, and the clip processing of the rendering model corresponding to the rendering sphere determined to be outside the clip area is performed. Can be performed, and the process for clipping the rendering model can be significantly simplified. As a result, rendering can be speeded up.

However, even if the rendering sphere is not clipped, not all the polygons and the like constituting the rendering model are inside the clip area. Therefore, in this case, for example, it is possible to perform clipping processing on polygons or the like that form the rendering model corresponding to the rendering sphere that has not been clipped. However, a grouping method using normal vectors, which will be described later, should be adopted. Is preferred.

FIG. 5 is a schematic diagram showing the relationship between the field of view approximated by a cone and the rendering sphere. In the figure, the viewpoint position is O = (X0, Y0, Z0), and the line-of-sight direction vector is V.
1 = (X1, Y1, Z1), viewing angle is θ1, center of rendering sphere is C = (Xc, Yc, Zc), radius of rendering sphere is r, rendering is from the viewpoint as the starting point and the center of the rendering sphere as the ending point The sphere vector is V2 = (X2, Y
2, Z2), and the angle formed by both vectors is set to φ.

In this case, the angle indicating the spread of the rendering sphere is given by equation 1. Further, the angle φ is given by the equation 2.

[0035]

[Equation 1]

[0036]

[Equation 2]

Therefore, the magnitude of the angle φ and the sum of the angles θ1 and θ2 is determined. If φ ≦ θ1 + θ2, the rendering sphere is in the field of view, and if φ> θ1 + θ2, the rendering sphere is out of the field of view. judge. This determination may be made for all of the rendering spheres, but preferably
It is preferable to perform the above determination in order from the sphere in the highest hierarchy to the rendering sphere, and omit the determination for the rendering sphere generated from the upper sphere determined to be out of the field of view. Then, as a result of the above determination, the rendering sphere that is determined to be outside the field of view is removed, and only the polygons of the rendering model that correspond to the rendering sphere that is determined to be within the field of view are drawn. It is possible to achieve high speed.

The case where the field of view is approximated by a cone has been described above. However, in the state where the field of view (view volume) is defined as a quadrangular truncated pyramid region in accordance with the display tube surface, four side surfaces of the quadrangular pyramid truncated region are defined. The distance between each and the center of the sphere may be calculated, and the size of the distance and the radius of the sphere may be determined to determine whether it is inside or outside the field of view.

In displaying a polygon or the like of the rendering model determined to be within the field of view in the above processing, the following method can be applied to achieve high-speed display. This method switches polygons and the like of the rendering model in the field of view according to the normal vector, so polygons and the like in the rendering sphere are divided into several groups based on the normal vector in advance,
The normal vector of each group is approximated by a cone (see FIGS. 6 and 7).

FIG. 6 is a schematic diagram showing a state in which the range of the normal vector of a polygon that is not drawn is approximated by a cone. Here, V2 and θ2 are the same as those in FIG. In this case, the spread angle θ3 of the cone is θ3 =
(Π / 2) −θ2. FIG. 7 shows two cones, one of which is the same as the cone of FIG. In the figure, the newly added center vector of the cone (vector starting from the apex of the cone and heading toward the center of the bottom of the cone) is V3 = (X
3, Y3, Z3), the spread angle of this cone is indicated by α, and the angle formed by both vectors V2, V3 is indicated by β.

In this case, the angle β becomes the expression 3.

[0042]

(Equation 3)

Therefore, the magnitude of the angle θ3 and the sum of the angles α and β is determined, and when θ3 ≦ α + β, a polygon or the like having a normal vector corresponding to the range of the corresponding cone is targeted for drawing, When θ3> α + β, a polygon or the like having a normal vector corresponding to the range of the corresponding cone is not a drawing target. By adopting this process,
Of the polygons etc. in the rendering sphere in the field of view,
It is possible to quickly extract polygons to be drawn.

Further, the polygons to be displayed are displayed in a level of detail (hereinafter abbreviated as LOD) display (“Adaptive Display Algor”).
itm for Interactive Frme
Rates During Visualization
no of Complex Virtual Envi
romments ", Thomas A. Funkh
user and Carlo H.S. Outside 1, COM
PUTER GRAPHICS Proceeding
s, Annual Conference Seri
es, 1993), the time required for drawing and displaying all polygons to be displayed can be shortened.

Specifically, as shown in FIG. 8, a plurality of LOD models such as polygons corresponding to each rendering model (a plurality of LOD models having different LOD layers,
A value obtained by subtracting the radius r of the rendering sphere from the length of the rendering sphere vector V2 (from the origin to the rendering sphere) in advance by creating D0, LD1, LD2 indicating the LOD layer). Either LOD layer is selected or the display is omitted according to the size of the minimum distance). The threshold value indicating which LOD layer is selected is set in advance based on an empirical rule or the like.

Therefore, in this case, polygons and the like that are distant from the start point can be displayed in a simplified manner, and a higher speed display of the rendering model can be achieved. 10 and 11 are flowcharts for explaining an example of a method for creating a hierarchical sphere model.

In the flowchart of FIG. 10, the diameter of the sphere of the lowest hierarchy is set in step SP1,
In step SP2, the surface of the object to be processed is mesh-divided by a conventionally known method such as the Delaunay triangulation method, the node group generated by the mesh division is stored, and in step SP3, the processing of the flowchart of FIG. 11 (sphere model creation) is performed. Processing), and the series of processing ends. The mesh division in step SP2 is set such that the length of each side of each mesh is less than or equal to the radius of the sphere in the lowest hierarchy.

In the flowchart of FIG. 11, it is determined in step SP1 whether or not there are a plurality of nodes. Then, when it is determined that the number of nodes is plural, in step SP2, a cube region (a rectangular parallelepiped or a cube including all the nodes) is generated based on all the corresponding nodes, and the minimum including the cube region is generated. A sphere having a radius is generated (see (A) in FIG. 12), and in step SP3, whether or not a node group is generated by the division (that is, the corresponding node group is generated by the division process described later). Whether or not) is determined. Then, when it is determined that the node group is not generated by the division, in step SP4, the minimum sphere generated in step SP2 is stored as the sphere of the highest hierarchy. On the contrary, when it is determined that the node group is generated by the division, step SP5
At step SP2, the minimum sphere generated in step SP2 is set to the step SP of the processing of the flowchart of FIG.
4 or is stored as a sphere in a layer below the smallest sphere stored in step SP5.

After the processing of step SP4 or step SP5, at step SP6, all the nodes are set to 2 on a plane which passes through the center of the smallest sphere stored immediately before and whose normal line is the longitudinal direction of the cubic region. After division (see (B) in FIG. 12), the process of the flowchart of FIG. 11 is performed on one of the divided node groups in step SP7, and the remaining divided node group is displayed in step SP8. The process of the flowchart of 11 is performed, and the process directly returns to the original process.

In the step SP1, the number of nodes is 1
If it is determined that the node group is generated by the division in step SP9 (that is, whether the corresponding node group is generated by the division processing described later). judge. Then, when it is determined that the node group is not generated by the division, in step SP10, the sphere of the lowest hierarchy is stored in a state where its center coincides with the node. On the contrary, when it is determined that the node group is generated by the division, step SP11
11, the sphere of the lowest sphere stored in step SP4 or step SP5 of the processing of the flowchart of FIG. 11 performed previously with the center of the sphere of the lowest sphere coincident with the node. Save as. After the processing of step SP10 or step SP11 is performed, the processing directly returns to the original processing.

Therefore, the sphere of the lowest hierarchy is shown in FIG.
When the mesh division of the surface of the object is performed as shown in the middle (A), for example, it is arranged as shown in (B) in FIG. 13 with respect to one side of the surface of the object. Therefore,
The error of the object shape represented by the sphere of the lowest hierarchy is less than the radius of the sphere of the lowest hierarchy. In addition, the spheres of the layers other than the lowest layer divide the node group into two on a plane that passes through the center of the sphere of the upper layer and the normal line is the longitudinal direction of the cube area, and Since it is generated as a sphere having a minimum radius including a cubic region determined based on, the sphere generation process can be simplified. Furthermore, since the sphere is generated only for the surface of the object, the sphere included in the object is not originally generated.
Therefore, there is no need to remove unnecessary spheres,
Moreover, since no useless sphere is generated, the number of spheres can be reduced.

The number of layers is automatically determined by the diameter of the sphere in the lowest layer and the size and shape of the target object. However, in the flowchart of FIG. 11, it is preferable to adopt a process of generating a sphere of minimum radius including all nodes, instead of a process of generating a sphere of minimum radius including a cubic area determined based on all nodes. The object occupancy rate in the sphere can be increased.

Further, in the flow chart of FIG.
It is preferable to use a plane whose normal is the maximum dispersion direction of all nodes, instead of a plane whose normal is the longitudinal direction of the cubic area, and which has a large object occupancy in the sphere including each divided node group. The probability of becoming can be increased.

[0054]

[Embodiment 2] FIG. 14 is a block diagram showing an embodiment of a high speed rendering apparatus of the present invention. However, in this block diagram, only the portion for performing the clipping process of the rendering model is shown. Therefore, a conventionally known device or a device described below is adopted as a part for performing other processing.

This apparatus selects a predetermined value from a hierarchical sphere model holding unit 1 that holds a hierarchical sphere model (center coordinate values of spheres for each hierarchy, radius, etc.) and spheres of all the hierarchies that make up the hierarchical sphere model. A rendering sphere setting unit 2 (rendering sphere designating means) for extracting a sphere having a radius equal to or less than the value of ??? A rendering model associating unit 3 (rendering model assigning unit) that associates a rendering model, a clip region holding unit 4 that holds data defining a clip region, and an inside / outside determination for determining whether each rendering sphere is inside or outside the clip region. Clip part 5 and the rendering model corresponding to the rendering sphere determined to be outside the clipping region. And a clip processing unit 6. Note that the operation of each component of the configuration is the same as the processing of the corresponding step of the flowchart in FIG. Of course, each of these components may be configured by hardware, but it is preferable that it is configured by a computer in which a program for performing the processing shown in FIG. 4 is incorporated.

Therefore, in the hierarchical sphere model, the inside / outside determination for the clip area is performed based on only the sphere set as the rendering sphere, and the clip processing of the rendering model corresponding to the rendering sphere determined to be outside the clip area is performed. Can be performed, and the process for clipping the rendering model can be significantly simplified. As a result, rendering can be speeded up.

However, the rendering model associating unit 3
The rendering sphere and rendering model are 1
When the one-to-one correspondence does not occur, the rendering model is divided into a one-to-one correspondence portion and a many-to-one correspondence portion based on an operator's instruction or the like to create a new rendering model,
It is preferable that the association is performed based on a new rendering model.

FIG. 15 is a block diagram showing the arrangement of an apparatus for determining whether the rendering sphere is inside or outside the visual field when the visual field is approximated by a cone. This apparatus includes a viewpoint position holding unit 11 that holds a viewpoint position O = (X0, Y0, Z0) and a line-of-sight direction vector V1 = (X
1, Y1, Z1) for holding the line-of-sight direction vector holding unit 1
2, a viewing angle holding unit 13 that holds the viewing angle θ1, a rendering sphere center holding unit 13a that holds the center C = (Xc, Yc, Zc) of the rendering sphere, and a radius holding that holds the radius r of the rendering sphere. A section 14, a rendering sphere vector calculation section 15 for calculating a rendering sphere vector V2 = (X2, Y2, Z2) having a viewpoint as a start point and a rendering sphere center as an end point, and a rendering sphere spread angle θ2 based on the equation 1. Spread angle calculation unit 16 for calculating
And a vector angle calculation unit 17 that calculates an angle φ formed by both vectors based on Equation 2, a magnitude determination unit 18 that determines the magnitude of the angle φ and the sum of the angles θ1 and θ2, and φ ≦ θ1 + θ
In the case of 2, the rendering sphere is in the field of view, and φ> θ1 +
In the case of θ2, the inside / outside determination unit 19 determines that the rendering sphere is out of the visual field and removes the corresponding rendering sphere. The processing by each of the above components is repeatedly performed for all rendering spheres.

Therefore, it is determined whether the rendering sphere is in or out of the visual field by calculating the angles φ and θ2 for all the rendering spheres and determining the magnitude of the sum of the angles φ and θ1 and θ2. A decision can be made. However, it is preferable to perform the inside / outside determination from the sphere of the highest hierarchy to the rendering sphere in order, and omit the above determination for the rendering sphere generated from the upper sphere determined to be out of the visual field. In this case, the processing for each rendering sphere by each holding unit, each calculation unit, and each determination unit may be performed on a sphere in a higher hierarchy than the rendering sphere. Then, as a result of the above determination, the rendering sphere that is determined to be outside the field of view is removed, and only the polygons of the rendering model that correspond to the rendering sphere that is determined to be within the field of view are drawn. It is possible to achieve high speed.

FIG. 16 is a block diagram showing the arrangement of an apparatus for determining whether or not the polygons of each group are objects to be drawn when the polygons included in the rendering sphere are grouped based on the normal vector. It is a figure. In this block diagram, a processing unit for performing grouping and a processing unit for approximating a set of normal vectors belonging to each grouped section by a cone are not shown. These processing units may perform necessary processing, for example, based on the input value set by the operator. In this case, for example, the spread angle of the cone and the central axis of the cone are defined in the set input value. The central axis vector and so on are included.

This device uses the center C = of the rendering sphere.
The rendering sphere center holding unit 21 that holds (Xc, Yc, Zc), the radius holding unit 22 that holds the radius r of the rendering sphere, and the rendering sphere vector V2 = (the starting point is the viewpoint and the ending point is the center of the rendering sphere). X2, Y2
Rendering sphere vector calculation unit 23 for calculating Z2)
And a first spread angle calculation unit 24 that calculates the spread angle θ2 of the rendering sphere based on Equation 1, (π / 2−θ
The second spread angle calculation unit 25 that calculates the spread angle θ3 of the cone that approximates the range of normal vectors of polygons that are not drawn by performing the calculation of 2), and the range of normal vector of polygons of any group. The center vector holding unit 26 holding the center vector V3 = (X3, Y3, Z3) of the cone defining the above, the third spread angle holding unit 27 holding the spread angle α of this cone, and A vector angle calculator 28 for calculating an angle β formed by the vector,
A magnitude determination unit 29 that determines the magnitude of the angle θ3 and the sum of the angles α and β, a polygon or the like corresponding to the case of θ3 ≦ α + β, and a polygon or the like corresponding to the case of θ3> α + β. Drawing determination unit 30 that determines that it is not a target
And have. The processing by each of the above components is repeatedly performed for all cones.

Therefore, it is possible to easily and quickly determine whether or not the polygons of each grouped group are the objects to be drawn, and it is possible to speed up the rendering. FIG. 17 is a block diagram showing an apparatus for LOD-displaying polygons and the like which form a rendering model. Although a processing unit that creates a plurality of LOD models is not shown in this device, a conventionally known device can be used as this processing unit.

This apparatus comprises, for each rendering model, an LOD model holding section 31 for holding a plurality of LOD models created in advance for corresponding polygons,
A distance calculation unit 32 that obtains a value obtained by subtracting the radius of the rendering sphere from the length of the rendering sphere vector V2, and a size determination unit 33 that determines the size of the calculated distance and a preset threshold value of 1 or 2 or more. And an LOD model selection unit 34 that selects the LOD model of the corresponding LOD layer from the LOD model holding unit in response to the size determination result by the size determination unit 33, and a display unit 35 that performs display based on the selected LOD model. have.

Therefore, by selecting a simplified LOD model for displaying polygons and the like,
The rendering model can be displayed faster.
In the rendering model, a large number of polygons and the like are included in the distant view, and these can be displayed in a simplified manner, so that a significant speedup can be achieved. FIG. 18 is a block diagram showing an example of the hierarchical structure sphere model creating apparatus.

This hierarchical structure sphere model creating apparatus includes a shape data holding unit 41 for holding shape data of a target object,
A diameter holding unit 42 that holds the set value of the diameter of the sphere of the lowest layer, and the shape data of the target object and the diameter of the sphere of the lowest layer are input, and the surface of the target object is divided into meshes by a conventionally known method. A mesh dividing unit 43 that sets the length of each side of each mesh to be equal to or less than the diameter, and a mesh data holding unit 44 that holds a node group obtained by mesh division (for example, holds coordinate values of each node). When,
The highest sphere sphere generation unit 45 that generates the smallest sphere including a cubic area determined based on all the nodes and assigns it as the sphere of the highest tier, and the most newly generated sphere that is included in the unprocessed sphere. A dividing unit 4 that divides this group of nodes into two on a plane that passes through the center of the sphere and that has the normal to the longitudinal direction of the cubic region, provided that the number is plural.
6, a node number determination unit 47 that determines whether or not the number of nodes belonging to each section divided by the division unit 46 is plural, and the division unit 46 divides the image on the condition that the number of nodes is plural. A minimum sphere including a cubic area determined based on all the nodes belonging to each section is generated, and an intermediate layer sphere generation unit 48 that allocates it as a sphere of the next layer, and the number of nodes are 1 It has a lowest layer sphere arranging section 49 for arranging the lowest layer sphere in a state where the nodes and the center are aligned.

The operation of each component is the same as the processing of the corresponding step in the flowcharts of FIGS. 10 and 11, and detailed description thereof will be omitted. Of course, each of these components may be configured by hardware, but is preferably configured by a computer in which a program for performing the processing illustrated in FIGS. 10 and 11 is incorporated. Further, a control unit for repeating the division processing by the division unit 46 and the processing of the intermediate hierarchical sphere generation unit 48 until the number of nodes becomes 1 is omitted. Also, the holding unit that holds each of the two-divided node groups is not shown.

[0067]

According to the first aspect of the present invention, the calculation of the positional relationship using the polyhedron representation is not required, and the interference of the rendering sphere can be calculated to achieve the clip of the render length model at high speed. Has the unique effect of being able to speed up.

According to the invention of claim 2, the spheres included in the object are not originally generated, and the number of generated spheres can be reduced without complicating the processing. The same effect as in claim 1 is obtained based on the hierarchical sphere model created. According to the invention of claim 3, the calculation of the positional relationship using the polyhedron expression is unnecessary, and the removal of the rendering sphere unnecessary for drawing can be easily performed by the high-speed conical-to-conical collision check. It has a unique effect that the detection can be achieved at high speed.

According to the invention of claim 4, the calculation of the positional relationship using the polyhedron representation is unnecessary, the polygon to be drawn can be selected at high speed from the polygons in the rendering sphere, and the rendering can be performed at high speed. There is a unique effect that it can be converted. The invention according to claim 5 has a peculiar effect that the drawing can be speeded up by adopting polygons having low drawing accuracy for some of the polygons, and thus the rendering can be speeded up.

According to the sixth aspect of the present invention, the calculation of the positional relationship using the polyhedron representation is not necessary, and the interference calculation using the rendering sphere enables the clip of the render length model to be achieved at high speed, and the high speed rendering. There is a unique effect that it can be converted. According to the invention of claim 7, the sphere included in the object is not originally generated, and the number of spheres generated can be reduced without complicating the processing, and the hierarchy generated in this way. The same effect as in claim 1 is obtained based on the sphere model.

According to the invention of claim 8, the calculation of the positional relationship using the polyhedron representation is unnecessary, and the removal of the rendering sphere unnecessary for drawing can be simplified by the high-speed cone-to-cone interference check. It has a unique effect that detection of a drawing polygon can be achieved at high speed. According to the invention of claim 9, the calculation of the positional relationship using the polyhedron expression is not required, the polygon to be drawn can be selected at high speed from the polygons in the rendering sphere, and the rendering can be speeded up. There is a unique effect that can be.

According to the tenth aspect of the present invention, by adopting polygons having low drawing accuracy for some of the polygons, it is possible to speed up the drawing, and it is possible to speed up the rendering.

[Brief description of drawings]

FIG. 1 is a schematic diagram showing a relationship between a hierarchical sphere model and a rendering model.

FIG. 2 is a schematic diagram illustrating a state in which the rendering model is separated into a one-to-one correspondence portion and a many-to-one correspondence portion when the rendering sphere and the rendering model do not have a one-to-one correspondence.

FIG. 3 is a schematic diagram showing a rendering model in a state where it is separated into a one-to-one correspondence portion and a many-to-one correspondence portion.

FIG. 4 is a flowchart illustrating clip processing of a rendering model in the high-speed rendering method.

FIG. 5 is a schematic diagram showing a relationship between a field of view approximated by a cone and a rendering sphere.

FIG. 6 is a schematic diagram showing a state in which a range of a normal vector of a polygon or the like that is not drawn is approximated by a cone.

FIG. 7 is a schematic diagram showing the relationship between two cones that approximate the range of a normal vector such as a polygon.

FIG. 8 is a schematic diagram showing a plurality of LOD models such as polygons corresponding to respective rendering models.

FIG. 9 is a diagram showing a relationship between the LOD layer and a value obtained by subtracting the radius of the rendering sphere from the length of the rendering sphere vector.

FIG. 10 is a flowchart illustrating a part of an example of a hierarchical sphere model creating method.

FIG. 11 is a flowchart illustrating the rest of the example of the method for creating a hierarchical sphere model.

FIG. 12 is a schematic diagram illustrating a generation process of a sphere having a minimum radius including a cubic region and a two-division process of all nodes.

FIG. 13 is a schematic diagram illustrating mesh division of the object surface and arrangement of spheres in the lowest hierarchy.

FIG. 14 is a block diagram showing an embodiment of a high-speed rendering device of the present invention.

FIG. 15 is a block diagram showing a configuration of an apparatus for determining whether a rendering sphere is inside or outside a visual field approximated by a cone.

FIG. 16 is a block diagram showing a configuration of an apparatus for determining whether or not polygons in each group are objects to be drawn when polygons included in a rendering sphere are grouped based on normal vectors. is there.

FIG. 17 is a block diagram showing an apparatus for LOD-displaying polygons and the like which form a rendering model.

FIG. 18 is a block diagram showing an example of a hierarchical sphere model creation device.

[Explanation of symbols]

2 rendering sphere setting unit 3 rendering model associating unit 5 inside / outside determination unit 6 clip processing unit 12 line-of-sight direction vector holding unit 13 viewing angle holding unit 15 rendering sphere vector calculation unit 16 spread angle calculation unit 18 large / small determination unit 19 inside / outside determination unit 24 First spread angle calculation unit 25 Second spread angle calculation unit 26 Center vector holding unit 27 Third spread angle holding unit 29 Large / small determination unit 30 Drawing determination unit 31 LOD model holding unit 32 Distance calculation unit 33 Large / small determination unit 34 LOD model selection Part 35 Display Part 43 Mesh Dividing Part 45 Highest Hierarchical Sphere Generating Part 46 Dividing Part 48 Intermediate Hierarchical Sphere Generating Part 49 Lowermost Hierarchical Sphere Generating Part

Claims (10)

[Claims] 1. A sphere of the highest hierarchy among spheres of a predetermined radius or less of a hierarchical sphere model is designated as a rendering sphere, and a rendering model is assigned corresponding to each rendering sphere. A high-speed rendering method characterized by determining whether the area is inside or outside a clipping area and clipping the corresponding rendering model when the rendering sphere is outside the clipping area. 2. A surface of an object to be modeled is mesh-divided, and the lengths of all sides of each mesh that is mesh-divided are set to be equal to or less than the preset diameter of the sphere of the lowest hierarchy, Generate the sphere of the lowest hierarchy centered on each node generated by the mesh division, then generate the smallest sphere including the cubic area determined based on all the nodes and make it the sphere of the highest hierarchy, then the previous hierarchy All nodes are divided into two in a plane that passes through the center of the sphere, and the smallest sphere that includes the cubic area that is determined based on all the nodes belonging to each divided segment is generated, and the sphere of the layer next to the layer before division is generated. Then, the process of dividing all nodes into two and the process of generating the smallest sphere including the cubic area determined based on all the nodes belonging to each segment and using it as the sphere of the next layer is performed. Repeat until The high-speed rendering method according to claim 1, wherein the hierarchical sphere model is created by creating the hierarchical sphere model. 3. A vector from the viewpoint to the center of the rendering sphere and a spread of the rendering sphere centered on this vector in a state where the visual field is defined by the line-of-sight direction vector and the viewing angle with reference to the viewpoint position. 3. The angle is calculated, and the corresponding rendering sphere is removed under the condition that the angle formed by the two vectors is larger than the sum of the viewing angle and the angle defining the spread of the rendering sphere. Fast rendering method. 4. A rendering sphere, at least a part of which is included in the field of view, divides polygons included in the rendering sphere into a plurality of groups based on directions of respective normal vectors, and modifies polygons belonging to each division. Approximate a set of lines with a cone and calculate a first spread angle that defines the spread of the rendering sphere centered on the vector from the viewpoint to the center of the rendering sphere. The central axis vector that defines the central axis and the second spread angle of this cone are calculated, and the angle between the vector going from the viewpoint to the center of the rendering sphere and the central axis vector is less than or equal to the sum of both spread angles. The high-speed render according to any one of claims 1 to 3, wherein a polygon belonging to a corresponding section is selected as a drawing target. Ring method. 5. A polygon corresponding to each rendering model is generated in advance with a plurality of polygons having different drawing accuracies corresponding to the size of a vector from the viewpoint to the center of the rendering sphere, and the actual vector of the vector is generated. The high-speed rendering method according to any one of claims 1 to 4, wherein a polygon having a corresponding drawing accuracy is drawn based on the size. 6. A rendering sphere designating means (2) for designating a sphere of a highest hierarchy among spheres having a predetermined radius or less in a hierarchical sphere model, and a rendering model corresponding to each rendering sphere. A rendering model assigning means (3) for assigning, an inside / outside determining means (5) for determining whether the rendering sphere is inside or outside the clip area, and an inside / outside determining means (5) for indicating that the rendering sphere is outside the clip area. A high-speed rendering device, comprising: clipping means (6) for clipping a corresponding rendering model in response to a determination result. 7. A mesh in which the surface of an object to be modeled is mesh-divided, and the lengths of all the sides of each mesh-divided mesh are set to be equal to or less than the preset diameter of the sphere of the lowest hierarchy. A dividing means (43), a lowest level sphere generating means (49) for generating a sphere of the lowest level centering on each node generated by the mesh division, and a minimum sphere including a cubic area determined based on all nodes. Top-level sphere generation means (45) for generating and forming the top-level sphere, division means (46) for dividing all nodes into two on a plane passing through the center of the sphere of the immediately preceding level, and each divided division Intermediate sphere generating means (48) and a dividing means (46) for generating a minimum sphere including a cubic area determined based on all the nodes belonging to the sphere and making it a sphere of a layer next to the layer before division.
7. The high-speed rendering device according to claim 6, further comprising: an iterative control unit that iterates the process according to 1. and the process by the intermediate layer sphere generating unit (48) until the number of nodes after division becomes one. 8. Holding means (12) (13) for holding a line-of-sight direction vector and a viewing angle set on the basis of a viewpoint position for defining a field of view, a vector from the viewpoint to the center of the rendering sphere, and Whether the calculation means (15) (16) for calculating the angle defining the spread of the rendering sphere around the vector and the angle formed by both vectors are larger than the sum of the viewing angle and the angle defining the spread of the rendering sphere. First magnitude determining means (18) for determining whether or not
And that the angle formed by both vectors is larger than the sum of the viewing angle and the angle that defines the spread of the rendering sphere.
Rendering sphere removing means (1) for removing the corresponding rendering sphere in response to the determination result of the size determining means (18).
The high-speed rendering device according to claim 6 or 7, further comprising 9). 9. A rendering sphere, at least a part of which is included in the field of view, and a partitioning unit that partitions a polygon included in the rendering sphere into a plurality of groups based on the directions of respective normal vectors, and a partitioning unit belonging to each partition. Approximating means for approximating a set of polygon normals with a cone, and first calculating means (24) for calculating a first spread angle that defines the spread of the rendering sphere centered on a vector from the viewpoint to the center of the rendering sphere. , Second calculating means (25) (26) (27) for calculating a central axis vector that defines a central axis of a cone that approximates a set of normals of polygons, and a second spread angle of this cone, and rendering from a viewpoint Second magnitude determination means (29) for determining whether or not the angle formed by the vector toward the center of the sphere and the central axis vector is less than or equal to the sum of both spread angles. And a polygon belonging to the corresponding section in response to the determination result of the second magnitude determining means indicating that the angle formed by the vector from the viewpoint to the center of the rendering sphere and the central axis vector is less than or equal to the sum of both spread angles. 9. The high-speed rendering device according to claim 6, further comprising a selection unit (30) for selecting a drawing target. 10. A polygon generation means (3) for generating a plurality of polygons having different drawing precisions in advance, corresponding to the size of a vector from the viewpoint to the center of the rendering sphere, for the polygon corresponding to each rendering model.
1) and a drawing means (32) (3) for drawing a polygon having a corresponding drawing accuracy based on the actual size of the vector.
The high-speed rendering device according to claim 6, further comprising 3), (34) and (35).