JP2001273523A - Device and method for reducing three-dimensional data - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize an efficient plotting processing and to provide a smooth three-dimensional walk-through by providing a means which reduces polygon data which is always invisible to a range where a viewpoint can be moved in the three-dimensional walk-through of a computer. SOLUTION: A polygon data reduction processing means detects polygon data invisible to all positions in the range, which is inputted from an input means and where the viewpoint can be moved, and deletes detected invisible polygon data and deletes a three-dimensional object, which has polygon data all deleted, from layout data. Thus polygon data which are always invisible to the range where the viewpoint can be moved is reduced in the three- dimensional walk-through of the computer.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a computer-based method for deleting invisible polygon data from three-dimensional data.
The present invention relates to an apparatus and a method for reducing dimension data.

[0002]

2. Description of the Related Art In a conventional real-time three-dimensional application for performing a three-dimensional walkthrough or the like, a culling process in which data unnecessary for drawing is excluded from calculation for each drawing frame in order to reduce the drawing time of three-dimensional data. It is carried out. For culling, back face culling that does not draw a polygon facing down from the viewpoint, view frustum culling that excludes objects that are not in view from the calculation target, and objects that are hidden by other objects are excluded from calculation There is a technique such as occlusion culling. Generally, the culling process time increases as the number of polygons and objects increases. Here, since the culling is performed every frame, if the culling processing time increases, the drawing time also increases.

The OpenGL Optimizer box uses a spatial grid algorithm and SRA (Success
ive Relaxation Algorithm). "OpenGL
According to the Optimizer Programmer's Guide, the spatial grid method is a method of setting a spatial grid that divides a three-dimensional space into a mesh and moving the vertices forming the three-dimensional object to the nearest grid point. Also, SRA is
This method checks the deviation between the polygon and the polygon when each vertex is deleted, deletes the polygon from the vertex having the smallest deviation, and terminates when the ratio of the number of polygons reaches a specified value. Neither technique is a technique for deleting data that is always invisible during a three-dimensional walkthrough.

In the "image processing apparatus" described in Japanese Patent Application Laid-Open No. H11-232485, when sending polygon data to a graphics pipeline, attention is paid to the order in which visible polygon data and invisible polygon data are lined up. Efficient rendering of three-dimensional data is performed while avoiding bottlenecks in lines.

[0005]

In the above-mentioned conventional culling, a culling process is executed each time a picture is drawn to increase the drawing speed. The three-dimensional object to be drawn or culled must be completely drawn, such as a polygon that always faces down from the viewpoint or a three-dimensional object that is completely hidden by other three-dimensional objects. There is also data that does not exist, and the drawing process and culling process on the data that is not drawn at all are useless processes. In addition, the culling processing time increases as the number of polygons and objects increases. In particular, CAD data often includes data inside objects that are not necessary for the 3D walk-through application, and measures to remove unnecessary data manually are required to increase the drawing speed. And

Further, in the above-described OpenGL Optimizer 3D data simplification method, since the relationship between the 3D data and the viewpoint position is not taken into account, the degree of detail of the entire 3D data is reduced, and The level of detail also decreases. In addition, data that is always invisible may remain. That is, unnecessary data may remain and necessary data may be reduced.

In the means described in Japanese Patent Application Laid-Open No. H11-232485, the invisible polygon causes a bottleneck in the graphics pipeline, but does not operate to delete the invisible polygon itself.

[0008] The present invention has been made in view of such problems of the prior art, and pre-processing capable of realizing a more efficient drawing process and a smoother three-dimensional walk-through than conventional devices and methods. In other words, it is an object of the present invention to provide a three-dimensional data reduction apparatus and method capable of deleting polygon data that is not drawn even when viewed from anywhere in the range in which the viewpoint can move.

[0009]

Means for Solving the Problems To solve the above problems, the present invention has the following features. That is, the present invention provides an input means for inputting a three-dimensional object, three-dimensional object arrangement data, and a viewpoint movable range, and a polygon from the three-dimensional object and the three-dimensional object arrangement data based on the input viewpoint movable range. In a three-dimensional data reduction apparatus having polygon data reduction processing means for performing data reduction processing, and output means for outputting a three-dimensional object and three-dimensional object arrangement data in which the polygon data has been reduced, the polygon data reduction processing means includes: Storage means for storing the three-dimensional object and the three-dimensional object arrangement data and the viewpoint movable range input from the input means, and invisible from all the ranges within the viewpoint movable range stored in the storage means To detect polygon data A polygon detection means visual, and having an invisible polygon data deleting means for deleting invisible polygon data detected by said invisible polygon data detecting means from the 3-dimensional object and the three-dimensional object arrangement data.

[0010] Another feature of the present invention is an input step of inputting a three-dimensional object, three-dimensional object arrangement data, and a viewpoint movable range to the three-dimensional data reduction device.
A polygon data reduction step of reducing polygon data from the three-dimensional object and the three-dimensional object arrangement data based on the input viewpoint movable range; and outputting a three-dimensional object and three-dimensional object arrangement data with reduced polygon data In the three-dimensional data reduction method having an output step, the polygon data reduction step stores the three-dimensional object, the three-dimensional object arrangement data, and the viewpoint movable range input from the input step in storage means, An invisible polygon data detecting step of detecting invisible polygon data from all ranges within the viewpoint movable range stored in the storage means;
An invisible polygon data deleting step of deleting the invisible polygon data detected by the invisible polygon data detecting step from the three-dimensional object and the three-dimensional object arrangement data.

According to the present invention, it is possible to reduce polygon data which is not drawn at all, without changing the appearance of a three-dimensional object when viewed from within the viewpoint movable range. By applying the present invention as pre-processing for three-dimensional walkthrough, rendering processing can be sped up without any change in appearance.

Further, the present invention is applied as preprocessing before reading a three-dimensional object and three-dimensional object arrangement data into a three-dimensional walk-through application. That is, according to the present invention, as the pre-processing of the three-dimensional walkthrough, it is possible to delete polygon data that is not drawn even when viewed from anywhere in the range in which the viewpoint can move. It works to shorten the calculation time for drawing a three-dimensional space, and realizes a smoother three-dimensional walk-through and can display more three-dimensional objects on a display monitor.

Further, according to the present invention, since the calculation processing in each viewpoint is independent of each other, it is possible to increase the speed by parallelizing the processing. Similarly, since the calculation processing in each line-of-sight direction is independent of each other, the processing can be parallelized to increase the speed.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A three-dimensional image generating apparatus and method according to an embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a functional block diagram of a three-dimensional data reduction device according to the present invention.

The three-dimensional data reduction apparatus according to the present invention comprises: an input means 1 for inputting a three-dimensional object, three-dimensional object arrangement data, and a movable range of a viewpoint; a polygon data reduction processing means 2 for reducing polygon data; Output means 3 for outputting a three-dimensional object and three-dimensional object arrangement data in which is reduced. The input means 1 is a means for inputting a three-dimensional object, three-dimensional object arrangement data, and viewpoint movable range data.

The polygon data reduction processing means 2 has a storage means 4, an invisible polygon data detecting means 5, and an invisible polygon data deleting means 6. The storage unit 4 is a unit that stores the three-dimensional object, the three-dimensional object arrangement data, and the viewpoint movable range input from the input unit 1. When the three-dimensional object stored in the storage means 4 is arranged in accordance with the three-dimensional object arrangement data stored in the storage means 4, the invisible polygon data detection means 5 is a polygon which is invisible from the viewpoint movable range. It is a means for detecting data. The invisible polygon data deleting unit 6 is a unit that deletes the polygon data stored in the invisible polygon data detecting unit 5 from the three-dimensional object stored in the storage unit 4.

The input means 1 can be realized by inputting a file. The output means 3 can be realized by file output. The storage unit 4 is realized by a hard disk or a memory. Invisible polygon data detection means 5
The method of implementing the invisible polygon data deleting means 6 will be described in detail below.

First, the operation of the three-dimensional data reduction device shown in FIG. 1 will be described with reference to FIG. First, in step 100, as the processing of the input means 1, input processing of a three-dimensional object, three-dimensional object arrangement data, and viewpoint movable range data is performed and stored in the storage means 4. The three-dimensional object and the three-dimensional object arrangement data are:
VRML97 (ISO / IEC 14772-1: 1997, Virtual Reali
3D data formats such as ty Modeling Language (VRML97)) can be used. The viewpoint movable range data is a set of three-dimensional point information including X coordinates, Y coordinates, and Z coordinates, and is expressed as shown in Table 1 below.

[0019]

[Table 1]

Next, in step 110, as processing of the invisible polygon data detecting means 5, invisible polygon data detection processing for detecting invisible polygon data from all positions within the movable range of the viewpoint is performed.

Next, as processing of the invisible polygon data deleting means 6, the polygon data detected by the invisible polygon data detecting means 5 is deleted from the three-dimensional object inputted by the input means 1 (step 120). If all the polygon data in is deleted, the three-dimensional object arrangement data is deleted from the three-dimensional object arrangement data input by the input means 1 (step 130). Next, as the processing of the output unit 3, the three-dimensional object after the step 120 of the invisible polygon data deletion unit 6 is applied and the three-dimensional object after the step 130 of the invisible polygon data deletion unit 6 is applied.
Output the dimensional object arrangement data (step 14)
0).

The input processing of the three-dimensional object, the three-dimensional object arrangement data, and the movable range of the viewpoint (step 100) in FIG.
The input processing of the three-dimensional object, the three-dimensional object arrangement data, and the viewpoint movable range data is performed.

The output processing (step 140) of the three-dimensional object and the three-dimensional object arrangement data in FIG.
The output processing of the three-dimensional object and the three-dimensional object arrangement data is performed using the file input / output processing.

FIG. 3 is a flowchart showing the detailed processing of the invisible polygon data detection processing (step 110). Detection of invisible polygon data is realized by tracing all lines of sight that come out of the movable range of the viewpoint. Here, paying attention to all gazes coming out of the viewpoint movable range pass through the surface of the viewpoint movable range, tracking all the gazes coming out of the viewpoint movable range requires a point on the surface of the viewpoint movable range. It can be said that it suffices to track the line of sight from. Therefore, first, the constituent surface of the viewpoint movable range is obtained from the viewpoint movable range data stored in the storage unit 4 and stored in the storage unit 4 (step 200).

From the storage means 4, a constituent surface of the unprocessed viewpoint movable range is extracted (step 210). next,
The viewpoint position for tracking the line of sight is determined as one point on the constituent surface of the movable range of the viewpoint extracted in step 210.

For example, considering the intersection of a plane forming the movable range of the viewpoint and a plane XZ parallel to the XZ plane as shown in FIG. 4, one line segment can be determined. Steps 230 and subsequent steps are applied while sequentially searching points on the line segment, for example, from left to right. When the processing for one line segment is completed, the plane XZ is moved along the Y axis with a sufficiently small width, a new line segment is similarly determined, and steps 230 and subsequent steps are sequentially applied to points on the line segment. When a plane constituting the movable range of the viewpoint is parallel to the XZ plane, X-
Similar processing is performed using the Y plane. Thus, all positions on the surface can be calculated as viewpoints (step 220).

Next, a line of sight from a point on the surface of the movable range of the viewpoint to the inside of the movable range of the viewpoint is always included in a line of sight from a point on the other surface of the movable range of the viewpoint.
Only the direction of the line of sight toward the outside of the movable range of the viewpoint is set as the tracking target with the viewpoint on the surface of the movable range of the viewpoint.

For example, as shown in FIG. 5, consider a local coordinate system in which the plane on which the viewpoint exists is an XY plane and the normal vector Z axis. Here, using the normal vector as the reference of the line-of-sight vector, a rotation of ± 90 degrees around the X axis and ± 90 degrees around the Y axis.
The range of the rotation of the degree is the line of sight vector of the tracking target. Therefore, the rotation angle around the X-axis and the rotation angle around the Y-axis are set at a sufficiently small interval while rotating the line-of-sight direction at step 24.
By applying 0 or less, it is possible to calculate for all the gaze directions that are tracking targets (step 230).

Next, the intersection between the line-of-sight vector and the polygon data of each three-dimensional object is determined, and a list of intersecting polygon data is created. Next, the list is sorted in ascending order from the viewpoint. Next, the angle between the line-of-sight vector and the normal vector of the polygon data is determined in the order from the viewpoint of the list.

Here, as shown in FIG. 6, since the angle between the line-of-sight vector and the normal vector of the polygon data is visible when it is less than 90 degrees, the angle between the line-of-sight vector and the normal vector of the polygon data is less than 90 degrees. Is determined as visible polygon data and stored in the storage means 4. Subsequent polygon data in the list is invisible because it is located behind the visible polygon data (step 240).

It is determined whether or not step 240 has been executed for all the line-of-sight vectors to be tracked. If step 240 has not been executed for all the line-of-sight vectors to be tracked, the process returns to step 230 to process the next line-of-sight vector (step 250).

It is determined whether or not step 230 has been executed for all points on the constituent surface of the viewpoint movable range where the viewpoint exists, and step 230 is executed for all points on the constituent surface of the viewpoint movable range where the viewpoint exists. If not, the process returns to step 220, and the process is performed for the next viewpoint position (step 250).

It is determined whether or not step 220 has been executed for all the constituent planes of the viewpoint movable range. If step 220 has not been executed for all the constituent planes of the viewpoint movable range, the process returns to step 210, and the next viewpoint movable can be performed. The process is performed on all constituent surfaces of the range (step 26)
0).

Next, an invisible polygon is determined by removing the visible polygon data stored in the storage means 4 in step 250 from all the polygon data constituting the three-dimensional object input by the input means 1 (step 2).
60).

Next, the processing in the flowcharts of FIGS. 2 and 3 will be described with reference to FIGS.

For example, by inputting a three-dimensional object, three-dimensional object arrangement data, and a viewpoint movable range (step 100), four cubes A, B,
When C, D, and E are input as three-dimensional objects, the viewpoint movable range S is also input as a cube, and three-dimensional object arrangement data arranged as shown in FIG. If it is as shown in FIG. 8 when viewed from the direction and as shown in FIG. 9 when viewed from the Y-axis direction, there are six constituent surfaces because the view point movable range S is a cube (step 200).

Here, the unprocessed viewpoint movable range S
Is determined (step 210). Here, the viewpoint P is one point on the constituent surface of the viewpoint movable range S as shown in FIG. 10, and the following processing is sequentially performed for each point on the constituent surface (step 220).

When the position of the viewpoint P is viewed from the Z-axis direction, FIG.
It looks like 1. Here, the direction of the line of sight to be tracked has a range of ± 90 degrees in the vertical and horizontal directions around the center of the component plane where the viewpoint P exists in the direction of the normal vector with the viewpoint P as the reference position.

FIG. 12 shows some examples of gaze directions to be tracked (step 230). Here, the three-dimensional objects located on the front side of the constituent surface where the viewpoint P exists are A and C.

When the line of sight is tracked, the position of the viewpoint P is located above the upper surfaces of the three-dimensional objects A and C. Therefore, as shown in FIG. Determined and saved (step 240). Since the three-dimensional objects B, D, and E are located behind the plane on which the viewpoint is located, there is no visible polygon data. It is determined whether or not tracking of all lines of sight has been completed, and if not, tracking of the next line of sight is performed (step 250). Once all gazes have been tracked,
It is determined whether or not processing for all viewpoints has been completed, and if not, processing is performed for the next viewpoint (step 26).
0).

When processing has been completed for all viewpoints, it is determined whether or not processing has been completed for all constituent surfaces. If processing has not been completed, processing for the next constituent surface is performed (step 27).
0).

When the processing of all the constituent surfaces is completed, FIG.
3D objects A, B, C, D
Since the polygon data of each surface becomes visible, the line of sight of the three-dimensional object E is always interrupted by the polygon data of the three-dimensional object D, and there is no visible polygon.

When the visible polygon data thus obtained is excluded from all polygon data, invisible polygon data that is invisible from all positions inside the view point movable range S is determined as shown in FIG. 280).

The invisible polygon data is deleted from all the polygon data (step 120), and the arrangement data of the three-dimensional object E from which all the constituent polygon data are deleted is also deleted (step 130). The dimensional object arrangement data is output (step 140).

When the output three-dimensional object and three-dimensional object arrangement data are read by the three-dimensional walkthrough application, the result is as shown in FIG. In this way, the data can be reduced without any change in appearance as seen from inside the viewpoint movable range S before and after the processing.

In this example, the input three-dimensional object has five cubes (hexahedrons) and a total of 30 polygons. However, the number of visible polygon data of the three-dimensional objects A, B, C, and D is increased by the above processing. Each of the polygons has 4 polygons, for a total of 16 polygons. That is, the data of 14 polygons could be reduced.

[0047]

As described above, according to the present invention,
As a pre-process of the three-dimensional walkthrough, it is possible to delete polygon data that is not drawn even when viewed from anywhere within the movable range of the viewpoint. As a result, the calculation time for rendering in a three-dimensional space at the time of executing the three-dimensional walk-through is reduced, realizing a smoother three-dimensional walk-through and displaying more three-dimensional objects on a display monitor. can do.

Further, according to the present invention, since the calculation processing in each viewpoint is independent of each other, it is possible to increase the speed by parallelizing the processing. Similarly, since the calculation processing in each line-of-sight direction is independent of each other, the processing can be parallelized to increase the speed.

[Brief description of the drawings]

FIG. 1 is a functional block diagram of a three-dimensional data reduction device according to an embodiment of the present invention.

FIG. 2 is a flowchart of a three-dimensional data reduction method according to an embodiment of the present invention.

FIG. 3 is a detailed flowchart of the invisible polygon data detection processing of FIG. 2;

FIG. 4 is a diagram illustrating an example of processing for determining a viewpoint position.

FIG. 5 is a diagram illustrating an example of a range of a line-of-sight vector to be tracked at a viewpoint P;

FIG. 6 is a diagram illustrating an example of a state in which a polygon is visible.

FIG. 7 is a diagram illustrating an example of a three-dimensional object and a movable range of a viewpoint.

8 is a diagram illustrating an example of a positional relationship between a three-dimensional object and a viewpoint movable range in FIG. 7 when viewed along a Z axis.

FIG. 9 is a diagram illustrating an example of a positional relationship between the three-dimensional object and a viewable range in FIG. 7 when viewed along a Y-axis;

FIG. 10 is a diagram illustrating an example of a viewpoint on a constituent surface of a viewpoint movable range in FIG. 7;

11 is a diagram illustrating an example of the viewpoint position in FIG. 9 viewed along the Z axis.

FIG. 12 is a diagram illustrating an example of a tracking line-of-sight direction at the time of FIG. 9;

FIG. 13 is a diagram showing an example of visible polygon data at the time of FIG. 11;

FIG. 14 is a diagram illustrating an example of visible polygon data when tracking of a line of sight has been completed for all viewpoints.

FIG. 15 is a diagram showing an example of invisible polygon data as viewed from all positions within a view point movable range.

FIG. 16 is a diagram illustrating an example of a three-dimensional object after deleting invisible polygon data.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Input means 2 ... Polygon data reduction processing means 3 ... Output means 4 ... Storage means 5 ... Invisible polygon data detection means 6 ... Invisible polygon data deletion means

Claims (2)

[Claims] An input means for inputting a three-dimensional object, three-dimensional object arrangement data and a viewpoint movable range, and polygon data from the three-dimensional object and the three-dimensional object arrangement data based on the input viewpoint movable range. A three-dimensional data reduction apparatus comprising: a polygon data reduction processing means for reducing the polygon data; and an output means for outputting a three-dimensional object and three-dimensional object arrangement data in which the polygon data has been reduced. A storage unit for storing the three-dimensional object, the three-dimensional object arrangement data, and the viewpoint movable range input from the input unit; and an invisible from all ranges in the viewpoint movable range stored in the storage unit. Invisible poly to detect certain polygon data A three-dimensional data reduction device, comprising: a gon detecting unit; and an invisible polygon data deleting unit that deletes the invisible polygon data detected by the invisible polygon data detecting unit from the three-dimensional object and the three-dimensional object arrangement data. . 2. An inputting step of inputting a three-dimensional object, three-dimensional object arrangement data, and a viewpoint movable range to a three-dimensional data reduction device, and the three-dimensional object and the three-dimensional object based on the input viewpoint movable range.
A polygon data reduction step of reducing polygon data from the three-dimensional object arrangement data; and a three-dimensional object reduction method having a three-dimensional object with reduced polygon data and an output step of outputting three-dimensional object arrangement data. A storage step of storing the three-dimensional object, the three-dimensional object arrangement data, and the viewpoint movable range input from the input step in a storage unit, and all the ranges within the viewpoint movable range stored in the storage unit. An invisible polygon data detecting step of detecting invisible polygon data from the object; and deleting the invisible polygon data detected by the invisible polygon data detecting step from the three-dimensional object and the three-dimensional object arrangement data. A three-dimensional data reduction method, comprising a step of deleting invisible polygon data.