JP5616198B2 - Method and apparatus for generating appearance display image of same feature having different level of detail - Google Patents

Description

  The present invention relates to a method and an apparatus for generating an appearance display image of the same feature having a different level of detail when forming a three-dimensional city model using a computer.

A three-dimensional city model includes (i) a polygon having a three-dimensional coordinate value representing a building, topography, etc., (ii) a texture to be pasted on the polygon, and (iii) a road center line and signal control information. It means data consisting of three elements: various information to be formed. Since this three-dimensional city model simulates a virtual city or a real city, it is created for use in a simulation field such as a driving simulator or a flood simulator.
Each model to be displayed on the monitor screen of the computer is displayed by switching a plurality of different levels of detail according to the distance to the virtual viewpoint for viewing the model, and has three-dimensional data for that purpose. The level of detail is higher as the distance of the viewpoint or the camera viewed from the feature is shorter, and the detail is lower.
As described above, the three-dimensional data is composed of polygons such as buildings and terrain and the texture of the surface image thereof. Each model polygon has three-dimensional data having a plurality of different levels of detail.
At this time, it is expected that it is extremely complicated and time-consuming to prepare textures of three-dimensional data with different levels of detail in advance for each polygon. Therefore, if the texture used for the model data with the high degree of detail is pasted on the model data with the low degree of detail, it is possible to save the trouble of creating the texture to be attached to all the polygons for all the details. However, model data with a high degree of detail cannot be simulated near reality unless it is possible to display even complex parts, and a different texture file is used for each such complex part. It takes a lot of work to edit and organize such texture files.

  The problem to be solved is to efficiently generate appearance display images of the same feature having different levels of detail.

  In order to solve the above-mentioned problem, the method for generating an appearance display image of the same feature having different levels of detail according to claim 1 is a three-dimensional model having an appearance display image with the highest level of detail among one target feature. A first process of arranging the three-dimensional model in the virtual space and immediately before the level of detail of the three-dimensional model is switched when photographing the three-dimensional model of the camera in the virtual space arranged to photograph the three-dimensional model. A second step of photographing the entire circumference of the three-dimensional model in a plurality of N times with an angle of view within a range in which the three-dimensional model of the feature is contained, and a tertiary image taken in the second step. The third process of rendering N times with the appearance display image with the level of detail at the time when the original model was photographed, and the level of detail of the 3D model is switched from the distance rendered in the previous process to a lower level of detail. Cut off The distance to the point immediately before switching is taken, and the entire circumference of the 3D model is photographed in multiple times with the angle of view within the range in which the 3D model of the feature is contained, and the image of the feature obtained in the previous process is taken. It is characterized by comprising a fourth process in which rendering using the texture of the three-dimensional model is repeated to a predetermined low level of detail.

In order to solve the above problem, an apparatus for generating an appearance display image of the same feature having different levels of detail according to claim 2 stores an appearance display image as texture data having the highest level of detail for the feature (surface image) A texture data storage unit for recording electronic data),
A 3D model data storage unit for storing a 3D model which is geometric information data simulating the feature based on the photographed feature;
Parameters for rendering the entire perimeter of the 3D model by shooting the 3D model placed in the virtual space with a virtual camera placed in the virtual space at a predetermined distance, position, orientation, and angle of view. Camera parameter editing means for determining the number N of combinations;
The 3D model read from the 3D model data storage unit is rendered with the texture read from the texture data storage unit according to the parameters determined by the camera parameter editing means, and an image is generated. An image for displaying the appearance of the same feature having different levels of detail, comprising image generation means for storing in the texture data storage unit together with the level of detail as texture data of the level of detail of the 3D model lower than that of the 3D model The camera parameter editing means is separated from the distance rendered by the image generation means to a distance immediately before the level of detail of the three-dimensional model is switched to a lower level and further switched to a level of lower detail. , The range in which the 3D model of the feature fits The entire circumference of the three-dimensional model is photographed in a plurality of times at an angle of view, and the image generation means renders using the obtained feature image read from the texture data storage unit as the texture of the three-dimensional model It is characterized by repeating this to a predetermined low level of detail .

  According to the method for generating an appearance display image of the same feature having different levels of detail according to claim 1, up to a distance immediately before the level of detail of the three-dimensional model is switched when the camera in the virtual space captures the three-dimensional model. Since it is rendered N times by the appearance display image having the detail level at the time of shooting, the appearance display image can be efficiently generated every time the detail level is different.

  According to the apparatus for generating an appearance display image of the same feature having different levels of detail according to claim 2, the tertiary read out from the 3D model data storage unit according to the parameter determined by the camera parameter editing unit in the image generation unit The original model is rendered to generate an image, and the generated image is stored in the texture data storage unit as texture data of a three-dimensional model having a lower level of detail than the current three-dimensional model. A display image can be generated efficiently.

It is a flowchart explaining the production | generation method of the image for a feature appearance display. It is a functional block diagram explaining the production | generation apparatus of the external appearance display image of a feature. It is a figure explaining the image for appearance display of a feature. It is a figure explaining the pasting of the texture from which a detail degree differs.

  Generation of the appearance display image of the feature is realized by a function of the computer.

  FIG. 1 is a flowchart illustrating an embodiment of the method of the present invention. FIG. 2 is a functional block diagram of an embodiment of the apparatus of the present invention. In FIG. 2, 101 is a texture data storage unit, 102 is a model data storage unit, 103 is camera parameter editing means, and 104 is image generation means.

  The texture data storage unit 101 stores electronic data of each appearance display image as texture data having the highest degree of detail for a predetermined number of features. The texture data is a texture obtained from a photograph taken from a distance at which the level of detail that is planned to be used in this system for each feature is highest. The three-dimensional model data storage unit 102 stores a three-dimensional model that is geometric information data simulating the feature based on the photographed feature. The 3D model data is geometric information data (polygons) that electronically simulates real features that have been captured in the photograph at the position where the level of detail is the highest in advance. Includes three-dimensional coordinate values such as latitude, longitude, height, and orientation. The camera parameter editing unit 103 reads a predetermined three-dimensional model read from the three-dimensional model data storage unit 102 and arranged in the virtual space with a predetermined distance, position, orientation, and image by a virtual camera arranged in the virtual space. The number N of combinations of parameters for photographing at the corner and rendering the entire periphery of the three-dimensional model is determined. The image generation unit 104 generates an image by rendering the 3D model read from the 3D model data storage unit 102 with the texture read from the texture data storage unit 101 in accordance with the parameters determined by the camera parameter editing unit 103 At the same time, the generated image is stored in the texture data storage unit 102 together with the degree of detail as texture data of a three-dimensional model having a degree of detail lower than that of the current three-dimensional model.

A three-dimensional model of a predetermined feature is read from the three-dimensional model data storage unit 102 and arranged in the virtual space (P101 in FIG. 1). The three-dimensional model data at this time is stored as the highest level of detail among the features. Along with the arrangement in P101, the 3D model also includes information on how to paste the texture. From the 3D texture data storage unit 101, the texture of the 3D model of the predetermined feature is stored. The appearance display image is also read at the same time. This texture has the highest level of detail corresponding to the 3D model.
Here, the camera parameter editing unit 103 determines a combination of parameters such as the camera position, orientation, and angle of view necessary for rendering in accordance with the size, such as width and height, of the read three-dimensional model. The total number N of combinations of these parameters is prepared so that the entire periphery of the three-dimensional model is sufficiently rendered. The virtual camera in the virtual space is separated to the distance immediately before the level of detail of the 3D model having the highest level of detail display image when the 3D model is photographed, and the 3D model of the feature The entire circumference of the three-dimensional model is photographed N times a plurality of times at an angle of view within a range (P102 in FIG. 1). This number N is determined by a combination of parameters for rendering the entire perimeter of the 3D model. The maximum number of renderings is the same as the number of surfaces perpendicular to the camera direction, that is, the model surfaces. The angle of view when rendering varies depending on the model shape.
As shown in Fig. 3, the relationship between the level of detail and the distance of the 3D model switches the level of detail of the 3D data to be displayed according to the distance between the model and the camera. Display three-dimensional data with a lower level of detail as the distance increases. Even if three-dimensional data with a high degree of detail is always displayed, if the distance is long, it will be displayed small on a monitor (not shown), and the details will be indistinguishable to the human eye, rendering processing will be wasted. . The present invention is to avoid this.
Data with a low degree of detail is data with a small number of polygons, and data with a high degree of detail is data with a large number of polygons. At this time, data with a high degree of detail may use different textures (appearance display images) for each part without using the textures combined into one sheet. The reason is
A. Since the unevenness is complicated, it is easier to use a plurality of textures for each polygon than to use all the polygonal textures together.
B. I want to share the same texture partially on different features.
Here, when it is desired to create data with a low level of detail, if a plurality of textures are used with data with a high level of detail, the texture cannot be used as it is for data with a low level of detail. The reason is that in the case of a normal texture mapping method in which texture coordinates are assigned for each vertex of a polygon, texture coordinates cannot be assigned to other than the vertex of the polygon. For example, if the data with low level of detail has 4 vertices and you want to paste two textures that use data with high level of detail here, as shown in FIG. Since all 8 vertices (5 polygons) are required, it is only possible to paste the two textures in a state where they are completely overlapped on data with a low level of detail.
Therefore, if you want to paste a texture for data with a high level of detail to data with a low level of detail, you need to edit those textures and create a new texture for the data with a low level of detail. Since it is cumbersome to perform manually for each feature, a texture for data having a low level of detail is generated without a manual operation using data having a high level of detail and a texture used therefor.
The image generation unit 104 reads the 3D model having the highest detail level of the feature read from the 3D model data storage unit 102 from the texture data storage unit 101 in accordance with the parameters determined by the camera parameter editing unit 103. most detailed level to produce an image by rendering a high texture (M1) (P103 in Fig. 1). The image generated for the entire periphery is stored in the texture data storage unit 101 together with the level of detail as texture data (M2) of a level of detail three-dimensional model lower than that of the current three-dimensional model.
When a lower level of detail is require lower three-dimensional models and textures, camera parameter editing unit 103, for example, switched low degree of detail is cut exchange despite further details of the level of detail of the three-dimensional model rendered in P103 process The distance to the immediately preceding distance is separated and the combination of parameters such as position, posture, and angle of view at that position is determined. The virtual camera in the virtual space shoots the entire circumference of the three-dimensional model in a plurality of N times in a state with the determined parameters. The image taken at this time is stored in the texture data storage unit 101 together with the detail level as a texture (M3). The image generation unit 104 generates an image by pasting and rendering the texture (M2) read from the texture data storage unit 101 on the three-dimensional model photographed by the camera parameter editing unit 103. Hereinafter, the same way sequentially repeated to as low as Ete necessary details of replacement cut verbosity (P104 in Fig. 1).
In this way, data with different levels of detail is created, and data with the optimal level of detail can be displayed according to the distance. At this time, a texture of data having a high degree of detail is pasted on data having a low degree of detail. Even if the level of detail of the three-dimensional data is low, it is preferable to make it as close as possible to the appearance of data with a high level of detail. Can get quite close.

101 texture data storage unit 102 three-dimensional model data storage unit 103 camera parameter editing means,
104 Image generation means

Claims (2)

A first process of arranging a three-dimensional model having an appearance display image with the highest level of detail among one target feature in a virtual space;
The camera in the virtual space arranged to photograph the three-dimensional model is separated to a distance immediately before the detail level of the three-dimensional model is switched when photographing the three-dimensional model, and the three-dimensional model of the feature is A second process of photographing the entire circumference of the three-dimensional model by dividing it into a plurality of N times with an angle of view within a range of fit;
A third step of rendering the three-dimensional model N times with an appearance display image having a degree of detail when the three-dimensional model taken in the second step is taken;
The level of detail of the 3D model is switched from the distance rendered in the previous process to a distance just before the level of detail switches to a lower level, and the angle of view of the 3D model is within the range of the 3D model of the feature. It consists of a fourth process in which the entire circumference is shot in multiple times, and rendering using the image of the feature obtained in the previous process as the texture of the 3D model is repeated to a predetermined low level of detail. A method for generating an appearance display image of the same feature having different levels of detail. A texture data storage unit for storing an appearance display image as texture data having the highest degree of detail for a feature (recording electronic data of a surface image);
A 3D model data storage unit for storing a 3D model which is geometric information data simulating the feature based on the photographed feature;
Parameters for rendering the entire perimeter of the 3D model by shooting the 3D model placed in the virtual space with a virtual camera placed in the virtual space at a predetermined distance, position, orientation, and angle of view. Camera parameter editing means for determining the number N of combinations;
The 3D model read from the 3D model data storage unit is rendered with the texture read from the texture data storage unit according to the parameters determined by the camera parameter editing means, and an image is generated. An image for displaying the appearance of the same feature having different levels of detail, comprising image generation means for storing in the texture data storage unit together with the level of detail as texture data of the level of detail of the 3D model lower than that of the 3D model A generator of
The camera parameter editing means is separated from the distance rendered by the image generating means to a distance immediately before the level of detail of the 3D model is switched to be lower and further level of detail is switched to be lower,
The virtual camera shoots the entire circumference of the three-dimensional model in multiple times at an angle of view within a range where the three-dimensional model of the feature fits,
The image generating means repeats rendering using the obtained feature image read out from the texture data storage unit as a texture of the three-dimensional model to a predetermined low level of detail, and has the same level of detail An apparatus for generating an appearance display image of a feature.

Images