JP6025615B2 - Image drawing device - Google Patents

Description

  The present invention relates to an image drawing apparatus that expresses a two-dimensional or three-dimensional shape using a polygon model.

  In computer graphics, a polygon model is widely used as a method of expressing a two-dimensional or three-dimensional shape. In the polygon model, a triangle is mainly used as a unit shape, and the shape is expressed by a combination thereof.

3D city models used in car navigation systems are often represented by polygons such as triangles, but if each building is represented by several hundred polygons or more, the polygons that are included in the field of view when viewing a wide area The number is very large, hundreds of thousands. On the other hand, since the drawing time of the polygon model depends on the number of polygons to be drawn, when the model is composed of a large number of polygons, the drawing time becomes long.
In such a case, LOD (Level Of Detail) technology is generally used to shorten the drawing time. The LOD technique is a technique for improving the drawing speed by keeping the number of polygons to be drawn small. For example, it is realized by preparing a simplified model of each model in advance and replacing a less important model with a simplified model according to the positional relationship between the viewpoint and the model.

  When the LOD technology is applied to the drawing of the 3D city model, since the number of buildings included in the 3D city model is very large, it is difficult to manually create a model in which each building is simplified. Yes, it must be generated automatically. For example, in Non-Patent Document 1, in order to generate a simplified model, first, the outline of a silhouette when a building is viewed in parallel projection from a total of three directions, ie, a vertical direction and two horizontal directions, is linearly approximated. Then, using the horizontal contour, height information is given while dividing the vertical contour. Furthermore, a simplified model is generated by polygonizing each divided area as a column. In addition, by controlling the resolution of the silhouette and the maximum error of the linear approximation, it is possible to easily generate models with different levels of detail.

Tsai, F.A. Lin, W .; J. et al. , And Chen, L .; C. , Multi-Resolution representation of digital terrain and building models. Advances in Geo-Spacial Information Science, 2012.

  However, in Non-Patent Document 1, since a silhouette is generated from three directions determined regardless of the direction and shape of the building, a simplified model shape is obtained particularly when the wall surface of the building is oblique to the horizontal direction. Is very different from the original building shape. Further, unnecessary polygons may be included because the vertical outline may be divided unnecessarily many times, and the columns that contact each other have wall surfaces. Furthermore, no means for generating a texture to be used for the generated model is described.

  The present invention has been made to solve the above-described problems. When a model configured with the same number of polygons as that of a model generated using the conventional technique is generated, the conventional technique is used. It is possible to obtain a model closer to the original model shape than the original model, to automatically generate a texture to be applied to the generated model, and to draw a three-dimensional building group at high speed using those models and textures An object is to obtain an image drawing apparatus.

  The image drawing apparatus according to the present invention receives building model group data having a texture and inputs a viewpoint determining unit that determines three or more viewpoints with respect to the original model of each building, and each determined by the viewpoint determining unit A model projection unit that projects an original model using a viewpoint to generate a silhouette, a contour generation unit that generates a contour outline of the silhouette generated by the model projection unit as a set of line segments, and a contour generated by the contour generation unit One of the two is divided using another contour, and a contour dividing unit that gives height, a contour combining unit that combines a part of the contour divided by the contour dividing unit, and a combination processing by the contour combining unit A contour polygon generating unit for generating a three-dimensional polygon model from the contour, a polygon reducing unit for deleting polygons unnecessary for drawing from the three-dimensional polygon model generated by the contour polygon converting unit, Generate a texture to be applied to the 3D polygon model from which polygons unnecessary for drawing are deleted by the reduction unit, and the original model and texture, as well as the 3D polygon model from which polygons unnecessary for drawing by the polygon reduction unit are deleted A texture generation unit that records the generated texture in the storage unit, an original model that is recorded in the storage unit, a three-dimensional polygon model, and a texture corresponding to each of the original model that is drawn according to input information or The image processing apparatus includes a drawing target determining unit that determines a three-dimensional polygon model, and a drawing unit that draws an image using the original model or the three-dimensional polygon model determined by the drawing target determining unit, and a texture corresponding to the original model.

  According to the present invention, when a model composed of the same number of polygons as a model generated using the conventional technique is generated, a model closer to the original model shape than that using the conventional technique is obtained. In addition, it is possible to automatically generate a texture to be applied to the generated model, and to draw a three-dimensional building group at high speed using the model and the texture.

It is a block diagram which shows the image drawing apparatus which concerns on Embodiment 1 of this invention. It is a figure which shows the example of the model simplified with the image drawing apparatus which concerns on Embodiment 1 of this invention. It is a figure which shows the viewpoint determined by the viewpoint determination part of the image drawing apparatus which concerns on Embodiment 1 of this invention. It is a figure which shows the outline which the silhouette image which the model projection part of the image drawing apparatus concerning Embodiment 1 of this invention produces | generates, and the outline production | generation part produce | generate. It is a figure which shows a mode that the outline division part of the image drawing apparatus which concerns on Embodiment 1 of this invention divides | segments a main outline. It is a figure which shows the coupling | bonding conditions of an outline in the outline coupling | bond part of the image drawing apparatus which concerns on Embodiment 1 of this invention. It is a figure which shows the state which couple | bonded all what can be combined among the main outlines in the outline coupling | bond part of the image drawing apparatus which concerns on Embodiment 1 of this invention. It is a figure which shows the texture image which the texture generation part of the image drawing apparatus which concerns on Embodiment 1 of this invention produces | generates. It is a figure which shows the simplified model produced | generated by the image drawing apparatus which concerns on Embodiment 1 of this invention.

Embodiment 1 FIG.
FIG. 1 is a diagram showing a configuration of an image drawing apparatus according to Embodiment 1 of the present invention. The image drawing apparatus includes a preprocessing unit 1, a runtime processing unit 2, a storage device 3, and an image output device 4. The pre-processing unit 1 receives building model group data subjected to texture mapping, and creates building model group data simplified with a plurality of details. Further, the input building model group data and the building model group data simplified with the plurality of details are output to the storage device 3. The storage device 3 stores the building model group data output from the preprocessing unit 1. The execution time processing unit 2 reads out the building model group data stored in the storage device 3 and, for example, the building model group having an appropriate level of detail according to input information such as a viewpoint position that is expressed in a scale or the like. Is output to the image output device 4. The image output device 4 outputs the building model group output from the runtime processing unit 2 to an image display device (not shown) such as a monitor. Note that the building model group data input to the preprocessing unit 1 may be three-dimensional building model group data, particularly, for example, three-dimensional building model group data composed of polygons having a texture.

The preprocessing unit 1 includes a viewpoint determination unit 11, a model projection unit 12, a contour generation unit 13, a contour division unit 14, a contour combination unit 15, a contour polygonization unit 16, a polygon reduction unit 17, and a texture generation unit 18.
The viewpoint determination unit 11 determines two or more viewpoints for each model in the building model group input as data. The model projection unit 12 generates a silhouette image by drawing each model by parallel projection using the viewpoint determined by the viewpoint determination unit 11. The contour generation unit 13 generates the contour of the silhouette included in the silhouette image generated by the model projection unit 12 as a set of line segments. The contour dividing unit 14 divides one of the contours generated by the contour generating unit 13 using another contour, and gives height information. The contour combining unit 15 combines a part of the contours divided by the contour dividing unit 14 according to conditions. The contour polygonizing unit 16 generates a three-dimensional polygon model from the contours processed by the contour combining unit 15. The polygon reduction unit 17 deletes polygons unnecessary for drawing from the three-dimensional polygon model generated by the contour polygon conversion unit 16. The texture generation unit 18 generates texture image data, and outputs it to the storage device 3 together with the building model group data and texture input to the viewpoint determination unit 11 and the data of the 3D polygon model reduced by the polygon reduction unit 17. .

  The runtime processing unit 2 includes a drawing target determining unit 21 and a drawing unit 22. The drawing target determination unit 21 reads out building model group data stored in the storage device 3 and determines a building model group to be drawn according to input information such as a viewpoint position. The drawing unit 22 draws the building model group determined by the drawing target determination unit 21 and outputs it to the image output device 4.

Next, the operation of the image drawing apparatus according to Embodiment 1 of the present invention will be described.
The operation of the viewpoint determination unit 11 constituting the preprocessing unit 1 will be described with reference to FIGS. FIG. 2 shows an example of a model to be simplified. The model shown in FIG. 2 is a model of detail when the building model group data input to the viewpoint determination unit 11 is displayed as it is on the display device, and is a model before simplification, that is, an original model.
The viewpoint determination unit 11 determines two or more viewpoints to be used for generating a model silhouette for each model in the building model group input as data. Hereinafter, one of the determined viewpoints is called a main viewpoint, and the other viewpoints are called sub-viewpoints. The main viewpoint is determined so as to look down on the model, and one or more sub viewpoints are determined as viewpoints other than the main viewpoint.

  As a method for determining these viewpoints, for example, the position of the main viewpoint and the direction of the line of sight can be a viewpoint in which the center of the bounding box of the model is viewed from vertically above. In addition, the orientation of the main viewpoint can be determined as follows. First, a model silhouette image and its contour are generated by parallel projection from the main viewpoint, with the east direction temporarily being the right-hand direction of the main viewpoint. At this time, as a XY coordinate axis of the silhouette image (contour), a two-dimensional orthogonal coordinate system in which the right-hand direction (eastward direction) of the main viewpoint is, for example, the X-axis direction and the upper direction (north direction) of the main viewpoint is the Y-axis direction. Set. Then, in the vector of each minute line segment constituting the outline (the direction of the vector may be any one of the two end points of the minute line segment, whichever is the starting point), it is a vector in the X and Y axis directions set. An inner product with each of vectors of equal magnitude (for example, X and Y axis unit vectors) is calculated, and a value having a smaller absolute value is selected from the two calculated inner product values. Further, the sum of the absolute values selected in each minute line segment is calculated.

Next, the right-hand direction of the main viewpoint is rotated a small amount on a plane perpendicular to the line-of-sight direction, that is, a horizontal plane, with the line-of-sight direction of the main viewpoint, that is, the vertical direction as the rotation axis. At this time, the XY coordinate axes set as described above are similarly rotated by a small amount in accordance with the right-hand direction of the main viewpoint. Then, the process of calculating the sum of the absolute values of the inner product values in the direction of the rotated main viewpoint and the XY coordinate axes is repeated until the right-hand direction of the main viewpoint and the XY coordinate axes rotate 90 degrees. In other words, a plurality of two-dimensional orthogonal coordinate systems are set for the silhouette image (contour), and in each of the two-dimensional orthogonal coordinate systems, processing for calculating the sum of the absolute values of the inner product values described above is performed.
Finally, when the right-hand direction of the main viewpoint and the XY coordinate axes are rotated by 90 degrees, the direction of the main viewpoint when the sum of the absolute values is minimized is obtained, and the direction of the main viewpoint is set as the direction of the main viewpoint. decide. That is, out of a plurality of set two-dimensional orthogonal coordinate systems, a two-dimensional orthogonal coordinate system that minimizes the sum of the absolute values is obtained, and the X-axis direction of the two-dimensional orthogonal coordinate system is the right-hand direction of the main viewpoint, the Y-axis The direction is the upward direction of the main viewpoint.
Note that the right-hand direction of the main viewpoint and the position of the rotation axis when rotating the XY coordinate axes by a small amount are arbitrary, and can be, for example, positions passing through the center of gravity of the model.

  The position, line-of-sight direction, and orientation of the sub-viewpoint are, for example, upward in the vertical direction, and the right-hand direction of the main viewpoint (the X-axis direction in the XY coordinate axes when the sum of the absolute values is minimized). The viewpoint that can be set as the sub-viewpoint A, the upper direction being the vertical direction, and the upper direction of the main viewpoint (the Y-axis direction in the XY coordinate axes when the sum of the absolute values is minimized) is the sub-viewpoint B. And can.

Further, sub-viewpoints D and C can be obtained by rotating the viewing directions of sub-viewpoints A and B by 45 degrees in the horizontal direction. FIG. 3 shows the main viewpoint and the sub viewpoint generated as described above. In this determination method, the line-of-sight directions of the respective viewpoints intersect at right angles or at 45 degrees, so that the calculation in the processing operation in each unit after the model projection unit 12 described below can be easily performed.
As will be described later with reference to the operation of the model projection unit 12, the viewpoint determined by the viewpoint determination unit 11 is used for parallel projection, so the viewpoint position is at infinity. In FIG. 3, the position of each viewpoint is shown for convenience.
Information of each determined viewpoint is output to the model projection unit 12.

  Note that the method of determining the main viewpoint is not limited to the above method, and if the line-of-sight direction is a direction looking down on the model and the area of the silhouette of the model obtained by parallel projection at that time is determined to be small. Any method can be used. For example, a method is conceivable in which a plurality of main viewpoint position candidates are set at random, and the one with the smallest silhouette area is determined as the main viewpoint.

The method for determining the sub-viewpoint is determined so that the number of polygons of the model generated by the polygon reduction unit 17 to be described later is small and the difference between the model shape and the original model shape before simplification is small. Any method can be used.
For example, a volume that intersects silhouettes of parallel projections from each viewpoint of the main viewpoint and two or more sub-viewpoints may be generated, and each viewpoint may be determined so as to minimize the volume. That is, a volume in which a plurality of viewpoint groups composed of a plurality of viewpoints having different positions and line-of-sight directions are set, and in one viewpoint group, silhouettes of models by parallel projection from each viewpoint constituting the viewpoint group are intersected. For all the set viewpoints. Then, the viewpoint group with the smallest volume may be selected, and the viewpoints constituting the viewpoint group may be determined as the main viewpoint and the sub viewpoint.

Further, the number of sub-viewpoints is also arbitrary, and may be changed according to the quality of the model generated by the contour polygonizing unit 16 to be described later. If the quality is low, a sub-viewpoint may be added. By increasing the number of sub-viewpoints, the quality of the generated model can be improved.
For example, the quality of the model is determined by comparing the volume of the model input to the viewpoint determining unit 11 and the model generated by the contour polygonizing unit 16, and the smaller the difference volume, the higher the quality. Or the like can be used.

  The operation of the model projection unit 12 that constitutes the preprocessing unit 1 will be described with reference to FIG. The model projection unit 12 draws the model as a binary image by parallel projection from the viewpoint determined by the viewpoint determination unit 11 to generate a silhouette image. FIG. 4 shows silhouette images generated from each viewpoint. 4 (a) is a silhouette image generated from the main viewpoint, (b) is a silhouette image generated from the sub viewpoint A, (c) is a silhouette image generated from the sub viewpoint B, and (d) is generated from the sub viewpoint C. A silhouette image (e) is a silhouette image generated from the sub-viewpoint D. However, in the figure, the contour generated by the contour generation unit 13 described later is also drawn. In the same figure, with respect to the silhouette image and contour generated from each viewpoint, the right-hand direction of the viewpoint can be set as the X-axis direction, and XY coordinates determined by the Y-axis direction orthogonal to this direction can be set.

The resolution of the silhouette image generated at this time is determined so that the length L in the three-dimensional space is P pixel. At this time, by increasing the ratio P / L of P with respect to L, the level of detail of the generated model increases, and by decreasing it, the level of detail of the model decreases. Therefore, P / L is a parameter that controls the level of detail of the generated model.
Information on the generated silhouette image is output to the contour generation unit 13.

The operation of the contour generation unit 13 constituting the preprocessing unit 1 will be described with reference to FIG. First, the outline of the silhouette generated from the main viewpoint is obtained. Then, the contour is approximated to a set of a minimum number of line segments with an error E of ₁ pixel or less. Note that the larger the error E, the rougher the model that is finally generated, and the greater the difference from the original model. This is a parameter that controls the level of detail of the model. Hereinafter, this contour is referred to as a main contour. FIG. 4A shows the main contour. Also, height information is given to each vertex of the main contour. This height is a distance in a direction opposite to the line-of-sight direction of the main viewpoint from the ground at each vertex constituting the main outline.

Next, a contour is generated by the following method for the silhouette generated from each sub-viewpoint. Hereinafter, this contour is referred to as a sub-contour. First, in the XY coordinates set in the silhouette generated from each sub-viewpoint as described above, the largest pixel included in the silhouette among x = i (0 ≦ i <image width−1, i is a natural number) of the silhouette image. The y coordinate is obtained as y (i). Next, the obtained coordinates (i, y (i)) are connected in order from i = 0. Then, the connected line segment group is approximated to the minimum number of line segment groups with an error E of ₂ pixels or less. However, in the approximation, when the line segment is traced from the end point of the contour where x = 0 to the other end point, the x component of the line segment is set to 0 or more. 4B to 4E show sub-contours A to D generated from the sub-viewpoints A to D shown in FIG. The generated contour information is output to the contour dividing unit 14.

The operation of the contour dividing unit 14 constituting the preprocessing unit 1 will be described with reference to FIG. The contour dividing unit 14 first generates a straight line obtained by extending each vertex of the sub-contour A in the line-of-sight direction, and projects it onto the main contour. Then, the main contour is divided using the projected straight line. Here, the height of the new vertex generated at the intersection of the projected straight line and the main contour is calculated by linearly interpolating the heights of the two vertices on the main contour connected to the vertex, and the projection at the intersection. Compare the height of the straight line and use the smaller one. Note that the height of the projected straight line at the intersection is the value of the y coordinate component at the point where the projected straight line forming the intersection intersects the sub-contour A. Thereafter, the main contour is similarly divided using the other sub-contours B to D. 5A shows a state in which the main contour is divided by the sub-contours A and B generated in FIG. 4, and FIG. 5B further shows sub-contours C and D from the state shown in FIG. A state in which the main contour is divided is shown.
In addition, when a straight line projected from one sub-contour intersects with a straight line projected from another sub-contour, the height at this intersection (vertex) is the smaller of the heights of the projected and intersecting straight lines. adopt.

Note that since the line of sight of the main viewpoint that is the basis of the main contour generated in FIG. 4 is in the vertical direction, the height of each vertex constituting the main contour matches the distance in the vertical direction from the ground. Since the line of sight of each sub-contour is in the horizontal direction, the straight line extending from the apex to the line-of-sight also matches the horizontal direction. It is possible to easily calculate the intersection point with the straight line projected in the direction and projected onto the main contour, and the height of the intersection point. Further, since the viewing directions of the sub-viewpoints A and B are determined to be perpendicular to each other and perpendicular to the viewing direction of the main viewpoint, the straight line is parallel to the X-axis direction or the Y-axis direction of the silhouette image that generated the main contour. Yes. Similarly, the straight line extended from the sub-contours C and D also coincides with the horizontal direction and is 45 degrees with respect to the X-axis direction or the Y-axis direction of the silhouette image that is the origin of the main contour, so the main contour is divided. In this case, it is possible to easily calculate the intersection of the main contour and the straight line projected on the main contour by extending each vertex of the sub-contour in the viewing direction and the height of the intersection.
Information on the main contour divided as described above is output to the contour combining unit 15.

The operation of the contour combining unit 15 constituting the preprocessing unit 1 will be described with reference to FIGS. 6 and 7. The contour coupling unit 15 detects and deletes unnecessary ones of the line segments constituting the divided main contour.
For example, as shown in FIG. 6A, when there are two contours C1 and C2 that are divided and touching each other, the vertex P0 has the same height on the line segment L1 side and the line segment L2 side, and the vertex P1 is The line segment L3 and the line segment L4 have the same height, the line segment L1 and the line segment L2 are in the same direction (the angle at which the line segments cross each other is 0 degrees), and the line segment L3 and the line segment L4 are in the same direction. In some cases, the contours C1 and C2 can be combined by deleting the line segment L5.

Here, the height on the line segment side will be described in a model in which the main contour and the sub-contour as shown in FIG. 6B are obtained. The vertex P0 has a height h0 on the line segment L1 side. It has a height of h1 on the minute L2 side. Similarly, the vertex P1 has a height h0 on the line segment L3 side and a height h1 on the line segment L4 side. In this case, the vertex P0 is not the same height on the line segment L1 side and the line segment L2 side. Similarly, the vertex P1 is not the same height on the line segment L3 side and the line segment L4 side. Therefore, in this case, the contours cannot be combined.
FIG. 7 shows the result of combining all the connectable segments of the line segments constituting the main contour divided as shown in FIG.
Information on the main contour in which unnecessary line segments are combined after the division is output to the contour polygonization unit 16.

  However, the method for determining whether or not the combination is possible is arbitrary, and more contours may be combined by loosening the determination condition. For example, as shown in FIG. 6A, in two contours C1 and C2 that are divided and in contact with each other, if the line segment L1 and the line segment L2, and the line segment L3 and the line segment L4 are in the same direction, the contour C1 And C2 are combined to form one contour, and the two segments C3 and C4 that are divided and in contact with each other as shown in FIG. 6C, line segment L5, line segment L6, and line segment L7 The positions and heights of the vertex P2 and the line segment P3 are shown so that the line segment L8 has the same orientation, that is, the sum of the inner angle formed by the contour C3 and the inner angle formed by the contour C4 at the vertex P2 and the vertex P3 is 180 degrees. When changed as shown in FIG. 6 (d) (note that line segment L7 and line segment L8 are originally in the same direction, so the displacement of P3 is 0), the displacement of the positions of vertex P2 and vertex P3 before and after the change That is, the contour C3 before and after the change If is less than a certain threshold amount of deformation of 4 may be a method such as binding by changing the contour.

  The operation of the contour polygonizing unit 16 constituting the preprocessing unit 1 will be described. In the contour polygonizing unit 16, the contour generated by the contour combining unit 15 is lifted from the ground in the direction of the main viewpoint according to the height of each vertex to obtain a column. Then, the upper surface and the side surface of each column are divided into a set of triangles, which are polygonized as a three-dimensional polygon model. Information on the generated three-dimensional polygon model is output to the polygon reduction unit 17.

The operation of the polygon reduction unit 17 constituting the preprocessing unit 1 will be described. The polygon reduction unit 17 examines all adjacent column bodies, and if one of the wall surfaces of one column body is included in one of the wall surfaces of the other column body, the polygon constituting the included wall surface Are deleted as polygons unnecessary for drawing. Information of a three-dimensional polygon model that is composed of a column and from which polygons unnecessary for drawing are deleted is output to the texture generation unit 18.
In parallel to the processing in each part constituting the pre-processing unit 1 described above, the processing in each part is not performed on the pre-simplified texture-mapped building model group data input to the viewpoint determination unit 11. In the state, delivery to each part is also performed sequentially.

The operation of the texture generation unit 18 constituting the preprocessing unit 1 will be described with reference to FIGS. The texture generation unit 18 draws the inputted texture-mapped model before simplification by parallel projection from several directions into an image, thereby applying a texture to be applied to the simplified model 3D polygon model Generate an image. For example, a texture image can be generated by drawing a texture-mapped model before simplification from a total of five directions, vertically upward and four horizontal directions. FIG. 8 is a texture image generated using viewpoints in which the main viewpoint and sub-viewpoints A and B and sub-viewpoints A and B shown in FIG. 3 are directed in opposite directions. 8A is a texture image generated from the main viewpoint, FIG. 8B is a texture image generated from the sub viewpoint A, FIG. 8C is a texture image generated from a viewpoint in which the sub viewpoint A is directed in the opposite direction, and FIG. Is a texture image generated from the sub-viewpoint B, and (e) is a texture image generated from a viewpoint with the sub-viewpoint B directed in the opposite direction.
FIG. 9 shows an example in which the texture image generated by the texture generation unit 18 is mapped and drawn on the three-dimensional polygon model generated by the processing up to the polygon reduction unit 17 described above, that is, a simplified model.

  In the above description, a case has been described in which the number of viewpoints determined by the viewpoint determination unit 11 is a total of five viewpoints, one main viewpoint and four sub viewpoints. When the number of viewpoints is other than five, for example, three, the accuracy of reproducing the original model shape is lower than the precision when the number of viewpoints is five as shown in FIG. Thus, a model closer to the original model shape can be obtained than a model generated with the same number of polygons.

  Finally, the texture generation unit 18 records the three-dimensional polygon model data and the texture data generated by the processing up to the polygon reduction unit 17 in the storage device 3 realized by an HDD or the like. The pre-simplification model and texture data input to the viewpoint determination unit 11 are also recorded in the storage device 3 at the same time.

The operation of the drawing target determining unit 21 constituting the runtime processing unit 2 will be described. Drawing object determination unit 21 from the storage device 3, front and front and original model group data which is an input of the processing unit 1, by changing the parameters P / L and E ₁ and E ₂ to be used in the preprocessing unit 1 By executing the processing in the processing unit 1, model group data simplified with a plurality of details and textures corresponding to the model group data are read out. Then, a model to be drawn is determined according to input information such as a viewpoint position. For example, depending on the distance between the viewpoint and the model, the original model is used when the input scale is short, i.e. the original scale is large, and the simplified model is used when the input distance is small, i.e. when the input scale is small. It can be determined by a method such as The model group data of the degree of detail determined as the drawing target is output to the drawing unit 22.

  The operation of the drawing unit 22 constituting the runtime processing unit 2 will be described. The drawing unit 22 draws the model group which is the drawing target by the drawing target determining unit 21 using the corresponding texture and outputs the model group to the image output device 4. Since the drawing method of the model conforms to general polygon drawing processing, it is omitted.

  As described above, according to the first embodiment of the present invention, when a model configured with the same number of polygons as a model generated using the conventional technique is generated, the model using the conventional technique is used. In addition, a model closer to the original model shape can be obtained, a texture to be applied to the generated model can be automatically generated, and a three-dimensional building group can be drawn at high speed using the model and the texture.

  In the present invention, any constituent element of the embodiment can be modified or any constituent element of the embodiment can be omitted within the scope of the invention.

  DESCRIPTION OF SYMBOLS 1 Pre-processing part, 2 Runtime processing part, 3 Storage device, 4 Image output device, 11 Viewpoint determination part, 12 Model projection part, 13 Contour generation part, 14 Contour division part, 15 Contour combination part, 16 Contour polygonization part , 17 Polygon reduction unit, 18 Texture generation unit, 21 Drawing target determination unit, 22 Drawing unit.

Claims (9)

A viewpoint determination unit that takes the building model group data having a texture as input and determines three or more viewpoints for the original model of each building;
A model projection unit that projects the original model using each viewpoint determined by the viewpoint determination unit and generates a silhouette;
An outline generation unit that generates an outline of the silhouette generated by the model projection unit as a set of line segments;
One of the contours generated by the contour generation unit is divided using another contour, and a contour division unit that gives a height;
A contour combining unit that combines a part of the contour divided by the contour dividing unit;
A contour polygonizing unit that generates a three-dimensional polygon model from the contours combined by the contour combining unit;
A polygon reduction unit that deletes polygons unnecessary for drawing from the three-dimensional polygon model generated by the contour polygon unit;
A texture to be applied to a 3D polygon model from which polygons unnecessary for drawing are deleted by the polygon reduction unit is generated, and polygons unnecessary for drawing are deleted by the original model and texture, and the polygon reduction unit. A texture generation unit for recording the three-dimensional polygon model and the generated texture in the storage unit;
A drawing target determining unit that reads the original model, the three-dimensional polygon model, and the texture corresponding to each of the original model and the three-dimensional polygon model to be drawn according to input information;
An image drawing apparatus comprising: an original model or a three-dimensional polygon model determined by the drawing target determining unit; and a drawing unit that draws an image with a texture corresponding to the original model or the three-dimensional polygon model. The viewpoint determination unit determines a first viewpoint for viewing the original model from vertically above, and second and third viewpoints for viewing the original model from a horizontal direction,
A plurality of two-dimensional orthogonal coordinate systems are set with respect to the contour of the original model generated from the first viewpoint, the vectors of the line segments constituting the contour of the original model generated from the first viewpoint, and the above One of the coordinate axes of the two-dimensional orthogonal coordinate system that minimizes the total value when calculating the sum of the values of the smaller absolute values of the inner product values of the vectors in the direction of the respective coordinate axes of the two-dimensional orthogonal coordinate system; The line-of-sight direction of the second viewpoint is parallel,
2. The image drawing apparatus according to claim 1, wherein the other coordinate axis of the two-dimensional orthogonal coordinate system is parallel to the line-of-sight direction of the third viewpoint.   3. The viewpoint determining unit determines a fourth viewpoint and a fifth viewpoint obtained by rotating the second viewpoint and the third viewpoint, respectively, by 45 degrees in the horizontal direction. Image drawing device.   The viewpoint determination unit sets a plurality of viewpoint groups composed of a plurality of viewpoints having different positions and line-of-sight directions, and in the plurality of set viewpoint groups, the original model generated from each viewpoint constituting the viewpoint group The image drawing apparatus according to claim 1, wherein each viewpoint that constitutes the viewpoint group in which the volume obtained using the silhouette is minimized is determined as a viewpoint.   2. The image drawing apparatus according to claim 1, wherein the viewpoint determination unit changes the number of viewpoints according to the quality of the three-dimensional polygon model generated by the contour polygon unit. In the two adjacent contours, the contour coupling unit is configured to determine whether the adjacent angle formed by one contour at the vertex shared by the two adjacent contours is 180 degrees when the sum of the inner angle formed by the other contour is 180 degrees. Combine two matching contours,
The two adjacent points when the shared vertex is moved so that the sum of the inner angle formed by one contour and the inner angle formed by the other contour at the vertex shared by the two adjacent contours is 180 degrees. 2. The image drawing apparatus according to claim 1, wherein it is determined whether or not a contour deformation amount is equal to or less than a specific threshold value. If the contour deformation amount is equal to or less than the threshold value, the two adjacent contours are deformed and combined.   2. The polygon that is deleted by the polygon reduction unit is a polygon that constitutes the other wall surface included in one of the two wall surfaces facing each other between adjacent column bodies. Image drawing device.   The texture generator generates the original model from vertically above the original model, the second and third viewpoints, and two viewpoints obtained by rotating the second and third viewpoints 180 degrees horizontally. 3. The image drawing apparatus according to claim 2, wherein a texture is generated by drawing the original model texture-mapped with a texture corresponding to the model.   The image drawing apparatus according to claim 1, wherein the drawing target determining unit determines to draw a three-dimensional polygon model or an original model having a degree of detail corresponding to a scale used at the time of drawing.

Images