JPH0660153A - Method for generating three-dimension graphic data - Google Patents

Abstract

PURPOSE:To provide the 3-dimension graphic data generating method able to automatically generate a solid model expressing a 3-dimension shape of an object from 2-dimension graphic data drawn on three drawings without limiting the shape of the object different from a conventional technology, generating an imaginary part or reconversion of data obtained once. CONSTITUTION:The method includes a processing 101 entering 2-dimension graphic data forming three drawings, processing sections 103-108 detecting a closed loop included in each of the entered three drawings, while generating an auxiliary line to convert a concaved closed loop into a convexed loop, calculating a feature quantity including coordinates of end points as to the convex closed loop, an inter-drawing processing 109 generating sets of the three convex closed loops making the convex closed loops obtained from the three drawings correspondent with each other based on the feature quantity calculated, a processing 110 generating a basic cubic body from the sets of the closed loops, and a processing 111 implementing mutual set arithmetic operation when the plural number of basic cubic bodies are generated.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for generating three-dimensional graphic data used in a drawing input system or the like, and particularly, as two-dimensional graphic data such as a front view, a plan view and a side view. The present invention relates to a method of generating three-dimensional graphic data used when, for example, automatically inputting a three-dimensional shape of an object represented in a CAD system.

[0002]

2. Description of the Related Art As a model for expressing a three-dimensional shape of an object, a wire frame model expressed by vertices which are shape feature points of the object and a ridge line connecting the vertices, and an object expressed by its surface shape Surface models that are often used. However, since these models cannot accurately represent objects and the outside, their applications are limited.

On the other hand, the solid model is intended to represent an object not only on the surface but also inside, and to accurately distinguish it from the outside. Boundary Representat is used to represent objects using this solid model.
ion (B-Reps) model and Constructive Solid Geo
There is a metry (CSG) model. The B-Reps model is a model that represents an object with a combination of the boundary surface between the object and the outside. FIG. 14 is a diagram for explaining an object representation method using the B-Reps model. In order to represent the object 1409 in FIG. 14, boundary surfaces 1401 to 1408 between the object 1409 and the outside may be given.

FIG. 15 is a diagram for explaining the representation of an object by the CSG model. In FIG. 15, the object 1503
In order to express, the primitives 1501 and 1502 that are basic shape solids are first obtained, and then the set arithmetic expression to be executed between them is obtained. In FIG. 15, the primitive 1501
By subtracting between 1502, the object 1503 can be represented.

[0005]

In order to represent an object using a solid model, it is necessary to find a boundary surface or a primitive to be combined in any of the above methods, and further find an arithmetic expression between them. Conventionally, human beings have to manually input and specify boundaries, primitives, and arithmetic expressions while considering the objects they want to express.
I was forced to do extremely complicated work.

In particular, an object such as a machine part is a three-view drawing by a normal projection method (three projection planes orthogonal to each other are defined in a space, and a figure obtained by parallelly projecting the shape contour of the object on the projection plane). , For example, a front view, a plan view, and a side view) have been widely used, and an enormous amount of drawing data has been accumulated. In order to make a database of these drawings and reuse them in a three-dimensional CAD system or to edit, process and analyze them, it is necessary to generate a solid model that represents an object, but let's input this manually by humans. If so, a huge amount of labor is required and it is extremely difficult to realize.

As one of the methods for automatically generating a solid model using two-dimensional graphic data such as a front view, a plan view and a side view, there is a method disclosed in Japanese Patent Laid-Open No. 62-202271. . In this method, the first, second, and third projections viewed from three directions orthogonal to each other are input,
A projection plane is projected in the depth direction to form a projection surface, and a polyhedron having a closed figure defined by the contour line and the projection line as each surface is formed, and a polyhedron obtained from each of the three projection views is logically formed. The product is processed to obtain a solid model of the object.

However, since this method uses only a polyhedron defined by a projection plane and a projection line obtained by projecting a projected outline in the depth direction, the shape of an object that can be expressed is limited to that which can be swept in the depth direction. Therefore, there is a problem that the shape of the applicable object is significantly limited.

Another method is, for example, the method disclosed in JP-A-61-60173. According to this method, in a projection view when a three-dimensional line segment that constitutes an object is projected on a plurality of projection planes, the line segment can be either horizontal, vertical, or diagonal with respect to the projection plane or coordinate axes. The relationship as to whether it appears as a dimensional line segment is registered in the table in advance. The projected figure of the object to be input is stored as two-dimensional line segment data, and it is calculated from the end point coordinates of this line segment data whether it is horizontal, vertical, or oblique with respect to the projection plane or coordinate axis, A three-dimensional line segment is obtained by referring to the table in which the above relationships are registered. The above-described processing is sequentially performed on all the two-dimensional line segments, and the obtained three-dimensional line segment is used to represent the object.

However, since the two-dimensional line segment and the three-dimensional line segment do not necessarily have a one-to-one correspondence, a three-dimensional line segment that should not actually exist appears, and as a result, the object also appears. The imaginary part that should not actually exist appears. Therefore, there is a problem that a correct object cannot be represented unless the process of removing the imaginary part is performed. Furthermore, since the obtained result is a three-dimensional line segment, the ridgeline of the object can be expressed, but the surface and the inside of the object cannot be expressed. Therefore, in order to obtain a solid model, the data once obtained is retransformed. There is a need.

[0011]

SUMMARY OF THE INVENTION The present invention for solving the above-mentioned problems of the prior art is a two-dimensional figure in which the shape of an object is drawn in three plan views such as a front view, a plan view and a side view. The process of inputting data, the process of detecting a closed loop from the line segments included in each of the three views, the process of adding an auxiliary line so that the detected closed loop becomes a convex figure, and the convex figure were made. A process of calculating a feature amount of a closed loop, a process of associating each of the closed loops in each plan, a process of generating a basic shape solid from a set of the associated closed loops, and a set operation of the generated basic shape solids And processing to perform.

[0012]

According to the present invention, by inputting the two-dimensional graphic data of three views such as a front view, a plan view and a side view, a closed loop representing an object contour in each view is detected and its feature amount is determined. By calculating and generating primitives that are basic shape solids from a set of closed loops that have a correspondence relationship between each of the drawings, and performing the set operation between the primitives, the shape of the three-dimensional object represented as two-dimensional figure data is calculated. A solid model to be expressed is automatically generated without being limited to the shape of an input target object as in the prior art, and without generating an imaginary part or reconverting data once obtained. You can According to the present invention, since the shapes of existing three-dimensional objects drawn in three views can be automatically stored in a database on a computer, it becomes possible to easily input them into a three-dimensional CAD system.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a conceptual diagram showing a method of generating three-dimensional graphic data according to an embodiment of the present invention by disassembling it into respective processing units. Each processing unit is realized by a processing routine that runs on an electronic computer, dedicated hardware, or the like.

In FIG. 1, the shape of the object to be input is determined by the two-dimensional figure data input unit 101 as two-dimensional figure data including a front view, a plan view and a side view whose main elements are an outline and a hidden line. Is read.

FIG. 2 is an example of a front view read by the two-dimensional data input unit 101. Table 1 shows the contents of the two-dimensional graphic data corresponding to FIG.

However, a: identifier of a plan (0: front view, 1: plan view, 2: side view) b: total number of line segments included in this plan c: identification number of each line segment d: each line segment Identifier for the type of (0: outline, 1:
Hidden line) e: Coordinates of start points of each line segment f: Coordinates of end points of each line segment g: Total number of other line segments connected to start point of each line segment h: Identification number of other line segment connected to start point of each line segment i : Total number of other line segments connected to the end point of each line segment j: Identification number of other line segment connected to the end point of each line segment

Such two-dimensional graphic data is a two-dimensional C
It may be created using AD, or may be created as a text file on an electronic computer using an editor. Alternatively, the drawing itself is input to an image processing device using an image input device such as an image scanner, the contour line and the hidden line are separated / extracted, the dimension value is recognized, and the lengths of the contour line and the hidden line are automatically determined. By doing so, it is also possible to generate two-dimensional graphic data.

Further, a model primitive label which is a pre-registered basic shape solid and a model primitive consisting of a front view, a plan view, a contour line and a hidden line in a side view showing the shape of this model primitive. The model primitive data consisting of a figure and its geometrical parameters is transferred to the model primitive data input unit 10 of FIG.
Read by 2. Such model primitives include, for example, rectangular parallelepipeds, prisms, pyramids, spheres, cylinders,
Cones and toroids are suitable.

Table 2 shows right-angled triangular prism model primitive data. The model primitive graphic is a label of a graphic composed of an outline and a hidden line when the shape of the model primitive is expressed by using a front view, a plan view, and a side view.

[0020]

However, a: identification number b: label of model primitive c: shape parameter d: total number of combinations of model primitive figures e: combination of labels of model primitive figures f: label of characteristic model primitive figures

The model primitive figure will be described with reference to FIG. The order of the front view, the plan view, and the side view is not taken into consideration, and only the combination of the types of figures appearing in each plan is focused on.

In FIG. 3, for the right-angled triangular prism 301, refer to A. One right triangle appears in any one of the views and one rectangle appears in each of the other two views (refer to the set of front view 302, plan view 303, and side view 304). One right triangle appears in any one of the views, and two rectangles sharing one side in the other two views (see the set of front view 305, plan view 306, and side view 307) C. When one right triangle appears in any one of the drawings, one rectangle appears in any one of the drawings, and two rectangles sharing one side with each other appear in any of the drawings (front view) 308, top view 309, side view 310)).

The names of figures such as right-angled triangles and rectangles in these combinations are referred to as labels of model primitive figures. The geometrical parameters are used to uniquely determine the shape of the right triangular prism,
For example, the height and the length of two orthogonal sides of a right triangle.

In the front view data retrieval unit 103 of FIG. 1, the contents of the two-dimensional graphic data input from the two-dimensional graphic data input unit 101 are inspected and only the front view data is extracted.
It is sent to the front view closed loop feature amount calculation unit 106. The front view feature amount calculation unit 106 calculates the closed loop feature amounts shown in Table 3 for all closed loops.

[0026]

However, a: identification number b: maximum and minimum values of abscissa coordinate in the drawing c: maximum and minimum values of ordinate coordinate in the drawing d: unevenness determination result e: shape determination result f: recursion Depth g: attribute

With respect to the plan view and the side view, the same processing as the front view is executed. The front view closed loop feature quantity calculation unit 106 will be described later in detail. In the cross-figure correspondence processing unit 109 of FIG. 1, a set of closed loops having a correspondence relationship with each other is extracted based on the closed-loop feature amount of each of the views.
FIG. 4 is a diagram for explaining a method of associating closed loops appearing in each plan.

In FIG. 4, 401 is a front view, 402 is a plan view, and 403 is a side view. Reference numerals 404, 405, and 406 are one of closed loops included in each plan. The conditions for associating the closed loops 404, 405, and 406 are, for example, that the existence range of the X-axis coordinate of the closed loop 404 and the existence range of the X-axis coordinate of the closed loop 405 match, and the existence range of the Z-axis coordinate of the closed loop 405 and the Z of the closed loop 406. The existence range of the axis coordinate matches and the closed loop
That is, the existence range of the Y-axis coordinate of 406 matches the existence range of the Y-axis coordinate of the closed loop 404. Whether or not this condition is satisfied can be determined, for example, by checking whether or not the minimum and maximum coordinate values in the plan shown in Table 3 for the closed loops 404 to 406 are equal.

In the candidate primitive generator 110 of FIG.
Each of the closed-loop combinations extracted by the inter-diagram correspondence processing unit 109 is matched with the model primitive data, the matched combination is saved as a candidate primitive, and the unmatched one is discarded. It Then, for the candidate primitive, the candidate primitive data shown in Table 4 is generated. The method of determining the recursive depth and the attribute of the candidate primitive in the data of Table 4 will be described in detail later.

[0031] However, a: identification number d position parameter b: label e recursion depth c: size parameter f attribute

In the set operation processing section 111 of FIG. 1, a set operation of a solid body in a three-dimensional space is executed based on the solid model of the candidate primitive and its attribute. The set operation of the candidate primitives is executed in the descending order of the recursive depth of the candidate primitives shown in Table 4 based on the operation expression determined according to the value of the attribute.

The calculation result of the set calculation processing unit 111 is taken in by the solid model generation unit 112, and the final result is generated as a solid model representing the shape of the object to be input and displayed on the screen of the TV monitor. It is stored in the data storage device as a solid model. Furthermore, by inputting the obtained solid model into a three-dimensional CAD system, it is possible to perform graphic display by shading and erasing hidden surfaces, and processing and editing of shapes.

Here, the processing of the front view closed loop feature quantity calculation unit 106 in FIG. 1 will be described in detail with reference to FIG. The plan view closed loop feature amount calculation unit 107 and the side view feature amount calculation unit 108 have the same configuration and processing contents as the front view closed loop feature amount calculation unit 106 except that input data and output data are different. , Front view closed-loop feature quantity calculator
The description of 106 represents the description of each figure feature amount calculation unit.

The outermost closed loop detection unit 501 in FIG. 5 determines the outermost position based on the end point coordinates and connection relationship between the contour line and the hidden line as shown in Table 1 which constitutes the two-dimensional figure data of the front view to be processed. Detection of existing closed loops is performed. In this closed loop detection, for example, a search tree is generated by tracing the connection relationships of line segments, and the endpoints of the line segments that have already been traced are extracted when returning to the endpoints of the line segment that started the search by the depth-first search. Any method may be used. Of course, other appropriate methods can be used. Whether or not it is the outermost circumference is determined, for example, when the minimum / maximum values of the coordinates in the drawing of the closed loop are equal to the minimum / maximum values of the coordinates in the drawing in the original two-dimensional figure data and the area becomes maximum. This can be easily determined by examining each closed loop.

The outermost closed loop unevenness thus detected is determined by the outermost closed loop unevenness determination unit 502. As a method of determining the unevenness, for example, a method of determining that the closed loop is convex if all arbitrary line segments connecting the two end points forming the closed loop are present inside the closed loop, and is concave otherwise is there. Of course, other appropriate methods can be used.

If the outermost closed loop is determined to be convex,
In the next outermost peripheral closed loop shape determination unit 512, it is determined whether the shape of the outermost peripheral closed loop is a model primitive figure. For example, in order to determine whether the line is a right triangle, there are three or more line segments in a closed loop, a line segment set exists on three straight lines that are not parallel to each other, and a line segment that shares an end point. It suffices to determine whether one set is orthogonal. In addition, it is possible to use an appropriate method for determining a more complicated shape by using the area or the perimeter which are the shape feature amounts of the closed loop.

When it is determined that the outermost closed loop can determine the shape, the original two-dimensional figure data is held as it is and input to the closed loop detection unit 504. If it is determined that the outermost circumference closed loop cannot be shape-determined, or if the outermost circumference closed loop unevenness determination unit 502 determines that the outermost circumference closed loop is concave, the outermost circumference closed loop auxiliary line generation unit 503.
At, an auxiliary line is generated so that the outermost closed loop becomes convex, and the generated auxiliary line is added to the original two-dimensional graphic data to generate new two-dimensional graphic data, which is input to the closed loop detection unit 504. .

FIG. 6 is a diagram for explaining a method of generating an auxiliary line by the auxiliary line generator 503. In FIG. 6, the end point having the maximum or minimum coordinate value and its coordinate axis are obtained from the end point coordinates of the line segment data forming the closed loop 601, and the closed loop 601 is included and each side is parallel or perpendicular to the coordinate axis in the plan view. A circumscribed rectangle is generated so that Line segments newly generated by this processing are auxiliary lines 602 and 603.
And the point where these auxiliary lines 602 and 603 intersect is the end point
It is 604. In this way, the new two-dimensional graphic data generated by the auxiliary line generation unit 503, that is, the auxiliary lines 602 and 603 and the end points 604, do not originally belong to the target object. On the other hand, the outline and the hidden line belong to the target object.

The closed loop detector 504 of FIG. 5 detects all closed loops from the two-dimensional graphic data. The detection of this closed loop is performed in the same manner as the above-described outermost peripheral closed loop detection method. If the detected closed loop overlaps with the already detected closed loop, the closed loop is sent to the closed loop discard processing unit 509.

The feature amount initial setting unit 505 of FIG. 5 initializes the feature amount for all the detected closed loops. That is, the recursion depth in Table 3 is initialized to 0, and the maximum and minimum values of the two-dimensional coordinates in the front view are calculated.

The detected unevenness of the closed loop is the unevenness determination unit 506.
It is judged by. The determination of the unevenness is performed by the same method as the determination method of the outermost closed loop described above. The shape determining unit 507 determines whether or not the closed loop determined to be convex by the unevenness determining unit 506 matches the pre-registered model primitive figure based on the coordinates of the end points of the line segments forming the closed loop. . This shape determination is performed in the same manner as the outermost closed loop determination method described above. The feature amount calculation unit 508 calculates the maximum value and the minimum value of the attribute and the two-dimensional coordinate in the plan for the closed loop determined by the shape determination unit 507 as the shape determination being possible.

A method of calculating the attributes included in the closed loop feature will be described in detail later. Closed loop discard processing unit 50
In 9, the shape determination unit 507 determines that the shape determination is impossible, and the closed loop detection unit 504
Closed loops that are redundantly detected in are discarded and excluded from the subsequent processing targets.

In the auxiliary line generation unit 510, an auxiliary line is added to the two-dimensional figure data forming the closed loop so that the closed loop is convex with respect to the closed loop determined by the unevenness determination unit 506 to be concave. . The process of adding the auxiliary line is similar to the process of the outermost closed loop auxiliary line generating unit 503 described above.

The recursive processing control unit 511 has an auxiliary line generation unit 510.
For 2D figure data with auxiliary lines added,
While recursively repeating a series of processes of the closed loop detection unit 504, the feature amount initial setting unit 505, the unevenness determination unit 506, the shape determination unit 507, the feature amount calculation unit 508, the closed loop discard processing unit 509, and the auxiliary line generation unit 510. Every time the processing of the auxiliary line generation unit 510 is executed for the closed loop to be processed, the recursion depth of this closed loop shown in Table 3 is increased by the unit amount 1. This recursion depth is used to recalculate the attributes of the closed loops associated with each other and to determine the operation order of the candidate primitives in the set operation processing, as described later in detail.

Here, the unevenness determination unit 506 and the shape determination unit 507
A method of setting an attribute performed by the closed-loop feature amount calculation unit 508 will be described for a closed-loop that is convex and is determined to be shape-determinable. The closed loop in the present invention refers to a set of contour lines of a surface that distinguishes an object from the other space, which is obtained when the shape contour of the object is parallel-projected onto three projection surfaces orthogonal to each other. The contour line that constitutes this closed loop is the contour line of the surface that can be seen directly from the viewpoint placed on the side opposite to the projection plane with respect to the object, and the hidden line is the contour of the surface that cannot be seen directly from the above viewpoint. It is a line.

By the way, 3 of the object using the CSG model
The three-dimensional shape representation method is a method of representing an object, which is an object having a target shape, by a set of primitives, which are objects having a simpler shape. Therefore, since the whole or a part of the shape of the primitive always appears in the shape of the object, it is composed of the outline of the surface of the object obtained when the shape of the object is parallel-projected onto a certain projection plane, that is, the outline and the hidden line. The closed loop is a contour line of all or part of the face of the primitive.

To construct an object from a set operation of primitives in the CSG model, whether each primitive has an attribute of "+" to be added or an attribute of "-" to be subtracted in this set operation. Have to decide. The present invention includes a method for automatically obtaining the attributes of each primitive from the attributes of a closed loop in a figure obtained by projecting the shape contour of an object on a projection surface, the principle of which will be described in detail below.

First, consider the attributes of "+" and "-" for the closed loop. What is the "+" attribute for closed loop?
This means that a primitive having this closed loop as a contour of a surface has a “+” attribute in one plan, that is, it can be determined that this primitive belongs to an object. Further, "-" regarding the closed loop means that a primitive having this closed loop as a contour of a surface in one drawing has the attribute of "-", that is, it can be determined that this primitive does not belong to an object.

Further, considering that it may not be possible to determine "+" or "-" from only one drawing,
Consider a third attribute of "0". The reason is originally 3
This is because when shape data having a dimensional information amount is projected onto a two-dimensional projection surface, it may be degenerated in two dimensions and the information amount may be lost. FIG. 7 is a diagram for explaining an example in which such three-dimensional shape information degenerates into two dimensions. Figure 7
In, if only 701 figures are given, not only the shapes of 702,703,704 can be inferred as the three-dimensional shape of the object, but also the shapes in which the irregularities are reversed such as 702 and 704. There is.

In order to set the attribute of the closed loop, in addition to the drawing rule, the following attribute setting rule is set by utilizing the property that the contour line and the hidden line belong to the object but the auxiliary line does not belong to the object. To do. This attribute setting rule is as follows: 1. Set the attribute “+” for the outermost closed loop, A closed loop other than the outermost closed loop, 2-1. Attribute "-" for those that include auxiliary lines as part
Set 2-2. The attribute "+" is set for those having a hidden line on one side, and this hidden line shares an end point with the contour line included in the outermost closed loop including this closed loop, and 2-2. 2-1. Also, the attribute "0" is applied to those that do not correspond to 2-2. By applying these attribute setting rules, closed loop attributes can be set uniquely.

Next, the attributes of the candidate primitives are determined by using the closed-loop pairs that can be associated in the inter-drawing correspondence processing unit 109 in FIG. 1 and the attributes of each closed-loop that constitutes this pair. The method will be described in detail.

For the set of closed loops for which the candidate primitives could be generated, the recursion depths of the closed loops forming this set are compared. As a result of this comparison, if the recursion depths are all the same value, the same value is held as the attribute of the closed loop. If any one of the recursion depths has a different value,
Based on the smallest recursion depth, the closed loop attributes are recalculated using the method described below.

It is assumed that integers i, j, and k that are equal to or greater than zero are the recursion depths of the corresponding closed loops in the front view, the plan view, and the side view, respectively, and s, t, and u are the attributes of each closed loop. If i <j (1) i <k (2), the closed loop recursive depth i of the front view is used as a reference value, and the closed loop attributes t and u included in the plan view and the side view are used.
Is changed to the following attributes t ′ and u ′ by using the difference between the reference value i of the recursion depth and the recursion depths j and k of the closed loop.

T ′ = t · (−1) ^ji (3) u ′ = u · (−1) ^ki (4) By this calculation, for example, the attribute t is “+” and (j
If -i) is an odd number, the attribute t'is "-". If the attribute t is "0" and (j-i) is an odd number, the attribute t'is "0".

Further, the attributes of the candidate primitives are calculated as shown in Table 5 by using the changed closed-loop attributes s, t ', and u'.

[0057]

Further, the recursion depth is also considered for the candidate primitive, and the maximum value of the recursion depth of the closed loop constituting this complementary primitive is set as the recursion depth of this candidate primitive.

Next, the description will be supplemented by taking as an example the case of being applied to a three-view drawing of a concrete object. First, a first specific example will be described with reference to FIGS. 8, 9 and 10. Note that FIG.
In, depth represents the recursive depth of the closed loop enclosed by the dotted line.

With respect to the front view 801, first, the outermost closed loop detection unit 501 detects the outermost closed loop 804 based on the outline which is the two-dimensional graphic data forming the front view 801. Since the outermost closed loop 804 is rectangular, the outermost closed loop unevenness determining unit 502 determines that it is convex.
In addition, since the shape of this rectangle can be determined, the front view
The two-dimensional figure data of 801 is the outermost closed loop shape determination unit.
When the rectangle is determined by 512, it is directly input to the closed loop detection unit 504. In this closed loop detection unit 504, all closed loops 804, 805, 806 are detected, and the recursive depth depth of each closed loop is set to an initial value 0.

The concave-convex determining unit 506 determines that the closed loop 804 is convex and the closed loops 805 and 806 are concave. The closed loop 804 that is determined to be convex is determined to be a rectangle in the shape determination unit 507, and the feature amount shown in Table 3 is calculated in the feature amount calculation unit 508. As for the attribute of the calculated feature amount, since the closed loop 804 is the outermost closed loop at the recursion depth 0, the rule 1. Based on, the attribute of "+" is set.

For the closed loops 805 and 806 that have been determined to be concave, the auxiliary line generation unit 510 generates auxiliary lines 850 to 854,
An auxiliary line 850 is added to the two-dimensional figure data of the closed loop 805, and an auxiliary line 851 to the two-dimensional figure data of the closed loop 806.
854 is added. In the recursive processing control unit 511, the recursive depth depth of each closed loop to which an auxiliary line is added is 0 as an initial value.
From the closed loop detection unit 50
The process after 4 is repeated recursively.

As a result, in the closed loop detector 504,
Closed loops 807 and 808 are detected from a rectangle with auxiliary line 850 attached to closed loop 805 and auxiliary line 851 is detected in closed loop 806.
Closed loops 809,810,811 are detected from the rectangle with ~ 854 added. Closed loops 807,808 and closed loops 809-811 are
In both cases, the unevenness determining unit 506 determines that the shape is convex, the shape determining unit 507 determines that the shape is rectangular, and the feature amount calculating unit 508 calculates the feature amount as shown in Table 3.

Regarding the attributes of the calculated feature quantities, the closed loop 807 corresponds to the outermost closed loop at the recursion depth 1, so that the attribute setting rule 1. The attribute “+” is set based on Regarding the closed loop 808, this does not correspond to the outermost closed loop at the recursion depth 1 and has an auxiliary line in part, so that the attribute "-" is set based on the attribute setting rule 2-1. Closed loop 80
For 9, this corresponds to the outermost closed loop at recursion depth 1, so the attribute “+” is set based on the attribute setting rule 1. For closed loops 810 and 811, both are the outermost closed loop at recursion depth 1. The attribute "-" is set based on the attribute setting rule 2-1 because it does not correspond to the above and has an auxiliary line in part.

As for the plan view 802 of FIG. 8, the outermost closed loop 812 is first detected by the outermost closed loop detection unit 501 from the outline of the two-dimensional graphic data forming the plan view 802. Since the outermost closed loop 812 is rectangular, it is determined to be convex by the outermost peripheral unevenness determination unit 502, and this plan view
The two-dimensional figure data of 802 is the outermost closed loop shape determination unit 51.
It is input to the closed loop detection unit 504 via 2 as it is, all closed loops 812 to 825 are detected here, and the recursion depth depth of each closed loop is set to an initial value 0.

In the unevenness judging section 506, the closed loop 812 ...
819 is determined to be convex, and the closed loops 820 to 825 are determined to be concave. The closed loops 812 to 819 determined to be convex are determined to be rectangular in the shape determination unit 507, and the feature amount calculation unit calculates the feature amounts shown in Table 3. Regarding the attribute among the calculated characteristics, the closed loop 812 is the outermost closed loop at the recursion depth of 0, and therefore the attribute “+” is set based on the attribute setting rule 1. Closed loop 813 ~
As for 819, all of them do not correspond to the outermost closed loop 812 at the recursion depth of 0 and are configured only by the outline, and therefore the attribute “0” is set based on the attribute setting rule 2-3.
Is set.

On the other hand, the closed loops 820 to 825 determined to be concave
Is generated by the auxiliary line generation unit 510.
The recursion depth is added in the recursion control unit 511.
Is incremented from an initial value of 0 to 1, and the processing after the closed loop detection unit 504 is recursively repeated. As a result, the closed loop detection unit 504 detects each of the closed loops 820 to 825 and the closed loops that are the same as the already detected closed loops 814 to 819. That is, for example, when the detection is repeated for the convex closed loop in which the auxiliary line is added to the closed loop 820, the closed loop 820 itself and the same closed loops as the closed loops 814 and 816 are detected. These closed loops overlap the already detected closed loops 820,814,816. This duplication of detection is the same for the same closed loop as closed loops 821-825. Therefore, in the closed loop discard processing unit 509, all the closed loops 820 to 825 are discarded and excluded from the processing target.

For side view 803, the two that make up this
From the outline of the dimensional figure data, first, the outermost closed loop detection unit 501 detects the outermost closed loop, and the outermost closed loop is determined to be concave by the outermost closed loop unevenness determination unit 502. Therefore, the auxiliary lines 855 and 856 are generated and added in the outermost closed loop auxiliary line generating unit 503.
The three-dimensional figure data is input to the closed loop detection unit 504.

In the closed loop detection unit 504, all closed loops 826 to 831 are detected, and the recursive depth d of each closed loop is detected.
epth is set to the initial value 0. Of the detected closed loops, 826 to 829 are all determined to be convex by the unevenness determination unit 506, are determined to be rectangular by the shape determination unit 507, and the feature amount calculation unit 508 calculates the feature amounts as shown in Table 3. To be done. Closed loop for attributes of calculated features
Since 826 corresponds to the outermost closed loop at the recursion depth of 0, the attribute “+” is set based on the attribute setting rule 1.

Since the closed loops 827 and 829 do not correspond to the outermost closed loop 826 at the recursion depth of 0 and have an auxiliary line in part, the attribute "-" is assigned based on the attribute setting rule 2-1. Since the closed loop 828 does not correspond to the outermost closed loop 826 and is configured only by the outline, the attribute “0” is set based on the attribute setting rule 2-3.
Is set.

On the other hand, since the closed loops 830 and 831 are concave, an auxiliary line is generated and added by the auxiliary line generation unit 510, and the recursion depth control unit 511 sets the recursion depth depth from 0 to 1 as an initial value.
And the processes after the closed loop detection unit 504 are recursively repeated. As a result, in the closed loop detection unit 504,
Closed loops 830-831 and previously detected closed loops 827 and 8
A closed loop identical to 29 is detected. For example, closed loop 8
The closed loop detection performed on the convex closed loop in which an auxiliary line is added to 30 detects the closed loop 830 itself and the same closed loop as the detected closed loop 829. Both of these overlap with the detected closed loop. The closed loop duplication detection is the same for the closed loop 831.
Therefore, the closed loops 830 and 831 are closed loop discard processing units 509.
Will be discarded and excluded from processing.

Through the processing for each of the above drawings,
Finally, from the front view 801, the closed loops 804 and 807-81.
1 is from the top view 802 and the closed loop 812 to 819 is from the side view
From 803, closed loops 826 to 829 are detected, respectively, and the feature quantities for these closed loops are calculated.

FIG. 9 shows the correspondence between the plans executed by the plan correspondence processing unit 109 of FIG. 1 with respect to the closed loop detected from the front view, the plan view and the side view shown in FIG. It is a figure for explaining. Table 6 shows the correspondence between the detected closed loops illustrated in FIG. 8 and the closed loops in FIG. 9.

[0074]

In the interplanar correspondence processing unit 109, the maximum and minimum values of the coordinates within the plan included in the feature quantities of the closed loops 901 to 909 shown in FIG. 9 are compared, and a set of closed loops [901,904,907]
], A closed loop set [902,905,908] and a closed loop set [903,906,909], respectively. The attributes of the associated closed-loop set are the closed-loop set [901,
904, 907], the recursion depth of each closed loop is 0, so the same attributes are retained.

For the closed loop set [902,905,908], only the closed loop 902 has a recursion depth of 1 and the others have 0.
Therefore, the original value “+” of the attributes of the closed loop 902 is inverted to “−”, and the original values of the attributes of the other closed loops 905 and 908 are held. For the closed loop set [903,906,909], only the closed loop 903 has a recursion depth of 1 and the others have 0, so the attribute of the closed loop 903 is "+" by inverting the original value "-", and the other closed loop The 906 and 909 attributes retain their original values.

With respect to the set of closed loops associated as shown in FIG. 9, the shape data of each closed loop is collated with the label or shape parameter of the model primitive figure,
Candidate primitives are generated based on this matching result.
The solid model generation unit 112 in FIG. 1 generates a solid model representing the shape of the input target object based on the set calculation of the generated candidate primitive solid models.

FIG. 10 is a diagram for explaining a method of generating a solid model from the solid model of the candidate primitive obtained as a result of the association shown in FIG. In FIG. 10, the candidate primitive 1001 is a rectangular parallelepiped model primitive generated based on the closed-loop set [901,904,907] shown in FIG. 9, and the shape of each closed loop and the coordinate values in the plan. It has a rectangle as a label for the model primitive figure.

The length of each edge of the candidate primitive 1001 is equal to the length of one side of the closed loop rectangle, and the position coordinates of each vertex in the three-dimensional space are the coordinates of the corresponding closed loop vertex in the plan view. It is obtained by expanding to the coordinates of space. Similarly, the candidate primitive 1002 is a rectangular parallelepiped model primitive generated based on the set of closed loops [902, 905, 908] shown in FIG. 9, the shape of each closed loop, and the coordinate values in the plan, and the candidate primitive 1003 is shown in FIG. It is a rectangular parallelepiped model primitive generated based on the set of closed loops [903, 906, 909] shown and the shape of each closed loop and the coordinate values in the drawing.

The length of the edge line of each candidate primitive is equal to the length of one side of the rectangle of the individual closed loop, and the position coordinates of each vertex in the three-dimensional space are the coordinates in the plan view of the corresponding closed loop vertex. It is similar to the case of the candidate primitive 1001 that it can be obtained by expanding the coordinates in the dimensional space. Further, the attribute of each candidate primitive is obtained as shown in Table 7 based on the closed-loop attribute shown in FIG. 9, that is, the closed-loop attribute shown in FIG. 8 and the attribute setting rule shown in Table 5.

[0081]

However, A: the number of the candidate primitive in FIG. 10 B: the number of the closed loop in FIG. 9 C: the label of the candidate primitive in FIG. 10 D: the attribute of the closed loop in FIG. 9 E: the attribute of the candidate primitive in FIG. Recursion depth of 10 candidate primitives

As a result, "+" is set for the candidate primitive 1001, "-" is set for 1002, and "+" is set for 1003. In the set operation processing unit 111 of FIG. 1, the set operation of candidate primitives is executed in descending order of recursion depth. That is, first, a spatial set operation is performed between candidate primitives 1002 and 1003 having a recursion depth of 1, and a solid model 1004 having the attribute “−” is generated as an intermediate result. next,
A spatial set operation is executed between the solid model 1004 and the candidate primitive 1001 having a recursive depth of 0 to generate a final solid model 1005.

The figure in which the solid model 1005 is drawn in three views corresponds to the front view, the plan view and the side view shown in FIG. That is, a solid model expressing the correct three-dimensional shape of the target object can be automatically generated from the two-dimensional figure data of the front view, the plan view and the side view of FIG. 8 according to the method of the present invention.

Another embodiment of the present invention will be described with reference to FIGS. 11, 12 and 13. In FIG. 11, dept
h represents the closed loop recursive depth surrounded by the dotted line. Regarding the front view 1101, the outermost closed loop detection unit 501 first detects the outermost closed loop from the outlines of the two-dimensional graphic data forming the front view 1101, and the outermost closed loop unevenness determination unit 502 determines that it is concave. Therefore, for the outermost closed loop, the outermost closed loop auxiliary line generating unit 503 generates and adds the auxiliary line 1151 by generating a circumscribed rectangle surrounding the outermost closed loop.

The two-dimensional figure data of the front view 1101 to which the auxiliary line 1151 is added is input to the closed loop detection unit 504, where all the closed loops 1104 to 1116 are detected, and the feature amount initial setting unit 505 recurses each closed loop. The depth depth is set to an initial value of 0. In the unevenness determination unit 506, the closed loop 11
04-1110 are determined to be convex, and closed loops 1111-1116 are determined to be concave.

Each of the closed loops 1104 to 1108 determined to be convex is determined to be a rectangle, a rectangle, a triangle, a trapezoid and a triangle in the same order in the shape determination unit 507, and the feature amount calculation unit
At 508, the feature amount as shown in Table 3 is calculated. The attributes of the calculated feature amounts will be described. For the closed loop 1104, since this is the outermost closed loop at the recursion depth 0, the attribute “+” is set based on the attribute setting rule 1.

Regarding the closed loops 1105, 1106, 1108, since these do not correspond to the outermost closed loop 1104 and have an auxiliary line in part, the attribute "-" is set based on the attribute setting rule 2-1. . The closed loop 1107 does not correspond to the outermost closed loop 1104 and has a hidden line that shares an end point with the contour line included in the outermost closed loop that includes them. Therefore, the attribute based on the attribute setting rule 2-2. “+” Is set.

Regarding closed loops 1109 and 1110, since these are not registered as model primitive figures,
The shape determination unit 507 determines that the shape cannot be determined,
It is sent to the closed loop discard processing unit 509 and discarded. Since all of the closed loops 1111-1116 are concave, auxiliary lines are generated / added in the auxiliary line generation unit 510, and the closed loop detection units of the closed loops 1111-1116 are passed through the recursive processing control unit 511.
The processing after 504 is recursively repeated. Of these closed loops, 1111-1114 are detected as overlapping with already detected closed loops including themselves, and are therefore excluded from the processing target in the closed loop discard processing unit 509.

On the other hand, for closed loops 1115 and 1116,
The auxiliary lines 1152, 1153 and 1154-1157 are generated and added, the recursion depth depth is incremented from an initial value of 0 to 1, and the processing after the closed loop detection unit 504 is recursively repeated. as a result,
In the closed loop detector 504, the auxiliary line 11
Convex closed loop with 52 and 1153 added to closed loop 1117
And 1118 are detected, and closed loops 1119 to 1121 are detected from the convex closed loop in which the auxiliary lines 1154 to 1157 are added to the closed loop 1116. Closed loop 1117, 1118 and closed loop 1119 ~ 1121
Both are determined to be convex by the unevenness determination unit 506. In the shape determination unit 507, the closed loops 1117, 1119, and 1120 are all determined to be rectangular, and the closed loops 1118 and 1121 are both determined to be triangular. For each closed loop, the feature quantity calculation unit 508 has the characteristics shown in Table 3. The amount is calculated.

The attributes of the calculated feature quantities will be described. For the closed loops 1117 and 1119, since these are the outermost closed loops at the recursion depth 1, the attribute “+” is determined based on the attribute setting rule 1. Is set. Further, the closed loops 1118, 1120, and 1121 do not correspond to the closed loop at the outermost periphery and have an auxiliary line in part, so that the attribute “−” is set based on the attribute setting rule 2-1. .

With regard to the plan view 1102 of FIG. 11, the outermost closed loop 1122 is first detected by the outermost closed loop detection unit 501 from the outline of the two-dimensional figure data forming the same, and this outermost closed loop 1122 is a rectangle. Therefore, the outermost peripheral closed-loop unevenness determining unit 502 determines that it is convex.
Therefore, the two-dimensional figure data of the plan view 1102 is directly input to the closed loop detection unit 504, where all the closed loop 1122 are input.
1134 is detected, and the recursion depth depth of each closed loop is set to an initial value 0 in the feature amount initial setting unit 505.

Of the closed loops 1122 to 1134, 1122 to 1128 are judged to be convex by the concave / convex determining section 506, and closed loops 1129 to 1134 are judged to be concave. The closed loops 1122 to 1128 determined to be convex are each determined to be rectangular by the shape determination unit 507, and the feature amount calculation unit 508 calculates the feature amounts shown in Table 3.
Describing the attribute of the calculated feature amount, the closed loop 1122 corresponds to the closed loop at the outermost circumference at the recursion depth of 0, and thus the attribute “+” is set based on the attribute setting rule 1. Regarding the closed loops 1123 to 1128, all of them do not correspond to the closed loop 1122 of the outermost circumference and are composed only of the outline, and therefore the attribute “0” is set based on the attribute setting rule 2-3.

On the other hand, closed loops 1129 to 1134 determined to be concave
With respect to, the auxiliary line generation unit 510 generates and adds an auxiliary line, and the recursive depth depth is 0 by the recursive processing control unit 511.
To 1 and the processing after the closed loop detection unit 504 is recursively repeated. As a result, for the closed loops 1129 to 1133, the same closed loops as the closed loops 1123 to 1128 which have already been detected are detected. Therefore, the closed loop discard processing unit 509 discards the closed loops 1129 to 1133 and excludes them from the processing target.

On the other hand, for the closed loop 1134, the auxiliary line 11
58 and 1159 are generated and added, the recursion depth depth is incremented from an initial value of 0 to 1, and the processing after the closed loop detection unit 504 is recursively repeated. As a result, the closed loops 1135 and 1136 are detected, and the closed loop feature amount calculation unit 508 calculates the feature amounts shown in Table 3. To explain the attribute of the calculated feature amount, the closed loop 1135 corresponds to the outermost closed loop at the recursion depth 1, so the attribute “+” is set based on the attribute setting rule 1. For the closed loop 1136, Since this does not correspond to the closed loop at the outermost periphery and has an auxiliary line in part, the attribute "-" is set based on the attribute setting rule 2-1.

Concerning the side view 1103, this constitutes 2
The outermost closed loop 1137 is detected by the outermost closed loop detection unit 501 from the outline and the hidden line of the three-dimensional figure data. The detected outermost closed loop 1137 is determined to be convex by the outermost closed loop unevenness determination unit 502, and is input to the closed loop detection unit 504 as it is. The closed loop detection unit 504 detects all closed loops 1137 to 1143 and sets the recursion depth depth of each closed loop to an initial value 0. Closed loop 11
For the closed loops 1137 to 1141 out of 37 to 1143, the unevenness determination unit 506 determines that they are all convex, the shape determination unit 507 determines that they are all rectangular, and the closed loop feature amount calculation unit 508 determines the characteristics as shown in Table 3. The amount is calculated.

The attribute of the calculated feature amount will be described. For the closed loop 1137, this is the recursion depth 0.
Since it corresponds to the outermost closed loop in, the attribute “+” is set based on the attribute setting rule 1. In addition, the closed loop 1139 does not correspond to the outermost closed loop and is composed only of the outline, so the attribute “0” is determined based on the attribute setting rule 2-3. For the closed loops 1140 and 1141, this is the outermost circumference. Although it does not correspond to the closed loop and has a hidden line as a part of the side, this hidden line is only a part of the side. Therefore, the attribute “0” is set based on the attribute setting rule 2-3. Regarding the closed loop 1138, this does not correspond to the outermost closed loop 1137, has a hidden line on one side, and
Since this hidden line shares an end point with the outline included in the outermost closed loop 1137, the attribute “+” is set based on the attribute setting rule 2-2.

On the other hand, since the closed loops 1142 and 1143 are concave, an auxiliary line is generated / added in the auxiliary line generation unit 510, the recursive depth depth is stepped from the initial value 0 to 1, and the closed loop detection unit 504 and subsequent steps are performed. The process is recursively repeated. As a result, in the closed loop detector 504, the closed loops 1142, 11
43 and the closed loops identical to the already detected closed loops 1138 and 1139 are detected, respectively, and both of them overlap with the detected closed loop, so that the closed loops 1142 and 1143 are discarded by the closed loop discard processing unit 509.

Finally, from the front view 1101, the closed loop 11
04-1108 and 1117-1121 are closed loop 11 from the plan view 1102.
22 to 1128, 1135, 1136 are closed loop 11 from the side view 1103
37 to 1141 are detected, and the feature amount is calculated for each.

FIG. 12 is a view for explaining the correspondence between the plan views executed by the plan view correspondence processing section 109 of FIG. 1 for the closed loop obtained from the front view, the plan view and the side view shown in FIG. FIG. 12, the relationship of the closed loop of FIG. 12 with respect to the closed loop of FIG. 11 is as shown in Table 8.

[0101]

In FIG. 12, by comparing the maximum value and the minimum value of the coordinates in the drawings in the feature quantities of the closed loops 1201 to 1209, the inter-drawing correspondence processing unit determines the set of closed loops [1201,1].
204, 1207] and a closed loop pair [1202, 1205, 1208],
A set of closed loops [1203, 1206, 1209] is associated.
With respect to the attributes of the associated closed loop sets, since the recursion depths of the individual closed loops are all equal (0) in all of these closed loop sets, the values are retained as they are.

From the set of closed loops associated as shown in FIG. 12, candidate primitives are generated by matching the shape data of each closed loop with the label or shape parameter of the model primitive figure, and the candidate primitive solid model is generated. A solid model representing the shape of the input target object is generated by the set operation.

FIG. 13 is a diagram for explaining a method of generating a solid model of an input target object from a solid model of a candidate primitive obtained from the correspondence result shown in FIG. In FIG. 13, the candidate primitive 1301
Is a set of closed loops [1201, 1204, 1207] in FIG.
And a rectangular parallelepiped model primitive having three rectangles as labels of the model primitive figure, which are generated based on the shape of each closed loop and coordinate values in the drawing.

The length of each edge is equal to the length of one side of the rectangle of the closed loop, and the position coordinates of each vertex in the three-dimensional space are the coordinates of the corresponding closed loop vertex in the drawing in the three-dimensional space. Obtained by expanding. Similarly, candidate primitive 13
02 is a set of closed loops [1202, 1205, 120 in FIG.
8] is a rectangular parallelepiped model primitive having three rectangles generated as a label of the model primitive figure based on the combination of [8], the shape of each closed loop, and the coordinate values in the plan.

The candidate primitive 1304 models two rectangles and one right triangle generated based on the combination of the set of closed loops [1203, 1206, 1209] in FIG. 12, the shape of each closed loop, and the coordinate values in the drawing. It is a right-angled triangular prism model primitive used as a label for a primitive figure. The length of each ridge is equal to the length of one side of a closed loop rectangle or triangle, and the position coordinates of each vertex in the three-dimensional space are the coordinates of the corresponding closed loop vertex in the three-dimensional space. Can be obtained by expanding the candidate primitives
It is similar to the case of 1301. Further, the attribute of each candidate primitive is obtained as shown in Table 9 using the closed loop attribute in FIG. 12, that is, the closed loop attribute in FIG. 11 and the attribute calculation method shown in Table 5.

[0107]

However, A: candidate primitive in FIG. 13 B: combination of closed loop in FIG. 12 C: label of candidate primitive D: combination of attributes of closed loop E: attribute of candidate primitive F: recursion depth of candidate primitive

As a result, the candidate primitives 1301 have the attributes "+", 1302 "-", and 1303 "-", and the recursion depths are all 0. You may start with a spatial set operation between them. For example, candidate primitive 13
The calculation of the spatial difference between the solid model of 01 and the solid model of 1302 is performed, and the solid model of 1304 is obtained as an intermediate result. Furthermore, a spatial difference calculation is performed between the solid model of 1304 and the solid model of the candidate primitive 1303, and the final solid model 1305
Is obtained.

When the solid model 1305 is expressed in three views, it corresponds to the front view, plan view and side view shown in FIG.
According to the present invention, a solid model representing the correct three-dimensional shape of the target object can be automatically obtained from the two-dimensional figure data of the front view, the plan view and the side view of FIG.

As described above, when the candidate primitive is generated,
The configuration in which the attributes corresponding to the sign of the set operation are simultaneously generated is illustrated. However, only candidate primitives that do not have such attributes are generated and displayed on the screen as candidate primitives 1001, 1002, 1004 shown in FIG. 10, and the operation signs “+” and “ By inputting ``-'' from the keyboard,
It is also possible to adopt a configuration in which the code for the set operation is specified.

[0112]

As described above in detail, according to the present invention, the closed loop is detected from the two-dimensional figure data of the three views showing the shape of the input target object and detected from these two-dimensional figure data. The feature amount is calculated for a convex closed loop, and each of the views is associated to generate a candidate primitive having the corresponding closed loop as the contour line of the constituent surface, and the attribute of the candidate primitive is configured for the primitive. By performing the spatial set calculation of the solid model of the candidate primitive using the attribute of the candidate primitive and the attribute of the candidate primitive, an imaginary figure is generated without limiting the shape of the input target object. It is possible to automatically generate a solid model that represents the three-dimensional shape of the input target object without having to reconvert the data once obtained. It made.

[Brief description of drawings]

FIG. 1 is a conceptual diagram illustrating a method of generating three-dimensional graphic data according to an embodiment of the present invention by disassembling it into processing units.

FIG. 2 is a diagram showing an example of two-dimensional graphic data of a front view.

FIG. 3 is a diagram showing a model primitive figure of a right-angled triangular prism.

FIG. 4 is a diagram for explaining a method of associating closed-loop diagrams with each other.

5 is a diagram for explaining the contents of the processing of the front view closed loop feature quantity calculation unit 106 in FIG. 1. FIG.

FIG. 6 is a diagram for explaining a method of generating an auxiliary line.

FIG. 7 is a diagram for explaining that a three-dimensional shape degenerates into two dimensions.

FIG. 8 is a diagram for explaining a first specific example to which the present invention is applied.

FIG. 9 is a diagram for explaining the first specific example to which the present invention is applied.

FIG. 10 is a diagram for explaining the first specific example to which the present invention is applied.

FIG. 11 is a diagram for explaining a second specific example to which the present invention is applied.

FIG. 12 is a diagram for explaining the second specific example to which the present invention is applied.

FIG. 13 is a diagram for explaining the second specific example to which the present invention is applied.

FIG. 14 is a diagram for explaining a conventional B-Rep model.

FIG. 15 is a diagram for explaining a conventional CGS model.

[Explanation of sign]

 101 Two-dimensional figure data input unit 102 Model primitive data input unit 103 Front view data search unit 104 Floor plan data search unit 105 Side view data search unit 106 Front view closed loop feature amount calculation unit 107 Plan view closed loop feature amount calculation unit 108 Side view Closed loop feature calculation unit 109 Plan correspondence processing unit 110 Candidate primitive generation unit 111 Set calculation processing unit 112 Solid model generation unit 501 Outermost closed loop detection unit 502 Outermost closed loop unevenness determination unit 503 Outermost closed loop auxiliary line generation unit 504 Closed loop Detection unit 505 Feature amount initial setting unit 506 Unevenness determination unit 507 Shape determination unit 508 Feature amount calculation unit 509 Closed loop rejection processing unit 510 Auxiliary line generation unit 511 Recursive processing control unit 512 Outermost closed loop shape determination unit

Claims (11)

[Claims] 1. A three-dimensional diagram obtained by projecting an object having a three-dimensional shape onto three planes orthogonal to each other 2
Two-dimensional graphic data input processing for inputting three-dimensional graphic data, closed loop detection processing for detecting a closed loop included in the two-dimensional graphic data for each of the three views, and each of the three views If the detected closed loop is concave, an auxiliary line adding process is performed to add a convex line to the detected closed loop, and the convex shape detected by the closed loop detecting means for each of the three views. Of the closed loop and the convex closed loop to which the auxiliary line is added, a feature amount calculation process of calculating a feature amount including coordinate values of end points, and convex closed loops obtained for each of the three views. Corresponding to each other on the basis of the feature amount calculated for each, to create a set of three corresponding closed convex loops, It is characterized by including basic shape solid generation processing for generating a basic shape solid from a set of three closed convex loops, and processing for performing mutual set operation when there are a plurality of generated basic shape solids. A method of generating three-dimensional graphic data. 2. The closed loop detection process according to claim 1, wherein, for each of the floor plans, a first closed loop detection process for first detecting an outermost peripheral closed loop, and then a closed loop included in the detected outermost peripheral closed loop. And a second closed-loop detection process for detecting the three-dimensional figure data. 3. The method for generating three-dimensional graphic data according to claim 2, wherein the second closed loop detection process is repeated while interposing the addition of the auxiliary line by the auxiliary line addition process. 4. The three-dimensional graphic data according to claim 3, wherein the second closed loop detection process includes a process of discarding a concave closed loop that has generated the same closed loop as the already detected closed loop. How to generate. 5. The method for generating three-dimensional graphic data according to claim 4, wherein the feature amount includes an attribute generated based on a predetermined attribute setting rule. 6. The method for generating three-dimensional graphic data according to claim 5, wherein the feature quantity includes, as a recursive depth, the number of iterations of adding an auxiliary line accompanying the second closed loop detection processing. 7. The predetermined attribute setting rule according to claim 6, wherein: 1. Set the attribute “+” for the outermost closed loop, A closed loop other than the outermost closed loop, 2-1. Attribute "-" for those that include auxiliary lines as part
Set 2-2. The attribute "+" is set for those having a hidden line on one side, and this hidden line shares an end point with the contour line included in the outermost closed loop including this closed loop, and 2-2. 2-1. In addition, the method of generating three-dimensional graphic data is characterized in that the attribute "0" is set for items that do not fall under 2 or 2. 8. The operation code “+” or another basic code according to claim 7, wherein the basic shape solid generation processing indicates that the basic shape solid to be generated is added to another basic shape solid in the set calculation processing. An operation code "-" indicating that the operation code is subtracted from a solid is generated based on the attributes of the associated three closed convex loops. Generation method. 9. The generation processing of an operational code based on the attribute of each closed loop according to claim 8, wherein at least one of the three associated closed closed loops has an attribute “−” or all of them. Has attribute "0"
The method for generating three-dimensional graphic data comprises the step of generating the operation code "-" when the above-mentioned is included, and the generation processing of the operation code "+" otherwise. 10. The generation processing of an operation code based on the attribute of each closed loop according to claim 9, wherein the recursion depth for each closed loop is an integer i of 0 or more,
j, k, where i ≦ j ≦ k, the original attribute is set for the closed loop with the minimum recursion depth i, and the original attribute is set to (−1) ^ji for the closed loop with the recursion depth j.
A method for generating three-dimensional graphic data, characterized in that an attribute obtained by multiplying an original attribute by (-1) ^Ki is used for a closed loop having a recursion depth k. 11. The basic shape solid generated according to claim 8 or 3,
Has the maximum of the recursion depth of each of the convex closed loop of the number as the recursion depth of this basic shape solid,
A method of generating three-dimensional graphic data, wherein the set operation of the generated basic shape solids is performed in order from a large recursive depth of each basic shape solid to a small recursive depth thereof.