JP4929057B2 - Diagram compression processing program - Google Patents

Description

  The present invention relates to a compression processing program or the like that can thin out vertices while maintaining the accuracy of a diagram and can compress the amount of data.

As a method for compressing the capacity of map data by thinning out the vertices of a diagram generated by connecting a plurality of vertices in map data in SVG (Scalable Vector Graphics) format or other formats, the Douglas-Peucker algorithm (hereinafter referred to as DP) (Referred to as non-patent document 1, for example).
The DP method will be described with reference to FIG. In a diagram to be thinned out, that is, a diagram generated by connecting a plurality of vertices between a vertex as a start point and a vertex as an end point, except for a vertex a as a start point and a vertex z as an end point For each vertex b, c, d... I, a threshold value ε is set as a criterion for determining whether or not the vertex can be thinned out. This threshold value ε is a value obtained by ε = A × B. A is the parameter of the scale of the target map, and B is the required accuracy determined for each map scale. Each vertex T (b, c, d,..., I) has a threshold value with a threshold value ε as a radial dimension centered on the vertex T as shown in FIG. A circle R is set. That is, in the diagram compression processing program using the DP method, a threshold value ε, which is a fixed distance from the vertex, is set at each vertex of the diagram, and is positioned on both sides of the target vertex to be judged as to whether or not thinning is possible. When the shortest length between the line connecting the vertices to be noticed and the target vertex is smaller than the threshold value, the computer is caused to function as a means for thinning the target vertex. That is, as shown in FIG. 17, whether or not the target vertex can be thinned is determined by whether the computer connects the threshold circle set as the target vertex and the line segment connecting the vertices located on both sides of the target vertex. If the threshold circle and the line segment intersect, it is determined that it is possible to thin out the target vertex, and if the threshold circle and the line segment do not intersect It is determined that the target vertex cannot be thinned out.
Hereinafter, the procedure of the DP method will be described with reference to FIG. First, as shown in FIG. 16B, the vertex a serving as the start point of the line segment and the vertex z serving as the end point are connected by a straight line a1. The longest one of the lengths of the orthogonal lines (perpendicular lines) orthogonal to the straight line a1 connecting each vertex in the line segment and the straight line a1 is obtained. Here, the length of the longest orthogonal line, that is, the shortest distance between the vertex c and the straight line a1 is larger than the threshold value ε of the vertex c, and the straight line a1 and the threshold circle R of the vertex c do not intersect. Therefore, the vertex c remains without being thinned out. As shown in FIG. 16C, the vertex c and the vertex a (start point) are connected by a straight line a2, and the vertex c and the vertex z (end point) are connected by a straight line a3. Since the shortest distance between the vertex b and the straight line a2 is larger than the threshold value ε of the vertex b and the straight line a2 and the threshold circle R of the vertex b intersect, as shown by the white circle in FIG. In addition, the vertex b is thinned out (deleted). Among the vertices between the vertex c and the vertex z, the longest vertex e having the length of the orthogonal line orthogonal to the straight line a3 is obtained. Since the straight line a3 and the threshold circle R of the vertex e do not intersect, the vertex e remains undecimated as shown in FIG. Vertex e and vertex c are connected by straight line a4, and vertex e and vertex z are connected by straight line a5. Since the straight line a4 and the threshold circle R of the vertex d intersect, the vertex d is thinned out as shown in FIG. Among the vertices between the vertex e and the vertex z, the longest vertex g of the length of the orthogonal line orthogonal to the straight line a5 is obtained. Since the straight line a5 and the threshold circle R of the vertex g do not intersect, the vertex g remains without being thinned out. Vertex g and vertex e are connected by straight line a6, and vertex g and vertex z are connected by straight line a7. Since the straight line a5 and the threshold circle R of the vertex f intersect, the vertex f is thinned out. Among the vertices between the vertex g and the vertex z, the longest vertex i having the length of the orthogonal line orthogonal to the straight line a7 is obtained. Since the straight line a7 and the threshold circle R of the vertex i do not intersect, the vertex i remains undecimated as shown in FIG. The vertex i and the vertex g are connected by a straight line a8, and the vertex i and the vertex z are connected by a straight line a9. Since the straight line a8 and the threshold circle R of the vertex h intersect, the vertex h is thinned out.
Also, in the DP method, since the starting point and end point of the figure cannot be known if the figure is closed, the thinning process cannot be performed. Therefore, if the target of the thinning process is a closed figure, the closed figure is converted into several diagrams. After the division, the thinning process is performed for each of the diagrams, and the diagrams after the thinning process are connected again.
That is, the procedure of the compression processing program by the DP method is as follows. First, whether or not the threshold circle of the target vertex intersects with the straight line by connecting the vertex that is the start point and the end point of the line segment with a straight line, and the vertex that is farthest from this straight line is the first target vertex In the case of intersecting, all the vertices of the line segment are thinned out. That is, only the start point and the end point remain, and the line segment after thinning becomes a straight line connecting the start point and the end point. When the threshold circle of the target vertex and the straight line do not intersect, the target vertex remains without being thinned out. When the vertex remains, the vertex and the start point are connected by a straight line, and the threshold circle of the focused vertex and the straight line intersect with this line and the farthest vertex as the focused vertex. When doing so, all the vertices between the target vertex and the start point are thinned out. When the threshold circle of the target vertex and the straight line do not intersect, the target vertex remains without being thinned out. The determination as to whether or not the threshold circle of the target vertex intersects the straight line as described above is performed until there is no target vertex.
Duglas D.D. and Peucker. T.A. , 1973, Algorithms for the reduction of the number of points required for representative a digitized line or character, Canadian Cart122, 10, 112-

However, according to the DP method, in order to use a constant threshold value ε at each vertex, in the thinning process, as shown in FIG. 18, the diagram 11A and the diagram 12A after the thinning process intersect, As shown in FIG. 19, the corner vertex 50 of the diagram may be removed. For example, in the case of the closed figure Z shown in FIG. 20A, it is not desired to thin out the vertices 51 and 52 in order to maintain the shape. But Do La, as shown in FIG. 20 (b) ~ (d) , when intended to make thinning process with DP method, the dividing points 60; all thresholds epsilon (teeth of each vertex of the graphic other than 61 Since the threshold circle R) is constant, the vertices 51 and 52 are also thinned, and the closed figure ZA after the thinning cannot maintain the shape of the original closed figure Z. That is, it is difficult to compress the data amount by thinning out the vertices of the diagram while maintaining the accuracy of the diagram .
The present invention has been made in view of the above-mentioned problem, and provides a compression processing program that allows compress the data amount can be thinned vertex while maintaining the accuracy of the diagram.

When compressing the amount of data by thinning out the vertices of a diagram generated by connecting a plurality of vertices, the compression processing program for the diagram of the present invention compresses a threshold value that is a fixed distance from the vertex to each vertex of the diagram. If the shortest length between the line connecting the vertices located on both sides of the target vertex and the target vertex is smaller than the threshold value, the target vertex is thinned out. In a diagram compression processing program that causes a computer to function as the shortest length between a vertex of one diagram and a segment of the other diagram of a pair of different diagrams is shorter than a predetermined value And a threshold value adjustment processing program for causing the computer to function as a means for correcting the threshold value to be small.
The threshold adjustment processing program includes an inequality triangle network generation processing program that causes a computer to function as a means for generating an inequality triangle network by connecting vertices of a pair of different diagrams, and an opposite side of a triangle facing a target vertex. If it is on the diagram, find the shortest length between the target vertex and the opposite side of the triangle as a correction value, and if the opposite side of the triangle facing the target vertex is not on the diagram, the target vertex in the triangle that includes the target vertex A correction value calculation processing program for causing a computer to function as a means for obtaining the length of the short side of the two sides connecting the vertices on the diagram not including the target vertex as a correction value; A correction threshold value calculation processing program for causing a computer to function as a means for calculating a correction threshold value multiplied by a coefficient of 5 or less; That the means for setting the smaller one compared with the positive threshold value as the threshold of the attention vertices and a threshold value setting processing program for causing a computer to function also features.
The line diagram compression processing program according to the present invention is such that the angle formed by one line segment connected from the target vertex to one vertex and the other line segment connected from the target vertex to the other vertex is closer to 90 °. A threshold adjustment processing program for causing a computer to function as a means for correcting the threshold value to be small is provided.
The line diagram compression processing program of the present invention is the length of the shorter one of the one line segment connected from the target vertex to one vertex and the other line segment connected from the target vertex to the other vertex. And the threshold value are compared, and the length of the line segment is smaller than the threshold value and the threshold value is increased as the difference between the length of the line segment and the threshold value increases. A threshold adjustment processing program for causing a computer to function as a means for performing the processing is provided.
The program for compressing a diagram of the present invention is a threshold adjustment for causing a computer to function as a means for setting a correction threshold value calculated by multiplying the threshold value of a target vertex by a correction coefficient as the threshold value of the target vertex A processing program, the threshold adjustment processing program includes a correction coefficient calculation program, and the correction coefficient calculation program is connected to one vertex from the target vertex and one vertex to the other vertex. The first correction coefficient set smaller as the angle formed with the other line segment is closer to 90 °, the one line segment connected from the target vertex to one vertex, and the other line segment connected from the target vertex to the other vertex A second positive coefficient that is set to be larger as the length of the shorter line segment is smaller than the threshold value and the difference between the line segment and the threshold value is larger. Of the attention vertex And characterized by causing a computer to function as means for calculating the serial correction coefficient.

According to the diagram compression processing program of the present invention, since the threshold adjustment processing program is provided, the vertexes can be thinned out while maintaining the accuracy of the diagram, and the data amount can be compressed.
Since the threshold adjustment processing program includes an inequality triangle network generation processing program, a correction value calculation processing program, a correction threshold calculation processing program, and a threshold setting processing program, If the target vertex in Fig. 2 is close to the other diagram, the threshold value of that target vertex can be made appropriately smaller than the original threshold value, making it difficult for the target vertex to be thinned out. Crossing between the diagram and the other diagram can be prevented, and the amount of data can be compressed while maintaining the accuracy of the diagram.
According to the line diagram compression processing program of the present invention, the closer the angle formed by one line segment connected from the target vertex to one vertex and the other line segment connected from the target vertex to the other vertex is closer to 90 °. Since the threshold adjustment processing program that causes the computer to function as a means for correcting the threshold value to be small is provided, the shape of the portion that needs to be maintained in the original diagram can be maintained and the data can be stored. It can be compressed. This is particularly effective when thinning out a diagram that has no difference in length between vertices.
According to the line diagram compression processing program of the present invention, the shorter line segment among the one line segment connected from the target vertex to one vertex and the other line segment connected from the target vertex to the other vertex. Threshold adjustment processing for causing a computer to function as a means for correcting the length to be larger as the length of the line is smaller than the threshold and the difference between the length of the line segment and the threshold is larger Since the program is provided, it is possible to maintain the shape of a portion that needs to be maintained in the original diagram and to compress data. This is particularly effective when thinning a diagram having many curves and few corners close to 90 °.
According to the line diagram compression processing program of the present invention, the closer the angle formed by one line segment connected from the target vertex to one vertex and the other line segment connected from the target vertex to the other vertex is closer to 90 °. The first correction coefficient set to be small and the length of the shorter one of the one line segment connected from the target vertex to one vertex and the other line segment connected from the target vertex to the other vertex As a means for calculating the correction coefficient of the target vertex using a second positive coefficient that is set to be larger as the difference between the length of the line segment and the threshold value is larger than the threshold value. Since the correction coefficient calculation program for functioning is provided, it is possible to appropriately maintain the shape of the part that needs to be maintained in the original diagram and to compress the data .

Best form 1
1 to 6 show the best mode 1 of the present invention. FIG. 1 shows a configuration of a compression processing apparatus, FIG. 2 shows a state in which vertices of a diagram are connected to generate an unequal triangle network, FIG. 3 explains a method of calculating a correction value, FIG. 4; FIG. The procedure of the thinning process is shown, and FIG. 6 shows a diagram after the thinning process.

  The diagram compression processing apparatus 1 according to the best mode 1 includes a threshold adjustment processing means 2 and a thinning processing means 3. The threshold adjustment processing unit 2 includes an inequilateral triangle network generation processing unit 4 (hereinafter referred to as a TIN (Triangulated Irregular Network) generation processing unit), a correction value calculation processing unit 5, a correction threshold calculation processing unit 6, and a threshold. Value setting processing means 7 is provided.

The compression processing device 1 is realized by a computer and a compression processing program. The thinning processing unit 3 is realized by a computer and a thinning processing program. If the shortest length between the line connecting the vertices located on both sides of the target vertex and whether the target vertex is to be determined on the diagram is smaller than the threshold, the thinning process program The computer is made to function as a means of thinning out the vertex of interest. The TIN generation processing means 4 is realized by a computer and a TIN generation processing program. The TIN generation processing program causes the computer to function as a means for generating an unequal triangular network by connecting vertices of a pair of different diagrams. The correction value calculation processing means 5 is realized by a computer and a correction value calculation processing program. When the opposite side of the triangle facing the target vertex is on the diagram, the correction value calculation processing program obtains the shortest length between the target vertex and the opposite side of the triangle as the correction value, and the opposite side of the triangle facing the target vertex is When not on the diagram, as a means to calculate the length of the short side of the two sides connecting the target vertex and the vertex on the diagram not including the target vertex as a correction value in the triangle including the target vertex Make the computer work. The corrected threshold value calculation processing means 6 is realized by a computer and a corrected threshold value calculation processing program. The correction threshold value calculation processing program causes the computer to function as means for calculating a correction threshold value εj obtained by multiplying the correction value ε ₁ or ε ₂ by a coefficient α of 0.3 to 0.5. The threshold setting processing means 7 is realized by a computer and a threshold setting processing program. The threshold setting processing program causes the computer to function as means for comparing the original threshold value ε ₀ of the target vertex with the modified threshold value ε j and setting the smaller one as the threshold value Ε of the target vertex.

  The computer referred to in the present invention is a computer comprising hardware such as a CPU, a main storage device, an external storage device and basic software such as an operating system. Refers to a device that realizes

Next, a diagram processing method will be described. First, when a range to be thinned out is specified in the map, the TIN generation processing unit 4 connects the vertices 13 of a pair of different diagrams 11; 12 in the range to generate an unequal triangle network. (See FIG. 2). Next, when the correction value calculation processing means 5 determines whether or not the opposite side of the triangle that faces the target vertex is on the diagrams 11 and 12, and then determines that there is the opposite side of the triangle that faces the target vertex, For example, as shown in FIG. 3, a perpendicular line 14 perpendicular to the opposite side 12a of the triangle facing the target vertex 13a and reaching the target vertex 13a is drawn, and the length x1 of the vertical line 14 (that is, the target vertex 13a and its opposite side 12a). Is calculated as a correction value ε ₁ . Then, the correction threshold value calculation processing means 6 calculates a correction threshold value εj = ε ₁ × α obtained by multiplying the correction value ε _{1 by} , for example, a coefficient α = 0.4. Next, the threshold setting processing means 7 compares the original threshold value ε and the corrected threshold value εj of the target vertex 13a and sets the smaller one as the threshold value の of the target vertex 13a. A threshold circle 15 having a radius of the threshold value の of the target vertex 13a is set around the target vertex 13a.
If the correction value calculation processing means 5 determines that the opposite side of the triangle facing the target vertex is not on the diagram 11; 12, the diagram 11; 12 does not include the target vertex and the target vertex in the triangle including the target vertex. It determined the length of two sides that connects the vertices of the upper, to calculate the length x2 of the short sides of the ones of the length of the two sides as the correction value epsilon _2, modified threshold value calculation processing means 6, the correction value epsilon For example, a corrected threshold value εj = ε ₂ × α obtained by multiplying ₂ by a coefficient α = 0.4 is calculated.

  FIG. 4A shows a state in which a threshold circle 15 is set for each vertex, with the threshold value の of each vertex set by the threshold setting processing means 7 as a radius.

  Then, the thinning processing means 3 performs the thinning processing as in the conventional case. This thinning process will be described with reference to FIGS. 4B to 4D, FIGS. 5A and 5B. Since the thinning process for the vertices of the diagrams 11 and 12 is the same, only the thinning process for the vertices of the diagram 11 will be described below. First, as shown in FIG. 4B, in the diagram 11, a straight line 16 that connects the start point and the end point is drawn, and a vertex 17 that is farthest from the straight line 16 is obtained. Since the threshold circle 15 of the vertex 17 and the straight line 16 do not intersect, the vertex 17 is not thinned out (not deleted). As shown in FIG. 4C, the vertex 17 and the starting point are connected by a straight line 19, and the vertex 20 farthest from the straight line 19 is obtained. Since the threshold circle 15 of this vertex 20 and the straight line 19 intersect, this vertex 20 is thinned out (deleted). Note that all vertices between the vertex 17 and the start point are thinned out. Further, the vertex 17 and the end point are connected by a straight line 23, and the vertex 24 farthest from the straight line 23 is obtained. The threshold circle 15 at the vertex 24 is smaller than the threshold circle R by the conventional original threshold ε. Conventionally, as shown in FIG. 18, since the threshold circle R of the vertex 24 and the straight line 23 intersect each other, the vertex 24 is thinned out, so that the straight line 23 remains and the straight line 23 and the straight line 23 are lined up. The vertex 25 in FIG. 12 intersected. On the other hand, in the present invention, since the threshold circle 15 of the vertex 24 and the straight line 23 do not intersect, the vertex 24 remains, and thus the vertex 25 of the diagram 12 and the diagram 11 do not intersect. Hereinafter, as shown in FIG. 4D, the threshold circle 15 of the vertex 26 and the straight line 26a do not intersect, so the vertex 26 remains, and as shown in FIGS. 5A and 5B, the vertex 27 Since the threshold circle 15 and the straight line 27a do not intersect with each other, the thinned vertex 27 remains. That is, the thinning process can be performed in which the vertex 25 and the diagram 11 which have been crossed do not cross each other and the accuracy of the map data is maintained. FIG. 6 shows diagrams 11A and 12A after the thinning process. As shown in FIG. 6, after the thinning-out process according to the best mode 1, the line 11A and the line 12A do not cross each other as in the prior art.

  According to the best mode 1, when the target vertex on one diagram is close to the other diagram, the threshold value Ε of the target vertex is made smaller than the original threshold value ε. Is less likely to be thinned out, and crossing between one diagram after the thinning process and the other diagram can be prevented. For example, it is possible to prevent an intersection between a diagram showing a road and a diagram showing a house. Therefore, the map data can be compressed while maintaining the accuracy of the map data.

Best form 2
Means for correcting the original threshold value ε to be small when the shortest length between the vertex of one of the pair of different diagrams and the line segment of the other diagram is shorter than a predetermined value As a threshold value adjustment processing program for causing a computer to function. If different predetermined values are determined depending on the scale of the map, the map data can be compressed while maintaining the accuracy of the map data as in the best mode 1.

Best form 3
If the figure is a closed figure, the starting point and the ending point of the figure are not known, so the thinning process cannot be performed. In order to obtain the start point and end point, after dividing the closed figure into several diagrams, as shown above, the diagram start point and end point are connected by lines to start the diagram thinning process. The process of thinning out the closed figure is performed by connecting the end point and the start point of the diagram after the process is finished and returning it to the closed figure. Conventionally, in this closed figure dividing process, when a closed figure is divided into n pieces, it is divided into n line diagrams formed by the number of vertices obtained by dividing the number of vertices constituting the closed figure by n. For example, when a closed figure formed by connecting 16 vertices is divided into four, a certain arbitrary vertex is determined as a dividing point, and the fourth, eighth, 12th from one arbitrary vertex toward one direction. Dividing into four diagrams by determining the vertices as the dividing points. In other words, four diagrams are formed by connecting four vertices, thinning processing is performed for each diagram, and the end points and starting points of the four diagrams after thinning processing are connected. Return to a closed figure. However, in such a division process, the closed figure after the thinning process is far from the original closed figure, and the shape of the part that needs to be maintained in the original closed figure cannot be maintained.

  In the best mode 3, a closed figure division processing program that can maintain the shape of a part that needs to be maintained in the original closed figure will be described. 7 to 9 show the best mode 3 of the present invention. FIG. 7 shows the configuration of the division processing apparatus, FIG. 8 shows the division procedure, and FIG. 9 is an explanatory diagram of determination elements.

  The closed graphic division processing device 30 includes a determination element calculation processing unit 31 and a division point determination processing unit 32. The division processing device 30 is realized by a computer and a division processing program. The determination element calculation processing means 31 is realized by a computer and a determination element calculation processing program. The division point determination processing means 32 is realized by a computer and a division point determination processing program.

The determination element calculation processing program is a program that causes a computer to function as means for calculating a determination element for determining whether or not each vertex of a closed figure is to be a division point. As shown in FIG. 9, the determination element calculation processing program connects one line segment L1 connected from the target vertex T1 to which the determination element is calculated to one vertex T2 and one target vertex T1 to the other vertex T3. Means for calculating the angle θ formed with the other line segment L2 as a first determination element, and the length of the one line segment L1 connected from the target vertex T1 to which the determination element is calculated to the one vertex T2 a computer to the length of the shorter segment lengths of the lengths l ₂ of l ₁ and target vertex T1 other segments L2 tethered to another vertex T3 from the means for calculating a second decision element Let The division point determination processing means 32 is realized by a computer and a division point determination processing program. The division point determination processing program is a program that causes a computer to function as means for determining a division point based on a determination element for each vertex. In this division point determination processing program, as the angle θ as the first determination element of the vertex is closer to 90 °, the vertex is ranked higher as the division point candidate, and the length of the vertex as the second determination element is longer As a result, a means for ranking the attention vertex as a candidate for a dividing point and a rank of the first determination element and a rank of the second determination element for each vertex are combined to determine a vertex having a higher overall rank as a division point. Make a computer function as a means.

  FIG. 8 shows an example in which the vertex up to the third highest overall rank is divided by setting the division point to “3”. Note that the numbers assigned to the sides of the vertices in FIG. 8B are the order of the overall rank. Here, among the vertices at the four corners of the closed figure Z, as shown in FIG. Three vertices Tx; Ty; Tz are determined as dividing points. The vertex Te of the overall rank 4th in the corner that has not become the dividing point has a shorter line segment length connected to the vertex of the 8th overall rank than the vertex in the corner that has become the dividing point. And since the angle is far from 90 °, the overall rank is low.

  As shown in FIG. 8C, the thinning process described above is performed for each of the three divided diagrams Y1; Y2; Y3. Here, the diagram is thinned out so as to be a straight line connecting the start point and the end point, the diagram is also thinned out to a state closer to an L shape, and these diagrams are again at the vertices that became the dividing points. By being connected, it becomes a closed figure closer to a rectangle. That is, the thinned closed figure can be obtained without breaking the rectangular shape that is the shape of the original closed figure. Therefore, after the thinning process, it is possible to obtain a closed graphic in which the shape of the part that needs to be maintained in the original closed graphic is maintained.

Best form 4
The division point determination processing program may be a division point determination processing program that causes a computer to function as means for determining a division point based on the first determination element for each vertex. In other words, this division point determination processing program calculates the vertex as the angle between one line segment connected from the target vertex to one vertex and the other line segment connected from the target vertex to the other vertex is closer to 90 °. The computer is caused to function as a means for ranking higher as a division point candidate. Even when this dividing point determination processing program is used, it is effective, for example, when dividing a closed figure having no difference in length between vertices.

Best form 5
The division point determination processing program may be a division point determination processing program that causes a computer to function as means for determining a division point based on the second determination element for each vertex. In other words, this division point determination processing program calculates the length of the shorter line segment of the one line segment connected from the target vertex to one vertex and the other line segment connected from the target vertex to the other vertex. The longer it is, the computer is made to function as a means for ranking the attention vertex as a candidate for the dividing point. Even when this dividing point determination processing program is used, it is effective, for example, when dividing a closed figure having many curves and few corners close to 90 °.

Best form 6
10 to 14 show the best mode 6 of the present invention. FIG. 10 shows the configuration of the compression processing apparatus, FIG. 11 shows the procedure of the thinning process, FIG. 12 shows the closed figure after the thinning, and FIG. 13 shows an example of a function equation and a function graph for obtaining the first correction coefficient β1. FIG. 14 shows an example of a function expression and a function graph for obtaining the second correction coefficient β2.

The diagram compression processing means 1 according to the best mode 6 includes a threshold adjustment processing means 2 and a thinning processing means 3. The threshold adjustment processing unit 2 includes a correction coefficient calculation processing unit 35 and a threshold setting processing unit 36.
The correction coefficient calculation processing means 35 is realized by a computer and a correction coefficient calculation program. The correction coefficient calculation program causes the computer to function as means for setting the correction threshold value calculated by multiplying the original threshold value ε of the target vertex by the correction coefficient as the threshold value of the target vertex. More specifically, the correction coefficient calculation program sets the smaller the angle formed by one line segment connected from the target vertex to one vertex and the other line segment connected from the target vertex to the other vertex is closer to 90 °. The first correction coefficient β1 and the length of the shorter line segment of the one line segment connected from the target vertex to one vertex and the other line segment connected from the target vertex to the other vertex are The computer is caused to function as means for calculating the correction coefficient β of the target vertex using the second correction coefficient β2 that is set to be larger as the difference is larger than the threshold value. The threshold setting processing means 36 is realized by a computer and a threshold setting processing program. The threshold value setting processing program causes the computer to function as means for setting the threshold value 求 め obtained by the correction coefficient β of the target vertex × the original threshold value ε as the threshold value の of the target vertex.

ｘ：頂点の角度（０度から１８０度）
ａ：傾き係数
ｘ０：角度フィルタ係数（Ｐ（ｘ０）＝０．５となるときの角度）
例えば、第２修正係数β２は、以下に示す関数Ｐ（ｌ´）で求める。
Ｐ（ｌ´）＝１−ｃ×ｌ´
ここで、
ｌ´：正規化辺長、ｌ´＝ｌ／ε（０＜ｌ＜ε、ｌ≧εならば、ｌ´＝ｌ）
ｃ：傾き係数（一般に、１−ｃ＝０．２が標準値）
そして、修正係数βを、√β１×√β２で求める。 The first correction coefficient β1 and the second correction coefficient β2 may be set so as to meet the above-described conditions. For example, the first correction coefficient β1 is obtained by the function P (x) shown below.
here,

x: vertex angle (from 0 to 180 degrees)
a: slope coefficient x0: angle filter coefficient (angle when P (x0) = 0.5)
For example, the second correction coefficient β2 is obtained by the function P (l ′) shown below.
P (l ′) = 1−c × l ′
here,
l ′: normalized side length, l ′ = l / ε (if 0 <l <ε, l ≧ ε, l ′ = l)
c: slope coefficient (in general, 1-c = 0.2 is a standard value)
Then, the correction coefficient β is obtained by √β1 × √β2.

  FIG. 11A shows a state in which a threshold circle 15A is set for each vertex, with the threshold value Ε of each vertex set by the threshold setting processing means 36 as a radius.

  Then, the thinning processing means performs the thinning processing as in the conventional case. This thinning process will be described with reference to FIGS. In this example, since a closed figure is taken as an example, the thinning process is performed after the closed figure is divided by the above-described division processing means before the thinning process. In FIGS. 11B to 11D, after the closed figure is divided at the vertices Tc; Td, thinning processing is separately performed for the diagrams XA and XB connecting the divided vertices. The thinning-out method is the same as in the best mode 1. The threshold values 各 of the vertices T11; T12; T13 of the diagram XA set by the threshold setting processing means 8 are all reduced by the correction coefficient β. 11B, since the straight line M1 and the threshold circle 15A of the vertex T11 do not intersect, the vertex T11 remains, and the threshold of the straight line M2 and the vertex T12 as shown in FIG. 11C. Since the circle 15A does not intersect, the vertex T12 remains, and as shown in FIG. 11D, the straight line M3 and the threshold circle 15A of the vertex T13 do not intersect, so the vertex T13 remains. The threshold value の of each vertex T21; T22; T23; T24 of the diagram XB set by the threshold setting processing means 36 is large because the second correction coefficient β2 is large, and the diagram XB is thinned out. By processing, the vertices T21; T22; T23; T24 are thinned out. FIG. 12 shows a closed figure obtained by connecting diagrams XA and XB after thinning.

  According to the compression processing apparatus 1 of the best mode 6, only the vertices T21; T22; T23; T24 are thinned out after the thinning process, and the shape is maintained in the original closed graphic of the original closed graphic. Can maintain the shape of the necessary part. Therefore, similarly to the best mode 1, the map data can be compressed while maintaining the accuracy of the map data. That is, it is possible to maintain the shape of the part that needs to be maintained in the original diagram and to compress the data. Further, since the correction coefficient β is calculated using the first correction coefficient β1 and the second correction coefficient β2, in particular, as shown in FIG. 11, one line segment connected from the target vertex to one vertex and the target Even if the angle between the vertex and the other line segment connected to the other vertex is close to 90 °, the one line segment connected from the target vertex to one vertex and the other line connected from the target vertex to the other vertex The vertices T21; T22; T23; T24 in which the length of the shorter one of the minutes is smaller than the threshold value can be thinned out. That is, when only the first correction coefficient β1 is used, the vertices T21; T22; T23; T24 may remain, but the use of the first correction coefficient β1 and the second correction coefficient β2 is unnecessary for maintaining the shape. Vertices T21; T22; T23; T24 can be thinned out.

Best form 7
As a threshold adjustment processing program, the threshold value is set so that the angle between one line segment connected from the target vertex to one vertex and the other line segment connected from the target vertex to the other vertex is closer to 90 °. A threshold value adjustment processing program that causes a computer to function as a means for correcting it to be smaller may be used. Even when this threshold value adjustment processing program is used, it is effective, for example, when thinning a diagram having no difference in length between vertices.

Best form 8
As a threshold adjustment processing program, the length of the shorter line segment of the one line segment connected from the target vertex to one vertex and the other line segment connected from the target vertex to the other vertex is described above. As a means for comparing the threshold value and making the threshold value larger as the length of the line segment is smaller than the threshold value and the difference between the length of the line segment and the threshold value is larger. A threshold adjustment processing program that causes a computer to function may be used. Even when this threshold adjustment program is used, it is effective, for example, when thinning a diagram having many curves and few corners close to 90 °.

Best 9
The flow of processing by the map data compression processing apparatus provided with the compression processing means 1 of the best mode 1; 6 and the division processing means of the best mode 3 will be described with reference to FIG. The map data compression processing apparatus inputs a command for designating the range of the map to be compressed, and reads the data in the designated range (step S1). Threshold correction processing according to the best mode 1 is performed (step S2). Threshold correction processing according to the best mode 6 is performed (step S3). Division processing according to the best mode 3 is performed (step S4). A thinning process is performed (step S5). The order of processing from step S2 to step S4 may be changed.

  The present invention can also be used for thinning out a diagram other than map data and processing for dividing a closed figure other than map data.

The block block diagram which shows a compression processing apparatus (best form 1). The figure which shows an unequal side triangular network (best form 1). Explanatory drawing of the correction value calculation method (best form 1). The figure which shows the procedure of a thinning process (best form 1). The figure which shows the procedure of a thinning process (best form 1). The figure which shows the diagram after a thinning process (best form 1). The block block diagram which shows a division | segmentation processing apparatus (best form 3). The figure which shows a division | segmentation procedure (best form 3). Explanatory drawing of the determination element (best form 3). The block block diagram which shows a compression processing apparatus (best form 6). The figure which shows the procedure of a thinning process (best form 6). The figure which shows the closed figure after thinning (best form 6). The figure which shows an example of the function type | formula which calculates | requires 1st correction coefficient (beta) 1, and a function graph (best form 6). The figure which shows an example of the function type | formula which calculates | requires 2nd correction coefficient (beta) 2, and a function graph (best form 6). The flowchart of the process by the map data compression processing apparatus (the best form 9). The figure which shows the procedure of the conventional thinning process. Explanatory drawing of the threshold value of the conventional thinning process. The figure which shows the conventional thinning-out process result. The figure which shows the conventional thinning-out process result. The figure which shows the conventional thinning-out process result.

Explanation of symbols

1 compression processing device, 2 threshold adjustment processing means, 3 thinning processing means,
4 TIN generation processing means, 5 correction value calculation processing means, 6 correction threshold value calculation processing means,
7 threshold value setting processing means, 30 division processing device, 31 determination element calculation processing means,
32 division point determination processing means, 35 correction coefficient calculation processing means, 36 threshold value setting processing means.

Claims (5)

  When compressing the amount of data by thinning out the vertices of the diagram generated by connecting multiple vertices, set a threshold value that is a fixed distance from the vertex to each vertex of the diagram, and determine whether thinning is possible A diagram compression processing program that causes a computer to function as a means for thinning out a target vertex when the shortest length between the target vertex and a line connecting vertices located on both sides of the target vertex is smaller than a threshold value , The threshold value is corrected to be smaller when the shortest length between the vertex of one of the pair of different diagrams and the line segment of the other diagram is shorter than a predetermined value. A program for compressing a diagram, characterized in that it comprises a threshold adjustment program for causing a computer to function as means.   The threshold adjustment processing program includes an inequality triangle network generation processing program that causes a computer to function as a means for generating an inequality triangle network by connecting vertices of a pair of different diagrams, and an opposite side of a triangle facing a target vertex. If it is on the diagram, find the shortest length between the target vertex and the opposite side of the triangle as a correction value, and if the opposite side of the triangle facing the target vertex is not on the diagram, the target vertex in the triangle that includes the target vertex A correction value calculation processing program for causing a computer to function as a means for obtaining the length of the short side of the two sides connecting the vertices on the diagram not including the target vertex as a correction value; A correction threshold value calculation processing program for causing a computer to function as a means for calculating a correction threshold value multiplied by a coefficient of 5 or less; A threshold value setting program for causing a computer to function as means for setting a smaller value as a threshold value of a target vertex by comparing with a positive threshold value. Compression processing program.   When compressing the amount of data by thinning out the vertices of the diagram generated by connecting multiple vertices, set a threshold value that is a fixed distance from the vertex to each vertex of the diagram, and determine whether thinning is possible A diagram compression processing program that causes a computer to function as a means for thinning out a target vertex when the shortest length between the target vertex and a line connecting vertices located on both sides of the target vertex is smaller than a threshold value , In which the threshold is made smaller as the angle formed by one line segment connected from the target vertex to one vertex and the other line segment connected from the target vertex to the other vertex is closer to 90 ° A diagram compression processing program comprising a threshold adjustment processing program that causes a computer to function as   When compressing the amount of data by thinning out the vertices of the diagram generated by connecting multiple vertices, set a threshold value that is a fixed distance from the vertex to each vertex of the diagram, and determine whether thinning is possible A diagram compression processing program that causes a computer to function as a means for thinning out a target vertex when the shortest length between the target vertex and a line connecting vertices located on both sides of the target vertex is smaller than a threshold value In the above, the length of the shorter one of the line segment connected from the target vertex to one vertex and the other line segment connected from the target vertex to the other vertex is compared with the above threshold. The threshold for causing the computer to function as a means for correcting the line segment to be larger as the length of the line segment is smaller than the threshold value and the difference between the length of the line segment and the threshold value is larger. Value adjustment processing pro Compression processing program of the diagram, characterized in that it comprises a ram. When compressing the amount of data by thinning out the vertices of the diagram generated by connecting multiple vertices, set a threshold value that is a fixed distance from the vertex to each vertex of the diagram, and determine whether thinning is possible A diagram compression processing program that causes a computer to function as a means for thinning out a target vertex when the shortest length between the target vertex and a line connecting vertices located on both sides of the target vertex is smaller than a threshold value A threshold adjustment processing program for causing a computer to function as a means for setting a correction threshold value calculated by multiplying the threshold value of the target vertex by a correction coefficient as a threshold value of the target vertex, The processing program includes a correction coefficient calculation program, and the correction coefficient calculation program is connected from the target vertex to one vertex and from the target vertex to the other vertex. The first correction coefficient set smaller as the angle formed with the other line segment is closer to 90 °, the one line segment connected from the target vertex to one vertex, and the other line segment connected from the target vertex to the other vertex A second positive coefficient that is set to be larger as the length of the shorter line segment is smaller than the threshold value and the difference between the line segment and the threshold value is larger. A diagram compression processing program for causing a computer to function as means for calculating the correction coefficient of the target vertex .

Images