JP7394193B1 - Road network generation device, road network generation method and program - Google Patents

Abstract

The present invention provides a road network generation device and the like that makes it possible to easily and highly accurately detect an area indicating a road network using the shape of a candidate area of the road network. [Solution] A road network generation device includes a generation unit that generates a road network candidate area based on topography data indicating the topography of a predetermined geographical area, and a generation unit that sets a circumscribing rectangle of the candidate area and It includes a calculation unit that calculates an index regarding the shape of the candidate region using the length, and a determination unit that determines whether the candidate region indicates a road network based on the index. [Selection diagram] Figure 4

Description

The present invention relates to a road network generation device, a road network generation method, and a program.

Techniques for extracting road networks from image data such as aerial photographs and topographical data such as DEM (Digital Elevation Model) are known. For example, Patent Document 1 describes a method of extracting a road candidate area based on pixel values of a color image showing a feature.

Japanese Patent Application Publication No. 2004-246554

There is a need to detect such road candidate areas more easily and with high precision.

The present invention has been made to solve the above-mentioned problems, and is a road network generation method that makes it possible to easily and highly accurately detect an area indicating a road network using the shape of a candidate area of a road network. The purpose is to provide a device, a route network generation method, and a program.

A road network generation device according to an embodiment of the present invention includes a generation unit that generates a candidate area of a road network based on topography data indicating the topography of a predetermined geographical area, and a generation unit that sets a circumscribing rectangle of the candidate area and generates a circumscribed rectangle. A calculation unit that calculates an index related to the shape of the candidate area using the length of the diagonal line of the candidate area, and a determination unit that determines whether the candidate area indicates a road network based on the index. do.

Further, it is preferable that the calculation unit calculates a roundness coefficient of the candidate area as an index regarding the shape of the candidate area.

The calculation unit further calculates the area ratio of the candidate area to the circumscribed rectangle or the aspect ratio of the circumscribed rectangle, and the determination unit determines whether the candidate area indicates a road network based on the area ratio or aspect ratio. It is preferable to judge.

Further, the determination unit determines whether the roundness coefficient of the candidate region is equal to or greater than a first threshold, and for the candidate region determined to have a roundness coefficient equal to or greater than the first threshold, the area ratio of the candidate region is determined to be equal to or less than a second threshold. Determine whether the area ratio is equal to or less than a threshold value and whether the aspect ratio of the candidate region is equal to or less than a third threshold value, and the candidate region whose area ratio is determined to be equal to or less than the second threshold value and whose aspect ratio is equal to or less than the third threshold value. It is preferable to determine the candidate area determined to be a candidate area indicating a road network.

A road network generation method according to an embodiment of the present invention is a road network generation method executed by a road network generation device, in which a road network candidate area is extracted from topographic data indicating the topography of a predetermined geographical area, Setting a circumscribing rectangle of the candidate area, calculating an index regarding the shape of the candidate area using the length of the diagonal of the circumscribing rectangle, and determining whether the candidate area indicates a road network based on the index. It is characterized by including.

A program according to an embodiment of the present invention extracts a candidate area for a road network from topographic data indicating the topography of a predetermined geographical area, sets a circumscribing rectangle of the candidate area, and uses the length of a diagonal line of the circumscribed rectangle. The present invention is characterized by causing a computer to calculate an index related to the shape of a candidate area, and to determine whether or not the candidate area indicates a road network based on the index.

A road network generation device, a road network generation method, and a program according to the present invention enable highly accurate detection of an area indicating a road network using the shape of a candidate area of a road network.

1 is a functional block diagram of a road network generation device 1. FIG. It is a diagram showing the data structure of terrain data T1. It is a figure showing the data structure of candidate area data T2. It is a flow diagram showing the flow of road network generation processing. FIG. 3 is a schematic diagram for explaining a method of generating candidate regions. FIG. 3 is a flow diagram showing the flow of determination processing. (A)-(C) are diagrams showing examples of candidate areas determined to show a road network, and (D)-(E) are diagrams showing examples of candidate regions determined not to show a road network. It is.

Hereinafter, various embodiments of the present invention will be described with reference to the drawings. It should be noted that the technical scope of the present invention is not limited to these embodiments, but extends to the claimed invention and its equivalents.

FIG. 1 is a functional block diagram of a road network generation device 1 according to an embodiment of the present invention. The road network generation device 1 generates candidate areas for a road network from topographic data indicating the topography of a predetermined geographical area. In the following explanation, the road network is assumed to be a forest road network (a road network including forest roads, forestry roads, and forest working roads).

The forest road network has a shape (wavy shape or curved shape) that extends in a predetermined direction while winding in order to prevent rainwater from concentrating and flowing on the road surface. In particular, forestry roads and forest working roads are constructed along the topography in consideration of economic efficiency, so they have more pronounced wavy or curved lines. In addition, forestry roads are maintained by branching off from main forest roads, and forestry roads are maintained by further branching off from forestry roads. The road network generation device 1 determines whether the candidate area indicates a road network based on the characteristic shape of such a forest road network. The road network generation device 1 is an information processing device (computer) such as, for example, a PC (Personal Computer), a server, a mobile phone, or a smartphone. The road network generation device 1 includes a storage section 11, a communication section 12, and a processing section 13.

The storage unit 11 is configured to store programs and data, and includes, for example, a semiconductor memory. The storage unit 11 stores, as programs, an operating system program, a driver program, an application program, etc. used in processing by the processing unit 13. The program is installed in the storage unit 11 from a computer-readable, non-temporary, portable storage medium such as a CD-ROM (Compact Disc Read Only Memory) or a DVD-ROM (Digital Versatile Disc Read Only Memory).

The communication unit 12 is configured to enable the road network generation device 1 to communicate with other devices, and includes a communication interface circuit. The communication interface circuit included in the communication unit 12 is a communication interface circuit such as a wired LAN (Local Area Network) or a wireless LAN. The communication unit 12 receives data from another device and supplies it to the processing unit 13, and also transmits the data supplied from the processing unit 13 to the other device.

The processing unit 13 is configured to collectively control the operation of the road network generation device 1, and includes one or more processors and their peripheral circuits. The processing unit 13 includes, for example, a CPU (Central Processing Unit). The processing unit 13 may include a GPU (Graphics Processing Unit), a DSP (Digital Signal Processor), an LSI (Large Scale Integration), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), and the like. The processing unit 13 controls the operation of each component and executes various processes so that various processes of the road network generation device 1 are executed in appropriate procedures based on the programs stored in the storage unit 11. do.

The processing unit 13 includes an acquisition unit 131, a generation unit 132, a selection unit 133, a calculation unit 134, a determination unit 135, and an output unit 136 as its functional blocks. Each of these units is a functional module realized by a program executed by the processing unit 13. Each of these parts may be implemented in the road network generation device 1 as firmware.

FIG. 2 is a diagram showing the data structure of the terrain data T1 stored in the storage unit 11. The terrain data T1 is data indicating the terrain of a predetermined geographical area. The terrain data T1 is data indicating the curvature and slope of the earth's surface, and a diagram generated by plotting these data is known as a CS three-dimensional diagram. The terrain data T1 stores rectangular areas, elevations, slopes, curvatures, etc. in association with each other.

The rectangular area is information that identifies the position of an area (mesh) obtained by dividing a predetermined geographical area including a road network into rectangles of a predetermined size, and is, for example, coordinates indicating the center position of the rectangular area. The altitude is a representative value of the altitude of the rectangular area, and is, for example, the altitude at the center position of the rectangular area. The slope is information indicating the slope of the ground surface in a rectangular area, and is calculated based on, for example, the altitude of the rectangular area and the altitude of a rectangular area adjacent to the rectangular area. The curvature is information indicating the curvature of the ground surface in a rectangular area, and is calculated based on, for example, the altitude of the rectangular area and the altitudes of two adjacent rectangular areas with the rectangular area in between.

The topographical data T1 is generated in advance based on a DEM, an aerial photograph, etc. acquired based on an aerial laser survey, etc., and is stored in the storage unit 11.

FIG. 3 is a diagram showing the data structure of candidate area data T2 stored in the storage unit 11. The candidate area data T2 is polygon data indicating the shape of the candidate area of the road network. The candidate area data T2 stores candidate area IDs, shapes, etc. in association with each other.

The candidate area ID is information that identifies a candidate area. The shape is information indicating the shape of the candidate area. The candidate area is a polygon formed by connecting a plurality of rectangular areas, and information in which the position coordinates of the vertices are arranged in the order of adjacency is stored as the shape.

The candidate area data T2 is generated based on the terrain data T1 in a road network generation process described later, and is stored in the storage unit 11.

FIG. 4 is a flow diagram showing the flow of the road network generation process executed by the road network generation device 1. The road network generation process is realized by the processing unit 13 cooperating with each component of the road network generation device 1 based on a program stored in the storage unit 11.

First, the acquisition unit 131 acquires terrain data T1 from the storage unit 11 (step S11).

Next, the generation unit 132 generates a candidate area for the road network based on the terrain data T1 (step S12). For example, the generation unit 132 generates candidate regions of the road network using the learned model.

The trained model outputs, when the terrain data T1 is input, the confidence that the rectangular area corresponds to the position of a road (that is, indicates a road network) for each rectangular area included in the terrain data T1. This is a model trained on The model is, for example, a convolutional neural network. The generation unit 132 acquires terrain data T1 from the storage unit 11. The generation unit 132 inputs the terrain data T1 to the trained model and obtains the confidence of each rectangular area output from the trained model. The generation unit 132 determines that the pixel value of a pixel corresponding to a rectangular area whose confidence level is greater than or equal to a predetermined value is "1", and the pixel value of a pixel corresponding to a rectangular area whose confidence level is less than a predetermined value is "0". Generate a binary image with .

The generation unit 132 extracts connected components of pixels whose pixel value is "1" from the binary image. The generation unit 132 generates polygon data indicating a polygon corresponding to the outer periphery of the extracted connected component as a candidate region of the road network. The generation unit 132 stores the generated candidate area in the storage unit 11 as candidate area data T2.

FIG. 5 is a schematic diagram for explaining a method of generating candidate regions from a binary image. The rectangles shown in FIG. 5 indicate pixels that constitute a binary image. In FIG. 5, areas where the pixel value is "1" in the binary image are shown in black, and areas where the pixel value is "0" are shown in white. In the example shown in FIG. 5, the generation unit 132 extracts connected components A1 and A2 from the binary image as connected components of pixels whose pixel value is "1". The generation unit 132 identifies the outer periphery of the connected components A1 and A2, and extracts the vertices of the outer periphery (indicated by white circles in FIG. 5). The generation unit 132 generates data in which the extracted vertices are arranged in order of adjacency as the shape of the candidate area. The generation unit 132 associates the generated shape of the candidate area with the identification information of the candidate area and stores it in the storage unit 11 as a candidate area table T2.

Returning to FIG. 4, the selection unit 133 selects one candidate area from the generated candidate areas (step S13).

Next, the calculation unit 134 and the determination unit 135 calculate an index regarding the shape of the selected candidate area, and execute a determination process of determining whether the candidate area indicates a road network based on the index (step S14 ). Details of the determination process will be described later.

Next, the selection unit 133 determines whether all the generated candidate areas have been selected (step S15). If all candidate areas are not selected (step S15-No), the process returns to step S13, and the selection unit 133 selects one candidate area from among the unselected candidate areas.

If all candidate areas have been selected (step S15-Yes), the output unit 136 outputs the determination result (step S16). For example, the output unit 136 associates the candidate area included in the candidate area data T2 with the result of the determination as to whether or not it represents a road network. The output unit 136 outputs the candidate area data T2 associated with the determination result by transmitting it to another device via the communication unit 12. With this, the road network generation process ends.

FIG. 6 is a flow diagram showing the flow of the determination process. The determination process is executed in step S14 of the road network generation process.

First, the calculation unit 134 calculates the roundness coefficient, area ratio, and aspect ratio for the selected candidate region (step S21). The calculation unit 134 sets a circumscribed rectangle of the candidate area, and uses the length of the diagonal of the circumscribed rectangle to calculate an index regarding the shape of the candidate area. The circumscribed rectangle is a rectangle with the smallest area among the rectangles in which all vertices of the candidate area are located inside or on the sides. The circumscribed rectangle may be a rectangle with the smallest circumference among rectangles in which all vertices of the candidate area are located inside or on the sides. The calculation unit 134 calculates the roundness coefficient of the candidate area using the length of the diagonal of the circumscribed rectangle. The calculation unit 134 also calculates the area ratio of the candidate area to the circumscribed rectangle and the aspect ratio of the circumscribed rectangle. The significance of each indicator will be explained later. Note that the roundness count of the candidate region is an example of an index regarding the shape of the candidate region.

Next, the determination unit 135 determines whether the roundness coefficient of the candidate area is equal to or greater than the first threshold (step S22). The first threshold value is experimentally set so that a candidate area indicating a road network can be appropriately determined, and is, for example, 9.0. If the roundness coefficient of the candidate region is less than the first threshold (step S22-No), the determination unit 135 determines that the candidate region does not represent a road network (step S23). With this, the determination process ends.

If the roundness coefficient of the candidate region is equal to or greater than the first threshold (step S22 - Yes), the determination unit 135 determines whether the area ratio of the candidate region to the circumscribed rectangle is equal to or less than the second threshold and the aspect ratio of the circumscribed rectangle. It is determined whether or not is less than or equal to a third threshold (step S24). The second threshold value is experimentally set so that a candidate area indicating a road network can be appropriately determined, and is, for example, 0.25. The third threshold value is experimentally set so that a candidate area indicating a road network can be appropriately determined, and is, for example, 0.25. If the area ratio is larger than the second threshold and the aspect ratio is larger than the third threshold (step S24-No), the determination unit 135 determines that the candidate area does not represent a road network (step S23). If the area ratio is less than or equal to the second threshold or the aspect ratio is less than or equal to the third threshold (step S24-Yes), the determination unit 135 determines that the candidate area indicates a road network (step S25). This completes the determination process.

In the following, an index related to the shape of a candidate area will be explained. As mentioned above, the forest road network has a wavy or curved shape. Furthermore, since roads generally do not constitute simple closed curves, candidate regions that constitute simple closed curves are considered not to represent a road network. On the other hand, since forestry roads and forest work roads diverge from forest roads, there is a possibility that some closed curves may be included in the forest road network. By using an index related to the shape of a candidate region as described below, candidate regions that constitute a simple closed curve are excluded, and candidate regions representing a road network are appropriately extracted.

ここで、候補領域の面積Ｓは、候補領域の外周を示す線分で囲まれる領域の面積である。すなわち、候補領域が直線形である場合には、候補領域の面積は候補領域の幅と長さとの積である。また、候補領域が閉曲線である場合には、候補領域の面積は、候補領域の幅と長さとの積に、候補領域に囲まれる閉領域の面積を加えたものである。例えば、候補領域が円環形である場合には、候補領域の面積は、候補領域の幅と長さとの積に、候補領域に囲まれる円形領域の面積を加えたものであるから、候補領域の外径を直径とする円の面積と等しい。 The roundness coefficient (roundness) of a candidate region is an index indicating whether the shape of the candidate region is close to a perfect circle or flat. The roundness coefficient takes a minimum value of 1 when the shape of the candidate region is a perfect circle, and takes a large value when the shape of the candidate region is linear extending in a predetermined direction, such as a road. The roundness coefficient is expressed by the following formula using the maximum distance (maximum length) lmax between the vertices of the candidate region and the area S of the candidate region.

Here, the area S of the candidate area is the area of the area surrounded by a line segment indicating the outer periphery of the candidate area. That is, when the candidate region is linear, the area of the candidate region is the product of the width and length of the candidate region. Further, when the candidate region is a closed curve, the area of the candidate region is the product of the width and length of the candidate region plus the area of a closed region surrounded by the candidate region. For example, if the candidate region is circular, the area of the candidate region is the product of the width and length of the candidate region plus the area of the circular region surrounded by the candidate region. It is equal to the area of a circle whose diameter is the outer diameter.

If the candidate region is considered to be a perfect circle with radius r, the maximum length lmax can be considered to be the diameter of the circle, 2r, so the roundness coefficient is 1.

Since the candidate region has a large number of vertices, calculating the maximum length for each candidate region increases the calculation load. On the other hand, since the maximum length lmax is the longest distance between the two vertices of the candidate area, it is considered to be shorter than and close to the diagonal length d of the circumscribed rectangle that circumscribes the candidate area. It will be done. Therefore, the roundness factor is estimated by the following formula:

The area ratio of the candidate area to the circumscribed rectangle is calculated by dividing the area of the candidate area by the area of the circumscribed rectangle. When the candidate area is a straight line or a rectangular simple closed curve, the circumscribed rectangle and the candidate area match, and the area ratio takes the maximum value of 1. Furthermore, when the candidate region includes a portion that is a closed curve, the area ratio increases because the area of the candidate region includes the area of the region surrounded by the closed curve, as described above. As the candidate region approaches a waveform shape or a curved line shape, the area of the circumscribed rectangle becomes larger, and thus the area ratio becomes smaller.

The aspect ratio of a circumscribed rectangle is calculated by dividing the short side length of the circumscribed rectangle by the long side length of the circumscribed rectangle. If the candidate area is linear, the circumscribed rectangle will also be a rectangle extending in one direction, and the aspect ratio will be small. The closer the candidate area is to a wavy linear shape, bent curve shape, or closed curve, the closer the circumscribed rectangle becomes to a square, and the larger the aspect ratio becomes.

FIGS. 7A to 7E are diagrams showing examples of determination results for candidate regions. In FIGS. 7A to 7E, candidate regions are shown by solid lines, and circumscribed rectangles are shown by broken lines.

FIG. 7A shows an example of a candidate area that is determined to represent a road network. The candidate area in FIG. 7(A) has a waveform linear shape and has the same shape characteristics as the forest road network.

Since the candidate region in FIG. 7A does not include a closed curve, the area of the candidate region is small. Therefore, the roundness coefficient is greater than or equal to the first threshold. Furthermore, since the candidate region in FIG. 7A is close to a straight line, the area of the circumscribed rectangle of the candidate region is also small. Therefore, the area ratio is greater than the second threshold. On the other hand, since the candidate area is close to a straight line, the aspect ratio of the circumscribed rectangle is less than or equal to the third threshold. From the above, the candidate area in FIG. 7(A) is determined to represent a road network.

FIG. 7(B) shows another example of a candidate area that is determined to represent a road network. A part of the candidate area in FIG. 7B forms a closed curve, and a plurality of waveform curves branch from the closed curve. Such a shape is, for example, the shape seen in a forest road network where a single forestry road is connected to multiple forest roads, so the candidate area in Fig. 7(B) has the same shape as the forest road network. It has characteristics.

A part of the candidate area in FIG. 7(B) constitutes a closed curve B1. As described above, in such a case, the area of the candidate area includes the area of the area B2 (indicated by hatching) surrounded by the closed curve B1, so the area of the candidate area is large. On the other hand, since the candidate region in FIG. 7B has a plurality of waveform curves B3 branching from the closed curve, the circumscribed rectangle is also large. Therefore, the roundness coefficient is greater than or equal to the first threshold. Further, the area ratio is less than or equal to the second threshold. On the other hand, since the extending directions of the plurality of waveform curves B3 branched from the closed curve are different from each other, the aspect ratio of the circumscribed rectangle is larger than the third threshold value. From the above, it is determined that the candidate area in FIG. 7(B) represents a road network.

The candidate area in FIG. 7(C) shows another example of a candidate area determined to represent a road network. The candidate area in FIG. 7(C) has a curved line shape and has the same shape characteristics as the forest road network, but is closer to a straight line than the candidate area in FIG. 7(A).

Similar to the candidate region in FIG. 7A, the candidate region in FIG. 7C does not include a closed curve, so the area of the candidate region is small. Therefore, the roundness coefficient is greater than or equal to the first threshold. On the other hand, since the candidate region in FIG. 7C is close to a straight line, the shape of the candidate region and the shape of the circumscribed rectangle approximately match. Therefore, the area ratio is greater than the second threshold. On the other hand, since the candidate region in FIG. 7C is close to a straight line, the aspect ratio of the circumscribed rectangle is less than or equal to the third threshold. From the above, it is determined that the candidate area in FIG. 7(C) represents a road network.

As shown in Figures 7 (A) to (C), if the candidate region does not include a closed curve, or if it includes a closed curve but other curves diverge from the closed curve, the area of the candidate region Since the diagonal length of the circumscribed rectangle increases, the roundness coefficient increases. Further, as shown in FIG. 7A, when the candidate region has a wavy line shape or a curved line shape, both the area ratio and the aspect ratio become small. As shown in FIG. 7B, when the candidate region is a closed curve that branches off from another curve, the aspect ratio may become large, but the area ratio becomes small. As shown in FIG. 7C, when the candidate area is close to a straight line, the area ratio may be small, but the aspect ratio is also small. That is, when the candidate region has the same shape characteristics as the forest road network, the roundness coefficient becomes large and at least one of the area ratio and aspect ratio becomes small. Therefore, the candidate area indicating the road network is appropriately extracted through the determination process.

FIG. 7(D) shows an example of a candidate area determined not to represent a road network. The candidate region in FIG. 7(D) is a substantially elliptical simple closed curve. In general, roads do not constitute simple closed curves, so the candidate area in FIG. 7(D) has different shape characteristics from the forest road network.

The candidate region in FIG. 7(D) is approximately elliptical and close to a perfect circle, so the roundness coefficient is less than the first threshold. Therefore, it was determined that the candidate area in FIG. 7(D) does not represent a road network.

FIG. 7E shows another example of a candidate area determined not to represent a road network. Figure 7(E) is approximately elliptical like the candidate area in Figure 7(D), and has different shape characteristics from the forest road network, but is flatter than Figure 7(D). .

Since the candidate region in FIG. 7E is a simple closed curve, its area is large. However, since the candidate region in FIG. 7(E) is flatter than the candidate region in FIG. 7(D), the roundness coefficient was greater than or equal to the first threshold value. On the other hand, since the candidate region in FIG. 7E is a simple closed curve, the area ratio is larger than the second threshold. Further, the aspect ratio of the circumscribed rectangle of the candidate area in FIG. 7(D) is also larger than the third threshold value. Therefore, it is determined that the candidate area in FIG. 7(D) does not represent a road network.

As shown in FIGS. 7(D) and (E), when the candidate region forms a simple closed curve, the roundness coefficient becomes smaller or the area ratio and aspect ratio become larger depending on the shape. . Therefore, candidate areas that do not represent a road network are appropriately excluded through the determination process.

As explained above, the road network generation device 1 calculates an index regarding the shape of the candidate region using the diagonal of the circumscribed rectangle of the candidate region, and determines whether the candidate region indicates a road network based on the calculated index. Determine. By determining whether the candidate area indicates a road network based on the shape of the candidate area, it is determined with high accuracy whether the candidate area indicates a road network. Furthermore, by using the diagonal length of the circumscribed rectangle to calculate the index regarding the shape of the candidate area, the index can be easily calculated. Therefore, the road network generation device 1 makes it possible to easily and highly accurately detect whether a road candidate area indicates a road network.

In the above description, in step S12 of the road network generation process, the generation unit 152 generates the candidate area based on the topographical data T1, which is a CS three-dimensional map, but the present invention is not limited to such an example. The generation unit 152 may generate a candidate area from image data such as an aerial photograph in which a rectangular area and pixel values (RGB values) are associated.

In the above description, the calculation unit 154 calculates the roundness coefficient, area ratio, and aspect ratio in step S21 of the determination process, but the calculation unit 154 is not limited to such an example. For example, the calculation unit 154 may calculate only the roundness coefficient and not the area ratio and aspect ratio. In this case as well, it is possible to calculate with a certain degree of accuracy whether or not the candidate area indicates a road network. Further, the calculation unit 154 may calculate an index different from the roundness coefficient, area ratio, and aspect ratio.

In the above description, the road network generation device 1 is assumed to generate candidate areas for the forest road network, but the present invention is not limited to such an example, and may also generate candidate areas for other road networks. Since it is a general property that roads do not constitute a simple closed curve, and is not limited to forest road networks, the road network generation device 1 can also detect areas indicating other road networks with high accuracy.

It will be appreciated by those skilled in the art that various changes, substitutions and modifications can be made thereto without departing from the scope of the invention. For example, the embodiments and modifications described above may be implemented in appropriate combinations within the scope of the present invention.

1 Road network generation device 132 Generation unit 134 Calculation unit 135 Determination unit

Claims (5)

a generation unit that generates a road network candidate area based on topographic data indicating the topography of a predetermined geographical area;
a calculation unit that sets a circumscribed rectangle of the candidate area and calculates a roundness coefficient of the candidate area using the length of a diagonal line of the circumscribed rectangle;
a determination unit that determines whether the candidate area indicates a road network based on the roundness coefficient ;
A road network generation device comprising: The calculating unit further calculates an area ratio of the candidate area to the circumscribed rectangle or an aspect ratio of the circumscribed rectangle,
The determination unit further determines whether the candidate area indicates a road network based on the area ratio or the aspect ratio.
The road network generation device according to claim 1 . The determination unit includes:
Determining whether the roundness coefficient of the candidate region is greater than or equal to a first threshold;
Regarding the candidate region whose roundness coefficient is determined to be equal to or greater than the first threshold, whether or not the area ratio of the candidate region is equal to or less than a second threshold, and whether the aspect ratio of the candidate region is equal to or less than a third threshold. Determine whether or not
Determining a candidate region in which the area ratio is determined to be less than or equal to the second threshold value and a candidate region in which the aspect ratio is determined to be less than or equal to the third threshold value as a candidate region indicating a road network;
The road network generation device according to claim 2 . A road network generation method executed by a road network generation device, the method comprising:
Generate road network candidate areas from topographic data showing the topography of a predetermined geographical area,
setting a circumscribing rectangle of the candidate area, calculating a roundness coefficient of the candidate area using the length of a diagonal of the circumscribing rectangle;
determining whether the candidate area indicates a road network based on the roundness coefficient ;
A road network generation method characterized by comprising the steps of: Generate road network candidate areas from topographic data showing the topography of a predetermined geographical area,
setting a circumscribing rectangle of the candidate area, calculating a roundness coefficient of the candidate area using the length of a diagonal of the circumscribing rectangle;
determining whether the candidate area indicates a road network based on the roundness coefficient ;
A program that causes a computer to perform certain tasks.

Images