KR101129490B1 - Method for generating pedestrian network using existing spatial datasets - Google Patents

Abstract

The present invention relates to a method for generating a pedestrian network using pre-built spatial information, wherein (a) the pre-built first spatial information is input to the main memory as input data, and the pre-built system is stored using a program in the main memory. Extracting a one-step walking space from one spatial information; (b) inputting pre-built second spatial information into the main memory as input data, and extracting a two-step walking space from the pre-built second spatial information using a program in the main memory; (c) generating a first-class basic network through linearization using a program in a main memory including a skeleton algorithm in the walking space extracted in step (a); (d) generating a class 2 basic network through linearization using a program in a main memory including a skeleton algorithm in the walking space extracted in step (b); and (e) step (c). In step (d), the network structure of the node-link that can be utilized in the PNS by merging the gateway information extracted from the basic network and the gateway layer is generated by using a program in the main memory, thereby constructing a pedestrian network. It not only facilitates network creation by reducing the time required, but also reduces construction costs and can be used as basic data for more precise network creation.

Description

Method for generating pedestrian network using existing spatial information {Method for generating pedestrian network using existing spatial datasets}

The present invention relates to a method for generating a pedestrian network using pre-established spatial information. More particularly, the present invention relates to a numerical map, a road name address base map or an internet map, an aerial photograph, satellite image, etc. The present invention relates to a method for generating a pedestrian network by extracting a pedestrian space and performing a two-step linearization process on the extracted pedestrian space.

In general, the Car Navigation System (CNS), which is widely used as a killer application of Location Based Service, provides drivers with the shortest route using a network based on a road center line. Providing. Recently, however, wireless network environments such as WiBro and Wi-Fi, as well as PMP (Portable Multimedia Player), Netbook, Smart Phone, and GPS (Global Position System) are received. The development of portable devices such as mobile phones has made it easy to use the navigation not only inside the vehicle but also from the outside.

However, at present, the CNS provides the shortest path based on the network information on the space where the car can move along the road center line.However, unlike cars, pedestrians have limitations in using the CNS network, which is free to move and only the road center line. . This means that new network information for pedestrians is needed.

The research related to the network considering the pedestrian movement characteristics can be largely divided into two types: the node-link graph structure and the non-graph structure. First, in the case of a graph structure, information about an object that can be moved by a pedestrian is added to a CNS network, or a network that considers a pedestrian's movement characteristics is created by using a space syntax. Next, we propose a network using lines and planes to overcome the limitations of constructing and updating existing graph structures without moving along the center line of the road. Ota (2004) is a two-dimensional complex (surface) in the downtown area, and narrow alleys in one dimension. Tang and Pun-Cheng (2004) argue that the centerline auto-generation process provided by most GIS commercial programs is limited to generating sophisticated information and is time-consuming and cumbersome. Therefore, a space object (a block surrounded by roads, sidewalks, and roads) that is defined as being able to walk with respect to a space is a face, and a limited pedestrian space (passage, crosswalk) or a walkable object (stair) on a road is a line. Structured. Walter et al. (2006) modeled urban areas or blocks with raster data, saying that large area vector data is useful only for roads and not for blocks surrounded by pedestrians or roads. However, the path is searched by creating a line or a center line indicating a direction in a network considering a line and a face, that is, a non-graph structure.

Basically, the vector data can contain more properties than the raster data structure. Based on this, the shortest path can be searched, so the former is used in the Pedestrian Navigation System (PNS). It is a situation. For example, in recent years, the US navigation mapping company, Navtek, has added information about vehicles, such as India, parks, overpasses, and auto-free pedestrian zones, to North America. A pedestrian map was created and commercialized as Nokia Ovi Maps3.0, which can be downloaded for free on Nokia mobile phones and the shortest path when walking through the Internet site Ovi Maps (http://maps.ovi.com). Providing. Figure 1 is an example of the route search service on the Internet Ovi Maps, the route from the Institute of Texan Cultures in San Antonio, Texas, USA to the Fairchild Park was explored by the drive and walking options respectively. FIG. 1 (a) shows a route searched with a drive option, and FIG. 1 (b) shows a route provided with a walking option, and crosswalk information is added to provide a different route.

The current network for pedestrians is being provided by adding information on the main road and pedestrian spaces that cross the main roads, such as pedestrian crossings, overpasses, and underpasses. In other words, after considering the space for pedestrians to move, the network is created and provided. The pedestrian network currently being serviced is built around the main roads, so the route search is provided only to the periphery of the building.

Therefore, it is necessary to reflect not only the CNS road center line or the main road-oriented network but also the space where pedestrians can move, that is, sidewalks and crosswalks, as well as spaces that can be walked in blocks surrounded by roads. In addition, the network for pedestrians in the future will need to provide a PNS linked to the network inside the building through the entrance as well as the path to the building entrance through the entrance of the individual building.

However, there is a problem that a lot of time and money are required to construct a network by examining objects for these pedestrians. For example, according to Google's performance report, Korea invested 2.4 trillion won in 2006, $ 1.9 billion in data center construction and operation costs, and $ 2.4 billion in 2007. Although the above costs are not all network building costs, it shows simply that high costs are required to build and manage data.

The present invention has been made to solve the above problems, an object of the present invention is to build a pedestrian network data automatically by using the already built spatial information to shorten the time required for construction to facilitate network creation In addition, it provides a pedestrian network generation method that can be used as basic data for more precise network creation by reducing construction costs.

To achieve these and other advantages and in accordance with the purpose of the present invention,

(a) pre-built first spatial information is input to the main memory as input data, and a pedestrian space for linearization around the main road is extracted from the pre-built first spatial information using a program in the main memory; Wow;

(b) Linearization of the pedestrian space in the block surrounded by the main road of the second-stage construction target area from the pre-built second spatial information by using the program in the main memory as input data to the main memory as input data. Extracting a walking space for;

(c) generating a first-class basic network through linearization using a program in a main memory including a skeleton algorithm in the walking space extracted in step (a);

(d) generating a second-class basic network through linearization using a program in main memory including a skeleton algorithm in the walking space extracted in step (b); and

(e) merging the basic network and the gateway information extracted from the gateway layer in steps (c) and (d) to generate a node-link network structure that can be utilized in the PNS using a program in the main memory. It is the basic characteristic to become.

The pedestrian network generation method using the pre-built spatial information of the present invention as described above not only facilitates network creation by shortening the time required for construction by automatically constructing a pedestrian network using the pre-built spatial information. In addition, it can be used as a basic data for creating more precise network by reducing construction cost, and by connecting the internal network to the entrance node of the pedestrian network, it is possible to provide PNS not only in the external environment but also in the internal environment.

1 is a view showing a screen of a pedestrian route search service currently provided by the Internet site (http://maps.ovi.com).
2 is a flowchart illustrating a process of generating a pedestrian network using pre-built spatial information.
3 is a diagram illustrating a feature classification system of a digital map, which is one of prebuilt spatial information;
4 is a conceptual diagram illustrating the generation of a pedestrian space for linearization around a main stage 1 road.
FIG. 5 is a flowchart illustrating a pedestrian space generation for linearization around a main stage 1 road; FIG.
FIG. 6 is a diagram illustrating a concept of generating a pedestrian space for linearization in a block surrounded by a main road of a two-step construction target area; FIG.
FIG. 7 is a flowchart illustrating the generation of a pedestrian space for linearization in a block surrounded by a main road of a two-stage construction target area; FIG.
8 is a flowchart illustrating the linearization of the walking space around the main road of the first stage.
Fig. 9 is a flowchart showing the linearization of pedestrian spaces within blocks surrounded by main roads of the two-stage construction target area.
10 shows a flowchart for creating a pedestrian network.
11 shows a detailed flowchart for creating a pedestrian network.
12 and 13 illustrate the results generated step by step by applying the method according to the invention.

When described in detail with reference to the accompanying drawings a preferred embodiment of the present invention configured as described above are as follows.

FIG. 2 is a flowchart illustrating the overall process of generating a pedestrian network by using previously constructed spatial information. As shown in FIG. 2, first, a walking space is provided free of charge by the national map and the Ministry of Public Administration and Security. The pedestrian space for the linearization of the first stage is extracted by analyzing the spatial information, such as the basic road map of the name of the road (S10). Next, it analyzes the spatial information, such as the digital map of the national basic map and the road name address basic map provided by the Ministry of Public Administration and Security for free, to extract the pedestrian space for two-step linearization (S20). Subsequently, linearization around the main stage of the first stage is performed on the walking space for the linearization of the first stage, which is extracted in step S10, to generate a first-class basic network (S30). Subsequently, linearization is performed in the two-stage block with respect to the walking space for the two-stage linearization extracted in S20 to generate a second-class basic network (S40). Finally, based on the basic networks generated at S30 and S40, data for the shortest path search service, ie, a pedestrian network, is generated (S50).

Before extracting pedestrian space from the pre-built spatial information, it is necessary to analyze the definition of the pedestrian space and the information of the pedestrian space that can be obtained from the pre-built spatial information. Therefore, first, the analysis of the walking space based on the previous research and the already serviced PNS, and an example of the analyzed items to create a pedestrian network are shown in Table 1 below.

As shown in Table 1 above, the pedestrian path search service that is provided in the previous research and service can be largely divided into a space in which a pedestrian can move and a space in which a pedestrian cannot move, that is, an obstacle. Looking at the detailed spatial objects included in each, pedestrians can include sidewalks, walkways, bicycle paths, squares, bridges (pedestrian bridges), pedestrian crossings, underpasses, parks, pedestrian spaces, building entrances, and stairs. Spaces that cannot be moved by pedestrians, such as obstacles, include buildings, roads, mountains, and rivers / lakes.In this case, the space inside the building is a walking space, but the building itself is an obstacle for pedestrians to go around. There is a space (india, pedestrian crossing, underpass, overpass, etc.), and the rest of the space is mainly a space where cars move, and pedestrians cannot move.

The pedestrians analyzed in this way should be analyzed whether or not the space that can be moved is built and managed in the existing spatial information, and should use the corresponding data (layer) to extract the walking space (S10, S20).

First, the pre-established spatial information to be used in the present invention mainly used the digital map, which is a national base map produced and sold by the National Geographic Information Institute, and the road name address basic map built by the Ministry of Public Administration and Security for the purpose of managing road name addresses. Target data from internet portal sites.

The digital map, which is the national base map, is produced based on the numerical mapping rules and consists of scales of 1: 1,000, 1: 5,000 and 1: 25,000. 3 is a digital map feature classification system, which is classified into eight major categories such as traffic, buildings, facilities, vegetation, aquatic landforms, boundaries, and cycles regardless of the scale, and the features of the same characteristics are classified into subclasses below. For example, in the case of large-scale traffic, 22 features such as road boundaries, road center lines, sidewalks, crosswalks, safety zones, overpasses, bridges, intersections, solid intersections, interchanges, tunnels, tunnel entrances, and stops Included. The features included in these subcategories are constructed in one layer for each, and spatial information (vector data structure) and attribute information are linked. For example, if a road center line is drawn in the road center line layer, the road center line contains the road number, the name of the road, the classification of the road, the start and end points of the road, the pavement material, the presence and absence of lanes, the number of lanes, the width of the road, and the one-way road. Attribute information such as related information is linked. Digital map features include A, B,... Major classifications are one letter of the alphabet, such as H, and small classifications are assigned a unique number with three digits (001-999).

The road name address map of the Ministry of Public Administration and Security is spatial information constructed for the purpose of managing the new address. It is updated every month and is currently distributed free of charge to applicants. The road name address base map does not have a classification system like a digital map, and spatial information and attribute information are linked to the corresponding layer for each spatial object, and additional attribute information for managing road name addresses is managed in a separate table. . For example, in case of building layer, the building management number, road section management number, building group management number, road name management number, underground part, building number main and minor, building name, detailed building name, building status, building Attribute information such as application codes, ground and underground levels is linked and managed. The constructed spatial information includes the actual road, road section, road section history, road nameplate, base section, building, building history, building group, entrance, connecting line, subway line, railway line, bridge, park, river, lake, tunnel , Subway stations, subway gates, railway stations, and other signs.

In the following Table 2, the corresponding layers were analyzed for each pre-established spatial information based on the pedestrian space and the non-walkable space (obstacles). India, promenade, bicycle path, plaza, bridge (viaduct), crosswalk, underpass, park, pedestrian space, building entrance, stairs, etc., which can be moved by pedestrians Information on bridges, viaducts, pedestrian crossings, underpasses, parks, and entrances to buildings can be used. Obstacles can utilize buildings, roads, mountains, and rivers / lakes. In the present invention, the castle, subway line, railway line, and tunnel information are added to the obstacle. Castles in the city are not pedestrian spaces, and underground subway and rail lines are also obstacles. In the case of tunnels, roads are classified as obstacles that cannot be walked due to the structure passing by them. In an unusual case, however, information on mountains (boundary information) or parks is not constructed in the digital map, and the park boundary containing the pedestrian space and mountain boundaries are managed in the park layer.

There is a layer managed in duplicate in the previously constructed spatial information, and the construction time of these spatial information is different. The digital maps are updated every five years, and the digital maps 1: 1,000 and 1: 5,000 were also constructed differently. As mentioned above, the road name address base map is updated on a monthly basis.

Therefore, the present invention intends to approach in a manner of utilizing additional information in the digital map based on the road name address base map. The road name address base map is for the purpose of managing the new address of the building unit, so the information related to the building can be regarded as the latest information. However, the road is not the main managed spatial object. The road-related information uses the 1: 5,000 numerical map, which is the most recent national base map. In addition, crosswalks, which are spaces for pedestrians to move on the road, are built only on 1: 1,000 digital maps. Table 3 below is an example showing the final pre-built spatial information used in the present invention.

A pedestrian network of good quality can be generated by the method proposed by the present invention only when the walking space is obtained at a level that can reflect the real world as much as possible. Pedestrian space is also extracted accordingly for two stages of linearization (step 1: linearization around the main road and step 2: linearization of blocks). Linearization around the first stage main road extracts pedestrian crossings and overpasses for pedestrians from sidewalks and roads, which are the pedestrian spaces around the main road in the construction area, and within the block surrounded by the main road for linearization of the second stage block unit. Extract the walking space.

1. Extraction of pedestrian space for linearization around the first stage main road ( S10 )

4 is a conceptual diagram for pedestrian space extraction (S10) for linearization around the main road in step 1, and in the first step, a basic network related to pedestrian crossings and viaducts crossing both main roads and main roads in both sides of the main road is generated. For this purpose, it removes the part where the pedestrian crossing and the pedestrian bridge overlaps on the main road (1) and newly defines the road boundary as much as the road width is subtracted from the actual road width. This road boundary is called a main road obstacle (3) as a space, that is, an obstacle that pedestrians cannot travel. Next, buffer the contour layer by a certain distance to generate a wide contour layer to apply the method proposed in the present invention (4), and remove the main road boundary (road width of the main road) from the large contour layer (5) do. Finally, by merging the main road obstacles 3 into the area 5 in which the road width of the main road has been removed, a difference in the width of the sidewalk and a walking space (white) in which pedestrian crossings and overpasses are removed, and linearization is performed Space (6).

5 is a flowchart for extracting a walking space (S10) for linearization around a main road in step 1, wherein the first spatial information, which is input data used in step 1, is a road center line layer of a digital map (1: 5,000). 100, the road boundary layer 200, the contour layer 300 of the numerical map (1: 1,000), the crosswalk layer 400 of the numerical map (1: 1,000) and the viaduct layer of the numerical map (1: 5,000) At 500 it is entered into the main memory. Here, the contour layer 300 is a mesh that manages data in a digital map, and the whole of Korea is divided into several maps to manage spatial information.

Article 16 (Pressure) of the Rules on Structural and Facility Standards of Roads states that the effective width of sidewalks can be at least 2 m, and inevitably 1.5 m or more. Rule on Installation Standards Article 9 (Road Classification) Pedestrian-only roads are defined as pedestrian-only roads with a width of 1.5m or more. However, for the installation of India, it is installed in consideration of the safety of pedestrians and smooth traffic such as cars, and the criteria for the road width and the minimum number of required lanes are not clear. Therefore, in the first step, the main road is defined as the road name address and road standard, that is, a road having a road width of 12 m or more (S111).

In this way, the main road centerline 1800 having a road width of 12 m or more is extracted from the road center line layer 100 (S112), and the main road boundary lines 2500 and S113 overlapping the main road centerlines 1800 and S112 are road boundary layers. Extract at 200.

In the main road center line (S112), the crosswalk layer 400 of the digital map (1: 1,000) and the pedestrian crossing layer 500 of the digital map (1: 5,000) are acquired and merged (Merge) (S114). After removing the crosswalk and the viaduct from the main line of the main road (S115). Pedestrian crossings and overpasses are spaces for pedestrians to move around, and are not obstructions. A main road obstacle is generated by performing a buffer (S116) at a distance obtained by subtracting the width of both sidewalks from the actual road width to the main road centerline generated in S115.

Skeleton algorithm to be applied in the linearization process (S30, S40) is characterized in that the Y-shape generated near the boundary of the region (partial region) to which the algorithm is applied. However, when linearizing the walking space, due to the limitation of the equipment (computer) to do this, the entire region is cut into partial regions and the skeleton algorithm is applied, so the linear lines generated by the skeleton for each partial region are partial. It should also be created in a straight line at the boundary of the area. Therefore, in the present invention, a partial region for applying a skeleton algorithm is defined as a contour boundary of a 1: 1,000 digital map, and solves the above-mentioned problem (a linear shape is generated at the boundary of the partial region in a Y shape). In order to perform linearization by applying a skeleton algorithm by creating an area wider than the contour boundary of the 1: 1,000 digital map, cut out the generated skeleton alignment into the contour boundary of the 1: 1,000 digital map. We try to solve the problem of the skeleton algorithm by cutting out the Y-shaped alignment at the partial region boundary. The buffer is applied to the contour layer 300 of the 1: 1,000 numerical map by a predetermined distance (S117) to generate a contour layer 1700 that is wider than the actual contour layer 300 (S118).

By removing the main road boundary (2500, S113) from the wide contour layer (S118) to create a region removed by the actual road width, merge the main road obstacles generated in the above (S116) (S119) of both main roads Create pedestrian crossings and viaducts across India's pedestrian spaces and main roads.

On the other hand, the walk space extraction (S10) process for linearization around the main road of the first step is performed by a program stored in the main memory by coding an algorithm directly through a programming language in order to perform this by a computer, or ARGIS, which is commercial software. This is done through the (ArcGIS) modeler.

2. Extraction of pedestrian space for linearization of pedestrian space in the block surrounded by main roads S20 )

FIG. 6 is a conceptual diagram of pedestrian space extraction (S20) for linearizing pedestrian spaces within a block surrounded by main roads of a two-step construction target area, and the second step is to extract pedestrian spaces within a block. In step 1), buffers are performed by a predetermined distance (2) to remove them from the wide contour layer (a wider range than the actual 1: 1,000 numerical map contour layer), thereby generating a wide block (3).

Next, an obstacle layer is generated (4) by overlapping layers such as a building group, a building, and a river / lake. The obstacle layer 4 thus generated is removed from the wide block 3 generated above to finally generate a walking space (blue) in the block (5).

FIG. 7 is a flowchart illustrating extraction of a walking space for linearization of a walking space in a block surrounded by a main road of a two-step construction target region. The second spatial information, which is an input data, is a castle layer (1: 5,000) of a numerical map (1: 5,000). 1100), building group layer 600 of road name address base map, park layer 700, building layer 800, subway line layer 900, railway line layer 1000, river / lake layer 1200, railway History layer 1300 and subway station layer 1400 are required, which are input into the main memory.

In addition, a wide contour layer (1700) and a main road centerline layer (1800) generated in the first stage linearization are required, and buildings and parks such as apartments, schools, and public institutions extracted from aerial photographs or satellite images provided on the Internet are required. An inner road layer 1600 is used.

The buffer is applied to the main road centerline layer 1800 by a predetermined distance (S211), and then removed from the wide outline layer 1700 (S212). The result is a wide block layer to which the skeleton algorithm will be applied.

The inner road layer 1600 of the building group and the park is a layer which extracts only an inner road that can remove the building group or the building group existing in the interior of the building group and the park. Since the inner road layer 1600 is a walkable road, a building group or a park in which the inner road layer exists when the obstacle is generated should be excluded. For this purpose, the building group existing in the park among buildings and building groups was not considered as an obstacle, and the park layer overlapping with the inner road 1600 was not considered as an obstacle. That is, since the building group and the park already use the inner road layer 1600 as a walking space, the building or the building group overlapping the park and the park layer having the inner road need not be considered as an obstacle. Therefore, the building group on the park is deleted by overlapping the building group layer 600 and the park layer 700 (S213).

The main road 1800 or the inner road 1600, which is a space for walking through the building group S213 generated above, is removed to form final building group layers 2100 and S214. It is often the case that buildings pass on the road because the base of the road name address and the 1: 5,000 digital map are different. Therefore, if the building layer 800 is present on the main road 1800 is removed (S216).

Also, in order to remove the building overlapping the park layer, the building included in the final building group 2100 in which the building group in the park generated above is deleted (S214) is deleted (S217) to generate the final building layer 2200. do.

A buffer is applied to the subway line layer 900 of the road name address base map, the railway line layer 1000, and the castle layer 1100 of the 1: 5,000 numerical map by a predetermined distance (S218), and arbitrary boundary information is applied to each layer. Create

Next, the final building group layer 2100 generated by removing the main road 1800 and the internal road 1600 passing through the obstacles generated in the step, that is, the building group (S214), and removing the internal road in the park (S215). The final park layer generated by deleting the building in the building group (2100), the pedestrian space is not present in the park, the final building layer (2200), buffer applied (S218) subway line, railway The obstacles in the block are generated by merging the track, the castle layer, the river / lake layer 1200, the railway station layer 1300, and the subway station layer 1400 (S219). The obstacle thus generated is removed from the wide block generated in S212 to generate a walking space in the wide block.

In addition, the pedestrian space extraction (S20) process for linearizing the pedestrian space in the block surrounded by the main road of the second stage construction target area is performed by a program stored in the main memory by coding an algorithm directly through a programming language to perform this by a computer. Or through the commercial software ArcGIS modeler.

Meanwhile, hereinafter, the linear network for generating the pedestrian network is performed by performing linearization in two steps (step 1: linearization around the main road and step 2: block linearization) on the walking space generated in S10 and S20. To generate (S30, S40).

As a process of creating a basic network divided into two stages, linearization around the main road in the first step is performed by applying a skeleton algorithm to sidewalks, pedestrian crossings, and overpasses, which are the pedestrian spaces around the main road generated in S10. The linearization process is performed, and the second step is a linearization process by applying a skeleton algorithm to the walking space in the block surrounded by the main road generated in S20. In the present invention, a skeleton algorithm applied to generate a linear shape in a walking space is a thinning technique, which erodes inward from the boundary of the foreground background in a binary image, while the center pixel of the target area is eroded. As described above, the algorithm is repeatedly performed until one line remains. As described above, a Y-shaped linearity occurs at the boundary of the region, and in particular, distortion occurs at the junction. Nevertheless, it is one of the widely used algorithms for extracting linearity from raster images.

3. Linearization of pedestrian space around the first stage main road ( S30 )

FIG. 8 is a flowchart of linearization around the main road of step 1 (S30). In S10, the skeleton algorithm considers a case where a Y-shaped linear is generated at the boundary of the target area, and thus, the actual target area. Create walking space for a wider area. The walk space is 0 (S313) according to whether the walking space 1900 is generated or not (S313). Otherwise, the obstacle space is allocated NODATA (S312) to obtain a binary image having a pixel size of 1m. Create 1m can be seen as the minimum space that the pedestrian can move to the shoulder width of the pedestrian, and the walking space is linearized by applying a skeleton algorithm (S314) to the walking space (0).

Therefore, in order to generate a network, a linear raster structure is converted into a vector structure (S315). At this time, one intersection should be generated at the intersection, but when linearized by applying the skeleton algorithm, a problem arises in that several intersections are created instead of one. In order to solve the problem that multiple intersections are generated at the intersection, it is determined whether each of the linear segments (a line connected to the point, that is, a line segment) is smaller than a threshold (pixel size) (S316). In operation S317, the segment that is smaller than the threshold value is integrated in operation S317, in operation S318, to generate a network of connected node-link structures.

In operation S318, linear simplification is performed by applying a Douglas-Peucker algorithm, which is frequently applied to linear data, to an alignment converted into a vector structure (S319), and unnecessary nodes existing in the middle of the generated linear ( vertex). In this way, basic network data 2300 having a road grade of 1 is finally generated (S320).

In the meantime, the linearization (S30) process around the main road in step 1 is performed by a program stored in the main memory by coding an algorithm directly through a programming language to perform this by a computer, or by using an arcGIS modeler, which is commercial software. Is done through.

4. Linearization of pedestrian space in the block surrounded by main roads S40 )

FIG. 9 is a flowchart illustrating linearization of a walking space in a two-stage block (S40). In S20, the skeleton algorithm considers a case where a Y-shaped linear is generated at the boundary of the target area. Create walking space for an area wider than the area. In addition, in step S20, the skeleton algorithm generates a walking space in a block wider than the actual block boundary in consideration of the case where a Y-shaped linear is generated at the block boundary. For this purpose, the boundary layer S411 of the block surrounded by the actual road is generated by removing the road boundary layer 200 from the wide outline layer 1700. The boundary layer (S411) of the block surrounded by the actual road is applied to the wide block boundary (S212) during the linearization process in the two-stage block. By cutting only as much as (S411), it is utilized to remove the Y-shaped distorted linearity generated at the block boundary.

Next, with respect to the walking space 2000 generated in the above step S20, the walking space is given with 0 (S414) and the obstruction is given with NODATA (S413) according to whether the walking space is S412. The binary image is generated, and the linearity of the generated raster structure is converted to the linearity of the vector structure by applying a skeleton (S415) (S416). At this time, one intersection should be generated at the intersection, but when linearized by applying the skeleton algorithm, a problem arises in that several intersections are created instead of one. In order to solve the problem of generating multiple intersections at the intersection, it is determined whether each of the linear segments (a line connected by a point or a line, that is, a line segment) is smaller than a threshold (pixel size) (S417). In operation S418, in operation S418, the removed segment is integrated (S419) to generate a network of connected node-link structures. In step S419, a linear simplification is performed by applying a Douglas-Peucker algorithm, which is frequently applied to linear data, to a linear converted into a vector structure (S420), and unnecessary nodes existing in the middle of the generated linear ( The process of removing the vertices is equivalent to one-step linearization.

In order to solve the problem that the Y-shaped linear is generated at the block boundary, the linear structure of the vector structure generated at S420 is cut out using the generated block boundary layer S411 (S421). As a result, it is possible to remove the distortion linearity that may occur at the block boundary and generate a second-class basic network 2400 (S422).

In addition, the linearization of the walking space in the two-stage block (S40) is also performed by a program stored in the main memory by coding an algorithm directly through a programming language in order to perform this by a computer, or arcGIS modeler which is commercial software. Is done through.

5. Create a pedestrian network ( S50 )

FIG. 10 is a flowchart for generating a pedestrian network (S50), and extracts the doorway information from the doorway layer 1500 of the road name address basic diagram inputted into the main memory as input data (S510), and generates the above (S30). Merges the first-class basic network 2300 and the second-class basic network 2400 generated in S40 (S520), and generates a link connecting the entrance and the network extracted from the S510 to the network (S520). In operation S530, a pedestrian network is generated.

FIG. 11 is a detailed flowchart for generating a pedestrian network (S50), each of the building group layer 2100 and the building layer 2200 generated and extracted in S20, and the entrance layer 1500 of the basic road name address map. Overlapping the building group or the entrance on the building to extract (S511-S512). The entrance (S511) on the extracted building group and the entrance (S512) on the building are merged (S513) to generate a final entrance layer.

The first-level basic network 2300 generated at S30 and the second-level basic network 2400 generated at S40 are cut into the contour layer 300 of the 1: 1,000 numerical map (S521-S522). Remove the Y-shaped distorted alignment created at the boundary of the region. In order to connect the first-class and second-class networks generated in the above steps (S521-S522), a point adjacent to the first-class network is extracted from the second-class network, and the point having the shortest distance from the extracted point to the first-class network is 1 Extract from the rating network. The first-class network connects the first-class network to the second-class network by creating a connection line (line segment) between the points adjacent to the first class network and the points above the first class network at the shortest distance.

In more detail, in the second-level basic network, a point S523 of the second-class basic network adjacent to the main road boundary layer 2500 generated in S10 is extracted. In the first-class basic network, the points extracted in the above-mentioned (S523) and the shortest points are extracted from the first-class network (S524), and between the points of the second-class basic network (S523) and the points of the first-class basic network (S524). Create a connection link (S525). As a result, one network is created by merging (S526) the first-class basic network (S522), the second-class basic network (S521), and the connection link (S525) of the two basic networks.

In order to provide a path to the doorway of the building, when the doorway generated in S513 is connected to the merged basic network S526, a network for pedestrians is created. Similar to the process of merging the first and second basic networks, this process also extracts the points at the shortest distance from the entrance (S513) to the merged basic network (S528) in the merged basic network (S526) (S531). . By generating a link connecting the point (S531) and the entrance (S513) at the shortest distance extracted in this way (S532) and merging it with the merged basic network (S526) (S533), it can be guided to the entrance of the building. Create the final pedestrian network.

In order to provide a route search service using the pedestrian network generated in S533, it is necessary to change the graph structure of the node-link. To this end, node (point) information is extracted from the pedestrian network generated in S533 (S541) to generate an object identifier (ID) for each node, and a link ID is assigned (S542). To each link of the pedestrian network generated in S533, an ID S542 of a start node and an end node constituting each link is added (S543).

In addition, Clink provides the distance between nodes needed in the process of calculating the shortest distance in order to provide the path search service providing the shortest path to the pedestrian, that is, the link length information and the link information adjacent to the discovered node during the path search. Information is generated (S544). In this way, a pedestrian network in consideration of a moving space of a free pedestrian capable of providing a path from the PNS to the entrance of the building is generated (S545).

On the other hand, the process of generating the pedestrian network (S50) is performed by a program stored in the main memory by coding the algorithm directly through a programming language in order to perform this by a computer or through a commercial software ArcGIS modeler (ArcGIS) modeler. .

12 and 13 are diagrams illustrating the results generated step by step by applying the method according to the present invention. For each description, (a) removes pedestrian crossings and viaducts from the center line of the main road, and the width of the guideway Represents a one-step obstacle S116 generated by performing a buffer as long as the length removed, and the part indicated in white in (b) removes the boundary of the main road in a wide contour layer, and the one-step obstacle (gray) in (a). ) Represents a one-step walk space (S119) generated by merging), and the solid red line in (c) represents a skeleton of the raster structure generated by applying a skeleton algorithm to the first-step walk space of (b), (d) is a first-class basic network (S522) which is converted into a linear structure of a vector structure by converting the skeleton of (c) into a linear structure, and then clipping it into a contour layer of a 1: 1,000 digital map.

In addition, the gray polygon in (e) represents a two-stage obstacle (S219) in which a building, a building group, a subway line, a rail line, a buffered castle, a river / lake, a park, etc. are merged, and the green polygon of (f) Denotes a two-stage walk space (S220) generated by removing the two-stage obstacle of (e) in the wide block layer, and the solid blue line in (g) represents a skeleton in the two-stage walk space of (f). The skeleton of the raster structure generated by applying the algorithm is represented, and (h) converts the skeleton of (g) into a linear structure of the vector structure, undergoes a linear simplification process, and cuts it into a block boundary layer (clip). It is a two-stage basic network (S521) generated by clipping again to the contour layer of the digital map.

And (i) the first-level basic network of (d), which is a solid red line, and the second-level basic network of (h), which is a solid blue line, and the shortest distance connection link between these first-level basic networks and a second-level network (black and thick The solid line) shows the merged state (S526), and (j) shows the merged link (solid solid line) between the basic network (gray solid line) merged in (i) and the entrance (S533). In (k), the solid gray line is a final pedestrian network (S545) which gives a link ID to the linearity of (j), finds the start node (red dot) and the end node (red dot) of each link, and assigns the node ID. In the link of the pedestrian network, the ID, start node ID, and end node ID of each link are stored, and the ID of each node is stored in the attribute table.

Although the above has been shown and described specific preferred embodiments of the present invention, the present invention is not limited to the above-described embodiment, those skilled in the art to which the present invention pertains the technical gist of the present invention. Various changes can be made without departing.

100: numerical map (1: 5,000) road center line 200: numerical map (1: 5,000) road boundary
300: Numerical map (1: 1,000) Outline 400: Numerical map (1: 1,000) Crosswalk
500: numerical map (1: 5,000) overpass 600: road name address base map building group
700: road name address base map park 800: road name address base map
900: road name address basic map subway line 1000: road name address basic map railway line
1100: numerical map (1: 5,000) Castle 1200: road name address basic map river / lake
1300: Road name address basic map railway station 1400: Road name address basic map subway station
1500: Road name address basic road entrance and exit 1600: Building group / park internal road
1700: wide road
1800: main road centerline created in one-step linearization
1900: Walking space for one-step linearization
2000: walking space for two-step linearization
2100: extracted building group 2200: extracted building
2300: Class 1 Basic Network 2400: Class 2 Basic Network
2500: Main road boundary created by one-step linearization

Claims (11)

(a) pre-built first spatial information is input to the main memory as input data, and a pedestrian space for linearization around the main road is extracted from the pre-built first spatial information using a program in the main memory; Wow;
(b) Linearization of the pedestrian space in the block surrounded by the main road of the second-stage construction target area from the pre-built second spatial information by using the program in the main memory as input data to the main memory as input data. Extracting a walking space for;
(c) generating a first-class basic network through linearization using a program in a main memory including a skeleton algorithm in the walking space extracted in step (a);
(d) generating a second-class basic network through linearization using a program in main memory including a skeleton algorithm in the walking space extracted in step (b); and
(e) merging the basic network and the gateway information extracted from the gateway layer in steps (c) and (d) to generate a node-link network structure that can be utilized in the PNS using a program in the main memory. Method for generating a pedestrian network using pre-established spatial information, characterized in that the.
The method of claim 1,
The first spatial information of step (a) includes the road center line layer 100 of the digital map (1: 5,000), the road boundary layer 200, the contour layer 300 of the digital map (1: 1,000), and the digital map ( 1: 1,000) pedestrian network generation method using the pre-established spatial information, characterized in that the crosswalk layer 400 and the overpass layer 500 of the digital map (1: 5,000).
The method of claim 1,
The second spatial information of step (b) includes the castle layer 1100 of the numerical map (15,000), the building group layer 600 of the basic road name address map, the park layer 700, the building layer 800, and the subway. Track layer 900, railway track layer 1000, river / lake layer 1200, railway history layer 1300, subway history layer 1400, interior layer of building group and park layer 1600, the step ( Method for generating a pedestrian network using the pre-built spatial information, characterized in that the wide contour layer (1700) and the main road centerline layer (1800) generated in a).
The method of claim 1,
The program in the main memory of step (a),
The road boundary where the main road center line 1800 whose road width exceeds a predetermined value is extracted from the road center line layer 100 which is the first spatial information, and the main road boundary 2500 that overlaps the main road center line is the first spatial information. Extracting the rare layer 200;
Acquire pedestrian crossing and viaduct information from the pedestrian crossing layer 400 of the digital map (1: 1,000), which is the first spatial information, and the viaduct layer 500, of the digital map (1: 5,000), which is the first spatial information, from the main road center line. Removing by;
Generating a main road obstacle by performing a buffer on the main road center line finally subtracted from the actual road width by subtracting both sidewalk widths (S120);
Generating a contour layer 1700 wider than the actual contour layer 300 by applying a buffer to the contour layer 300 of the numerical map (1: 1,000), which is first spatial information, by a predetermined distance, and
By removing the main road boundary from the wide contour layer, an area removed by the actual road width is created, and the main road obstacles generated above are merged to cross the sidewalks and the main roads on both sides of the main road. Pedestrian network generation method using the pre-established spatial information, characterized in that it comprises a step of generating a sidewalk and overpass.
The method of claim 1,
The program in the main memory of step (b),
Generating a wide block layer by applying a buffer to the center line layer 1800 as the second spatial information by a predetermined distance and removing the same from the wide contour layer 1700 as the second spatial information;
After the building group on the park is deleted by overlapping the building group layer 600, which is the second spatial information, and the park layer 700, the main road 1800 or the inner road 1600 passing through the building group, which is the second spatial information, is deleted. Removing the final building group layer 2100;
If the building layer 800, which is the second spatial information, exists on the main road 1800, the building is deleted and the building existing in the final building group 2100 generated above to remove the building overlapping the park layer. Removing the final building layer 2200;
A buffer is applied to the subway line layer 900, the rail line layer 1000, and the castle layer 1100 of the numerical map (1: 5,000) of the second spatial information by a predetermined distance (S218). Generating boundary information for each layer;
Final building group layer 2100, final park layer created by removing internal roads in the park, final building layer 2200, buffered subway lines, railway lines, castle layers, and river / lake layers, which are second spatial information. (1200) merging the railway history layer (1300), subway history layer (1400) to generate obstacles in the block, and
And a step of generating a walking space by removing the obstacle from the wide block generated in the above block.
The method of claim 1,
The program in the main memory of step (c),
Generating a binary image having a predetermined pixel size by allocating a zero walking space and an obscure space according to whether the walking space generated in the step (a) is capable of walking;
Applying a skeleton algorithm to the walking space (0) so that the walking space is linearized and a linear raster structure is converted into a vector structure;
When multiple intersections are generated at the intersection by linearizing by applying a skeleton algorithm, segments smaller than the threshold (pixel size) of each segment of the linear are removed, and segments which are smaller than the threshold are removed, and
Performing a linear simplification by applying a Douglas-Peucker algorithm to a linear transformed vector structure, and finally generating a basic network 2300 having a road grade of 1 Pedestrian network generation method using constructed spatial information.
The method of claim 1,
The program in the main memory of step (d),
Generating a boundary layer of the block surrounded by the actual road by removing the road boundary layer 200 from the wide contour layer 1700 for the walking space of step (b) generated in the block wider than the actual block boundary;
Generating a binary image having a constant pixel size by giving a walking space of 0 and an obstacle of NODATA according to whether the walking space is generated in the walking space generated in step (b);
Converting the linearity of the raster structure generated by applying a skeleton algorithm to the walking space 0 into the linearity of the vector structure;
If multiple intersections are created at the intersection by linearizing by applying the skeleton algorithm, segments smaller than the threshold (pixel size) are removed from each segment of the alignment, and those segments smaller than the threshold are merged. Where Douglas-Peucker linear simplification is performed for, and
And generating a second-class basic network 2400 by cutting out the linearity of the vector structure generated by the generated block boundary layer.
The method according to any one of claims 4 to 7,
The program in the main memory of step (e),
(f) extracting the doorway information from the doorway layer 1500 of the road name address base map inputted into the main memory as input data;
(g) merging the first-class basic network and the second-class basic network generated in steps (c) and (d), respectively, and
(h) generating a pedestrian network by generating a link connecting the doorway information extracted in step (f) and the basic network merged in step (g). How to create a pedestrian network.
The method of claim 8,
The step (f) overlaps the building group layer 2100 and the building layer 2200, which are generated and extracted in the step (b), with each of the building layer 2200 and the doorway layer 1500 of the basic road name address map. Extracting and merging the doorway on the building group with the doorway on the building to create a final doorway layer.
The method of claim 8,
Step (g) is,
Cutting the first-level basic network 2300 generated in step (c) and the second-level basic network 2400 generated in step (d) into the contour layer 300 of the 1: 1,000 numerical map;
Extracting points of the second-class basic network adjacent to the boundary layer 2500 of the main road generated in the step (a) from the second-class basic network cut and formed by the contour layer 300;
Extracting, from the first-class basic network, the points extracted at the shortest distance from the points extracted from the second-class basic network;
Generating a connection link between the points of the second-class basic network and the points of the first-class basic network, and
Leverage the pre-established spatial information, which consists of the steps of creating a network of 1st and 2nd basic network by merging connection link connecting 1st class basic network, 2nd class basic network and 2nd basic network How to create a pedestrian network.
The method of claim 8,
Step (h) is,
Extracting a point at the shortest distance to the basic network merged at the entrance and exit;
Generating a link connecting a point and an entrance at the shortest distance to the merged basic network and merging it with the merged basic network to generate a pedestrian network;
Extracting node (point) information from the generated pedestrian network, generating an object identifier (ID) for each node, and assigning an ID of a link (linear);
Adding the node ID generated above to a start node and an end node constituting each link of the generated pedestrian network; and
C link that provides distance between nodes, that is, link length information and link information adjacent to the discovered node during path search, to calculate the shortest distance in order to provide the path search service that provides the shortest path to pedestrians. A method for generating a pedestrian network using pre-established spatial information, characterized in that the step of generating information.

Images