KR101190882B1 - Apparatus and method of cloud computing for online visualization of noise map at high speed - Google Patents

Abstract

Cloud computing can be used to generate three-dimensional noise maps in u-city. In this paper, we describe a 3D noise map generation method using cloud computing and evaluate its performance. Three-dimensional image processing using GIS data requires a large amount of computing resources because it uses complex and massive spatial information. Cloud computing solves this problem in an easy and transparent way. Hadoop, which consists of HDFS and MapReduce, was used for distributed parallel processing of massive 3D GIS data. Computing clusters can be used to reduce processing time.

Description

Apparatus and method of cloud computing for online visualization of noise map at high speed}

The present invention is part of the national research and development project, the project unique number: "10561", the research project name: "Seoul City industry-university cooperation project (2005 technology-based construction project)", the research title: "Intelligent for smart (ubiquitous) city City information convergence system development ".

The present invention relates to a cloud computing device and method for visualizing noise maps in ubiquitous-city, and more particularly, to a distributed parallel processing technology of cloud computing in a GIS application for ubiquitous-city (hereinafter referred to as "U-city"). Application relates to an apparatus and method for visualizing a noise map.

The Act on the Construction of Ubiquitous Cities, etc. [Law No. 9763, Enforcement 2010. 3.10] refers to ubiquitous cities based on ubiquitous urban technology to improve the competitiveness and quality of life of cities. "A city that provides ubiquitous urban services anywhere, anytime, through facilities," he says.

In other words, u-city is a city that is created, managed and utilized by the convergence of information communication, construction, and urban engineering. It provides services that are internally complex but well integrated, provided by converged technology and infrastructure. The term U-City itself has been led by Korea and is spreading around the world, and the U-City sector is led by Korea. Although some countries in the world are working on high-tech city projects that emphasize some areas, our U-City name and form are rarely carried out except in Korea. On the other hand, starting in Korea a few years ago and now 50 U-Cities are in progress, it can be said to lead the world. It is still in the early stages, and efforts are being made in Korea to define all aspects of current concepts, methodologies, necessary technologies, etc. To this end, they feel a great need for research and development. This research, which is the basis of this patent, has been conducted by the Intelligent City (U-City) research group since 2005, and is a large-scale national project selected and supported by Seoul City. In its form, the world's first systematic study of U-City middleware systems and related technologies has been conducted. As of 2008, the U-city project, which has been carried out by local governments, was inconsistent in each direction and criticized about its utility. It aims to make the U-City project a core new industry that will be a driving force for the development of the country while being consistent and effective. To this end, efforts are being made to give a desirable direction.

NIST recently defined the definition of cloud computing as "a cloud computing model that provides a fast, affordable, and minimally manageable pool of computing resources (network, server, storage, applications, and services) anywhere, anytime, into a convenient consumer-centric network. ". By using cloud computing, users can access services from the Internet without having to know or control the technical infrastructure they support. Initial expenditure is low because it is provided as a service that pays for software or other computer resources when needed. The availability of computers is high because virtualization technology and distributed computing technology provide server services to users who need them by tying or dividing server resources. You can implement a consistent user experience through abstracted services. You can keep your data safe by keeping it on a reliable server.

The recent development of Geographic Information System (GIS) technology has made it possible to produce noise maps. For example, the German state of Rhineland-Palatinate has released two-dimensional noise maps to the public over the Internet since 2008. The UK uses two-dimensional noise maps to focus on the assessment of noise hazards. Three-dimensional noise map is an area that is going to develop in the future of each country in the world. Since the EC adopted the standard for noise maps, noise research in many countries has become more active. Recently, each noise expert has combined the noise map and Visualization GIS technology to make the noise map more accurate and realistic.

However, noise maps using GIS technology use 2D maps and do not guarantee real time. The 2D noise map does not consider altitude. However, with the complexity of the city and the construction of tall buildings, 3D noise maps with altitude are needed. But urban 3D noise maps require a lot of computing power. In addition, the noise map for showing the real-time state of the noise to the user in real time also requires a lot of computing power. To date, there have been no studies of distributed parallel processing of noise maps using cloud computing. The patent also supports both two-dimensional and three-dimensional noise maps.

[1] DIRECTIVE 2002/49 / EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 25 June 2002 relating to the assessment and management of environmental noise, "Official Journal of the European Communities", 2002. [2] Kurakula, V., "A GIS-Based Approach for 3D Noise Modeling Using 3D City Models", MSc proposal, University of Southampton, UK, 2007. [3] J. Dean and S. Ghemaway, "Map Reduce: Simplified Data processing on Large Clusters", Communications of the ACM, vol. 51, January, 2008, pp. 107-113, doi: 10.1145 / 1327452.1327492. [4] Apache Hadoop Homepage [online], October 2010, Available from: http://hadoop.apache.org/common/. [5] The noise map in Europe [online], October 2010, Available from: http://www.xs4all.nl/~rigolett/ENGELS/maps/ [6] Cho, D. S., J. H. Kim, and D. Manvell. "Noise mapping using measured noise and GPS data". Applied acoustics, vol. 68 no. 9, pp. 1054-1061, 2007, doi: 10.1016 / j.apacoust. 2006.04.015. [7] Oh, S., I. LEE, S. Kim, and K. Choi. "Generation of a Spatial city model using a Digital Map and Draft Maps for a 3D Noise Map". Korean journal of remote sensing, vol. 24 no. 2, 2008, pp. 3-14. [8] N. Golpayegani and M. Halem, "Cloud Computing for Satellite Data Processing on High End Compute Clusters", Proceedings of the 2009 IEEE International Conference on Cloud Computing, 2009, pp. 88-92, doi: 10.1109 / CLOUD. 2009.71. [9] Bertram Ludaescher, Ilkay Altintas, Chad Berkley, Dan Higgins, Efrat Jaeger, Matthew Jones, Edward A. Lee, Jing Tao, and Yang Zhao, Scientific workflow management and the Kepler system: Research Articles. Concurr. Comput. Pract. Exper. 18 (10), pp. 1039-1065, 2006. [10] M. Weiser, "The Computer for the Twenty-First Century", Scientific American, 1991, pp. 94-101.

Accordingly, an object of the present invention is to provide a device and method that can significantly reduce the noise map production time by distributed parallel processing using cloud computing in the production of noise map.

One aspect of the present invention for achieving the above object relates to a cloud computing method for visualizing a noise map at high speed. The noise map visualization method includes a noise database generation step of collecting noise data to generate a noise database, a city model generation step of generating a visualized city model using a predetermined cloud computing technique, and a noise database using a cloud computing technique. And a noise map visualization step of generating noise maps by integrating the noise values read from the city model. The noise database generating step may include collecting the noise value received from the remote sensor through the ubiquitous sensor network USN, and generating the noise database using the collected noise value. The urban model generation step includes dividing the digital map data into grid cells, modeling the terrain using the MapReduce method using Hadoop, generating a digital elevation model, and building model using the MapReduce method. generating a building model, and integrating the divided grid cells to generate a city model. In addition, the numerical elevation model and the building model are generated as a three-dimensional urban model. Furthermore, the noise map visualization step uses the MapReduce method using Hadoop to convert each noise value to RGB values, partitions the ID of the city model output and groups it according to the ID, and maps the partitioned result to the MapReduce method. The method may include applying to a texture of the city model using.

을 이용하여 연산되며, 여기서 NC는 컬러 인덱스, Nmax는 소음레벨의 최대값, Nmin은 최소값, C는 전체 색의 수 및 N은 각 그리드의 소음 레벨을 나타내는 것을 특징으로 한다. Another aspect of the present invention for achieving the above objects relates to a cloud computing device for visualizing a noise map at high speed. The noise map visualization apparatus includes a noise database generator that collects noise data and generates a noise database, an urban model generator that generates a visualized city model using a predetermined cloud computing technique, and a noise database using a cloud computing technique. And a noise map visualization unit for generating noise maps by integrating the noise values read from the city model. The noise database generator is adapted to perform the operation of collecting the noise value received from the remote sensor through the ubiquitous sensor network USN, and generating the noise database using the collected noise value. The urban model generation unit divides the digital map data into grid cells, applies a MapReduce method using Hadoop to model a terrain and generates a digital elevation model, and a building model using the MapReduce method. (building model) is adapted to perform the operation of generating a city model by integrating the divided grid cells. In particular, the noise map visualization unit uses the MapReduce method using Hadoop to convert each noise value to an RGB value, to assign an ID to the output of the city model, and to group it according to the ID, and to use the MapReduce method. To adapt the texture to the urban model. Specifically, the RGB value is

Calculated by using N, where NC is the color index, Nmax is the maximum value of the noise level, Nmin is the minimum value, C is the total number of colors and N is the noise level of each grid.

The present invention has the following effects. First, in the existing method, the noise map was created by taking several hours to several days or more. In the present invention, in the case of distributed parallel processing of the noise mapping process using cloud computing, the noise map is made close to real time and visualized. can do. Second, it is possible to automate the process by eliminating the probability of failure of a job by using the MapReduce technique among the distributed parallel processing technologies of cloud computing.

1 is a diagram illustrating a digital elevation model.
2 is a diagram illustrating a building model.
3 is a diagram of a cloud computing process for generating a 3D noise map.
4 is a diagram illustrating an execution process of map reduce.
5 is a diagram schematically illustrating a processing process for a distributed parallel processing technique of a noise map using map reduction according to the present invention.
Figure 6 shows an example of the visualization of the three-dimensional noise map as a snapshot.

In order to fully understand the present invention, operational advantages of the present invention, and objects achieved by the practice of the present invention, reference should be made to the accompanying drawings and the accompanying drawings which illustrate preferred embodiments of the present invention.

Hereinafter, the present invention will be described in detail with reference to the preferred embodiments of the present invention with reference to the accompanying drawings. However, the present invention can be implemented in various different forms, and is not limited to the embodiments described. In order to clearly describe the present invention, parts that are not related to the description are omitted, and the same reference numerals in the drawings denote the same members.

Throughout the specification, when a part is said to "include" a certain component, it means that it may further include other components, without excluding the other components unless otherwise stated. In addition, the terms "... unit", "... unit", "module", "block", etc. described in the specification mean a unit for processing at least one function or operation, which means hardware, software, or hardware. And software.

Generally, noise maps are drawn using 2D maps and 3D GIS models. A 2D Noise Map, drawn using a 2D map, is visualized using a value at a certain height. However, as the city becomes more complex and more skyscrapers are built, noise at different altitudes is also becoming important. In order to display the noise level according to the altitude, a noise map using urban 3D modeling is required. To create a three-dimensional noise map, three steps are required:

1) Create a noise database

2) Create 3D city model

3) Apply noise value to 3D city model

First, in step 1), the noise data in the U-City is collected in a database through a remote USN (Ubiquitous Sensor Network). You can also use the interpolation method to create this database. In other words, it is possible to measure the noise value of an important point and predict the noise value of the unmeasured area by using a predictive equation using the data of the near measured area of the unmeasured area.

In step 2), urban 3D modeling includes generating a terrain model and generating a building model shown in FIG. 2.

1 is a diagram illustrating a digital elevation model, and FIG. 2 is a diagram illustrating a building model.

Urban 3D modeling consumes a lot of computing power to create a building model because of the complexity of the building model. To solve this problem, we used cloud computing.

In step 3), a noise value is applied to the 3D city model. Since the data of the 3D city model is very large, mapping each noise level to an RGB value and mapping the RGB value to the texture file of the 3D city model also consumes a lot of computing power. In this third phase of work, cloud computing was used. Therefore, processing time is reduced to a time close to online processing.

Hereinafter, cloud computing will be described with reference to FIGS. 3 and 4.

3 is a diagram illustrating a cloud computing process for generating a 3D noise map, and FIG. 4 is a diagram illustrating a process of executing MapReduce.

Cloud computing processes as shown in Figure 3 to create a three-dimensional noise map. To process the three-dimensional data, the data of the digital map was cut into grids and made into small sections. Because data from digital maps generates large files, Hadoop's MapReduce is one of the cloud computing technologies for massively distributed parallel processing. Hadoop consists of HDFS and MapReduce. HDFS uses replication to distribute and store copies on Hadoop clusters to ensure data safety. This can solve the data fault problem when the data is large.

Since MapReduce uses a single program multiple data (SPMD) model, we use a method of cutting and merging three-dimensional GIS image data into a grid (see [10]). As shown in Fig. 4, a 3D urban model is created using MapReduce, and this model is reused as an input of 3D noise mapping.

5 is a diagram schematically illustrating a processing process for a distributed parallel processing technique of a noise map using map reduction according to the present invention.

In order to generate a 3D noise map, MapReduce must be executed twice. The first map reduce is to produce 3D city model, and the second is to apply noise map value to 3D city model. In order to produce a 3D city model, a digital elevation model (DEM) with terrain data and a building model with building data are required. Digital map and draft map are used as input of map function to generate digital elevation model and building model. At this time, the electronic map enters the map function as a key-value pair of ((plane coordinate value), (height, and feature unique value)), and the degree of drawing is ((plane coordinate value), ( Height, building unique value)) into the input value of the map function. In order to make a digital elevation model, the height data of the terrain was extracted from the electronic map and the Dohwawon map, and a digital elevation model was produced using the extracted data.

The building model was constructed by extracting the boundary coordinates of the two-dimensional buildings of the electronic map and extracting the height values of the buildings that match each other using the electronic map and the drawing map. In the present invention, the target region is divided into several smaller regions (hereinafter referred to as subregions), and each has a unique ID value. Therefore, each local coordinate is given the ID value of a specific subregion when it is processed in the map function. Therefore, each subregion can be distinguished from each other. In other words, the key value is (plan coordinates, subregion specific values).

And before the result value of the map function is input to the reduce function, we use a partitioner to group and sort the coordinate values by the eigenvalues we have given among the keys of the map function result. By defining the number of reduce workers as many as the grouped subregions, one reduce function processes one subregion.

This grouping, which is the input value of the reduce function, and the coordinates of the numerical elevation model and the building model aligned with each other, make it possible to create a map of the terrain and the building through the coordinate values corresponding to each other in the reduce function. This result is then used as the map function input for the next step. The second map reduce also uses two kinds of map functions. The first map function is the input value of the 3D city model, which is the result of the previous map reduction process, and the second map function is a file that contains the coordinate values necessary for drawing the noise map and the noise level according to the coordinates. It is done.

That is, in the latter case, the key-value pair of ((plane coordinate value), (height, noise level)) is used as an input value of the map function. The former map function is responsible for processing the output of the previous MapReduce, and the latter map function is responsible for converting noise values into RGB values and dividing them into subregions. Noise level values should be converted to RGB values.

The resultant region was divided into several subregions, and the unique values of the subregions were matched with the subregions of the first MapReduce. Therefore, if the coordinate value of the target area in the first map reduce and the coordinate value of the target area in the second map reduce are equal, the same eigenvalue is given to the coordinate. The resulting value of the map function is in the form of a key-value pair of ((plane coordinate value, subregion unique value), (height, RGB value)). The result values obtained through the two kinds of map functions are grouped and sorted through the partitioner before the input values of the reduce function are matched with the unique values among the key values in the map function results.

These grouped and sorted key-value pairs are the inputs to the reduce function. Since the input values of the current reduce function are the RGB values of each point and the 3D city model, when the RGB values are simply added to the 3D city model, the map is generated with the dots on the 3D city model. The exterior walls of the building were created by interpolating between the points to look like walls. The result from reduce is the result file of the subregion. Therefore, combining the result files of these reduce functions becomes the final noise map.

This processing process will be described in detail as follows.

As mentioned above, MapReduce was used to create a three-dimensional urban model. We used two map functions: map_1 serves to create the Digital Elevation Model (DEM), and map_2 creates a building model with information about the topology and height of the object, as shown in FIG. Reduce is a fusion of DEM and building models. This process is called reduce_1, but it is obvious that this is for convenience of explanation only and does not limit the present invention. The result of Reduce_1's processing is a three-dimensional city model. Table 1 describes three functions: map_1, map_2, and reduce_1.

To create a building model, the getBld () method of map_2 extracts the coordinates of the boundary of a two-dimensional building from a digital map and extracts the height of the building from the draft map. The test area was divided into sub-areas and each was given an ID. The test area is given the ID of the detail area in the Map function. Therefore, the key value of all map functions is <sub-area-ID, x, y-coordinate>.

The results of Map_1 and Map2 are sorted and grouped by subdivision ID by the partitioner. The result of the partitioner is the input to reduce_1. This means that the results of the DEM and the building model are sorted and grouped by subregion IDs and become the input to reduce_1. Reduce_1 calls the generate3DCity () method to build a three-dimensional city model. Reduce is loaded to handle the reduce_1 function, which equals the number of subregions. The result of Reduce_1 is used as input for map_4 of the next map reduce.

Hereinafter, a cloud computing technique for generating a noise map will be described.

First, the noise level is combined with the three-dimensional city model to generate the three-dimensional noise map. We used two kinds of map functions. Map_3 is the input to reduce_2, which handles the noise level of the building. The result of Map_3 is expressed as << building ID, x, y coordinates> and <z coordinates, value of noise level >>. Map_4 transforms the results of reduce_1 for processing in reduce_2. Table 2 describes three functions: map_3, map_4 and reduce_2.

The inputs to Map_3 are the coordinates and building ID and noise levels needed to create a noise map. It can use the coordinates to map noise levels to a three-dimensional urban model. As the key of Map_3 we also used the building ID to identify each building.

The results of Map_3 and map_4 were sorted by the partitioner by subregion ID. Reduce_2 finally generates a three-dimensional noise map by combining the noise values of each region received from map_3 with the three-dimensional city model received from map_4. Here we use the getRGB () method to create an RGB color index. Equation 1 is used to create a color index using noise levels.

In Equation 1, nc is a color index, Nmax is a maximum value of noise level, and Nmin is a minimum value. C represents the number of colors and N represents the noise level of each coordinate.

After the noise level is converted into RGB values, a noise map is generated. First, determine the noise level on each wall according to points close to each point on the surface. Second, the point not found in the first step is to determine the RGB value using interpolation. After generating image values of each surface, they are combined into a single file to create a three-dimensional noise map. The resulting result is shown in FIG. 6.

Figure 6 shows an example of the visualization of the three-dimensional noise map as a snapshot.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is evident that many alternatives, modifications and variations will be apparent to those skilled in the art. The present invention can be implemented using the u-city middleware, but u-city middleware is not essential for implementing the present invention.

Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

The present invention can be applied to the technology of visualizing urban noise.

Claims (12)

In the cloud computing method for high-speed noise map visualization used in the noise map generation and visualization process by using a cloud computing technique to visualize the noise map online using computer graphics,
A noise data collection and storage step of collecting and storing noise data for online visualization using computer graphics;
A city model generation step of generating a visualized city model using a cloud computing technique that speeds up processing speed and making the generated city model available online to visualize it online using the computer graphic;
In order to visualize online using the computer graphics, the cloud computing technique is used to integrate the noise value read from the noise data into the city model to generate the noise map at high speed, and to generate the noise map online. Generating a noise map for use as a mobile device; And
Allowing the user to specify the generated noise map, and selectively visualizing the designated portion at high speed and online using the cloud computing technique,
The city model generation step using the cloud computing technique,
Dividing the digital map data into grid cells;
Generating a two-dimensional and three-dimensional digital elevation model for each of the divided grid cells by performing a terrain modeling by applying a divisional cloud computing technique using a constant cloud computing tool;
For each of the divided grid cells, generating a two-dimensional and three-dimensional structure model using the split cloud computing technique,
The noise map generation step using the cloud computing technique,
And applying the divisional cloud computing technique to generate the noise map by integrating the divided grid cells to combine the city model and the stored noise data. delete delete delete The method of claim 1, wherein the noise map generating step,
Converting the noise data into an RGB value by applying the divisional cloud computing technique; And
And using the split cloud computing technique in a partitioning process of assigning an ID to the output of the city model and grouping according to the ID. delete In the cloud computing device for the high-speed sound map visualization used in the noise map generation and visualization process by using a cloud computing technique to visualize the noise map online using computer graphics,
A noise database generator for collecting and storing noise data for online visualization using computer graphics;
A city model generator configured to generate a visualized city model at high speed using a cloud computing technique for speeding up the processing speed and to use the generated city model online to visualize it online using the computer graphic;
In order to visualize online using the computer graphics, the cloud computing technique is used to integrate the noise value read from the noise data into the city model to generate the noise map at high speed, and to generate the noise map online. And a noise map visualization unit for allowing the user to designate the generated noise map and selectively visualizing the designated portion at high speed and online using the cloud computing technique.
The city model generation unit using the cloud computing technique,
Two-dimensional and three-dimensional digital elevation models are generated by dividing the digital map data into grid cells and terrain modeling by applying a segmented cloud computing technique using a constant cloud computing tool to each of the divided grid cells. For each of the divided grid cells, generate a two-dimensional and three-dimensional structure model using the divided cloud computing technique,
The noise map visualization unit using the cloud computing technique,
And applying the divisional cloud computing technique to generate the noise map by integrating the divided grid cells to combine the city model and the stored noise data. delete delete delete The method of claim 7, wherein the noise map visualization unit,
High-speed noise, characterized in that the partitioning cloud computing method is applied to convert the noise data into RGB values, assign the ID to the output of the city model and grouping according to the ID in the partitioning process Cloud computing device for visualizing maps.
delete

Images