KR101293770B1 - apparatus for outputting data and method thereof - Google Patents

Abstract

The present invention relates to a data presentation system and a data presentation method. According to an embodiment of the present invention, a client requesting a visualization task for global environment data for a specific region, receiving the requested visualization task request, dividing the specific region into a plurality of visualization block region units, and performing the division A server controlling to simultaneously perform visualization operations on at least two block area units; And assigning at least one visualization task among the visualization tasks for the divided block region units from the server, and using the original data obtained by dividing and storing the global environment data for the specific region into a plurality of region units. It provides a data presentation system comprising a plurality of slave nodes for performing the visualized operation.

Description

Apparatus for outputting data and method

The present invention relates to an invention that can easily display a large amount of calculation input data or calculation result data, and relates to a data display system and method for easily processing a large amount of input data to display.

The data on the global environment, such as the atmosphere, the hydrosphere, the rock zone, the biosphere, and the ice zone, are enormous and require a lot of computational resources to systematically process and display the data. The development of computational resources is increasing the resolution of the observation or processing data of the earth environment data that the user requires, but when processing, processing, analyzing, and displaying the data according to such resolution, the user's desired result is limited due to the limitation of the data processing speed. It took a lot of time and effort for the user to get it, and the process for controlling and expressing each step was also inconvenient because the user had to manage, input, and control himself.

For example, the user must manually select the inputs according to the memory required by each system to obtain the desired model results, assign the inputs to each system, collect the results directly, and then use the data presentation tool. Visualization. However, as the large data is handled, the above operation takes a long time, resulting in inefficient and inconvenient problems. In the case of the earth environment data, it is four-dimensional data that changes in time and space, and it is not easy for the user to visualize it because it is a huge time variable data according to the time-space resolution.

In order to quickly and accurately analyze the observation data of global environmental data including meteorological data such as typhoon and flood or the model result data based on it, an effective method of processing a large amount of time variable data is needed.

 It is an object of the present invention to provide a data display system and method for extracting large amounts of data quickly by distributed processing.

 Another object of the present invention is to provide a data presentation system and method for efficiently processing, collecting and presenting extracted data.

 It is still another object of the present invention to provide a data presentation system and method for easily obtaining the result data for a user's desired data.

According to an embodiment of the present invention, a client requesting a visualization task for global environment data for a specific region, receiving the requested visualization task request, dividing the specific region into a plurality of visualization block region units, and dividing the specific region A server controlling to simultaneously perform visualization operations on at least two block area units; And assigning at least one visualization task among the visualization tasks for the divided block region units from the server, and using the original data obtained by dividing and storing the global environment data for the specific region into a plurality of region units. It provides a data presentation system comprising a plurality of slave nodes for performing the visualized operation.

The server may include: a master node allocating the visualization task to the plurality of slave nodes at the same time, and collecting result data of the visualization task performed by the slave nodes; And receiving the result data of the collected visualization task from a master node, and providing the received visualization task result data to the client as web-based data.

The data presentation server may receive from the client at least one of an identifier of the specific region, a minimum coordinate of the specific region, and a maximum coordinate of the specific region.

The data presentation server receives from the client a sampling interval of the numerical elevation of the specific area or an ortho-image ratio of the specific area from the client, and requests the visualization operation to the master node according to the received sampling interval or the ortho-image rate. Can be.

The data presentation server may adjust the resolution of the visualization operation result data using the presentation resolution of the client received from the client, and provide the visualization operation result data with the resolution adjusted to the client.

The server may divide the visualization operation according to the degree of redundancy between the divided at least two block area units and the plurality of area units of the original data.

The plurality of slave nodes may perform the visualization operation using polygon data and texture of the specific region.

The slave nodes may read the original data in the block region unit and cache data corresponding to the block region unit.

According to another embodiment of the present invention, the method may further include: requesting a visualization operation for global environment data for a specific region, receiving the requested visualization operation request, and dividing the specific region into a plurality of visualization block region units. At least two of the visualization operations assigned to at least one of the visualization operations for the divided block region units and using the original data storing the specific region in a plurality of region units. It provides a data presentation method comprising the step of performing a visualization operation for the unit at the same time, and collecting the result data of the visualization operation performed at the same time, and providing the collected result data.

 According to the present embodiment, a large amount of data can be quickly extracted through distributed processing, the extracted data can be efficiently processed and displayed, and the result data for the data desired by the user can be easily obtained.

1 illustrates an embodiment of a data presentation system according to the present invention.
2 is a diagram illustrating information each master node or data presentation server receives from a data presentation server or a slave node, respectively.
3 is a diagram illustrating an example in which a master node divides a visualization task into a slave node.
4 is a view showing a visualization working area in the area of FIG.
5 illustrates an example in which a master node divides a visualization task.
6 is a diagram illustrating a data acquisition method for visualizing at a master or slave node.
FIG. 7 illustrates that a master node creates tasks into a work queue and distributes the visualization tasks to slave nodes.
8 is a diagram illustrating an example of rendering a visualization operation result according to distributed parallel processing;
9 illustrates pseudo code for performing parallel processing performed by a master node.
10 shows an example of pseudo code for causing a slave node assigned to a task by the master node to process the task.
FIG. 11 is a diagram exemplarily illustrating a relationship between cache partitioning, a visualization area request, and a result of visualization
12 is a diagram illustrating parallel processing time of a visualization operation according to the size of a detail block to be partitioned for visualization.
FIG. 13 is a diagram illustrating parallel processing time performed for a visualization operation according to a overlapping ratio of a detailed block region and a visualization region requested by a user.
14 is a view showing an orthoimage in which nine patches corresponding to regions used for actual visualization work are connected;
15 and 16 are diagrams showing examples of providing visualization result data according to an interactive visualization method in the data presentation system described above.
FIG. 17 illustrates another example of interactively visualizing geographic data consisting of a plurality of patches using a distributed parallel visualization system in a data presentation system including a client-server. FIG.
18 to 21 illustrate the results of interactively implementing geographic data and precipitation data composed of a plurality of patches for a region displayed by the data display system in FIG. 17.
22 is a view showing the results of visualizing the flooded area according to sea level change

Hereinafter, an embodiment of a data presentation system and method thereof according to the present invention will be described with reference to the drawings. According to an exemplary embodiment of the present invention, only the data desired by a user is extracted quickly through distributed processing, and a large amount of data is hierarchically accessed to minimize data access time, and a large amount of data is easily accessed through a web-based interface. Can be expressed.

To this end, it provides a high performance distributed processing algorithm of large data, and provides a user interface that allows the user to receive the result in an efficient manner by using a client-server based parallel rendering system. As an example of distributing large amounts of data, a user may make a request to the client to specify the desired global environmental data. In this regard, the data presentation server is connected to the distributed processing nodes, distributes the large data based on the similarity between the data temporarily stored by the distributed processing nodes and the large data stored in the data server, and collects the results. Can be.

1 discloses an embodiment of a data presentation system in accordance with the present invention. One embodiment of the data presentation system may include a client 100, a data presentation server 200, at least one cluster system. Each cluster system may include a plurality of nodes, at least one of the plurality of nodes serving as the master nodes 310 and 410, and the remaining nodes 350 and 450 serving as slave nodes. . The cluster system may include file servers 370 and 470, respectively, which may store input data or output data of the data presentation server. The data presentation server 200 and the master nodes 310 and 410 may be one server.

The user uses the client 100 to display the data to be visualized, the range of visualization, the resolution of the visualization data, the coordinate values for the region to be visualized for the visualization of the global environment data, the sampling interval for the region, and the magnification of the displayed image. You can choose. The client 100 may transmit the memory size of the client or resolution information that can be expressed by the client to the data presentation server 200 in order to express a visualization result suitable for the user.

The client 100 requests the visualization data requested by the user to the data presentation server 200, receives a result according to the visualization request from the data presentation server 200, and provides the result to the client 100. The client 100 may display the data display server 200 as a web server and display the result in the form of a web browser, and the user may visualize the data interactively with the data display server 200 through the client 100. Can be.

The data presentation server 200 is a server of a web-based data presentation system, which divides the visualization task requested by the client 100 and delivers distributed tasks to each cluster system.

The master nodes 310 and 410 receive the memory size, displayable resolution information, etc. of the client 100 from the data display server 200, and determine values such as resolution of the visualization data and level of details (LOD) according to the visualization operation. Change and transmit the optimized results to the data display server 200.

The master nodes 310 and 410 divide the visualization task requested by the user to the slave nodes 350 and 450 of the cluster system by using a parallel processing interface, and transfer the divided tasks to the slave nodes 350 and 450. do. The master nodes 310 and 410 may divide the entire work W into work Wis which will be performed by the nodes including the M slave nodes 350 and 450 or the nodes including the master nodes 310 and 410 and the slave nodes 350 and 450. .

For example, if visualization of the global environment data is performed, the visualization work Wi will be performed by each node may be a task proportional to the size or resolution of each region according to the resolution.

A large amount of data may be stored as original data (patch data) in units of divided regions, in which case the master nodes 310 and 410 divide the visualization request region into blocks in order to decompose the requested visualization task into concurrent tasks. Visualization can be partitioned in parallel by dividing and distributing data for each divided block area to the slave nodes 350 and 450.

The divided block area data may be included in the original data which is one or more storage area units. The master nodes 310 and 410 may include the block area and the area unit of the original data that the slave nodes 350 and 450 need to calculate. In consideration of redundancy, the visualization work for each block area may be distributed to the slave nodes 350 and 450. If the visualization is performed in parallel using the data of the divided region unit, the visualization may be efficiently processed.

In this case, when the master nodes 310 and 410 read the original data into the cache of each node to perform the visualization operation, the master nodes 310 and 410 may distribute the work such that the data of the adjacent block area is included in the same original data. have.

Alternatively, the master nodes 310 and 410 may distribute the visualization work to the slave nodes 350 and 450 by checking which original data includes the data of the block area that each slave node 350 and 450 finally loads in the cache. have.

Whether the data in the block area unit is included in the original data in the same or adjacent area unit may be determined according to an algorithm set in the master nodes 310 and 410. Slave nodes 350 and 450 can read the data for the distributed block area at once with a cache structure, and the data contained in the area to be visualized but included in the other source data is obtained from the original data of the file server 370 and 470. I can read it. For example, the slave nodes 350 and 450 may refer to a unit that reads data on the distributed block area as a cache block.

The master nodes 310 and 410 collect the work result data performed by the slave nodes 350 and 450 for visualization and transmit the collected data to the data display server 200. The master nodes 310 and 410 collect the rendering results of the global environment data of the slave nodes 350 and 450 and transmit the collected results to the data display server 200.

In this case, the master nodes 310 and 410 may divide the entire visualization area in which the visualization is performed in the form of a block by using the interval of the numerical elevation of the area to be visualized and the magnification of the orthoimage. The magnification of the orthoimage may be determined according to the available memory size of the client 100, and filtering and scaling of data processed by the slave nodes 350 and 450 may be performed by using the same.

The master nodes 310 and 410 may query the slave nodes 350 and 450 for processing progress of the distributed work and determine whether there is a slave node in a resting state. The master nodes 310 and 410 may have information about the similarity of slave tasks, the similarity of block region data to be processed, or the proximity of block regions so that each slave node 350 and 450 can process tasks for similar block regions. Can be. The master nodes 310 and 410 create and execute a task queue for the slave nodes 350 and 450 to visualize using this similarity information.

Slave nodes 350 and 450 cache specific unit data (cache blocks) to be used for visualization in order to minimize the delay time incurred when accessing the target data of the visualization operation, and the data stored in the cache block. Can be used for visualization tasks. The slave nodes 350 and 450 may visualize the data in the distributed block area in parallel operations according to the commands of the master nodes 310 and 410. When the slave nodes 350 and 450 receive a job from the work queue, the slave nodes 350 and 450 receive the request area, the interval between the numerical elevation, and the magnification of the orthoimage, and the master nodes 310 and 410 use the information to determine the slave nodes 350 and 450. You can manage the working status of). If each of the slave nodes 350 and 450 performs the visualization of the block area, the data for the block area is extracted from the original data in the area unit, and the data is filtered according to the numerical elevation interval and the magnification of the orthogonal projection. To do the visualization.

2 exemplifies information that a master node or a data presentation server receives from a data presentation server or a slave node, respectively. Master nodes assign visualization tasks to slave nodes based on the data illustrated in FIG. 2. In this case, the similarity of the allocated task may be determined based on the data illustrated in FIG. 2, and the visualization task may be distributed by the slave node. When the client requests a visualization task, the client may transmit the information illustrated in FIG. 2 to the master node through the data presentation server 200.

For example, if global environmental data for a particular area is visualized, identification of the requesting area may be used for distribution of visualization work assignments. The identification information of the request region may be an encoded region index of the visualization request region among the entire regions. In addition, the minimum and maximum coordinates of the visualization target area of the visualization request area may be transmitted to the master node. Alternatively, the sampling interval or the magnification of the orthogonal image of the visualization target region in the request region may be transmitted to the master node.

When the visualization operation is requested, the master node may compare the data of FIG. 2 inputted for parallel visualization operation with the original data in area units, and request the slave node to visualize the corresponding block area. The master node divides into block-level areas that should be used to visualize the visualization areas requested by the client. Then, the source data in the area unit including the block unit areas are identified, and the block unit data is read and read in the cache block unit from the source data.

In this case, the block unit area may be divided in consideration of the numerical elevation interval input for the visualization operation and the magnification of the orthoimage, and the magnification of the orthoimage may be determined according to the available memory size of the client system used. The master node performs data filtering and scaling according to the magnification of the orthoimage.

3 is a diagram illustrating an example in which a master node divides a visualization task into a slave node. In FIG. 3, regions indicated by red, green, blue, and purple denote regions in which patch data is divided and stored. That is, data performed for the visualization operation may be divided and stored according to regions. The master node divides the target area on which the housekeeping operation is to be performed into a detailed block area (black border block) in FIG. 3 using the numerical elevation interval and the magnification of the orthogonal image. The color corresponding to the area data covered by each patch data and the number of encoded indexes constituting each patch data are illustrated in the right table. For example, the patch data in the red region includes 3361005 encoded indexes.

4 is a view showing a visualization working area in the area of FIG. 3. When the user selects the area requesting the visualization operation as the minimum coordinate (blue point) and the maximum coordinate (red point), the master node may proceed with the slave node to perform the visualization operation on the corresponding area (shaded). In this case, data for visualization may be loaded in each node in detail block units. The visualization task is divided into simultaneous tasks based on the detailed block of FIG. 4, and each disassembled task is assigned to a master node or a slave node. In this case, an area included in the same patch data may be performed by the same node or may be divided by several nodes. When the master node divides the visualization area (shading) into several block areas, it allocates each task to slave nodes that can optimally access the original data, but visualizes them so that multiple block areas can be visualized in parallel at the same time. Allocates to slave nodes block by block.

5 illustrates an example in which a master node divides a visualization task. When there is a visualization work request from the client, the master node divides the visualization work request area into detailed block units. If the master node splits the visualization task, the master node first retrieves its master hash. If patch data for the specific block area exists in the master hash of the master node (cache hit), the master node reads the original data for the detailed block area from its memory and performs visualization of the block area. . If there is no data for the detailed block area divided in the master hash, the hash of the slave node is searched. If patch data of the corresponding sub block exists in the hash of the slave node (cache hit), the visualization task of the area corresponding to the sub block is allocated to the slave node having the data. In this case, the slave node reads the corresponding data from its slave memory area and performs visualization. The degree of similarity as to whether the patch data of the detailed block area exists in each node may be determined by the master node. If the data corresponding to the detail block is not retrieved from other slave hashes, the master node allocates a visualization task for the detail block region to any slave node. Then, the slave node assigned the task reads the data included in the corresponding detail block from the file server and performs the visualization task.

6 illustrates a data acquisition method for a master node to visualize at a master or slave node using a hierarchical cache structure. The master node searches the hash of the master memory using the file name containing the patch data to obtain patch data of the corresponding detail block. If data for a desired file name exists in the master memory, the master node can obtain data for the corresponding detailed block area by using a memory address in which the data for the detailed block area is stored. On the other hand, when the master node finds the data of the block area from the slave hash from the coded file name, the visualization operation may be performed by generating an identifier of the slave node that can directly use the slave hash.

FIG. 7 is a diagram illustrating that a master node creates a work queue for a divided visualization task to perform simultaneous visualization tasks, and distributes tasks to slave nodes in order of work in the task queue. The number stored in the work queue may be a work number, which may be the same as or correspond to the number of the block area that subdivided the visible area. That is, a series of numbers in this figure represent the work order of the generated work queue, and the work number in the work queue may correspond to the block number of the visualization request area so that the slave node can perform a similar work. In addition to the work received from the work queue, the slave nodes receive request region information, the height of the numerical elevation, and the magnification of the orthoimage. This information can be used by the master node to manage the work status of the slave node. The master node may query the slave node about the degree of processing of the distributed job to determine whether there is a slave node in a resting state. If the visualization task requested from the client remains in the task queue, the master node reassigns the task to the slave node in the idle state. This job is repeated until all jobs are removed from the job queue. When all jobs are processed in the job queue, the master node collects only the data corresponding to the visualization request area and provides it to the client through the data presentation server. The master node reduces the time for the slave node to access the file server based on similar data, and judges the similarity between the data stored in the cache and the data in the subdivided block area so that each node can efficiently perform the work. Create a work queue based on that similarity.

8 shows an example of rendering a visualization operation result according to distributed parallel processing. Assume that the master node uses 4 slave nodes to perform visualization tasks for the data requested by the client. If patch data for a detailed block area exists in a master node or a slave node, the corresponding node performs the corresponding visualization. Visualization of each node may be performed using parallel processing means such as a message passing interface (MPI). Patch data for each region is stored in a storage medium, and this figure shows an example of reading patch data stored in an external shared memory. In this case, each node may share a memory device having a shared memory structure, and the shared memory may be connected to each slave node and a file server through a network such as Ethernet. When the slave node performs a visualization operation on a block area in which an area where a visualization operation is performed is divided, the visualization operation is performed by caching patch data included in the corresponding data. The slave node delivers the visualized result to the master node, and the master node renders the slave node and the visualization work performed by the node to the data presentation server or rendered to the client when the master node acts as the data presentation server. Visualization results can be provided to web-based applications such as Java. As described above, the master node data display server may be configured as a server, and the master node may serve as a data display server.

9 illustrates pseudo code for performing parallel processing performed by a master node. In order for the slave nodes to perform visualization tasks in parallel, the master node receives the index of the work area, the minimum coordinate, the maximum coordinate, the numerical elevation interval, and the magnification of the orthoimage (line 1). The area is divided into block-based areas of work (line 2), and a work queue is created in the manner described above (line 3). According to the number of processors n (line 4) and the number of jobs m (line 5) of the slave node, divided visualization tasks are allocated to the slave nodes (line 6 to line 16). In this case, based on the similarity between the original data and the data of the divided detailed block region, the visualization of the block region is allocated to the slave nodes.

First, the processors of each slave node are allocated in the order of work queues (line 7 to line 8). Receive the work result from the slave node that has been completed (line 11), and if the work queue is not empty, the work queues remaining in the slave node that have been completed are sequentially allocated (line 12 to line 14). When the slave nodes have completed all the work included in the work queue, the master node merges the visualization result data for the block area corresponding to the area where the visualization work is requested (line 18).

10 shows an example of pseudo code for causing a slave node assigned to a task by the master node to process the task. If there is no work request, the slave node is switched to the standby state, and when the work assignment is requested, the work illustrated in FIG. 8 may be sequentially performed. The slave node may receive a job index, a minimum coordinate and a maximum coordinate of the visualization request area according to the order of the job from the master node (line 1). The visualization work request area is divided into N block areas including the patch data (line 2). The visualization work area is prepared for the visualization request area having the original data among the divided block areas. The visualization target region may be divided into N sub-blocks, and the slave node extracts data necessary for visualization from the original data for the visualization operation (line 3) for each divided region (line 3) (line 4 to line 6). The slave node scales the visualization result data according to the set numerical elevation interval and the magnification of the orthoimage (line 8), and then transfers the scaled data to the master node (line 7).

11 exemplarily illustrates a relationship between cache partitioning, a visualization area request, and a result of visualization. The large-capacity data may be divided and stored in an area of a cacheable unit, and loaded into the memory of each node by a cache operation of each node. In this figure, the patch block, which is a unit in which the original data is stored, is green, and the cache block, which is a unit that each node caches, is displayed in red. That is, the patch block represents a unit in which the original data is stored, and the cache block may be a block area divided into detailed blocks for visualization. Therefore, when a user requests visualization of a specific area, the master node searches for a node that caches data for the block area in the cache block unit according to the size of the cache block and allocates a visualization task. The slave node to which work is assigned performs visualization for the block area. The visualization result can be collected by the master node or the data presentation server and the request result can be provided to the user. In this figure, the visualization area request is overlapped with a plurality of patch blocks and a plurality of cache blocks, and the slave node may generate a visualization result (request result) for the request area and provide the same to the client.

Figure 12 illustrates the parallel processing time of the visualization operation according to the size of the detail block to divide for visualization. The smaller the detailed block area that can be a unit of data cache, the more likely it is to be included in one original data unit. Therefore, the smaller the size of the detail block divided to the area that the user wants to visualize, the less time the visualization parallel processing task takes. Therefore, when the visualization operation is performed at one node for each divided block area, the visualization operation may be efficiently performed. The master node may transmit a visualization operation to each slave node in consideration of the size of the block area.

FIG. 13 is a diagram illustrating a parallel processing time performed for a visualization operation according to the overlapping ratio of the detailed block region for the visualization operation to the region requested by the user and the visualization region requested by the user. The higher the rate at which the user-requested area and the detailed block area performed by one node overlap each other, the less time the visualization parallel processing takes. Accordingly, the master node can quickly obtain the visualization result by dividing the visualization task and assigning the visualization task to the slave node so that the overlapping ratio of the detailed block region for the visualization task and the user's visualization request region is increased.

This client-server-based data presentation system enables distributed parallel visualization to interactively visualize large-scale data on low-performance platforms such as simple mobile devices, terminals, or netbooks. That is, even if the client is a terminal that does not have high performance such as a mobile device, a large amount of data visualization can be requested to the data presentation server through a web-based interactive system. The data presentation server can visualize the requested visualization data in parallel using a distributed parallel visualization system including master-slave nodes, and then process and transmit the result to the client. The data presentation server may provide visualization result data performed by the master node to the client by supporting an adaptive LOD and a cache memory structure to enable interactive visualization.

An example of providing visualization result data according to an interactive visualization method is described. In the example disclosed, the data used for visualization is a terrain model of approximately 1 million polygons and orthoimages with 15,604x18,564 resolution, and interactive visualization is possible using multiple LODs and cache memory. Large-scale geographic data and submerged simulation data were selected as global environmental data, and each data is shown in Table 1.

The test data used in this embodiment is divided into a total of nine patch data, and each patch data includes numerical elevation data and ortho images corresponding to the patch.

FIG. 14 shows an orthogonal image in which 9 patches corresponding to the entire area are connected to an area corresponding to 4,611m × 13,875m used in actual visualization work.

15 and 16 illustrate an example of providing visualization result data according to an interactive visualization method in the described data presentation system. Referring to each visualization result in FIGS. 15 and 16, the user may select a region to be visualized from the web menu. You can choose the name of the area you want to visualize, or you can choose the width or height of the area you want. In addition, the visualized area may be specified by selecting minimum and maximum coordinates for the visualized area. You can also choose the detail level of the texture for that visualization and the detail level of the polycone depending on the number of polycones.

When the user selects a desired region, the desired region may be displayed on a map on a portion of the screen. In FIG. 15 and FIG. 16, the red box of the lower right is shown. The user can check a part of the area to be visualized from the client on the map, and request the visualization operation to the data display server by displaying the area to select the visualizing operation on the map with a selection means such as a mouse. When the visualization area is selected, the visualization area may be displayed according to the elevation of the visualized area, the texture-mapped screen, or the selection option to display in the polycon mode. In addition, an interface for allowing a user to view a desired result in detail by expanding, rotating, or reducing the visualization result may be provided. 15 and 16 show visualization results of the Hallasan portion of Jeju Island, and FIG. 16 provides an enlarged visualization result of a partial region of FIG. 15.

17 shows another example of a result of interactively visualizing geographic data composed of a plurality of patches using a distributed parallel visualization system in a data presentation system including a client-server. If the client has a low-performance platform, it is possible to enable interactive visualization, which shows the result of visualizing the variable time data using about 600,000 polygons in the figure.

The data presentation system may display or analyze results of global environmental data. Observation data or modeling results of the regional environmental data of the region can be visualized together so that the prediction of the global environmental data can be efficiently displayed.

18 to 21 illustrate a result of interactively implementing geographic data and precipitation data composed of a plurality of patches for the region visualized by the data display system in FIG. 17. Here, the time variable data represents the flooding phenomenon caused by the flooding of the river due to the typhoon. If data analysis systems are used to visualize or analyze inundation data, users can perform more accurate data analysis or forecasts. In FIGS. 18 to 21, the time progresses from left to right and from top to bottom and may realistically represent the flooded area or part of the corresponding area.

22 shows the results of visualizing the flooded area according to sea level change. Similarly, time goes from left to right and from top to bottom. Therefore, incorporating a model result representing sea level change into time-volume variable data can realistically reproduce the flooding phenomenon in the region, which can greatly help in predicting or analyzing the phenomenon.

100: client
200: data presentation server
310, 410: master node
350, 450: slave node
370, 470: data server

Claims (10)

A data presentation server including a master node receiving a request for visualization of global environment data and at least one slave node performing a visualization operation assigned by the master node; And
And a storage medium storing original data of the visualization area according to the visualization task request.
Here, the master node divides the visualization area in which the visualization task request is performed into cache blocks, and the master node and the at least one slave node divide original data of a corresponding cache block among the divided cache blocks. Temporarily read each of the storage media,
The master node divides a target area of a visualization operation desired by the user among detailed visualization areas into detailed blocks, and temporarily stores data for a cache block having the highest rate of overlapping with the divided detailed blocks among the cache blocks. And assigning each visualization operation for the detail block to a node so as to parallelize the visualization operation desired by the user. The method of claim 1,
The data presentation server collects the result data of the visualization operation performed by the master node and the at least one slave node and provides the collected visualization result data as a web-based data. The method of claim 1,
And a target area of the visualization operation desired by the user is determined by any one of a minimum coordinate and a maximum coordinate of the target region, a visualization region name, and a region arbitrarily selected by a user through selection means. The method of claim 1,
The data display server receives from the user a sampling interval of the numerical elevation of the target area of the visualization operation or an orthoimage ratio of the target region of the visualization operation from the user, and performs the visualization operation according to the received sampling interval or the orthoimage ratio. Data presentation system performed. The method of claim 1,
And the data presentation server adjusts the resolution of the visualization operation result data using the presentation resolution received from the user, and provides the visualization operation result data with the resolution adjusted. The method of claim 1,
And the at least one slave node performs the visualization by using polygon data and texture of a specific region. Receiving a request for visualization of global environment data for a specific region;
Dividing a visualization area in which the visualization task request is performed into cache blocks, and storing master data and at least one slave node by reading original data of a corresponding cache block among the divided cache blocks from a storage medium; ;
The master node divides a target area of a visualization operation desired by the user among detailed visualization areas into detailed blocks, and temporarily stores data for a cache block having the highest overlapping ratio with the divided detailed blocks among the cache blocks. Allocating a visualization task for the detail block to each node;
Parallel processing of a visualization operation desired by the master node and the slave node; And
Collecting data according to the parallel processing to visualize the global environment data according to the visualization operation; a data presentation method comprising a. delete delete delete

Images