CN103268221A - Method and device for three-dimensional display of meteorological data volume based on WEB technology - Google Patents

Abstract

The invention provides a method and a device for performing high-fidelity volume partitioning three-dimensional demonstration on mass weather data on a WEB client through an Internet technique. In the method and the device, the conventional way of only processing a small quantity of data such as observation data of stations and only performing planar demonstration or fake three-dimensional demonstration on a single computer is changed. The method comprises the following steps: S1, preprocessing weather data to be demonstrated; S2, optimizing the preprocessed weather data, and dynamically pushing the optimized data onto a WEB server side; and S3, outputting the three-dimensional weather data on the WEB server side to the WEB client for performing volume three-dimensional demonstration. Due to the adoption of the method and the device, mass heterogenous isomerous weather data, including station observation data, satellite inversion data, numerical forecasting data and the like can be processed rapidly, and optimized weather data are subjected to three-dimensional processing and demonstration on the WEB client by using relevant techniques such as Internet transmission.

Description

Method and device for three-dimensional display of meteorological data volume based on WEB technology

technical field

The invention relates to the field of meteorological data processing, in particular to a method and device for three-dimensional display of meteorological data volume based on WEB technology.

Background technique

Meteorological data includes station observation data, satellite retrieval data, numerical forecast data, etc., and its content includes atmospheric humidity, air pressure, air temperature, atmospheric water temperature profile, cloud cover, precipitation, solar radiation and other climatic elements. The expression methods, organization forms, and temporal and spatial resolutions of data are different, which brings difficulties to the unified display of meteorological data. At the same time, with the development of earth observation technology, more and more meteorological observation data are obtained through satellites, and the data obtained every day reaches TB level. People urgently need to browse and access such a large amount of meteorological data through increasingly developed Internet technology , in order to serve your daily life and travel. In addition, in all aspects of meteorological data application, such as scientific research, government emergency relief decision-making, meteorological knowledge popularization, public user experience, etc., the past situation of single color tone, single data type, and flat display has become increasingly difficult to meet the needs. There is an increasing need for high-fidelity, body subdivision, and realistic 3D display effects.

Contents of the invention

The present invention provides a method and a device for performing high-fidelity three-dimensional display of massive meteorological data on a WEB client through Internet technology, which changes the process of only processing a small amount of data such as station observation data in the past, and only performing on a single computer. The condition of flat presentation or false three-dimensional presentation. This patent is of great significance for the industry application of the meteorological department and the expansion of the service field of meteorological data.

In order to achieve the above object, the technical scheme adopted in the present invention is as follows:

A method for three-dimensional display of meteorological data volume based on WEB technology, comprising the following steps:

S1, preprocessing the meteorological data to be displayed;

S2, optimizing the preprocessed meteorological data, and dynamically pushing the optimized data to the WEB server;

S3. Outputting the 3D meteorological data on the WEB server to the WEB client for volumetric 3D display.

Preferably, the preprocessing in S1 includes: performing format conversion processing, projection transformation processing, and data consistency check on the original meteorological data; the optimization in S2 includes: performing resampling, layering, Blocking; the output described in S3 includes: using progressive transmission method, data organization and compression method, download balance method, multi-level caching method, progressive rendering method of LOD multi-level of detail model and volume three-dimensional expression method of Hedgehog method, will optimize The final weather data is displayed on the WEB client.

Preferably, S2 is specifically:

S21, resampling the result data multiple times to form a pyramid structure with multiple layers and different resolutions for images in the same range;

S22, adopting a dynamic self-adaptive block strategy, performing block processing on the image in each layer; obtaining climate element data blocks;

S23, pushing and storing the resampled, stratified and block processed data on the WEB server.

Preferably, the dynamic adaptive block strategy is specifically:

Let V={v ₁ , v ₂ ,...v _n } be a group of finite point sets in a plane area D in the plane R ² , and after dividing the plane domain D according to the preset point threshold, a The set R of plane sub-fields={R ₁ , R ₂ ,...,R _n }, for each plane sub-field R _i , record the discrete point data and the total number of points in this sub-field, and record the data of each sub-field at the same time The coordinates of the four corner points are to obtain the data value of each sub-block.

Preferably, the preset point threshold is the minimum number of data points allowed for each sub-block.

Preferably, the coordinates of the four corners are respectively:

LeftUpX, LeftUpY, RightBottomX, RightBottomY.

Preferably, S3 is specifically:

S31, using progressive transmission method, data organization and compression method, download balance method, multi-level caching method to cache the meteorological data to be displayed to the WEB client and display it by the following method;

S32, using a field data organization model based on nodes and layers to simplify scene management;

S33, comprehensively use the pyramid data compression storage strategy based on linear quadtree, multi-resolution model, scene simplification method based on LOD and matching method based on "skeleton + skin" to simplify and speed up the rendering of 3D scenes;

S34, adopt the method of "scene loading management based on visibility buffer" to draw the scene, and apply the dynamic loading method to optimize the loading speed at the same time, so as to realize parallel loading;

S35, using the quadtree data structure to represent the climatic elements in the scene, each node in the tree covers the corresponding area in the climatic elements, dynamically select the climatic element nodes on the basis of meeting the preset error threshold to realize the Continuous multiresolution representation of climate element models;

S36, using a volumetric three-dimensional expression method of the Hedgehog method to display the meteorological data;

Wherein, the execution order of S32-S36 is in no particular order.

Preferably, S35 is specifically:

Divide the entire scene into blocks, calculate the location of the climate elements according to the location of the climate elements in the scene, and establish a linked list index to form a location query table;

If the number of climate elements on each scene exceeds a preset threshold, the climate elements are subdivided until the number is less than the preset threshold or the subdivision level is greater than the preset level threshold; so that the entire scene is organized into an n-ary tree, The whole scene is the root node, and the climate element is the intermediate node;

When dynamically roaming the scene, directly locate a certain scene according to the location lookup table, and then link to a new block node according to the block to which the target point belongs.

Preferably, S3 also includes the following steps:

For each quadtree node that meets the error threshold, the diagonal division method is used as the center pointing from the four corner points to the parent node; the quadtree network construction is attributed to the direct diagonal division form and the midpoint form; And the way of retaining the midpoint is divided into four basic representations.

A three-dimensional display device for meteorological data volume based on WEB technology, comprising:

The meteorological data preprocessing device is used to preprocess the meteorological data to be displayed; the preprocessing includes: performing format conversion processing, projection transformation processing, and data consistency inspection on the original meteorological data;

The meteorological data optimization device is used to optimize the preprocessed meteorological data and dynamically push the optimized data to the WEB server; the optimization includes: resampling and layering the preprocessed meteorological data ,Block;

The meteorological data output display device is used to output the three-dimensional meteorological data on the WEB server to the WEB client for display, and the output includes: using progressive transmission methods, data organization and compression methods, download balance methods, and multi-level caching methods , LOD multi-level of detail model progressive rendering method and Hedgehog method of volumetric three-dimensional expression method to display the optimized meteorological data on the WEB client.

The beneficial effects of the present invention are:

Through the technical solution of the present invention, it is possible to quickly process massive amounts of heterogeneous meteorological data, including station observation data, satellite inversion data, numerical forecast data, etc., and use Internet transmission and other related technologies to transfer the optimized meteorological data to the WEB client The three-dimensional processing and display on the terminal can make it very convenient for different departments and the public to access and browse meteorological data through the Internet. The high-fidelity three-dimensional display effect can give users a good visual experience and greatly expand the application field of meteorological data. At the same time, the technical solution of the present invention greatly shortens the processing time and improves the processing efficiency in the processing process.

Description of drawings

Fig. 1 is a schematic diagram of the image pyramid model after segmentation;

Fig. 2 is a schematic diagram of a quadtree data structure;

Figure 3 is a schematic diagram of the "crack" phenomenon;

Fig. 4 is a schematic diagram of a diagonal section;

Figure 5 is a form of diagonal division: direct diagonal division;

Figure 6 is another form of diagonal subdivision: keep the midpoint subdivision;

Fig. 7 is a schematic structural diagram of a three-dimensional display device for meteorological data volume based on WEB technology of the present invention.

Detailed ways

In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings. It should be understood that the specific embodiments described here are only used to explain the present invention, and are not intended to limit the present invention.

As shown in Figures 1-6, the present invention discloses a method for three-dimensional display of meteorological data volume based on WEB technology, which is characterized in that it comprises the following steps:

S1, preprocessing the meteorological data to be displayed; the preprocessing includes: performing format conversion processing, projection transformation processing, and data consistency inspection on the original meteorological data;

S2, optimizing the preprocessed meteorological data, and dynamically pushing the optimized data to the WEB server; the optimization includes: resampling, layering, and blocking the preprocessed meteorological data;

S2 is specifically:

S21, resampling the result data multiple times to form a pyramid structure with multiple layers and different resolutions for images in the same range;

S22, adopting a dynamic self-adaptive block strategy, performing block processing on the image of each layer; obtaining climate model data blocks;

S23, pushing and storing the resampled, stratified and block processed data on the WEB server.

The dynamic adaptive block strategy is specifically:

Let V={v ₁ , v ₂ ,...v _n } be a group of finite point sets in a plane area D in the plane R ² , and after dividing the plane domain D according to the preset point threshold, a The set R of plane sub-fields={R ₁ , R ₂ ,...,R _n }, for each plane sub-field R _i , record the discrete point data and the total number of points in this sub-field, and record the data of each sub-field at the same time The coordinates of the four corner points are to obtain the data value of each sub-block.

The preset point threshold is the minimum number of data points allowed for each sub-block.

The coordinates of the four corner points are respectively: LeftUpX, LeftUpY, RightBottomX, RightBottomY.

S3, output the three-dimensional meteorological data on the WEB server to the WEB client for three-dimensional display, the output includes: using progressive transmission method, data organization and compression method, download balance method, multi-level caching method, LOD multi-level The progressive rendering method of the level of detail model and the volumetric three-dimensional expression method of the Hedgehog method display the optimized meteorological data on the WEB client.

S3 is specifically:

S31, using progressive transmission method, data organization and compression method, download balance method, multi-level caching method to cache the meteorological data to be displayed to the WEB client and display it by the following method;

S32, using a field data organization model based on nodes and layers to simplify scene management;

S33, comprehensively use the pyramid data compression storage strategy based on linear quadtree, multi-resolution model, scene simplification method based on LOD and matching method based on "skeleton + skin" to simplify and speed up the rendering of 3D scenes;

S34, adopt the method of "scene loading management based on visibility buffer" to draw the scene, and apply the dynamic loading method to optimize the loading speed at the same time, so as to realize parallel loading;

S35, using the quadtree data structure to represent the climatic elements in the scene, each node in the tree covers the corresponding area in the climatic elements, dynamically select the climatic element nodes on the basis of meeting the preset error threshold to realize the Continuous multiresolution representation of climate element models;

S35 is specifically:

Divide the entire scene into blocks, calculate the location of the climate elements according to the location of the climate elements in the scene, and establish a linked list index to form a location query table;

If the number of climate elements on each scene exceeds a preset threshold, the climate elements are subdivided until the number is less than the preset threshold or the subdivision level is greater than the preset level threshold; so that the entire scene is organized into an n-ary tree, The whole scene is the root node, and the climate element is the intermediate node;

When dynamically roaming the scene, directly locate a certain scene according to the location lookup table, and then link to a new block node according to the block to which the target point belongs.

S36, using a volumetric three-dimensional expression method of the Hedgehog method to display the meteorological data;

Wherein, the execution order of S32-S36 is in no particular order.

The following steps are also included in S3:

For each quadtree node that meets the error threshold, the diagonal division method is used as the center pointing from the four corner points to the parent node; the quadtree network construction is attributed to the direct diagonal division form and the midpoint form; And the way of retaining the midpoint is divided into four basic representations.

The invention also discloses a three-dimensional display device for meteorological data volume based on WEB technology, including:

The meteorological data preprocessing device is used to preprocess the meteorological data to be displayed; the preprocessing includes: performing format conversion processing, projection transformation processing, and data consistency inspection on the original meteorological data;

The meteorological data optimization device is used to optimize the preprocessed meteorological data and dynamically push the optimized data to the WEB server; the optimization includes: resampling and layering the preprocessed meteorological data ,Block;

The meteorological data output display device is used to output the three-dimensional meteorological data on the WEB server to the WEB client for display, and the output includes: using progressive transmission methods, data organization and compression methods, download balance methods, and multi-level caching methods , LOD multi-level of detail model progressive rendering method and Hedgehog method of volumetric three-dimensional expression method to display the optimized meteorological data on the WEB client.

In order to realize the three-dimensional display of massive meteorological data (including ground-based observation data, satellite retrieval data, model simulation data, etc.), so as to serve the purpose of meteorological scientific research departments and public application departments. Specifically, the meteorological data are processed in the following ways:

1) Data organization of climate elements:

During data processing, all data is usually read into memory at one time for processing. When the amount of data is relatively small, this processing method is simple and effective, and the processing speed is relatively fast. However, when the amount of data is large, such as global remote sensing data with a resolution of 1KM, if it is read into the memory at one time, there will often be insufficient memory space, exhausted computer resources, or even crashes. In order to solve this contradiction, a certain amount of data processing has been carried out. Vertically, a hierarchical structure of pyramids is adopted. In the horizontal direction, automatic block and fast search technology are adopted.

(1) Data stratification method of climate elements

Vertically, it is necessary to generate an image pyramid for the result data of climate elements, and use images with different resolutions according to different display requirements. The method of constructing the pyramid is to take the climate element data as the bottom of the pyramid, and use a certain resampling method on the climate element data to create a series of images with different levels of detail, among which the original image has the highest resolution, and the resampled image The resolution gradually decreases with the increase of the number of pyramid layers, but the range of representation remains the same. The block image pyramid model is shown in Figure 1.

(2) Adaptive block strategy for climate element result data

When the amount of data is too large, adaptive block technology should be considered. The data is divided into different data blocks (DataBlock), and the size of the data blocks can be the same or different. According to the adjacency relationship between blocks, they can be divided into three categories: structured blocks, unstructured blocks and hybrid blocks. Structured blocks can be stored in multi-dimensional arrays in computer languages, and the adjacency between blocks can be determined by corresponding array subscripts, which is convenient for data organization on the computer ^[11] ; unstructured blocks need to specify additional adjacency, use Splicing between blocks. Hybrid blocks are a combination of structured and unstructured blocks.

When using chunking techniques, you may encounter two problems. That is, according to what principle should the data be divided into blocks, and what should be the size of each block? Second, when displaying or roaming, what strategy should be adopted to automatically splice different blocks so that the user cannot see the traces of the block? In order to solve the above problems and take into account the speed, this topic starts from the principle of block, abandons the traditional static block method, and adopts a dynamic adaptive block strategy, that is, the scalable dynamic window method. In the area where the data points are sparse, the retrieval window is appropriately enlarged, otherwise the retrieval window is reduced. The size of the data retrieval window is determined by a specified threshold (minimum number of data points per block).

The adaptive block strategy to be adopted is expressed as follows:

Let V={v ₁ , v ₂ ,...v _n } be a group of finite point sets in a certain plane area D in the plane R ² , according to the specified threshold (minimum data points per block) for the plane domain D After dividing ∑, a set R={R ₁ , R ₂ ,...,R _n } of a planar subfield is obtained. For each planar subfield R _i , record the discrete point data and the total number of points in this subfield, At the same time, record the four corner coordinates (LeftUpX, LeftUpY, RightBottomX, RightBottomY) of each subfield, that is, get the data value of each subblock.

2) Dynamic display and quick visualization of climate elements:

According to the above adaptive block strategy, using the powerful functions provided by Visual C++, the solution is as follows: Write the GetBlock() function to realize the dynamic block technology. The function of this function is to obtain the data that the user is interested in according to the different parameters passed in. Calling this function can realize the pyramid structure and automatic acquisition technology respectively. The Get Block() function is designed with seven parameters. Among them, dwX0 and dwY0 indicate the starting position of the acquired data, and dwWidth and dwHeight specify the size of the data block. These four parameters are passed in by ClipRect in the OnDraw() function when calling, which ensures the flexibility of data acquisition. The size of the data is not fixed, but is automatically given by the redrawing area. The step size nStep implements the pyramid structure, which specifies the decreasing level of data volume between layers. The larger the value of nStep, the faster the rate of data volume decreasing. The parameter *pnBands represents the data source to get the data. If the pyramid structure is to be realized, *pnBands represents the data of the pyramid structure of the upper layer; if the automatic acquisition technology is to be realized, *pnBand s represents the data of the pyramid structure of the current layer. In this way, on the vertical level, every time the GetBlock() function is called, a layer of pyramid data is formed, and then stored in an array (such as a custom array TowerArray); on the horizontal level, every time the GetBlock() function is called It means that the pyramid data of this layer has been automatically acquired once as needed. nNumberOfBands represents the number of bands in an image.

Part of the code is explained as follows:

3) Fast network transmission technology

For network-based 3D scene model rendering systems, since a large amount of geometric data needs to be transmitted to the client via the network, network bandwidth often becomes the bottleneck of such systems, so how to adopt compact and flexible data representation and efficient network transmission Strategies have always been a difficult point in 3D rapid display.

(1) Com component technology: a three-tier browser/server architecture organically combined with Com component technology is adopted, that is, the system data is centrally stored in the data server, and the system application server is mainly responsible for realizing related GIS spatial analysis and network services function, and then transmit the analysis and processing results to the client users through the Internet.

(2) Progressive transmission: With the mode of downloading and displaying at the same time, the server continues to send optimization information for the data, and the display effect of the client is gradually refined until the transmission is completed or the user is satisfied with the fineness of the model.

(3) Data organization and compression technology: effectively organize and efficiently compress image and texture data to minimize the amount of data in the system.

(4) Download balance technology: During the download process, the system will maintain multiple download queues, and each queue will automatically balance the network bandwidth to ensure the optimization of download efficiency.

(5) Multi-level cache technology: Three-level optimization of hard disk cache, memory cache and video memory cache has been established to effectively improve the client's ability to process data.

Through the above network transmission methods, the transmission delay of meteorological data on the network can be greatly shortened, so that the meteorological data can be output to the WEB client cache more quickly. Display while caching to improve overall efficiency.

4) 3D data optimization

For the climate model, the scene mainly includes the undulating terrain and the climate elements corresponding to different terrains, and their optimization is the most important content in the scene optimization.

(1) Optimization of "batch"

Batch is one of the most important concepts in scene optimization. It refers to a rendering call (DP), and the batch size is the number of polygons (climate elements) that can be rendered in this rendering call. Each batch call will consume a certain amount of CPU time. For graphics cards, the number of polygons in a batch is far from the maximum number of draws. Therefore, it is the basic principle of batch optimization to render as many polygons as possible in one batch to reduce the number of batches and ultimately reduce CPU time. However, things are often not as expected, and in some cases the original batch will be broken, causing additional overhead, such as texture changes or different matrix states. In response to these problems, we can use some methods to avoid it as much as possible, and to maximize the batch size.

Merge multiple small textures into one large texture: In a certain scene, there are more than ten different climate elements on the ground, and their rendering states are the same except for their different textures. We can then pack their textures into one big texture, and then specify a "batch" for each weather element, so that we can render all objects with a single render call, reducing the number of batches from more than ten to one. This method is more suitable for objects that do not require high texture accuracy and do not have too many faces.

(2) Rendering state management

The rendering state is used to control the rendering behavior of the renderer, we can set the texture state, depth writing and so on. Changing the rendering state is a time-consuming task for the graphics card, because the graphics card must execute the API strictly according to the rendering path. Time consumption (CPU has to wait), the greater the rendering state changes, the more floating point operations are required. Therefore, effectively managing the rendering state and reducing its changes as much as possible has a huge impact on rendering performance.

(3) Scene management optimization

The optimization of scene management includes scene segmentation, visibility elimination, etc. When we find that the process of traversing the tree is slow during performance evaluation, there may be two reasons. One is that the depth setting of the tree is unreasonable, and the other reason is that a large number of small objects are allocated nodes, resulting in redundancy in the number of nodes. The solution is to divide these small objects into the larger nodes where they are located.

Visibility culling is the most common optimization method, and occlusion clipping is also a frequently used method. The common one is horizon clipping. But in some cases, the effect of occlusion and clipping is not obvious. For example, when the CPU usage is already 100%, the CPU side is the bottleneck. At this time, occlusion and clipping calculation consumes CPU time, and the effect is not obvious. However, in some cases, some pre-generated information methods are used to reduce the complexity of occlusion and clipping calculations, improve the efficiency of occlusion and clipping calculations, and improve scene performance to a certain extent.

5) Based on "importance"-driven visual correlation fast display technology and high-resolution multi-level expression method

Under the current hardware conditions, it is almost impossible to realize the real-time visualization of massive data according to the conventional visualization technology. Therefore, some special processing strategies must be adopted for the scene data itself. Specifically include:

(1) Effective scene management: The system adopts a scene data organization model based on nodes and layers, which greatly simplifies scene management.

(2) Efficient scene simplification technology: the system comprehensively uses the pyramid data compression storage strategy based on linear quadtree, multi-resolution model technology, scene simplification technology based on LOD, and matching technology based on "skeleton + skin", which greatly simplifies, Speeds up the drawing of 3D scenes.

(3) Efficient visibility technology: the system adopts the "visibility buffer-based scene loading management" technology to speed up the drawing of scenes, and the efficient dynamic loading technology not only greatly improves the browsing speed but also improves the loading speed, Parallel loading is realized, that is, browsing and loading are performed at the same time. Parallel loading makes the user not aware of any pauses caused by loading, so it can also be called "zero time loading".

(4) Parallel graphics optimization technology: An important branch of the development of 3D graphics technology is the development of graphics hardware technology. In order to support and give full play to the advantages of graphics hardware technology, the system will provide optimized support for the parallel processing functions provided by the hardware. As shown in the figure below, the range and accuracy of the 3D scene displayed smoothly on an ordinary PC machine.

There are many kinds of algorithms for real-time rendering of realistic graphics, and each of these algorithms has its scope of application. For example, in the fast blanking algorithm, the hierarchical Z-buffer algorithm and the hierarchical occlusion map algorithm are very effective for scenes with high occlusion rate, but they have no advantage for scenes with low occlusion complexity requirements; the scene model simplification algorithm can reduce the scene The purpose of complexity, but the highly complex scene is simplified but still can not achieve real-time; In addition, many algorithms are more efficient in static scenes, they generally generate scene occlusion trees in the preprocessing stage, such as BSP tree, octree, K-D tree, etc., but it is slow in dynamic scenes and the real-time performance is not high.

According to the limitations of the above algorithms, the concept of "importance" is introduced. "Importance" is determined according to the types of climate elements included in the terrain, their distribution density and the three-dimensional accuracy requirements. It describes the shortest distance between the viewpoint and the object point where the geometric error can be ignored during the rapid rendering of spatial terrain data.

The specific method is: apply the synthesis algorithm, and the specific data organization is: first divide the whole scene into blocks, calculate the longitude and latitude of the climate elements according to the location of the climate elements, and establish a linked list index. Then decide whether to subdivide for the next step according to the "importance". If the number of data on each scene exceeds the preset threshold, the data is subdivided until the number is smaller than the specified value or the subdivision level is larger than the specified value. If the 3D model is complex, it can be further subdivided and the bounding box data of the subdivided climate elements can be saved. This way the entire scene is organized into an n-ary tree, forming a tree. The whole scene is the root node, and the data block is the intermediate node. In this way, when dynamically roaming the scene, you can directly locate a certain scene according to the lookup table, and then link to the new data block node according to the data block to which the target point belongs.

The basic way of using the algorithm: use the quadtree data structure to represent the climatic elements, and each node in the tree covers a corresponding area in the climatic elements. Because the upper node has fewer sampling points, it has higher drawing efficiency; however, the fewer sampling points, the greater the error of climate element representation. On the basis of satisfying a given error threshold, feature nodes are dynamically selected to achieve continuous multi-resolution representation of climate element models. In order to improve the efficiency of simplification, the data nodes should be preprocessed first, so as to determine the execution sequence of node operations. Its implementation process can be represented by the figure below. Among them, the root node C0 covers the entire area, followed by secondary nodes, tertiary nodes... . (as shown in picture 2)

When multi-resolution representation of climate elements, the phenomenon of "cracks" usually appears (as shown in Figure 3). Node C ₁ has a lower resolution, while its adjacent nodes C ₂ and C ₃ have higher resolution. This leaves uncovered areas when diagonalizing, creating cracks when drawing. As shown in Figure 3 (the side where the crack is generated is indicated by a bold line).

Elimination of stitching seams: The most common method to solve the "crack" phenomenon is to construct a restricted quadtree. In the past, the algorithm for constructing a restricted quadtree either failed to maintain the topology of the original model, or caused discontinuity on the model surface.

The present invention, on the basis of constructing the restricted quadtree in the past, handles the subdivision of the quadtree as follows:

(1) For each quadtree node that satisfies the error threshold, its diagonal division method is from the four corner points to the center of the parent node. (As shown in Figure 4)

(2) The network construction of the quadtree is mainly attributed to two forms of directly performing diagonal division (as shown in Figure 5) and retaining the midpoint (as shown in Figure 6). And the way of retaining the midpoint can be divided into four basic representation ways (as shown in FIG. 6 ).

6) Volume 3D expression technology based on Hedgehog method

A vector is displayed as a three-dimensional directed line whose direction and magnitude are determined by the data. When Hedgehog is projected onto the two-dimensional viewing plane, the length of the line segment is shortened in the view of the observer, so the length of the line segment is not only related to the data, but also related to the position of the observer.

By adopting the above-mentioned technical scheme disclosed by the present invention, the following beneficial effects are obtained:

Through the technical solution of the present invention, it is possible to quickly process massive amounts of heterogeneous meteorological data, including station observation data, satellite inversion data, numerical forecast data, etc., and use Internet transmission and other related technologies to transfer the optimized meteorological data to the WEB client The three-dimensional processing and display on the terminal can make it very convenient for different departments and the public to access and browse meteorological data through the Internet. The high-fidelity three-dimensional display effect can give users a good visual experience and greatly expand the application field of meteorological data. At the same time, the technical solution of the present invention greatly shortens the processing time and improves the processing efficiency in the processing process.

The above is only a preferred embodiment of the present invention, it should be pointed out that, for those of ordinary skill in the art, without departing from the principle of the present invention, some improvements and modifications can also be made, and these improvements and modifications can also be made. It should be regarded as the protection scope of the present invention.

Claims (10)

1. the weather data body 3 D displaying method based on the WEB technology is characterized in that, may further comprise the steps:

S1 carries out pre-service with weather data to be showed;

S2 is optimized pretreated weather data, and the Data Dynamic that will optimize is pushed on the WEB server end;

S3 outputs to the WEB client with the weather data on the described WEB server end and carries out the body three-dimensional display.

2. the weather data body 3 D displaying method based on the WEB technology according to claim 1 is characterized in that pre-service described in the S1 comprises: original weather data is carried out format conversion processing, projective transformation processing, data consistency checks; Optimize described in the S2 and comprise: to pretreated weather data resample, layering, piecemeal; Output described in the S3 comprises: use progressive transmission method, data tissue and compression method, download balance method, multi-level buffer method, the progressive method for drafting of LOD detail model and the body three-dimensional expression method of Hedgehog method, the weather data after optimizing is showed in the WEB client.

3. the weather data body 3 D displaying method based on the WEB technology according to claim 1 is characterized in that S2 is specially:

S21 repeatedly resamples to described result data, forms the pyramid structure of the multilayer different resolution that is directed to same scope image;

S22 adopts the dynamic self-adapting partition strategy, the described image of each layer is carried out piecemeal handle; Obtain the climatic elements data block;

S23 pushes resampling, layering and piecemeal data processed and be stored on the WEB server.

4. the weather data body 3 D displaying method based on the WEB technology according to claim 3 is characterized in that described dynamic self-adapting partition strategy is specially:

If V={v ₁, v ₂... v _nBe plane R ²In one group of finite point set among a certain plane domain D, by the default threshold value of counting plane domain D is divided after the ∑, obtain the set R={R of a plane subdomain ₁, R ₂..., R _n, to each plane subdomain R _i, the discrete points data and the point that record this subdomain are total, record four angular coordinates of each subdomain simultaneously, namely obtain the data value of each sub-piece.

5. the weather data body 3 D displaying method based on the WEB technology according to claim 4 is characterized in that, the minimum data that the described default threshold value of counting allows for each sub-piece is counted.

6. the weather data body 3 D displaying method based on the WEB technology according to claim 4 is characterized in that, described four angular coordinates are respectively: LeftUpX, LeftUpY, RightBottomX, RightBottomY.

7. the weather data body 3 D displaying method based on the WEB technology according to claim 1 is characterized in that S3 is specially:

S31 uses progressive transmission method, data tissue and compression method, download balance method, multi-level buffer method that weather data to be shown is cached to the WEB client and adopts following method that it is displayed;

S32 adopts the field data organize models based on node and figure layer, simplifies scene management;

S33, the drafting of three-dimensional scenic is simplified, accelerated to integrated use based on the pyramid compression storing data strategy of linear quadtree, multi-resolution models, based on the scene short-cut method of LOD with based on the matching process of " skeleton+skin ";

S34 adopts " scene based on the observability buffering loads management " method to carry out the drafting of scene, uses dynamic loading method optimization simultaneously and is written into speed, thereby realize parallel being written into;

S35, utilize the quaternary tree data structure that the climatic elements in the scene is represented, each node all covers zone corresponding in the described climatic elements in the tree, and Dynamic Selection climatic elements node is realized the continuous multi-resolution representation to the climatic elements model on the basis of satisfying default error threshold;

S36, the body three-dimensional expression method of utilization Hedgehog method displays described weather data;

Wherein, the execution sequence of S32-S36 in no particular order.

8. the weather data body 3 D displaying method based on the WEB technology according to claim 7 is characterized in that S35 is specially:

To the whole scene piecemeal, according to scene mesoclimate key element residing position calculation climatic elements belonging positions and set up the chain table index, form the position enquiring table;

If climatic elements outnumbers predetermined threshold value on the every scene, then to climatic elements segmentation, up to number less than described predetermined threshold value or segmentation level greater than default level threshold value till; So that whole scene is organized into n fork tree, whole scene is root node, and climatic elements is intermediate node;

When scene dynamics is roamed, directly navigate to certain scene according to described position enquiring table, then according to the difference of piece under the impact point, be linked in the new piece node.

9. the weather data body 3 D displaying method based on the WEB technology according to claim 8 is characterized in that, and is further comprising the steps of among the S3:

Satisfy the quadtree's node of error threshold at each, with the diagonal angle partition patterns as the center of pointing to father node from four angle points; The network forming of quaternary tree is summed up as directly carries out diagonal angle subdivision form and keep the mid point form; And the mode that keeps mid point is divided into four kinds of basic expression modes.

10. the weather data body three-dimensional display apparatus based on the WEB technology is characterized in that, comprising:

The weather data pretreatment unit is used for weather data to be showed is carried out pre-service; Described pre-service comprises: original weather data is carried out format conversion processing, projective transformation processing, data consistency checks;

Weather data is optimized device, be used for pretreated weather data is optimized processing, and the Data Dynamic that will optimize is pushed on the WEB server end; Described optimization comprises: to pretreated weather data resample, layering, piecemeal;

Weather data output exhibiting device, be used for the three-dimensional weather data on the described WEB server end is outputed to the WEB client shows, described output comprises: the weather data after the body three-dimensional expression method of utilization progressive transmission method, data tissue and compression method, the progressive method for drafting of downloading balance method, multi-level buffer method, LOD detail model and Hedgehog method will be optimized is showed in the WEB client.

Images