JP2003208597A - Image converter and image conversion method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To continuously smooth scrolling and enlargement/reduction of an image in a display system for an image such as a map using a communication network. <P>SOLUTION: An image converter creates sub image data on a plurality of scales of stepwise different pixel sizes from original image data, and outputs sub image data on each scale in division into fragment image data for many small areas. The set of many pieces of fragment image data created is stored in an image server on a network, and when a client tries to display a desirable section of the original image, fragment image data necessary for the display are selectively transmitted from the image server to the client. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system for a client computer to receive and display image data from an image server through a communication network such as the Internet, and more particularly to a system of a special type suitable for such a system. The present invention relates to an image conversion technique for creating image data.

[0002]

2. Description of the Related Art In a conventional image display system of this type, for example, a map display system, a user performs an operation for changing a map display range such as scrolling or enlarging / reducing a map displayed on a screen to a client computer. Then, the client computer requests the map image of the new map display range from the map server. The map server reads the map data of the new map display range from the database of the server, draws the bitmap map image of the new map display range from the map data, and draws the bitmap map image on the client computer. Send to. The client computer receives and displays the bitmap (raster type) map image.

[0003]

The problem with this conventional system is that scrolling and enlarging / reducing a display image such as a map cannot be continuously and smoothly performed. That is,
When the user scrolls or enlarges or reduces the map, the server waits for a while until the server finishes drawing the map image in the new map display range and sends it back. Switch to a map image with a wide map display range. This is a discontinuous transition, just like jumping one page to another in a web browser. Then, after waiting for a while and the map images on the screen are switched, the user knows where the scroll destination is, or what the scale factor after scaling is. This makes it very difficult for the user to obtain the operational feeling of scrolling or enlarging / reducing the map. This problem becomes more prominent when the waiting time becomes long due to a situation in which many users are accessing the server or the communication line is busy.

The same problem applies not only to map images but also to other images. For example, when distributing bitmap images (raster type images) such as newspaper pages, product catalogs, and aerial photographs over a network, make it possible to clearly read detailed explanations of small letters and small buildings in large-area images. If so, similar problems occur with respect to scaling and scrolling.

Therefore, an object of the present invention is to make it possible to smoothly and smoothly scroll or scale an image in an image display system using a communication network.

Another object of the present invention is to provide a special type of image data that enables the above.

[0007]

An image converter according to one aspect of the present invention comprises means for creating a plurality of sub-image data having stepwise different pixel sizes from original image data, and the plurality of sub-images. It comprises means for dividing each piece of data into a large number of fragment image data pieces, and means for outputting the large number of fragment image data pieces for all of the plurality of sub-image data pieces.

If a large number of chunks of fragment image data created by this image converter are stored in an image server on the network, when a client tries to display a desired portion of an original image, it is necessary for that display. The fragment image data can be selectively transmitted from the image server to the client. At that time, the fragment image data of the sub-image data having a small pixel size can be preferentially transmitted. Thus, progressive display is performed in which images are sequentially displayed from a coarse image having a small pixel size to a fine image having a large pixel size. In addition, since it is sufficient to transmit only a small amount of fragment image data with a small data size, the waiting time from when the user scrolls or enlarges or reduces the display image until the target image is displayed is short. I'm done. In this way, it becomes possible to scroll and scale the display image smoothly and at high speed.

In a preferred embodiment, a large number of fragment image data which make up each sub-image data are stored in a hierarchical folder. This facilitates finding the required cross-sectional image data.

An image data structure according to another aspect of the present invention includes a plurality of pieces of sub-image data, each of which represents a common target image and which have stepwise different pixel sizes, and each sub-image data includes a large number of small areas. It is divided into fragment image data.

[0011]

1 shows a schematic overall configuration of an embodiment of an image display system to which the present invention is applicable.
This example is for displaying a map.

A map server 10 that provides a map providing service
And a large number of client computers 30, which are respectively used by a large number of users, can communicate with each other through a communication network such as the Internet 20. The map server 10 has a map database 11 that stores map data of a wide area. The client computer 30 has a client program 31 for accessing the map server 10. The client program 31 is provided with an image viewer 32 for sending a specific map data request to the map server 10 and displaying the map data received from the map server 10, for example, in the form of a plug-in. The client computer 30 includes a display device 33 and a mouse 34.
It also has a controller like.

In this embodiment (although not necessarily), the map server 10 is a WWW server, the client program 31 in the client computer 30 is a WWW browser, and both are HTTP.
Request map data and exchange map data by protocol.

FIG. 2 illustrates the types of maps that the map server 10 has and the order in which the image viewer 32 in the client computer 30 requests map data from the map server 10.

The map database 11 of the map server 10 stores data of a plurality of types of maps 110 to 112 having different scales. In the example shown in the figure, there are three types, one of which is a large scale map of 1/2500 (so-called city map) 110, one of which is a medium scale map of 1 / 25,000, and one of which is a small scale map of 1/20000. Stored (of course, more types of maps may be stored).

Since the map is used here, the term "scale" is used, but this is replaced with the term "pixel size" or "pixel number" of the image (that is, the number of pixels in the vertical and horizontal directions of the image). Can also be explained. In the following description, the term “depth” is used to indicate the level of “scale” or “pixel size” of a plurality of image data in which “scale” or “pixel size” is gradually reduced. We also use terms.

Now, a plurality of maps 1 having different scales as described above.
Each of 10 to 112 is divided into map data in units of a small rectangular area (hereinafter referred to as "mesh") obtained by finely dividing a wide area covered by the map in the directions of latitude and longitude. The map data is transmitted from the map server 10 to the image viewer 32 in units of one mesh. That is, the map data of one mesh is requested from the image viewer 32 to the map server 10 in one HTTP session between them, and the map data of the one mesh is returned from the map server 10 to the image viewer 32. It Therefore, the smaller the mesh size (area), the smaller the amount of mesh map data and the shorter the time required for one HTTP session, and the shorter the time required for displaying one mesh. However, the smaller the size (area) of the mesh, the larger the number of meshes included in the map display range desired by the user. Therefore, the number of HTTP sessions performed between the image viewer 32 and the map server 10 increases. . From this point of view, in order to end the display of the map display range desired by the user at high speed (or to make the user feel that way), the map data amount of one mesh is appropriate. The size of one mesh, in other words, the data of one mesh, so that the time required for one HTTP session is not too long and the number of HTTP sessions required for the display range of the map desired by the user is not too large. It is important to set the amount appropriately. For this reason, different map scales result in different mesh sizes. That is, the smaller the scale (that is, the larger the denominator of the scale), the larger the area area that can be covered by the same map data amount, so the mesh size is also set large (however, the data amount of one mesh is Not all maps 110-112 are the same).

The image viewer 32 in the client computer 30 grasps the display range of the map desired by the user in accordance with the user's operation of scrolling or enlarging / reducing the map with the mouse 34 or the like, and overlaps the display range (that is, the display range). , Which mesh is included in the display range or includes the display range) is determined for each scale map 110 to 112. For example, in the case of the example of the display range of the map shown in the display device 33 in FIG. 2, the meshes that overlap the display range are 16 meshes 110A to 110P of the large scale map 110 (indicated by broken lines in the figure). Boundary is shown) and medium scale map 111
4 meshes 111A to 111D (the boundaries are shown by solid lines in the figure) and one mesh 112A of the large scale map 112. These meshes 110A-11
After determining 0P, 111A to 111D, and 112A, subsequently, the image viewer 32 requests the map data of the meshes from the map server 10. At that time, the image viewer 32 preferentially starts from the small scale map, that is, first, the mesh 112A of the small scale map 112, the meshes 111A to 111D of the medium scale map 111, and finally the large scale map 110. Mesh Mesh 110A ~
The mesh data request is issued to the map server 10 in the order of 110P. However, if the display range determined by the user's enlargement / reduction operation is wide and therefore it is not necessary to display a map of a scale smaller than a certain scale (for example, the large scale map 110), the unnecessary map is displayed. It does not require a mesh (eg, large scale map 110).

The map server 10 reads the map data of each mesh in the requested order from the map database 11 and sends it back to the image viewer 32. As a result, the image viewer 32 can usually receive and display the mesh data of a smaller scale map first (although the order may sometimes change depending on the circumstances of the communication network). As described above, the smaller the map, the larger the size of one mesh, so that the entire display range desired by the user can be covered with a smaller number of meshes. Therefore, a smaller scale map is displayed at high speed over the entire display range desired by the user. After that, a larger scale map is displayed. Larger scale maps represent more detailed features. Therefore,
In the display range desired by the user, a rough feature display of a small-scale map first appears at high speed, and then a detailed feature display of a larger-scale map appears.

By adopting the progressive map display method using a plurality of types of maps having different scales as described above, a rough small scale map is displayed immediately when the user scrolls or enlarges or reduces the map. Therefore, the user no longer feels the discontinuity of the map after waiting for a while, as in the past.
It is possible to obtain an operational feeling that the map can be scrolled or enlarged and reduced continuously and smoothly.

The map data held by the map server 10 is mainly composed of vector data in which polygons, lines, character codes, etc. are described.
Transmits the requested mesh map data to the image viewer 32 as vector data without converting it to bitmap data (that is, drawing a map image). Then, the image viewer 32 draws a map image from the received vector data. in short,
The map server 10 simply repeats the simple operation of acquiring the data requested by the client from the database and returning it to the client. As a result, the load on the map server 10 is reduced, and the map server 1
0 can immediately return the requested map data to each client even when receiving requests from many clients. This also contributes to the effect that it is possible to obtain an operational feeling that the map can be scrolled or enlarged and reduced continuously and smoothly.

FIG. 3 shows the file structure of each type of map that the map server 10 has.

As already described with reference to FIG. 2, the map server 10 has a plurality of types of maps 110 to 112, and each of these maps 110 to 112 is a file as shown in FIG. Have a configuration. As shown in FIG. 3, one type of complete map 200 includes a plurality of layers having different types of data, for example, a polygon layer 210,
It is divided into a line layer 220 and a character layer 230.
The polygon layer 210 is a collection of polygon data representing, for example, a building or a site. Line layer 220
Is a collection of line data representing roads, railways, rivers, and the like. The character layer 230 is a collection of character code strings representing, for example, place names and building names, and data of symbol marks such as icons and map symbols.

As described above, the entire area covered by the map is divided into a large number of meshes subdivided in the directions of latitude and longitude. Therefore, the polygon layer 210, the line layer 220, and the character layer 230 are each divided into data 211, 221, and 231 for each mesh. Then, the data of one mesh of one layer is configured as one file.

Each mesh is assigned a mesh number corresponding to the position coordinate of each mesh on the whole map. As shown in FIG. 3, the mesh number of each mesh is the coordinate value in the longitude direction (mesh number in the longitude direction) Xi.
And the coordinate value in the latitude direction (lattice mesh number) Yj
And a set (Xi, Yj). Then, as shown in FIG. 3, polygon files 211, 211, ... For each mesh included in the polygon layer 210, line files 221 and 2 for each mesh included in the line layer 220.
.. and the character files 231, 231, ... for each mesh included in the character layer 230 are given the same file name “XiYj” as the mesh number (Xi, Yj) of each mesh. Therefore, when you want the map data of a certain mesh, if you know the mesh number (that is, the coordinate values in the latitude and longitude directions) of that mesh, that mesh number will represent the file name of the desired map data. . With this, the image viewer 32 determines whether the mesh is necessary for the display range desired by the user, and when requesting the data of those meshes from the map server 10, the file name to be requested is immediately determined from the mesh number. You can

As shown in FIG. 3, (although not necessarily) in this embodiment, the polygon layer 21
0, the line layer 220 and the character layer 230,
It is stored in different directories in the map database 11, for example, a polygon layer directory 121, a line layer directory 122 and a character layer directory 123.

FIG. 4 is a map database 1 shown in FIG.
1 shows a hierarchical structure inside each of the directories 121 to 123 for each layer within 1.

As described above, the size of one mesh is set to an appropriately small size so that one HTTP session can be completed in a short time without giving the user the feeling of waiting. Therefore, the entire area of the map is divided into a huge number of meshes. If this huge number of mesh files are simply stored in parallel in one directory, when a request for a certain file arrives, it takes a very long time to search that file. Therefore, in this embodiment, a huge number of mesh files are managed in a hierarchical directory as shown in FIG.

That is, continuous position coordinates (that is, mesh numbers, that is, file names) as shown in FIG. 4C.
File of 64 meshes of 8 × 8 array with
.. are grouped into one group, and the directory (hereinafter referred to as “folder”) one layer above (second layer) shown in FIG.
, 132, 132, ... The folder name (directory name) of each folder 132 shown in FIG. 4B is the longitude component “Xi” of the file name “XiYj” of the 64 mesh files 131, 131, ... Stored in each folder. The name represented by the quotient obtained by dividing the latitude component "Yj" by the number of files "8" in the longitude and latitude directions, that is,

[0030]

[Equation 1] Has become. Therefore, each folder 13 of this second hierarchy
It can be said that the folder name of 2 corresponds to the position coordinates of the 8 × 8 mesh area covered by the 64 mesh files stored in each folder 132.

Then, 64 second second arrays of 8 × 8 array having consecutive folder names as shown in FIG. 4B (that is, position coordinates of the 8 × 8 mesh area shown in FIG. 4C). The folders 132, 132, ... Of the hierarchy are collected into one group and are shown in FIG.
Folder 13 one level higher (first level) shown in (A)
It is stored in one of 3, 133, .... And FIG.
The folder name (directory name) of each folder 133 shown in (A) is the longitude component “X” of the folder names of the 64 folders 132, 132, ... Stored in each folder 133.
i / 8 ”and the latitude component“ Yj / 8 ”are divided by the number of folders“ 8 ”in the longitude and latitude directions, respectively,
That is

[0032]

[Equation 2] Has become. Therefore, each folder 13 of the first hierarchy
The folder name of 3 corresponds to the position coordinates of the 64 × 64 mesh area covered by the 8 × 8 folders (that is, 64 × 64 mesh files) of the second hierarchy stored in each folder 133. be able to.

Further, 64 first-level folders 1 in an 8 × 8 array having continuous folder names (that is, position coordinates of a 64 × 64 mesh area) as shown in FIG. 4A.
, 33, 133, ... are grouped into one group, and a directory of a higher hierarchy not shown (in the case of a polygon layer, for example, the polygon layer directory 121 shown in FIG. 3)
It is stored in.

Note that in this embodiment, FIG.
As shown in (C), a directory having a three-level structure is used and one directory manages 64 files or folders, so that 64 × 64 × 64 = about 2 in total.
It is possible to manage 60,000 mesh files.
However, this is an example, and the number of layers may be further increased as the number of meshes further increases. Further, the reason why the 64 files or folders are managed in one directory is the OS environment of Microsoft Corporation in the United States.
Map server 10 in a Windows (registered trademark) environment (in this OS environment, a high search speed can be obtained when the number of files or folders in one directory is less than 100).
This is because it has been determined that 64 are easy to handle when activating.

What is important in the hierarchical directory shown in FIG. 4 is that the name of each hierarchical directory can be easily calculated from the file name of a specific mesh which is the final search object. Is. That is, "X
When you want a file named "iYj", divide the longitude component Xi and the latitude component Yj of the file name "XiYj" by 8 to easily determine the folder name of the higher-level folder and arrange them. By doing so, the path to the desired mesh file is

[0036]

[Equation 3] It will be decided easily. By the way, if it is an n-level directory, the path to the target file is

[0037]

[Equation 4] As such, it is easily determined by calculation. The above calculation method of dividing the longitude component Xi and the latitude component Yj of the file name “XiYj” by 8 is only an example.
Other calculation methods may be adopted.

Whatever calculation method is used, the path to the file can be calculated by a predetermined numerical operation from the file name of the target mesh. That is, the file name of each mesh is the mesh number (position coordinate) of that mesh. ), The mesh file required for display by the image viewer 32 is stored in the map server 1.
This greatly simplifies the time required to request 0, and contributes to speeding up map display.

FIG. 5 shows a functional structure of the image viewer 32.

As shown in FIG. 5, the image viewer 32
Has a display unit 320 and a downloader 322. The main function of the display unit 320 is to determine a range to be displayed on the map in response to a user's operation of scrolling or enlarging / reducing the map using a mouse or the like, and determine a mesh file name (mesh number) required for the display range. ) And request the file of the mesh to the downloader 322,
That is, a map image to be displayed is drawn from the vector data included in the mesh file acquired from the downloader 322. The main role of the downloader 322 is to receive a file request from the display unit 320 and download the requested file from the map server 10.

The display unit 320 has a memory cache 321 for storing the data of the files acquired in the past in the memory in the priority order according to the recent access frequency. The downloader 326 lists the file name of the file to be downloaded (path to the file in the map server 10), the download request list 325 (the display unit 320 deletes the file name in this list at any time). And a download flag 326 for notifying the display unit 320 whether or not a new file has been downloaded, and a plurality of HTTP session threads 324A to 324D for performing each HTTP session when downloading each file.
(There are four threads in the illustrated example, but the number is variable), and a disk cache 323 for storing the downloaded file in a large-capacity nonvolatile storage such as a hard disk drive. There is.

FIG. 6 shows the flow of operation of the display section 320.

When the user performs an operation of changing the display range of the map such as scrolling or enlarging / reducing the map with the mouse or the like (step S1), the display section 320 responds to this and performs the following operation. That is, all the file names described in the download request list 325 are deleted (S2), the download flag 326 is reset (S4), and a new display range after change is determined and its display is displayed. The file name (mesh number) of the mesh required for the range is determined (S5). In step S5, the display unit 320 sets the optimum scale as the scale of the map to be displayed in the display range based on the display scale determined by the scaling operation (for example, if the display scale is large, a large scale,
Decide on a medium scale if it is medium and a small scale if it is small, and determine the file name required for the new display range for the map of that optimal scale (the file of the optimal scale determined here will be referred to as "necessary" below). File "). In addition, each scale that is smaller than the optimal scale (that is, the denominator is larger) is a temporary alternative display (that is, the progressive display described above) when the required file of the optimal scale is not immediately available. The file name of the mesh that overlaps with the new display range is also determined for the map of (the file of the map with a smaller scale for this alternative display is hereinafter referred to as "alternative file").

Subsequently, the display unit 320 draws and displays the map image in the new display range in the order of the polygon layer, the line layer, and the character layer.
The following processing is entered.

First, at the stage where the display of the polygon layer is incomplete (NO in S6), a file of the polygon layer having the file name (mesh number) determined in step S5 is read from the memory cache 321. Search (S9). At this time, for each mesh, first find the required file with the optimum scale, and if it does not exist, look for an alternative file with a smaller scale than that, and if not,
In addition, look for an alternative file of one-step smaller scale. in this way,
The file with the optimum scale is searched for with the highest priority, and if there is no such file, an alternative file with a smaller scale is searched step by step.
If the target file (necessary file or alternative file) is found in the memory cache 321 (Y in S9)
ES), a polygon map image is drawn from the vector data of the file and displayed in the corresponding mesh area of the display screen (S10). At this time, if you want to display a larger scale map image where a smaller scale map image was already displayed as a substitute, delete the smaller scale map image that was displayed as a substitute. Then, a larger scale map image is newly displayed.

If all the necessary files are found in the memory cache 321 (NO in S11), the display of the polygon layer map is completed. Next, unless the display range changes (NO in S17), the map is not yet displayed. In order to draw and display the map of the completed line layer (NO in S7), the processing in step S9 and subsequent steps described above is started. As for the line layer, if all the necessary files are found in the memory cache 321 (NO in S11), next, unless the display range is changed (NO in S17), the map drawing of the character layer which is still incomplete is performed. In order to display (NO in S8), the above-described processing of step S9 and subsequent steps is started. When the map image of the line layer is already displayed and the map image of the line layer is displayed next time, or when the map image of the character layer is further displayed, the map image of a smaller scale is displayed as a substitute. Unlike the case where a large-scale map image is displayed, the map image of the polygon layer that is already displayed is not erased, the map image of the line layer is displayed on top of it, and the text image is superimposed on it. Display map images of layers.

On the other hand, when the map of the polygon layer is being displayed, if there is only one required file that cannot be found from the memory cache 321, (S1
1), the display unit 320 then displays the missing required file and the alternative file for the missing required file in the memory cache 3
The downloader 322 is requested to find the files that are not found from 21 (the necessary files and the alternative files that are not found are collectively referred to as “missing files” below) to the downloader 322 (S12). Also at this time, the display unit 320 first requests the necessary file of the optimum scale with the highest priority, and subsequently requests the alternative file of a smaller scale stepwise.

As will be described later, the downloader 322 is
When a request for a missing file is received from the display unit 320, in the order requested (that is, as in the case where the display unit 320 searches the memory cache), the required file is first searched. Disk cache 3 for each missing file (in order of scaled alternate files)
23, and if found, the path to the file found in the disk cache 323 is displayed on the display unit 3
If no file is found, an answer that there is no file is returned to the display unit 320. The display unit 320 receives the path to the file found in the disk cache 323 from the downloader 322 (YE in S13).
S), the file is read from the disk cache 323 using the path and the memory cache 321
(S14), and a polygon map image is drawn from the vector data of the file and displayed in the corresponding mesh area of the display screen (S15).
As mentioned above, when trying to display a larger scale map image where a smaller scale map image has already been displayed as a substitute, the map image of the smaller scale that was displayed as a substitute is displayed. The map image is deleted and a larger scale map image is newly displayed.

If all the necessary files are found in the disk cache 323 (NO in S16), the display of the polygon layer map is completed. Next, unless the display range changes (NO in S17), the explanation has already been given. The process of drawing and displaying the map of the line layer in the same manner as (S9 and below)
to go into. Also for the line layer, if all required files are found from the disk cache 323 (N in S16)
O), since the display of the map of the line layer is completed, next, unless the display range is changed (NO in S17), similarly, the process of drawing and displaying the map of the character layer is started (S9 and thereafter). As described above, when the map image of the line layer is already displayed and the map image of the line layer is displayed next time or the map image of the character layer is further displayed, the map image of a smaller scale is displayed. Unlike the case where a large scale map image is displayed due to the place where was displayed as an alternative, the map image of the line layer is displayed over it without deleting the map image of the polygon layer that is already displayed, Further, the map image of the character layer is displayed overlaid on it.

In this way, if all the necessary files of all layers are present in the memory cache 321 or the disk cache 323, the display of the map of the new display range is completed instantly. When the display of the map in the new display range is completed (YES in S8), the process returns to the beginning.

On the other hand, when the downloader 322 is requested for the missing polygon layer file, the downloader 32
If there is even one required file for which a reply that there is no file is returned from 2 (YES in S16), the process of the display unit 320 returns to the beginning. Then, the display unit 320
Waits until the download flag 326 is set (S2) unless the display range is changed (NO in S1). As will be described later, when the downloader 322 returns the answer that there is no file for a certain file, the downloader 322 starts the work of downloading the file from the map server 10, and when the download of the file ends, the file is downloaded to the disk / disk. Store in cache 323 and set download flag 326. When the display unit 320 learns that the download flag 326 has been set (YES in S2), the display unit 320 repeats the processing in and after step S3, which is the same as that performed when the display range is changed. That is, all the file names in the download request list 325 are erased (S2), the download flag 326 is reset (S4), the display range is recalculated, and the mesh file name required for the display range is calculated. (Mesh number)
Is determined (S5), and the processing in and after step S9 for acquiring the file in the order of the polygon layer, the line layer, and the character layer to draw and display the map image is repeated. As a result, the display unit 320 draws the same map image again for the mesh that has already been drawn and displayed, and downloads the mesh for which the file has been downloaded. The file is read from the disk cache 323, a map image is drawn, and displayed in the area of the mesh.

Each time one missing file is downloaded, the display section 320 repeats the processing from step S3, so that the display area on the display screen approaches one mesh by mesh one by one. When the map image of the entire display area on the display screen is completed (Y in S8
S), the display unit 320 returns to the first step S1, waits for the display range to be changed by the user, and if changed, repeats the processing of step S3 and thereafter.

Also, when the display range is changed by the user before the map image of the entire display area on the display screen is completed (YES in S17), the display section 320 is displayed.
Repeats the processing from step S3 onward.

FIG. 7 shows the operation flow of the downloader 322.

As shown in FIG. 7, the downloader 322
Receives a file request from the display unit 320 (S2
1) First, the requested file is searched from the disk cache 323 (S22). If the target file is found from the disk cache 323 (S22
YES), the downloader 322 informs the display unit 320 of the path of the file in the disk cache 323 (S23). If all the files requested from the display unit 320 are found in the disk cache 323 and their paths can be notified to the display unit 320 (NO in S24), the downloader 322 returns to the first step S21 and returns to the display unit. Wait for the file request from 320 again.

On the other hand, if there is a file not found in the disk cache 323 among the files requested by the display unit 320 (YES in S24), the downloader 322 displays that there is no file for the missing file. The file name (including the path to the file in the map server 10) of the file not found is written in the download request list 325 while notifying the section 320 (S25) (S2).
6). At this time, download request list 3
25 is displayed in advance by the display unit 320 which has already been described.
Since all the file names written in the past are erased by the processing of step S3, only the file names of the files not found at this time are written. At this time, the downloader 322 preferentially writes the map file of a smaller scale among the files that cannot be found in the download request list 325 (that is, the smaller scale files are written closer to the top of the list). ).

Subsequently, the downloader 322 places the file names written in the download request list 325 in the idle state in order from the beginning of the list 325 (that is, preferentially from a map file with a smaller scale). An HTTP session thread 324A-324
Allocate to D (S28). HTTP session
The threads 324A to 324D execute an HTTP session with the map server 10 as soon as they are assigned file names, and download the respective assigned files from the map server 10. HTTP session thread 3 from smaller scale map file first
Since I assigned to 24A-324D, download
The smaller scale map file is completed first.

Each time each file is downloaded (Yes in S29), the downloader 322 erases the file name of the file from the download request list 325 (S30) and saves the file in the disk cache 323. (S31), and
The download flag 326 is set (S32). Then, the downloader 322 returns to the first step S21 and waits for another file request from the display unit 320. When the download flag 326 is set, the display unit 320 issues a file request again, as described above.
Downloaded files to disk cache 3
It reads from 23 and draws and displays the map image.

The above display section 320 and downloader 322
The downloader 322 simply repeats the process of downloading a missing file. In addition, the display unit 320 simply attempts to acquire all the files necessary for the display range each time the display range is changed or the file download is completed, and the process of drawing the map using only the acquired files is simple. Just repeat. Every time the display range changes or file download is completed,
There is a seemingly redundant work of getting all the necessary files, including already drawn files, and redrawing them from the beginning, but in return, which files have been drawn and which files have been drawn. It is not necessary to perform a troublesome determination process such as determining whether or not drawing has been completed and selectively drawing only undrawn files. Further, the display unit 320 and the downloader 322 operate asynchronously, and neither of the troublesome synchronization processing such as matching the operation timings of the two is performed. As a result, the map image can be displayed at high speed as a result.

Since the client computer 30 draws, the map server 10 is not burdened with drawing, which contributes to speeding up the response of the map server 10 and displaying the map image at high speed.

In addition, instead of trying to complete the desired map image at once, the polygon layer,
A method of progressive display of geographical elements with different appearances, such as sequentially displaying map images in the order of line layer and character layer, and downloading a smaller scale map file first and displaying it first, By using the method of progressive display of map images of different scales (that is, alternative display with a map of a smaller scale when the map of the optimum scale is not in hand), it is possible for the user to delay the map display due to the download time. It is possible to realize a continuous and smooth scrolling operation and a feeling of operation of enlarging / reducing without making the user feel substantially.

FIG. 8 shows an example of this progressive display.

It is assumed that the user scrolls the map from a state in which a large-scale map of a certain area as shown in FIG. 8A is displayed. Then, a complete large-scale map of the scroll destination area is finally displayed as shown in FIG. 8D, but by the time the final screen of FIG. 8D is reached. , Screens at intermediate stages shown in FIGS. 8B and 8C are sequentially displayed.

The screen of FIG. 8B is that immediately after scrolling, and the broken line in the drawing indicates the boundary of the mesh. In the screen immediately after this scroll, one mesh 41 on the upper right is displayed.
Only, the large scale map polygon layer file was already in the cache, so the large scale map polygon image is displayed. The remaining three meshes 42 ~
For No. 44, the large scale map polygon layer file is still being downloaded, so the large scale map image is not displayed at all, but since the medium scale map file is in the cache, the medium scale map is not displayed. The polygonal images of the large buildings 431 and 441 on the map are displayed instead.

Waiting for a while, the remaining three meshes 42-4
With respect to No. 4, the file of the polygon layer of the large scale map is completely downloaded, and the screen shown in FIG. 8C is displayed. On this screen, the display of the polygon layer of the large scale map is completed, but the display of the next line layer is only completed for the one mesh 41 at the upper right. Regarding the remaining three meshes 42 to 44, since the line layer file of the large scale map is still being downloaded, the image of the line layer of the large scale map is not displayed at all, but the file of the medium scale map is not displayed. Was in the cache (or the download was completed first), so the main road 43 on the medium-scale map
The line images of 2,433 are displayed as alternatives.

After waiting for a while, the files of all layers of the large scale map are gathered and the final screen as shown in FIG. 8D is completed (however, the image of the character layer is not shown in the figure). Available).

According to the prior art, when scrolling is performed from the state shown in FIG. 8A, the map server side creates and sends a final screen as shown in FIG. After waiting for a long time after doing
It will switch from (A) to FIG. 8 (D) at once.
On the other hand, in the present embodiment, when the scroll operation is performed, the screen as shown in FIG. 8B is immediately displayed, and the final screen shown in FIG. You can have the feeling that the map display has started immediately, and you can immediately determine where the scroll destination is, so you can get a continuous scroll feeling (same as for scaling). In addition, since the polygon image of a building is displayed first, the user can easily grasp the map immediately.

Further, even at the stage where an incomplete image as shown in FIG. 8B is displayed, the user can know which region it is, so whether the scroll should be further advanced or not. You can make immediate decisions and, if necessary, continue to move forward without stopping scrolling.

FIG. 9 shows another embodiment of the present invention in which a plurality of map servers are provided to distribute the load.

A plurality of map servers 50, 50, ... Are connected to a communication network such as the Internet 60, and a large number of client computers 70, 70 ,.
Can access any of the plurality of map servers 50, 50, .... The plurality of map servers 50, 50, ... Have map databases 51, 51 ,.

FIG. 10 shows a client computer 70.
Shows a method for requesting a plurality of map servers 50, 50, ... For a mesh file required for a display range desired by the user.

As shown in FIG. 10 (A), each client computer 70, in principle, stores all files of a plurality of meshes 81, 81, ... That overlap the display range 80.
Rather than requesting one map server 50, the first mesh is sent to the A server, the next mesh is sent to the B server, the next mesh is sent to the C server, and so on. , ... are distributed almost evenly to the request. As a result, the load is evenly distributed among the plurality of servers 50, 50, ..., It becomes possible to display the map at higher speed. Further, since each of the plurality of servers 50, 50, ... Has the same map data, the servers 50, 5 ,.
Maintenance of 0, ... Is easy.

If a plurality of servers 50, 50,
It is assumed that a failure occurs in any of the ..., For example, the C server. In this case, each client computer 70
When I know that the normal response is not returned from the C server,
As shown in FIG. 10B, the remaining normal A servers and B
Request evenly distributed files from the server.
As described above, since all of the plurality of servers 50, 50, ... Have the same map data, even if a failure occurs in some of the servers, another server can take the alternative measures. .

The form of load distribution by a plurality of servers includes
Various other than the above can be adopted. For example, the server may be divided for each area. For example, one or a plurality of servers are in charge of the metropolitan area, which is frequently accessed, and another one or a plurality of servers are in charge of the remaining area.
Alternatively, the servers may be divided according to different scale maps.
For example, a large scale map having a large amount of data is handled by one or a plurality of servers, and another scale map is handled by another server or a plurality of servers. Also, the server may be divided for each layer.

The system described above handles maps, but the types of images to which the present invention can be applied are not limited to maps, but cover all types of images. When the present invention is applied to a general image, for example, a plurality of bitmap image data having different pixel sizes are prepared for the image, the bitmap image data of each size is divided into meshes, and the mesh data is As described above, it is possible to deliver from the server to the client through the network and display it on the client.

Next, an embodiment of the image converter according to the present invention will be described.

This image converter has different scales (pixel size, depth) which are respectively mesh-divided so as to be stored in the server 10 of the above-mentioned image display system.
It is used to create multiple image data with and upload it to the server. When this image converter is used for the above-mentioned map display system, for example, it creates the map image data 11 as described above, but the map is merely an example for explanation, and all contents are included. Bitmap images can be processed. This image converter is implemented by executing a computer program on a computer in the following embodiments, but the invention is not limited to this and can be implemented using dedicated hardware.
The computer program that realizes this image converter is installed in, for example, the client computer 30 illustrated in FIG. 1 in the following embodiments, but the present invention is not limited thereto and can be installed in another computer.

FIG. 11 shows the flow of processing performed by this image converter.

As shown in FIG. 11, the image converter is
In step S101, the original image data to be converted is selected according to a user's instruction. The original image data is basically bitmap (raster type) image data, but it is not limited to this and may be vector type image data. In that case, the image converter converts the vector type image data. It is read and converted into bitmap (raster type) image data.

In steps S102 and S103, the image converter selects the mesh size and the compression method according to the user's instruction. Here, the mesh size is the number of pixels on one side of each mesh data (square in this embodiment). For example, one of the three mesh sizes of 64, 128 and 256 pixels can be selected by the user. For example, in an image in which minute changes cannot be overlooked, such as an aerial photograph used as the basis of a map, 6
A relatively small mesh size, such as 4 pixels, may be selected, and conversely, a relatively large mesh size, such as 256 pixels, may be selected for an image such as an illustration in which minute changes are small. For normal photographic images, such as landscapes and people, you can choose a medium mesh size, such as 128 pixels.

The compression method is a compression method performed when the created mesh data is converted into a file. For example, a user can select from two types of a specific lossy compression method and a specific lossless compression method. You can choose the one you want. The lossless compression method may be suitable for an image in which minute changes cannot be overlooked, and the lossy compression method may be sufficient for applications such as ordinary photographic images.

Upon receiving the execution request from the user in step S104, the image converter reads the original image data and performs mesh formation processing in step S105. A specific flow of the mesh creating process will be described later with reference to FIG. By executing the mesh creation processing, the original image data is converted into a plurality of pieces of image data having different scales (pixel size, depth) which are mesh-divided, and the converted image data is as shown in FIG. Client computer 3 in the form of a large number of mesh data files classified and stored in hierarchical folders
0 is stored in a predetermined directory.

In step S106, the image converter
A template HTML file is selected according to a user's instruction.
Here, the template HTML file is the image viewer 32 shown in FIG. 1 (for example, Ac of Microsoft Corporation in the United States).
It is an HTML file in which a script for executing a plug-in program using the tiveX technology and a background definition of the display screen are described.

When a preview request is received from the user in step S107, the image converter activates the WWW browser 31 shown in FIG.
To execute a preview (preliminary display) of the converted image data. That is, the image converter reads the template HTML file described above,
The script for executing the image viewer in the template HTML file is rewritten so that the converted image data described above becomes a display target, and the rewritten HTML file is newly saved in a predetermined directory, and , WWW browser 31 is started, and the saved HTML file after rewriting is WWW browser 3
Open to 1. The WWW browser 31 opens the HTML file, executes the image viewer 32 according to the script therein, and displays the converted image data stored in the above-mentioned predetermined directory. As already described, scrolling and scaling of images can be done smoothly.

In step S108, the image converter
According to a user's instruction, each upload destination (for example, a URL indicating a predetermined location in the server 10 shown in FIG. 1) of the above-mentioned rewritten HTML file and the converted image data is designated. When the upload execution is requested by the user in step S109, the image converter uploads the rewritten HTML file and the converted image data to each designated upload destination through a network such as the Internet.

FIG. 12 shows a specific procedure of the mesh creating process in step S105 described above.

As shown in FIG. 12, step S201.
Then, the image converter reads the selected original image data, and based on the pixel size of the original image data,
Determine the depth number n. Here, the number of depths n is the number of pieces of bitmap image data that are created as a result of converting the original image data and that have different pixel sizes in stages. For example, if the number of depths n is 3, as a result of conversion, for example, the same pixel size as the original image data, a pixel size one step smaller than that, and a pixel size one step smaller, a total of three depths are obtained. Bitmap image data of (pixel size) will be created. For determining the depth number n, for example, the number of pixels on the long side of the vertical and horizontal sides of the original image data,
Alternatively, the total number of pixels (that is, the number of vertical pixels × the number of horizontal pixels) is compared with a plurality of predetermined thresholds, and depending on the comparison result, for example, n = 3 if the threshold is less than the minimum threshold, and the second threshold or more than the minimum threshold. If it is less than n, n = 4, if it is more than the second threshold value, n = 5, and so on.

In step S202, the image converter
According to the determined depth number n, n images having different depths (pixels) are created using the original image data. For example, in the case of n = 3, the image data of the first depth is the original image data, and the image data of the second depth is created by reducing the original image data by one step in the number of pixels.
The image data having the depth of 3 can be prepared as three pieces of image data by a method in which the original image data is reduced by two steps in the number of pixels. In this case, it is possible to use a constant k less than 1 such as k = 0.8 as the reduction ratio for one step, and to reduce by a stepwise power multiplication factor of this constant k, or An optimal reduction ratio preset for each time may be used.

In step S203, the image converter
Each pixel size of n image data for each depth,
From the selected mesh size, the number m of meshes formed when each image data is divided into meshes is calculated. Then, based on the number m of meshes, the image converter determines the hierarchical structure of folders for storing mesh data, the name of each folder, which folder each mesh data file should be placed in, and so on. A folder group having such a hierarchical structure is created in a predetermined directory in the client computer 30.

In step 204, the image converter divides each of the n pieces of image data for each depth into the mesh data of the mesh number m calculated for each image data. Then, the image converter, in step S205,
Each of the divided mesh data is made into a file compressed by the selected compression method, and at step 206, the mesh data file is stored in an appropriate folder in the folder group of the hierarchical structure. The above process is performed for all the n image data, and the conversion is completed.

The embodiments of the present invention have been described above, but these embodiments are merely examples for explaining the present invention, and the scope of the present invention is not limited to the above embodiments. Therefore, the present invention can be implemented in various forms other than the above embodiment.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic overall configuration of an embodiment of an image display system to which the present invention is applicable.

FIG. 2 is a diagram illustrating the types of maps that the map server 10 has and the order in which the image viewer 32 in the client computer 30 requests map data from the map server 10.

FIG. 3 is an explanatory diagram showing a file configuration of each type of map that the map server 10 has.

FIG. 4 is an explanatory diagram showing a hierarchical structure inside each of the directories 121 to 123 for each layer in the map database 11 shown in FIG.

FIG. 5 is a block diagram showing a functional configuration of an image viewer 32.

FIG. 6 is a flowchart of the operation of the display unit 320.

FIG. 7 is a flowchart of the operation of the downloader 322.

FIG. 8 is a screen transition diagram showing an example of progressive display.

FIG. 9 is a block diagram showing an embodiment in which the load is distributed among a plurality of map servers.

10 is an explanatory diagram showing a method when the client computer 70 requests a plurality of map servers 50, 50, ... For a mesh file required for a display range desired by the user. FIG.

FIG. 11 is a flowchart of the overall processing performed by the image converter according to the embodiment of the present invention.

FIG. 12 is a flowchart of mesh division processing of the image converter according to the embodiment of the present invention.

[Explanation of symbols]

10 Map Server 20 Internet 30 Client Computer 31 Client Program 32 Image Viewer 34 Mouse 131 Mesh File 132 Second Level Folder 133 First Level Folder 211, 221, 231 Mesh

Continued front page    (72) Inventor Daisuke Tabuchi             4-20-3 Ebisu, Shibuya-ku, Tokyo Ebisu             Garden Place Tower 13th floor Dreamte             Knologies Co., Ltd. (72) Inventor Ichiro Nakajima             4-20-3 Ebisu, Shibuya-ku, Tokyo Ebisu             Garden Place Tower 13th floor Dreamte             Knologies Co., Ltd. (72) Inventor Yasuo Hinata             4-20-3 Ebisu, Shibuya-ku, Tokyo Ebisu             Garden Place Tower 13th floor Dreamte             Knologies Co., Ltd. (72) Inventor Yuki Ito             4-20-3 Ebisu, Shibuya-ku, Tokyo Ebisu             Garden Place Tower 13th floor Dreamte             Knologies Co., Ltd. F term (reference) 2C032 HB03 HB31 HC24 HC32                 5B050 AA08 BA07 BA10 BA17 CA07                       CA08 EA03 EA12 EA19 FA02                       FA09 FA19 GA08                 5B057 CA12 CA17 CB12 CB16 CD05                       CE08 CE09 CH12                 5C076 AA17 AA21 AA22 AA36 BA04                       BB42 CB05

Claims (5)

[Claims] 1. A means for creating a plurality of pieces of sub-image data having stepwise different pixel sizes from the original image data, and each of the plurality of pieces of sub-image data is divided into a plurality of small-area fragment image data. And a means for outputting the large number of fragment image data for all of the plurality of pieces of sub-image data. 2. The image display system according to claim 1, wherein the large number of fragment image data for each sub-image data is stored in a folder having a hierarchical structure. 3. A step of creating a plurality of pieces of sub-image data having stepwise different pixel sizes from the original image data, and each of the plurality of pieces of sub-image data is divided into a plurality of small-area fragment image data. And a step of outputting the large number of fragment image data for all of the plurality of pieces of sub-image data. 4. A step of creating a plurality of pieces of sub-image data having stepwise different pixel sizes from the original image data, and each of the plurality of pieces of sub-image data is divided into a large number of fragment image data pieces of small areas. And a step of outputting the large number of fragment image data for all of the plurality of pieces of sub-image data, the computer program causing the computer to perform the image conversion method. 5. A plurality of pieces of sub-image data, each of which represents a common target image, and which have different pixel sizes in stages, and each of the plurality of pieces of sub-image data is fragment image data of a large number of small areas. The image data structure that is split.