JP4143341B2 - Image generation device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image generation apparatus, and more particularly to an image generation apparatus that generates an image based on input language data.
[0002]
[Prior art]
Conventionally, several techniques for providing information with high visibility by generating an image based on data described in a natural language and displaying the image have been proposed. For example, Japanese Patent Application Laid-Open No. 8-22554 describes a technique for generating an image corresponding to a sentence by extracting the sentence from a sentence described in a natural language and performing semantic analysis. In the technique described in the publication, for example, a sentence “the sun is shining red” is extracted from a sentence “the sun is shining red. The moon is also white”. , Generating an image that displays “shining” in “red”.
[0003]
[Problems to be solved by the invention]
However, in the technique described in the above publication, when there are two objects “sun” and “moon” to be displayed in the image as in the above example, the positional relationship between the two is not necessarily specified. Therefore, the “sun” and “moon” cannot be arranged at appropriate positions. When the position is specified, for example, when the direction in which the sun exists is specified as “east”, the position of “east” is fixedly set in the image. Therefore, even if you want to change the direction of the center of the image to “south” or “north” and change the position of “east” in the output image, the position of “east” in the fixed image Cannot be moved.
[0004]
On the other hand, Japanese Patent Application Laid-Open No. 11-232274 describes a technique for complementing the discontinuity of a phenomenon described in a natural language and generating a corresponding moving image (animation). However, in the technique described in the publication, attention is paid only to the time series of phenomena when generating an image, and the arrangement of the phenomena is not considered. For this reason, the display position of the display target has meaning only when the action is described in a relative relationship with other things, for example, "the operator moves in front of the whiteboard". become. Therefore, for example, there is a problem that it is not suitable for generating an image in which it is important to arrange a plurality of independent display objects at appropriate positions, such as weather information for each prefecture.
[0005]
The present invention solves the above problems and can generate an image in which an object described in data is arranged at an appropriate position based on data described in a natural language such as Japanese or English. An object is to provide an image generation apparatus.
[0006]
[Means for Solving the Problems]
In order to achieve the above object, an image generation apparatus according to the present invention analyzes a document including document type information indicating a document type and object information having position information and semantic information, and an object described in the document. An image generation apparatus for displaying a first image on a screen, analyzing a document, determining a background image on the screen based on document type information, and based on position information of the extracted object A display position determining unit that determines a position of an object to be displayed in the background image; and a display shape determining unit that determines a shape of the object to be displayed in the background image based on the extracted semantic information of the object; It is characterized by providing.
[0007]
In the image generating apparatus of the present invention, a background image is selected based on document type information included in a document and designating a display target of the document, and position information and semantic information described in the document are included in the background image. Are synthesized and output as an image. For this reason, it is possible to generate a display-type image suitable for the input document, display an image in which the positional relationship is appropriately indicated for a plurality of objects, and display information in the document (language) as visual information ( Image information). The background image may be plain including a white background.
[0008]
In the image generation apparatus of the present invention, it is preferable that the shape of the object and the generated image are made of vector graphic data. In this case, even if the image is enlarged or reduced, details and sharpness are not lost.
[0009]
In the image generation apparatus of the present invention, it is preferable that the background image determination unit determines the background image according to a background image rule that associates document type information with a background image. In this case, it becomes easy to search for a background image corresponding to the document type information.
[0010]
In the image generation device of the present invention, the display position determination unit is based on a position information mapping rule that associates the position information and one area of the plurality of areas in the background image corresponding to the document type information. Then, the position of the object may be determined, or the position of the object may be determined based on a position information mapping rule that associates the position information with the relative position coordinates in the background image.
The position where the object is arranged in the background image is determined with reference to the position information mapping rule prepared corresponding to the document type. For this reason, if the document types are different, the position information of the object described in the same expression can be arranged in different regions or positions in the background image.
[0011]
In the image generation device of the present invention, the display shape determination unit may determine the shape of the object based on a display shape mapping rule that associates the semantic information and the shape of the object corresponding to the document type information. preferable. In this case, if the document types are different, the semantic information of the object described in the same expression can be synthesized with different shapes in the background image.
[0012]
The image generation apparatus of the present invention preferably further includes a database that stores a plurality of combinations of the background image rule, the position information mapping rule, and the display shape mapping rule. In this case, the background image rule, the position information mapping rule, and the display shape mapping rule to be used can be determined by searching the database using document type information as a key.
[0013]
The image generation apparatus of the present invention has a plurality of the combinations for one document type, and the combination of the background image rule, the position information mapping rule, and the display shape mapping rule to be used based on designation from the user. Is preferably determined.
Different images can be obtained from the same document by preparing a plurality of combinations of the background image rule, the position information mapping rule, and the display shape mapping rule corresponding to one document type. At that time, the user notifies the image generation device of client information such as user preferences and user terminal attributes using HTTP request header information, CC / PP (Composite Capabilities / Preference Profiles), etc. You can specify which combination to use.
[0014]
In the image generating apparatus of the present invention, it is preferable that the document includes temporal information, and the display position and / or display shape of the displayed object changes with time. In this case, the temporal information included in the document is described as the positional information and semantic information of each object at time 1 and the positional information and semantic information of each object at time 2, for example, time 1 and time 2 In this order, each object can be animated in the background image. In addition, the semantic information of the object includes temporal information, and the display shape can be changed with time.
[0015]
In the image generation apparatus of the present invention, it is preferable that the document is described in an XML language. In this case, since the XML language can uniquely define attribute information and logical structure of data using a unique tag, it is easy to describe a document or an object.
[0016]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, with reference to the drawings, the present invention will be described in more detail based on exemplary embodiments of the present invention. The image generation apparatus 10 includes a syntax analysis unit 11, a data extraction unit 12, a display area determination unit 13, a display shape determination unit 14, a mapping rule selection unit 15, a database 17 that stores a plurality of mapping rules 16, and graphic data output Part 18. The image generation apparatus 10 inputs XML (Extensible Markup Language) data 55 as language data, converts the XML data 55 using a style sheet such as XSLT (Extensible Stylesheet Language Transformations), and generates SVG (scalable Vector Graphic) as image data. ) Output vector graphic data 56 such as data.
[0017]
The XML data 55 includes language data described in a natural language, and the language data describes one or more objects (objects) to be combined with an output image, including a pair of semantic information and position information. . The XML data 55 defines a document type for data identification. The syntax analysis unit 11 extracts a document type from the input XML data 55. The data extraction unit 12 extracts semantic information and position information of the objects included in the XML data 55. The mapping rule selection unit 15 extracts and selects the mapping rule 16 for determining the display position and display shape of the background image and the object prepared in association with the document type of the XML data 55 from the database 17. .
[0018]
The display area determination unit 13 determines the display area of the object corresponding to the position information extracted by the data extraction unit 12 with reference to the selected mapping rule 16. The display shape determination unit 14 searches the selected mapping rule 16 to determine the display shape of the object corresponding to the semantic information extracted by the data extraction unit 12. The graphic data output unit 18 creates vector graphic data 56 based on the display area determined by the display area determination unit 13 and the display shape determined by the display shape determination unit 14, and is selected based on the document type. This is output as image data in which an object is arranged in the background image.
[0019]
The mapping rule 16 includes area shape data 21 obtained by dividing a background image to be combined with an object into a plurality of areas, a position information mapping rule 22 that associates the position information of the object and the display area of the area shape data 21, and the meaning of the object. There is a semantic information mapping rule 23 that associates information with the display shape on the output screen. The database 17 stores the document type and the mapping rule 16 corresponding to the document type in association with each other. There is no need for one-to-one correspondence between the plurality of position information and the display area, and the plurality of position information may be associated with the same display area. Further, the association between the semantic information and the display shape is not necessarily 1: 1, and a plurality of semantic information may be associated with the same display shape.
[0020]
FIG. 2 shows a specific example of XML data input to the image generation apparatus of FIG. In the XML data shown in the figure, the document type is defined in the second line. As language data, two objects described in the <data> tag are described between the third line and the twelfth line. For example, a “<language data>” tag indicates a start tag of language data, a </ language data> tag indicates an end tag of language data, and language data includes <language data> to </ language data>. Described in between.
[0021]
3A and 3B show examples of mapping rules used in the document type 1. FIG. 3A shows region shape data, FIG. 3B shows position information mapping rules, and FIG. 3C shows semantic information mapping rules. In the XML data 55 of the document type 1, for example, “position 1” or “position 2” which is the position information of the object described in the <position information> tag in the <data> tag is shown in FIG. According to the conversion table, it is converted into the position of the area A of the background image shown in FIG. Further, “maru” or “shiromaru”, which is the semantic information of the object described in the <semantic information> tag, is converted into a circle according to the conversion table shown in FIG.
[0022]
FIG. 4 shows an operation procedure of the image generation apparatus of FIG. FIG. 5 shows an example in which the objects included in the XML data shown in FIG. 2 are arranged on the image according to the mapping rule shown in FIG. Hereinafter, an image generation method according to this embodiment will be described with reference to FIGS. 1 to 5.
[0023]
When the XML data 55 as shown in FIG. 2 having a pair of semantic information and position information as language data is input to the image generation apparatus 10 shown in FIG. 1, the syntax analysis unit 11 reads the XML data 55 of the input XML data 55. Parsing is performed, and the document type of the XML data 55 is acquired from the DOCTYPE declaration included in the XML data (step S1). In the XML data 55 of FIG. 2, the document type 1 is designated in the second line as the document type.
[0024]
The mapping rule selection unit 16 selects the mapping rule 16 corresponding to the document type acquired by the syntax analysis unit 11 (step S2). The data extraction unit 12 extracts an object described in the <data> tag included in the <language data> tag (step S3). In the XML data 55 of FIG. 2, the language data includes two objects, object A and object B, and the data extraction unit 12 first extracts the object A.
[0025]
The display area determination unit 13 searches the mapping rule 16 selected in step S2 to determine the display area corresponding to the position information of the object described by the <position information> tag (step S4). In the XML data 55 of FIG. 2, the display area determination unit 13 searches the position information mapping rule 22 shown in FIG. 3B using the position information “position 3” of the object A as a key, and determines the display area of the object A. The region B shown in FIG.
[0026]
The display shape determination unit 14 searches the mapping rule 16 selected in step S2 to determine the display shape corresponding to the semantic information of the object described by the <semantic information> tag (step S5). In the XML data 55 of FIG. 2, the display shape determination unit 14 searches the semantic information mapping rule 23 shown in FIG. 3C using the semantic information “Kuromaru” of the object A as a key, and determines the display shape of the object A. Decide on “●”.
[0027]
When the display shape and display position of the extracted object are determined, the data extraction unit 12 determines whether or not the end tag of the language data has appeared, that is, whether or not all the objects have been extracted from the language data (step S6). ). If there are still objects to be extracted, the process returns to step S3 to extract the objects again. In the example of the XML data 55 in FIG. 2, the process returns to step S 3, the object B is extracted, the display position relative to the position information 8, the region F in FIG. decide.
[0028]
If it is determined in step S6 that all objects have been extracted, the display shape and display area of all objects included in the XML data 55 are determined, and the graphic data output unit 18 performs mapping as data for combining images. The region shape data 21 is collected from the rule selection unit 21, the display shape of each object is collected from the display shape determination unit 14, and the display region of each object is collected from the display region determination unit 13. When the data for synthesizing the images is collected, the graphic data output unit 18 has each shape having the shape specified by the display shape at a predetermined position in the region specified by the display region on the background of the region shape data 21. Image data obtained by combining the objects is generated as vector graphic data (step S7). In the example of the XML data 55 shown in FIG. 2, as shown in FIG. 5, a vector graphic image in which ● is arranged in the region B and ☆ is arranged in the region F is generated.
[0029]
In this embodiment, vector graphic data 56 that is image data is generated based on XML data 55 described as language data. At this time, the object described in the XML data 55 is arranged in a display area determined according to the conversion rule of the positional information mapping rule 22 in a display shape determined according to the conversion rule of the semantic information mapping rule 23. By displaying an object at an appropriate position and in an appropriate display shape, it is possible to provide information with high visibility as compared to providing information in a language.
[0030]
Since the mapping rule 16 can be applied universally to XML data 55 of the same document type, image generation from the frequently updated XML data 55 can be easily performed. In addition, since position information and semantic information can be associated with display positions and display shapes in an image for each document type, position information and semantic information described in the same expression can be displayed with different display positions and A display shape can also be used. For example, when “east” is described as the position information, the right end of the image is the “east” display area in the document type 1, and the left end of the image is the “east” display area in the document type 2. Also good.
[0031]
As a specific example of the XML data 55 input to the image system 10, data having weather information for each prefecture can be considered. In this case, the document type is defined as “weather in various parts of Japan”. As mapping data 16 corresponding to “weathers in various parts of Japan”, area shape data 21 obtained by dividing a Japanese map as a background image into units of prefectures, and a location information mapping rule that associates prefecture names with areas obtained by dividing a map of Japan 22 and a semantic information mapping rule 23 that associates semantic information such as sunny and rain with shape data such as sunny and rain marks are prepared.
[0032]
For example, the object in the XML data 55 includes a pair of location information specifying a place name such as “Tokyo” or “Osaka” and semantic information specifying a weather information such as “sunny” or “rain”. It is described as. The image generation apparatus 10 extracts a pair of position information and semantic information from the input XML data 55, and outputs vector graphic data 56 obtained by synthesizing weather for each prefecture on a map of Japan as a background. Weather information from all over Japan is displayed. When the weather information is updated, new weather information can be displayed as visual information using the same mapping rule 16 as before the update. When XML data 55 in which “Kanto weather” is specified as a different document type is input, the mapping rule 16 prepared for the document type “Kanto weather” is used to display the map data in the Kanto region. In addition, the vector graphic data 56 in which the weather in the Kanto region is combined is output.
[0033]
In the above embodiment, an example in which three rules of area shape data, position information mapping rule 22 and semantic information mapping rule 23 are prepared for each document type of XML data 55 has been described. Rules can be shared by multiple document types. FIG. 6 shows an example in which a rule is shared by a plurality of document types. The mapping rule 16 designates the area shape data 21, the position information mapping rule 22, and the semantic information mapping rule 23 of A or B for each document type of the XML data 55. In this case, when the rules to be used overlap, the rules can be shared between the XML data 55 of different document types. Further, the overlapping portion is divided like the location information mapping rule C and the semantic information mapping rule C, and these are used as a common portion in combination with the location information mapping rule 22 and the semantic information mapping rule 23 of A or B. You can also
[0034]
In the area shape data 21, the output plane is divided into a plurality of adjacent display areas. However, the display areas may overlap each other and belong to any display area in the output plane. The parts that don't fit may be right. The display position of the display shape determined from the semantic information may be the center of gravity of the display area, or may be an arbitrary place determined from the display area. When a plurality of objects are arranged in the same area, the objects can be arranged at different positions in the area. The area shape data 21 may be directly associated with the coordinates on the output plane instead of associating the position information with the display area. Further, the output plane can be regarded as a three-dimensional virtual space, and for example, a country name (position information) can be mapped to a country (display area) on the globe to generate a spatial image.
[0035]
The XML data 55 can also include information about time. In this case, for example, time information described in a <time information> tag can be added to each object. The time information is described as “Today”, “Tomorrow”, or “Tomorrow”. The image generation device 10 generates a moving image that displays objects having the same time information on the same image. For example, when data describing the weather information of “Today” and “Tomorrow” throughout the country is input as the XML data 55 data, an animation for alternately switching between today's weather and tomorrow's weather An image can be generated.
[0036]
In addition to or instead of adding <time information> to the XML data 55 described above, a temporal element such as “cloudy when sunny” or “cloudy after sunny” is included as semantic information of the object. It can also be described. In this case, the display shape determination unit 14 prepares a clear mark and a cloudy mark as display shapes corresponding to the semantic information, and generates a moving image based on time information such as “sometimes” or “after”. For example, if “sunny and cloudy” is described as the semantic information, an animation image that switches between a sunny mark and a cloudy mark can be generated.
[0037]
Moreover, although the example which switches the mapping rule 16 by the document type of the XML data 55 was shown in the said embodiment example, the several mapping rule 16 can also be prepared with respect to the same document type. In this case, the mapping rule selection unit 15 selects the mapping rule 16 from a plurality of prepared for the same document type according to the attribute of the terminal that outputs the vector graphic image 56 and the user's preference.
[0038]
As a method for notifying the image generation apparatus 10 of client information such as terminal attributes and user preferences, a method using HTTP request header information, CC / PP (Composite Capabilities / Preference Profiles), or the like is applied. it can. For example, in the case of using CC / PP, depending on the CC / PP determined from the client information, the mapping rule selection unit 15 selects the mapping rule 16 to be used from among a plurality prepared for the same document type. By doing so, an image suitable for the terminal attribute can be generated. For this reason, XML data 55 of the same document type can be displayed in different display shapes, display positions, etc., and more effective information can be provided.
[0039]
As described above, the present invention has been described based on the preferred embodiment. However, the image generation apparatus of the present invention is not limited to the above embodiment, and various modifications and changes can be made to the configuration of the above embodiment. An image generating apparatus that has been modified is also included in the scope of the present invention. For example, the background image may be plain such as white.
[0040]
【The invention's effect】
As described above, the image generation apparatus of the present invention has a background image determined by document type information included in a document described in a natural language, a display shape determined by semantic information at a display position indicated by position information. Generate an image that combines objects. For this reason, it is possible to provide highly visible information.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of an image generation apparatus from language information according to an embodiment of the present invention.
FIG. 2 is a description example showing an example of XML data input to the image generation apparatus of FIG. 1;
3A and 3B show examples of mapping rules, where FIG. 3A is a plan view showing a state of area division, FIG. 3B is a table associating position information with display areas, and FIG. 3C is associated with semantic information and display shape doors. Each table is shown.
FIG. 4 is a flowchart showing a method for generating an image from language information according to an embodiment of the present invention.
5 is a plan view showing an image generated from the XML data shown in FIG. 2. FIG.
FIG. 6 is a block diagram showing how a mapping rule is shared by a plurality of document types.
[Explanation of symbols]
10: Image generation device 11: Syntax analysis unit 12: Data extraction unit 13: Display region determination unit 14: Display shape determination unit 15: Mapping rule selection unit 16: Mapping rule 18: Graphic data output unit 21: Region shape data 22: Position information mapping rule 23: Semantic information mapping rule 55: XML data 56: Vector graphic data

Claims (10)

Analyzing a document including document type information indicating a document type and object information having position information and semantic information described in a natural language, and generating an image for displaying an image of the object described in the document on the screen A device,
A background image determination unit that analyzes a document and determines a background image on a screen based on document type information;
In the background image , based on the position information of the object extracted from the document and the position information mapping rule corresponding to the document type information and defining the correspondence between the position information and the display position in the background image A display position determination unit for determining the position of an object to be displayed;
An image generation apparatus comprising: a display shape determination unit that determines a shape of an object to be displayed in the background image based on semantic information of the object extracted from the document.   The image generation apparatus according to claim 1, wherein the shape of the object and the generated image are made of vector graphic data.   The image generation apparatus according to claim 1, wherein the background image determination unit determines the background image according to a background image rule that associates document type information with a background image.   The display position determination unit determines the position of the object based on a position information mapping rule that associates the position information corresponding to the document type information with one of a plurality of areas in the background image. The image generation apparatus according to claim 3. The said display position determination part determines the position of the said object based on the positional information mapping rule which links | relates the said positional information and the positional coordinate in the said background image corresponding to the said document type information. Image generation device.   The image according to claim 4 or 5, wherein the display shape determination unit determines the shape of the object based on a display shape mapping rule that associates the semantic information and the shape of the object corresponding to the document type information. Generator.   The image generation apparatus according to claim 6, further comprising a database that stores a plurality of combinations of the background image rule, the position information mapping rule, and the display shape mapping rule.   The combination of the background image rule, the position information mapping rule, and the display shape mapping rule to be used is determined based on designation from the user, wherein the combination includes a plurality of combinations for one document type. The image generating apparatus described.   The image generating apparatus according to claim 1, wherein the document includes temporal information, and a display position and / or a display shape of the displayed object changes temporally.   The image generation apparatus according to claim 1, wherein the document is described in an XML language.

Images