CN1838112B - translation device, translation method - Google Patents

Abstract

A translation device, comprising: a character recognition unit for recognizing text data in a text region of an input image; a translator for translating the text data of the text region; and a layout structure processor for generating data containing the translated text data and graphics in the input image, wherein the layout of the input image is maintained in the image layout of the data generated by the layout structure processor.

Description

translation device, translation method

technical field

The present invention relates to a technique for generating translation data after translating text contained in a document from one language to another. the

Background technique

Various types of translation devices have been proposed which receive image data of a document with or without a picture, translate text contained in a text area of the image data, and generate a translated document containing the translated text or generate a document containing the translated text. Documentation of post text and original graphics. the

It is known to propose a technique of using layout analysis to distinguish text areas and graphics of input data, and to recognize characters in the text areas for translation. Then the obtained translated text volume is compared with the size of the existing text area, so that the text area can be re-formed according to the comparison result. However, as a result of reformatting the text area, if the graphics area can no longer be allocated on the same page, the graphics area is allocated in the next page. Therefore, it may be difficult for readers to read the translated document due to the changed allocation of text area and graphics area. the

In addition, since commonly used translation devices output the original document and the translated document in different regions of the same page or in different pages, it is often difficult for users to find the correspondence between the original text and the translated text. It is known to propose a technique of arranging translated text between lines of an original text, thereby reducing the trouble caused by the user to find a correspondence between the original text and the translated text. the

However, the number and type of characters contained in the translated text will be different from those in the original text; as a result, the length of the character string contained in a line of translated text is not equal to the length occupied by a line of original text. the

Contents of the invention

The present invention has been made in view of the above circumstances, and provides a system for generating translation data including a translated text portion and an original graphic portion while preserving the original layout of a text area and a graphic area. Furthermore, the present invention provides a system for enabling translation data to be generated, and by using the system, a user can easily associate an original document with a translated document, improving viewability. the

In one aspect, the present invention provides a translation device, comprising: a character recognition unit for recognizing text data in a text area of an input image; a translator for translating the text data in the text area; and a layout structure processor for for generating data including translated text data of a text area and graphics in an input image by changing the character size of the translated text data, wherein the layout of the input image is maintained in the image layout of the data generated by the layout structure processor, wherein The layout structure processor generates the data by changing the size of the graphics area. the

According to an embodiment of the present invention, it is possible to generate translated text and graphic data whose text parts are translated while maintaining the layout of the text area and the graphic area of the input data. the

In another aspect, the present invention provides a translation method comprising the steps of: a character recognition step of recognizing text data in a text area of an input image; a translation step of translating the text data in the text area; and a layout structure a processing step of generating data including the translated text data of the text area and graphics in the input image by changing the character size of the translated text data, whereby the image layout of the generated data maintains the layout of the input image, wherein by changing the graphics area to generate the data. the

According to the embodiment of the present invention, since the character strings of the translated text and the character strings of the original text have the same length and are placed in parallel, the user can easily make the correspondence between the original text and the translated text. As a result, readability is greatly improved. the

Description of drawings

Embodiments of the present invention will be described in detail below according to the accompanying drawings, in the accompanying drawings:

Fig. 1 is a block diagram showing the structure of a translation device according to a first embodiment of the present invention;

Fig. 2 is a flowchart showing the process of the document translation program according to the first embodiment;

The diagram of Fig. 3 shows a part of the process of the document translation program according to the first embodiment;

The figure of Fig. 4A to 4D has described the effect of the first embodiment compared with conventional technology;

Fig. 5 is a flowchart showing the process of a document translation program according to a second embodiment of the present invention;

The diagram of Figure 6 describes processing according to another embodiment;

The diagram of Figure 7 describes processing according to another embodiment;

Fig. 8 is a block diagram showing an image generating device according to a third embodiment of the present invention;

9 is a block diagram showing the structure of a translation processing unit in an image generating device according to a third embodiment;

Fig. 10 is a flowchart showing the processing performed by the translation processing unit;

The diagrams of Figures 11A and 11B show images illustrating the processing performed in the translation processing unit; and

12A and 12B are diagrams showing character strings illustrating processing of correcting the length of character strings (ie, processing performed in the translation processing unit). the

Detailed ways

Hereinafter, embodiments of the present invention will be described with reference to the drawings. the

first embodiment

FIG. 1 is a block diagram showing the basic structure of a translation apparatus according to a first embodiment of the present invention. The structure of the translation device is similar to a multifunction machine that combines scanning functions, copying functions, printing functions, and facsimile functions. This translation device has an image reading device 1 with an ADF (automatic document feeder), a printing device 2, a communication interface 3, a display unit 4, an operation unit 5, a volatile memory 6, a nonvolatile memory 7, and control CPU8 for each unit. Under the control of the CPU 8, multiple functions are realized in the translation device: the printing device 2 prints the image read by the image reading device 1 to realize the copying function, and the image is transmitted to the corresponding device through the communication interface 3 and the network. The fax machine to realize the fax function and so on. the

In the non-volatile memory 7, a dedicated program of this embodiment and a control program for making the CPU 8 perform control to realize various functions provided by this type of multifunctional device are stored. document translation program for translating text contained in received images into another language, and the specialized program generates for output translation data with graphics containing translations in place of images representing original text post text data. A typical process performed by a document translation program with pictures is shown in the flowchart of FIG. 2 . In order to avoid redundant descriptions, details of this process will be described later in this embodiment when a description of the operation of this embodiment is provided.

The translation technique according to the present embodiment has such a structure that a document translation program with pictures can be realized in the following manner:

a. In the image reading device 1, an image of a document containing text and graphics written in one language is read, and the image is stored in the volatile memory 6 so that the image can be translated using a document translation program with pictures The processing is performed and the translated text and graphic data obtained through the processing are output as images using the printing device 2 . Alternatively, using the communication interface 3 to transmit the translated text and graphic data obtained through the process to the user who needs the translated data by fax or email; and

b. Receive an image of a document containing text and graphics written in a language via the communication interface 3 and store the image in the volatile memory 6 so that it can be processed using a document translation program with pictures and printed using The device 2 outputs the translated text and graphic data obtained through this processing as an image. Alternatively, the translated text and graphic data obtained through this process are transmitted to users who need the translated data by fax or e-mail using the communication interface 3 . the

In the nonvolatile memory 7 is stored a control program that enables transmission of images, which are input information of the document translation program with pictures, and translated text and graphic data, which are output information of the document translation program with pictures. Through commands provided via the operating unit 5 or the communication interface 3 , detailed instructions are provided to cause the control program to perform information transmission. the

Next, a description will be given of the operation of this embodiment. Once the image to be processed with one or more pages is input through the image reading device 1 and the communication interface 3, it is then stored in the volatile memory 6, and the CPU 8 executes the document translation program with pictures, the flow of which is shown in FIG. 2 in. In a layout analysis process 101 , which is the first process of the document translation program with pictures, the CPU 8 analyzes the layout of the input image stored in the volatile memory 6 . Then CPU8 obtains text area 201, graphic area 202 and margin area 203 as shown in FIG. At this stage, the text area 201 and graphics area 202 are empty. the

The CPU 8 then sequentially performs character recognition processing 102, translation processing 103, and text amount calculation processing 104. In the character recognition process 102, character recognition is performed on an image included in the text area 201 of the input image 200, thereby generating text area text data. The text area text data contains information on the type of recognized characters and setting information such as character size, line spacing, page margin, and the like. In the character recognition process 102, the image contained in the graphic area 202 is also processed to generate graphic area text data. The graphic area text data contains information about the type, position and size of the characters contained in the graphic area 202 . the

In the translation process 103, the text data in the text area and the text data in the graphics area are translated into another language to generate the translated text data 204 in the text area and the translated text data 205 in the graphics area, which are respectively assigned to the stored A text area 201 and a graphics area 202 for translated text and graphics data in the work area. The translated text data 204 of the text area contains information about the type of characters forming the translated text and setting information inherited from the original text area text data; Information of the type of the character and information indicating the position and size of the character inherited from the original graphics area text data. The language of the text contained in the input image 200 and the language of the translated text are respectively designated by a command provided through the operation unit 5 or the communication interface 3, and translation is performed in the translation process 103 according to the command. the

In FIG. 2, the grouping by dashed lines forms a layout structure processor 500 for generating translated text and graphics data containing translated text data 204 in its text area and translated text data 204 in its graphics area. Contains the translated text data 205 of the graphics area and the data in the graphics area 202 of the input image 200 . In the text amount calculation process 104 , the data amount of the translated text represented by the text area translated text data 204 is calculated and stored in the nonvolatile memory 7 . The sizes (ie, areas) of the text area 201 and the graphics area 202 are calculated in area size calculation processing 105 and stored in the volatile memory 6 . The area size calculation process 105 may be performed at any time after the layout analysis process 101 and before the text amount/area size comparison process 106 . the

In the text amount/area size comparison process 106, the character amount of the translated text obtained by the text amount calculation process 104 is compared with the size of the text area 201 obtained by the area size calculation process 105, and the comparison result is stored in the volatile Sex memory 6. Specifically, in the text amount/area size comparison process 106, the area occupied by an image obtained by imaging the translated text represented by the text area translated text data 204 and the size of the text area 201 to accommodate the image are calculated The ratio between, and the calculated ratio is stored in the volatile memory 6. the

The CPU 8 performs text size scaling processing 107 or graphics scaling processing 108 based on the result obtained by performing text size/area size comparison processing 106 . The text size scaling process 107 controls the setting of the translated text data 204 as a setting control means for the text area in the translated text and graphic data, thereby inheriting the layout of the input image 200 as the text area of the translated text and graphic data 201 and the layout of the graphics area 202. Specifically, in the text size scaling process 107, the ratio of the area occupied by the translated text to the size of the text area (this ratio has been obtained in the text volume/area size comparison process 106) is used to calculate the size of the translated text. The size of each character is such that when the translated text data 204 of the text area is imaged, the image can be accommodated in the text area 201 . the

When it is determined that the image of the translated text can be accommodated in the text area 201 without changing the character size if the size of the graphics area 202 is reduced (or enlarged), the graphics scaling process 108 is performed to enlarge (or expand) within the maximum allowable limit. shrink) the size of the text area 201. This determination is made based on the ratio obtained in the text amount/area size comparison process 106 . In the graphics scaling process 108, the size of the text area 201 required to accommodate the image obtained by imaging the data 204 is obtained without changing settings such as the size of characters forming the translated text data 204 of the text area. In addition, in the case of changing the size of the text area 201 according to the obtained size, the size of the graphics area 202 is also obtained, so that the text area 201 and the graphics area 202 in the same page of the input image 200 can still be accommodated in the translated text on the same page as the graph data. Then, the scaling factor of the graphics is obtained, and the scaling factor is a ratio of the size before and after the change. the

In the reconstruction process 109 , the translated text and graphics data are reconstructed using the results obtained from the text size scaling process 107 and the graphic scaling process 108 , and the data are stored in the work area of the volatile memory 6 . the

In the case that the text size scaling process 107 is performed but the graphics scaling process 108 is not performed, in the reconstruction process 109, the image of the graphics area 202 of the input image 200 that does not include the text image is first stored in the work area. Graphics data in graphics area 202. Next, the graphic area translated text data 205 obtained from the graphic area 202 of the input image 200 is stored in the graphic area 202 of the translated text and graphic data. When the data stored in the graphics area 202 is reproduced as an image, the translated text in the graphics area 202 will have the same size and occupy the same position as the original text of the input image 200. Subsequently, the text area translated text data 204 obtained from the text area 201 of the input image 200 is stored in the text area 201 of translated text and graphics data. The information specifying the character size contained in the text area translated text data 204 shows the size after the text size scaling process 107 is applied. As a result, when the text area translated text data 204 is imaged, the image fits the text area 201 exactly. the

On the other hand, in the reconstruction processing 109, in the case where the graphics scaling processing 108 is performed instead of the text size scaling processing 107, the text area of the translated text and graphics data is changed according to the result obtained by executing the graphics scaling processing 108 201 and graphics area 202. According to the scaling factor obtained by the graphic scaling process 108, the image of the graphic area 202 of the input image 200 after removing the character image is enlarged or reduced to be stored in the graphic area 202 of the translated text and graphic data. the

Also, the information indicating the character size contained in the graphics area translated text data 205 is modified to show a value obtained by multiplying the scaling factor obtained by the graphics scaling processing 108 . Then, according to the scaling factor, the information of the character position in the translated text data 205 in the display graphics area is modified. This modification is made so that when the graphics area translated text data 205 is imaged, in the graphics area 202, the images of the characters of the translated text occupy the same positions as the original characters of the input image. Next, the translated text data 204 of the text area is stored in the text area 201 of the translated text and graphic data. The text size scaling process 107 is not performed on information specifying the character size contained in the text area translated text data 204 , but the text area 201 is changed by the graphics scaling process 108 . As a result, when the text area translated text data 204 is imaged, the image of the translated text fits accurately in the text area 201 . the

Therefore, the translated text and graphics data are stored in the work area of the volatile memory 6 after the translated text replaces the text contained in the original input image. Then, the translated text and graphic data are printed on recording paper through the printing device 2, or transmitted to external users who need the translation result through the communication interface 3. the

4A to 4B are diagrams illustrating the effects of the present invention in comparison with conventional techniques. In this example, as shown in FIG. 4A, an input image of a document with pictures in a foreign language is stored in the volatile memory 6, and the text portion of the input image is translated into Japanese. the

When using the conventional technique, in the case where the translated text of the character string in the text area of the input image can no longer be accommodated in the original text area, as shown in Figure 4B, the translated text is placed in the enlarged text area, the result graphics area is moved to the next page of the original page. As a result, in translated text and graphics output using conventional techniques, the layout of the text and graphics areas looks very different from the original foreign-language document with the input image. This makes the translated text difficult to read. the

On the contrary, in this embodiment, the translated text and graphics as shown in FIG. 4C or 4D are obtained. Figure 4C shows the translated text and graphics obtained by performing the text size scaling process 107 without the graphics scaling process 108; Translated text and graphics. As shown in these figures, the layout of the text area and the graphics area of the translated text and graphics obtained according to the present embodiment is the same as that of the input image (FIG. 4C) or not very different from the layout of the input image (FIG. 4D). Therefore, translated text and graphics obtained according to the present embodiment can be easily understood by users when compared with translated text and graphics using conventional techniques. the

Second embodiment

The first embodiment is also effective for a case where an input image having a plurality of pages is to be processed. In the case where an input image having multiple pages is to be processed, according to the first embodiment, while maintaining the text area, graphic area, and margin of each page as much as possible, translated text and graphic data are generated and translated from the input image The translated text data is stored in the text area of each page of the translated text and graphics data in the text area of the same page as the input image. However, when this rule is strictly applied, it may cause the translated text density in the text area of each page of translated text and graphic data to be inconsistent from page to page. In this embodiment, the translated text data of the text area can be transferred between pages within the maximum allowable limit to reduce the inconsistency of the translated text density. In other words, for example, in the case where the text volume of the translated text is large relative to the size of the text area on a particular page, but the text volume of the translated text on the next page is small relative to the size of its text area, the Strings in the last part (which are unlikely to fit in the previous page) are passed to the next page. Conversely, in the case where the text volume of the translated text is small relative to the size of the text area of a particular page, but the text volume of the translated text on the next page is large relative to the size of its text area, the The string in the front section of the is sent to the previous page. the

Inconsistencies in the density of the translated text between different pages are thus reduced; however, this creates another problem. That is, when a character string transmitted to another page contains a label of a graphic contained in the page where the transmitted character string was previously located, the user must troublesomely turn to the previous page to confirm that the text containing the label Referenced graphics. This embodiment also prevents such inconvenience to the user. the

FIG. 5 is a flowchart showing the procedure of the document translation program with pictures according to the present embodiment. In this figure, processes corresponding to those shown in FIG. 2 are given the same reference numerals. In the document translation program according to the present embodiment, the same figure number search process 110 is also provided, and, according to the result obtained by executing the same figure number search process 110, it is determined whether to carry out the text amount grouped together by dashes. Calculation processing 104, text amount/area size comparison processing 106, text size scaling processing 107, and graphics scaling processing 108 (hereinafter referred to as group processing 111 for convenience). In other words, the same figure label search process 110 includes: translated text data transfer control means for transferring the translated text data of the text area within the maximum allowable limit between the text areas of different pages of the translated text and graphic data ; and transfer control/setting control switching means for enabling the setting control means to control the setting of the page in question in the following cases, instead of transmitting the translated text data by the translated text data transfer control means: a) graphics for identifying graphics The identification information is included in the translated text data in the graphic area; b) before the translated text data is transmitted by the translated text data transmission control means, it is included in the same page as the graphic identification information contained in the translated text data in the graphic area The text data after the translation of the text area includes the same graphic identification information as the graphic identification information contained in the text data after the translation of the graphic area; and c) if the text data after translation is transmitted by the text data transmission control means after translation, then Graphical identification information will be transferred to different pages. Hereinafter, this will be described in detail. the

The same figure number search process 110 is executed when it is possible to transfer the character string of the translated text between different pages. Specifically, the same figure number search process 110 is performed in each page from which a specific character string may overflow to another page, and it is determined whether such overflow is permitted. the

In the case where a character string flowing from a text area of a particular page does not contain any graphic identification information such as a label and a title, or even if it contains such graphic identification information, it does not correspond to graphic identification information in a graphic area of the same page Next, allow strings to flow like this into another page. Conversely, in the case where a character string flowing out from a text area of a certain page contains graphic identification information and it corresponds to graphic identification information in a graphic area of the same page, the character string is not allowed to flow to another page. the

In the same figure number search process 110, the above determination is made for each page from or to which character strings of translated text are transferred when transferring between different pages to determine whether transfer of these characters is permitted. The transmission of these characters is performed only for allowed pages, and the translated text to be assigned to the text area of each page is determined. As a result, the text areas of some pages may be filled to overflowing because character strings are not allowed to flow out to another page, or the text areas of some pages may contain blanks because character strings are not allowed to flow in from another page. In the same figure number search process 110, it is determined that group processing 111 should be performed on some of such pages and that group processing 111 should not be performed on other pages. the

The group processing 111 is executed only when the above-described identical figure number search processing 110 determines that the group processing 111 needs to be executed. The details of the group processing 111 and the reconfiguration processing were described in the first embodiment, so repeated explanations will be omitted. the

According to this embodiment, the text size is only done if it is preferable to have the label in the same page as the figure it refers to, and the figure label in the text area may be transferred to another page due to translation Scaling process 107 and figure scaling process 108 to prevent transfer of a label to another page where the figure referenced by the label is not located. Therefore, since the user does not have to refer to different pages when reading text with reference to graphics, and since the size of characters and/or graphics is not changed except when necessary, the user will easily read the translated text and graphics obtained according to the present invention. the

other embodiments

The first embodiment and the second embodiment have been described above, but the present invention can also have the following embodiments. the

1) In the first and second embodiments, the line spacing can be changed so that the translated text fits in the text area without performing the text size scaling process 107 . the

2) In the first and second embodiments, the size of the text area can be adjusted by enlarging or reducing the size of the margin so that the translated text can be accommodated therein.

3) Both graphics scaling processing 108 and text size scaling processing 107 can be performed to reconstruct the layout. Some patterns can be envisioned in this context. In a first mode, upper and lower bounds for the character size scaling factor may be provided, and the text size scaling process 107 occurs within the provided limits. In the case where the image of the translated text in the text area cannot be accommodated in the same page as the original page even if the character size is reduced or enlarged according to the scaling factor within the limit, the graphics scaling process 108 is further performed to reduce or enlarge the image by reducing or enlarging Graphics area to reserve the area required to accommodate the image of the translated text data in the text area. In a second mode, upper and lower bounds for the graphics area scaling factor are provided, and the graphics scaling process 108 occurs within the provided limits. In the case where the image of the translated text data in the text area cannot be accommodated in the same page as the original page even after the graphics area is reduced or enlarged according to the scaling factor within the limit, the text size scaling process 107 is performed so that the text area An image of translated text data can be accommodated in the text area. the

4) In some cases, the user can designate so-called "N-Up" printing in which images of N pages are output on one page as a combined image. In this case, the same figure number search process 110 may make the following determinations. A description will be given below of an example of "2-Up" printing. the

As shown in FIG. 6, assume a case where translated text and graphics data 400-1, which is 2-Up mode data of input images 200-1 and 200-2 to be output on the same page, will be generated. In the example shown in FIG. 6 , the input image 200 - 1 contains in its text area 201 graphic identification information corresponding to the graphic identification information contained in its graphic area 202 . In the translated text and graphic data 400-1, in some cases, the number of characters is changed due to translation, and the translation of the graphic identification information "Fig. 1" contained in the text area 201 of the page corresponding to the input image 200-1 may not fit in the same text area 201. In this case, since the figure identified by the figure identification information "Fig. 1" exists in the same page, translation of "Fig. 1" from the one corresponding to the input image 200-1 is allowed in the same figure label search process 110. The text area of the page is transferred to the text area 201 of the page corresponding to the input image 200-2. Furthermore, in the same figure number search process 110, it is determined whether or not it is necessary to perform group processing 111 on the page containing the translation "FIG. 1" first. the

As shown in FIG. 7 , assume a case where translated text data 400-1 (which causes data corresponding to input images 200-1 and 200-2 to be output in the same page in 2-Up mode) and translated text data 400-1 will be generated. Post text data 400-2 (which causes data corresponding to the input images 200-3 and 200-4 to be output in the same page in 2-Up mode). In the example shown in FIG. 7, for the input images 200-1 and 200-2, the graphic identification information corresponding to the graphic identification information existing in the graphic area 202 of the input image 200-1 exists in the text of the input image 200-2. District 201. In the translated text and graphics data 400-1, in some cases, since the number of characters is changed by the translation process, if the character sizes are the same, the text area 201 of the page corresponding to the input image 200-2 will be accommodated. The translation of the graphic identification information "Figure 1" cannot be accommodated in the text area 201. If the translation of the graphic identification information "Fig. 1" is allowed to be transferred to the next page (i.e., the page corresponding to the input image 200-3), the graphic identified by the graphic identification information "Fig. 1" will not be compatible with the graphic identification information exist on the same page. In order to avoid such a result, in the same figure number search process 110, the transfer of a character string is not permitted, and it is determined that the group process 111 should be performed on the data corresponding to the input image 200-2. According to this embodiment, the same effects as those of the second embodiment described above can be obtained. the

5) In each of the above-mentioned embodiments, the translated text data is stored as text data in the text area and the graphic area of the translated text and graphic data, so that the translated text and graphic data are generated, and the data has both text data and images data. Alternatively, the translated text data may be imaged to map the imaged translated text to a text area and a graphics area, thereby generating translated text and graphics data, the data comprising the entire image data. the

6) In each of the above-mentioned embodiments, a picture-to-file translation program is installed in the multi-function machine. However, the embodiments of the present invention are not limited to this mode, but the document translation program of the above-described embodiments may be stored in a computer-readable recording medium such as a CD-ROM for distribution to general users. Alternatively, the document translation program with pictures can be sent to ordinary users through the network. the

third embodiment

A description will be given below of a third embodiment of the present invention with reference to the drawings. the

FIG. 8 is a block diagram showing an image forming apparatus 100 according to an embodiment of the present invention. As shown in the figure, the image forming apparatus 100 has a translation processing unit 1 , an operation unit 2 , a network I/F unit 3 , a memory storage unit 4 , a printing unit 5 , and an image reading unit 6 . the

The printing unit 5 has a photoreceptor, an exposure unit, a transport unit, and a fixing unit. The printing unit 5 generates a toner image based on the image data supplied from the translation processing unit 1 to fix the toner image on a sheet of paper which is a recording medium. The operation unit 2 has a display unit formed of a liquid crystal display (not shown) and various buttons so that instructions from the user can be received. The user selects paper to be used, inputs various print settings, and the like using the operation unit 2 . the

The image reading unit 6 scans the data of the source document, and outputs it as image data. The memory storage unit 4 stores image data scanned by the image reading unit 6 and other data. Data communication between the translation processing unit 1 , the operation unit 2 , the memory storage unit 4 , the printing unit 5 , and the image reading unit 6 is enabled through the network I/F unit 3 . the

As shown in FIG. 9 , the translation processing unit 1 includes a CPU (Central Processing Unit) 11 , a RAM (Random Access Memory) 12 , and a ROM (Read Only Memory) 13 . The translation processing unit 1 controls each unit of the image forming apparatus 100 , and performs each image processing and calculation necessary for translation processing of input image data. Image data is temporarily stored in RAM 12 to be processed. Various image processing programs PRG and translation programs PRG necessary for image data processing and translation processing are stored in the ROM 13 . the

Next, an example of operations performed by the image forming apparatus 100 having the above-described structure will be given based on the flowchart shown in FIG. 10 . In this example, the image of the source document is read by the image forming apparatus 100, and Japanese contained in the image is translated into English. the

A source document is placed on a scanning plate (not shown) of the image reading unit 6, and the image reading unit 6 is caused to start reading the document. As a result, the image reading unit 6 reads the image existing in the scanning area (step S01). the

FIG. 11A shows an example of an image read by the image reading unit 6 . As shown in the figure, the scanned image G is an image including an original text J (including a portion indicated by a circle in the figure) and a graphic Z including Japanese. the

Referring to the display screen of the operation unit 2 showing an image such as that shown in FIG. 11A, the user operates an input device such as a mouse or a keyboard to move a pointer or cursor to designate an area of the scanned image G for translation (step S02). In this example, the entire area of the image G (the entirety of the image G shown in FIG. 11A ) is designated as the translation area. the

When the user specifies the entire area as the translation area, the translation processing unit 1 recognizes the text of the image data in the entire area specified in the image G. The recognized text is extracted from all image data (step S03). the

The translation processing unit 1 then performs OCR (Optical Character Reader) processing on the image data contained in the extracted text area, converts the image data of the text in the area into text data, and reads the converted text The data is used as text data of the original text J (step S04). the

The CPU 11 of the translation processing unit 1 then obtains the language information stored in the ROM 13 and compares the language information with the text data of the original text J to recognize the language of the original text J (step S05). In this example, after comparing the text with the language information in the ROM 13, the CPU 11 recognizes that the language of the original text J is Japanese. the

Further, the CPU 11 of the translation processing unit 1 runs and executes the translation program from Japanese to English from the translation processing program PGM stored in the ROM 13, thereby generating text data of the translated text E in English (step S06). the

After generating the translated text E from the original text J, the translation processing unit 1 compares the lengths of the character strings of the translated data E and the original text J (step S07). the

In this embodiment, determine the horizontal length of the image of the semantic part (such as a clause or a whole sentence) of the original text J as the length of the character string; The length obtained by summing the points of the spaces between the characters is determined as the length of the character string. The determined length of the character string of the translated text E is compared with that of the original text J. Alternatively, the character string length of the original text J may be determined as the sum of the points of the characters in the horizontal direction of the semantic part. the

The translation processing unit 1 determines appropriate processing from correction processing such as changing the font and/or point, equal spacing, etc., as a result of the comparison between the character string lengths of the original text J and the translated text E, as to be applied to the translated text E. Correction processing of text E. The translation processing unit 1 then performs the determined correction processing to make the character string length of the translated text E equal to the length of the original text J (step S08). the

It should be noted that when the character string length of the translated text E falls within a predetermined range of points (visually, the user can regard it as the same), the translation processing unit 1 compares the character string lengths of the translated text E and the original text J determined to be equal to each other. the

FIG. 12A shows a character string of the Japanese original text J and a character string of the translated text E. FIG. As shown, the string of the translated text E is longer than the string of the original text J. In this case, the translation processing unit 1 changes the font and reduces the number of points of the character size according to the comparison result of the character string length between the original text J and the translated text E as correction processing. The translation processing unit 1 changes the font and character size points to reduce the character string length, thereby changing the character string length of the translated text E to within the same predetermined point range as the character string length of the original text J as shown in FIG. 12B . the

Next, the translation processing unit 1 adds the text data of the translated text E that has been corrected to the image data of the entire image G, so that the translated text E whose character string length is corrected is parallel to and located below the original text J (step S09). the

When the user has input a printing instruction through the operation unit 2, the image data processed in the translation processing unit 1 is output to the printing unit 5, and in the printing unit 5, an image representing the image data is printed on paper (step S10 ). the

As a result, as shown in FIG. 11B , the translated text E in English (shown with an X mark in the figure) of an appropriate character string length is located below the existing Japanese original text J in the entire portion of the limited range. the

Therefore, according to this embodiment, by changing the points and fonts of the characters of the translated text E, the translated text E can be placed parallel to the original text J with the same character string length as that of the original text J. As a result, the correspondence between the original text J and the translated text E can be easily recognized based on the printouts representing the translated text E and the original text J placed in this manner; thus, the viewability of the printouts is greatly improved sex. the

Furthermore, since the text data in the image data is taken out before translation, only the characters or text of the image including graphics can be surely translated. the

In the above-mentioned embodiments, an explanation has been given of the case where the character string of the translated text is longer than that of the original text. In the case where the character string of the translated text is shorter than the original text, the translation processing unit 1 performs the same processing as above to arrange the translated text at equal intervals relative to the original text so that the length of the translated text is equal to that of the original text , and the translated and original texts are placed parallel to each other. the

Furthermore, in the above-described embodiment, since the length of the character string is corrected by changing the points or font of the characters in the translated text, it is alternatively possible to change the characters of the translated text into ruby-type characters to correct the string length of the translated text. the

In the above-described embodiments, a description has been given of an example of the process of adapting the length of a character string to an original text. Obviously, the character string length of the original text or the character string lengths of the original text and the translated text may be changed to have lengths equal to each other. In the case of changing the character string length of the original text, after converting the original text into characters, the font or point of the characters may be changed, or the converted characters may be arranged at equal intervals relative to the translated text. Alternatively, the size of the image data of the original text may be enlarged or reduced for adjustment. the

As long as the number of points in the lateral direction of characters and in the space between characters is predetermined, the length of the character string can be determined not only by summing the number of points but also by using the number of characters. Alternatively, the length of the character string may be determined according to the length (mm) of the printout. the

Obviously, the languages of the original text and the translated text are not limited to those described in the above embodiments, and may be German, French, Russian, Spanish, Chinese, Korean and other languages. the

In the above-described embodiments, descriptions have been given assuming that the present invention is applied to the image forming apparatus 100, but the present invention is not limited thereto. For example, it is also possible to provide a translation device or an image processing device having only the functions of the translation processing unit 1 in the image forming device 100 described above. In this case, the translation device or the image processing device may be an ASIC (Application Specific Integrated Circuit) having the function of the translation processing unit 1 described above. Furthermore, the present invention can also be provided by recording the translation processing program PRG for performing the above translation processing on a recording medium such as a magnetic disk, a floppy disk, a CD (Compact Disk), a DVD (Digital Versatile Disc), and a RAM. the

As described above, the present invention provides a translation device including: a character recognition unit for recognizing text data in a text area of an input image; a translator for translating the text data in the text area; and a layout structure processor, for generating data including translated text data of a text area and graphics in an input image, wherein the layout of the input image is maintained in the image layout of the data generated by said layout structure processor. the

According to an embodiment of the present invention, the layout structure processor can change the setting of the translated text data according to the amount of the translated text data and the size of the text area. the

According to another embodiment, the layout structure processor may generate the translated text data by changing the size of characters of the translated text data.

According to another embodiment, the layout structure processor may generate the translated text data by changing the line spacing of the translated text data. the

According to another embodiment, the layout structure processor may generate translated text data by changing margins around the translated text data. the

According to another embodiment, the layout structure processor may generate data by changing the size of the graphics area. the

According to still another embodiment of the present invention, the translation device may further include a graphic reformation unit for reforming the graphic area in the following cases: by reforming the text area in a predetermined area, the translated text data can be combined with The same character size of the text data is accommodated in the reformulated text area. the

According to another embodiment, the translating apparatus further includes a graphic reforming unit for reforming the text area when the setting of the translated text data is changed in a predetermined manner and the translated text data cannot be accommodated in the text area. and graphics area. the

According to another embodiment, the character recognition unit may also perform character recognition processing on the image of the graphic area of the input image to output graphic area text data representing the type, position and size of the character in the graphic area; the translation processor may also performing a translation process on the graphic area text data to output graphic area translated text data representing the type, position and size of characters to be accommodated in the graphic area; and the layout structure processor further includes: a translated data transfer controller, for transferring the translated text data of the text area between text areas of different pages of the translated text and graphic data in a predetermined manner; and a switching unit for controlling the setting of the page in the case of The graphic identification information of the graphics is transmitted: the graphic identification information is included in the translated text data of the graphic area of the page; before the translated data transmission controller transmits the translated text data, the included graphic identification information The translated text data in the text area of the same page contains the same graphic identification information as the included graphic identification information; and if the translated data transmission controller transmits the translated text data, the graphic identification information will be transmitted into a text area on another page. the

According to another embodiment, the character recognition unit can also perform character recognition processing on the graphic area image of the input image to output graphic area text data representing the type, position and size of the character in the graphic area; The graphic area text data performs translation processing to output the graphic area translated text data representing the type, position and size of characters to be accommodated in the graphic area; and wherein the layout structure processor may include: translated data transfer control a device for transferring the translated text data of the text area between text areas of different pages of the translated text and graphic data in a predetermined manner; In the case of translated text and graphic data of N-up structure, the setting of the page is controlled in the following cases, instead of transmitting the graphic identification information of the identification graphic: said graphic identification information is included in the translation of the graphic area arranged in the page In the post-text data; before the translated data transmission controller transmits the translated text data, the text area of the same page containing the contained graphic identification information contains the same graphics as the contained graphic identification information in the translated text data identification information; and if the translated data transmission controller transmits the translated text data, the graphic identification information will be transmitted to the text area of another page. the

As described above, the present invention provides a translation device including a translation processor for translating an original text and correcting so that the lengths of the original text character string and the length of the translated text character string are substantially the same as each other, wherein the original Text and translated text are placed in parallel. the

According to an embodiment of the present invention, the translation processor may change a character length of at least one of the original text and the translated text. the

According to another embodiment, the translation processor may determine the length of the character string by summing the spaces between characters and the points of the characters. the

According to another embodiment, the translation processor may change the point size of the characters. the

According to another embodiment, the translation processor may change character fonts. the

According to another embodiment, the translation processor may change the characters into phonetic characters. the

In one aspect, the present invention provides a translation method comprising the steps of: identifying text data in a text area of an input image; translating the text data in the text area; and generating data comprising the translated text data in the text area and graphics in the input image , where the layout of the input image is maintained in the image layout of the data generated in the generation step. the

On the other hand, the present invention also provides a translation method, comprising the steps of: translating the original text, and correcting so that the length of the character string of the original text and the length of the character string of the translated text are substantially the same as each other, wherein the original text and the translated text The text is placed in parallel.

In yet another aspect, the present invention also provides a computer-readable storage medium, which stores such a program whose instructions can be executed by a computer to realize the translation function, and the function includes: a character recognition function for recognizing Text data in the input image text area; a translation processing function for translating the text data in the text area; and a layout structure processing function for generating data including the translated text data in the text area and graphics in the input image, wherein The layout of the input image is maintained in the image layout of the data generated by the layout structure processing function. the

In addition, the present invention also provides a computer-readable storage medium, which stores a program whose instructions can be executed by a computer to realize a translation function, and the function includes: a translation processing function for translating an original text , and correct so that the lengths of the character strings of the original text and the character strings of the translated text are substantially the same as each other, wherein the original text and the translated text are placed in parallel. the

The foregoing description of the embodiments of the invention has been presented for purposes of illustration and description. It is not exhaustive, nor is it intended to limit the invention to the precise forms disclosed. Obviously, various modifications and variations will be apparent to those skilled in the art. The above-described embodiments were chosen and described in order to best explain the principles of the invention and its practical application, thereby enabling others skilled in the art to understand various embodiments of the invention and to understand the various embodiments of the invention for the particular use contemplated. Various appropriate modifications. The scope of the invention is defined by the following claims and their equivalents. the

The entire contents of Japanese Patent Application No. 2005-82047 filed on March 22, 2005 and Japanese Patent Application No. 2005-90179 filed on March 25, 2005, including specification, claims, Figures and abstract.

Claims (10)

1. A translation device, said translation device comprising: A character recognition unit for recognizing text data in the text area of the input image; a translator for translating the text data in the text area; and a layout structure processor for generating data including the translated text data of the text area and graphics in the input image by changing the character size of the translated text data so that in the image layout of the data generated by the layout structure processor maintain the layout of the input image, wherein the layout structure processor generates the data by changing the size of the graphics area, and Wherein the layout structure processor reformats the graphic area in the following case: By reformulating the text area within a predetermined reformat size limit, the translated text data can be accommodated in the reconstructed text data at the same character size as the untranslated text data. in the text area. 2. The translation apparatus according to claim 1, wherein the layout structure processor changes the setting of the translated text data according to the amount of the translated text data and the size of the text area. 3. The translation apparatus according to claim 1, wherein the layout structure processor generates the translated text data by changing a line spacing of the translated text data. 4. The translation apparatus according to claim 1, wherein the layout structure processor generates the translated text data by changing a margin around the translated text data. 5. The translation apparatus according to claim 1, wherein said layout structure processor changes the setting of the translated text data within a predetermined reformation size limit in a case where the translated text data cannot be accommodated in the text area, and then Form text area and graphics area. 6. A translation method, said method comprising the steps of: A character recognition step, identifying text data in the text area of the input image; a translation step of translating the text data in the text area; and a layout structure processing step of generating data including the translated text data of the text area and graphics in the input image by changing the character size of the translated text data, thereby maintaining the layout of the input image in the image layout of the generated data, wherein said data is generated by varying the size of said graphics area, and Wherein the layout structure processing step reformats the graphic area in the following case: By reformulating the text area within a predetermined reformation size limit, the text data after translation can be accommodated in the reformulation with the same character size as the text data before translation. in the text area. 7. The translation method according to claim 6, wherein the layout structure processing step changes the setting of the translated text data according to the amount of the translated text data and the size of the text area. 8. The translation method according to claim 6, wherein the layout structure processing step generates the translated text data by changing the line spacing of the translated text data. 9. The translation method according to claim 6, wherein the layout structure processing step changes margins around the translated text data to generate the translated text data. 10. The translation method according to claim 6, wherein said layout structure processing step changes the settings of the translated text data within a predetermined reformation size limit in a case where the translated text data cannot be accommodated in the text area, and then Form text area and graphics area.

Images