JPH0756924A - Bilingual device - Google Patents

Abstract

PURPOSE:To reduce memory quantity and work quantity by storing a connected word and an idiom in the same structure as a word at the time of retrieving a dictionary for translating a language which is optically recognized and outputting a retrieved translated word in an arbitrary system. CONSTITUTION:When a bilingual mode is selected in an operation part 5, an original picture which is read by a scanner 2 is binarized in a picture processing part 3. An OCR part/spelling check part 8 executes a character recognition processing in an area recognized to be a character area by an area identification part 7. When the original picture is English, the English word which is character recognition-processed is spelling-checked. A dictionary retrieval part 9 retrieves the dictionary 11 for the recognized word and idiom for translating the English sentence on the original into Japanese. CPU for bilingual processing 6 retrieves the word and the idiom, which are connected by a mark at the time of retrieving the dictionary, and the connected word/idiom are stored in the memory 13 in the same structure as the word. Then, they are outputted in the arbitrary system by using the retrieved translated word.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a bilingual apparatus for reading an English sentence on a manuscript, translating it into Japanese, and outputting it.

[0002]

2. Description of the Related Art A digital copying machine for translating an arbitrary language into an arbitrary different language is disclosed in Japanese Patent Laid-Open No. 4-77965. However, in this digital copying machine, it was not possible to output a compound word composed of a plurality of words together with the translated language (hereinafter referred to as bilingual format output). Further, in this digital copying apparatus, the divided English words connected by symbols such as hyphens are stored separately from general English words, so that there is a problem that both the memory amount and the work amount increase.

When a bilingual format output is made for a manuscript having a large number of characters per sheet, the number of translated characters also becomes large.
Output was performed over more than one sheet of output paper. More specifically, first, the document shown in FIG. 10 (1) is divided into regions which are recognized, and the regions are translated in the order of region A (101) to region E (105). Then, the translated characters are the area A '(106) and the characters of the manuscript, respectively.
Region B '(107), Region C' (108), Region D '
(109) and area E '(110), and this area A'
The areas (106) to E '(110) are successively output to the vacant spaces of the output paper, as shown in FIG.
The parallel translation output is as shown in (3).

However, since the bilingual format output as shown in FIGS. 10 (2) and 10 (3) has a completely different layout from the original shown in FIG. 10 (1), which part of the original corresponds to which output. There was a problem that it was difficult to understand what was going on.

Further, although the accuracy of the recent OCR using the optical character recognition technology is close to 100%,
Misrecognition will always occur depending on the state of the read document and the form of the character font. In that case, if a misrecognized word is accidentally found in the dictionary, a translated word that is completely different from the original text may be output, and it is not possible for general users to know where the error occurred. was there.

Further, there is a problem in that it is not possible to provide a satisfactory translated word for a specialized document if only a widely used dictionary is installed.

Further, there is a problem that it is not possible to perform parallel translation processing on white characters and outline characters.

[0008]

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and its first object is to:
To reduce the amount of memory and work.

A second object is to be able to output translations of both the idiom and the words that compose it.

A third object is to make it easy to refer to the original and the bilingual output.

A fourth object is to make it easy to understand where a mistake has occurred in the process of outputting a translation.

A fifth object is to prevent a translation error and a translation failure.

A sixth object is to enable translation of white characters and outline fonts.

[0014]

In order to achieve the first object, the invention of claim 1 is an image reading means for reading an image of a document, an image storing means for storing read image information, and a read image. In a digital image copying machine provided with image processing means for performing image processing on information and image recording means for recording an image signal output by the image processing means on a recording medium, a character area is identified from an image read by the reading means. Identifying means, a recognizing means for performing optical character recognition of the identified character area, a dictionary searching means for performing a dictionary search to translate a language composed of the recognized characters into a different language, and the dictionary search. Means for retrieving words and phrases that are linked by symbols; means for storing the concatenated words and phrases with the same structure as the words when storing the retrieved translated words; And means for outputting in any format using been translation.

The invention of claim 2 is provided with a means for marking and outputting the searched idiom in the first invention.

According to a third aspect of the present invention, in the first aspect of the present invention, the original sentence, the translated word, and the optically recognized character are provided on the same recording paper in a predetermined format.

The invention according to claim 4 is the first aspect of the invention, further comprising means for distinguishing the idiom from the word by a line drawing, color, etc. and means for outputting the idiom book.

According to a fifth aspect of the present invention, in the first aspect, the dictionary selection means and the means for simultaneously outputting a plurality of translated words for one word in the selected dictionary are provided.

According to a sixth aspect of the invention, in the first aspect of the invention, the image processing apparatus further comprises means for designating an image processing range and image processing means.

[0020]

According to the present invention, the image reading means for reading the image of the original, the image storage means for storing the read image information, the image processing means for performing image processing on the read image information, and the image processing means. In a digital image copying machine provided with an image recording means for recording an image signal output by a recording medium on a recording medium, an identifying means for identifying a character area from an image read by the reading means, and an optical character of the identified character area A recognition means for recognizing; a dictionary search means for performing a dictionary search for translating a language composed of the recognized characters into a different language; a means for searching words and phrases that are linked by symbols in the dictionary search; When storing the translated word, a means for storing the concatenated word and the idiom in the same structure as the word, and a method for outputting the searched translated word in an arbitrary format. Because comprising the door, amount of memory, amount of work decreases.

According to the invention of claim 2, the parallel translation device according to claim 1 is provided with a means for marking and outputting the retrieved idiom, so that both the idiom and the words constituting it are translated. Is output.

According to the third aspect of the present invention, the parallel translation apparatus according to the first aspect includes means for outputting the corresponding translated word below the original sentence in a layout similar to the original document in an arbitrary output format. It becomes easier to refer to the bilingual output.

According to the invention of claim 4, the parallel translation apparatus according to claim 1 is provided with means for outputting the recognized character as well as the corresponding translated word below the original sentence in an arbitrary output format. , It is easy to understand which error occurred in the process of outputting the translation.

According to the invention of claim 5, in the bilingual apparatus according to claim 1, as a dictionary to be used, a means that uses both a dictionary that is generally widely used and a dictionary corresponding to a technical term, and a dictionary. Since there is a means for selecting the usage mode of, there is no translation error or translation failure.

According to the invention of claim 6, in the bilingual apparatus according to claim 1, since the input range designating means is provided,
You can translate white characters and outline fonts.

[0026]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The configurations of a digital copying machine and a bilingual apparatus will be described with reference to FIGS. The cable 400 is used for communication of commands between the digital copying machine 100 and the parallel translation device 200, and input / output of images. When the parallel processing mode is selected on the LCD 300, which is the operation unit, the digital copying machine 10
The multi-valued image read by the scanner of 0 is converted into binary data and then stored in the IC card 202 which is the memory of the parallel translation device 200 via the AI base board 201.
Then, the binarized data is subjected to parallel translation processing in the AI base board 201 and output to the digital copying machine 100. The control unit 203 controls the parallel translation processing and the timing control with the digital copying machine.

The configurations of the digital copying machine and the parallel translation device will be described in more detail with reference to FIG.

First, the digital copying machine will be described. The digital copying machine is a CP for controlling the digital copying machine.
U1, a scanner 2 for reading an original image, an image processing unit 3 for performing image processing on the read original image, a printer 4 for recording the input image on a transfer paper, various modes such as a bilingual mode, etc. It has an operation unit 5 such as an LCD to be designated.

In the bilingual apparatus, the bilingual processing CPU 6, which is a control unit for timing, etc., the area identifying unit 7 for recognizing whether the input image area is a character area or a picture area, and the area identifying unit 7 for recognizing the character area. OCR section for performing character recognition processing in the designated area, spell check section 8 for spell-checking English words subjected to character recognition processing when the original image is an English sentence, and dictionary is searched for the recognized words and idioms. When a character is generated in the dictionary search unit 9, a drawing unit 10 that combines a manuscript image, a translated language character, and a character recognized character, a dictionary 11 in which at least two languages are associated, and the drawing unit 10 Used fonts Stores the stored fonts 12, document images, character recognition results and images at the time of composition, and is used as a work memory for parallel translation processing. With a memory (RAM) 13 to or.

An outline of the control operation of the CPU of the digital copying machine and parallel translation apparatus of the present invention will be described with reference to FIG.

When the parallel translation mode is selected on the operation section 5, the original is read by the scanner 2 (step 5). The read document image is binarized by the image processing unit 3 and then input to the bilingual apparatus (step 7). It is recognized whether the input document image is a character area or a picture area (step 10). The number of lines of the character in the character area, the position of the line, the number of characters in the line, the position of the character, and the division into words are performed (step 20). The characters in the character area are recognized using OCR (step 30), and the processing results of steps 10, 20, and 30 are stored in the memory 13 in a predetermined format (step 35). Step 1
Using the processing results of 0, 20, and 30, a dictionary search for words consisting of the recognized characters is performed (step 40). Drawing data is created in a predetermined format from the processing result stored in step 35 and the dictionary search result and stored in the memory 13 (step 50). An output image is created based on the drawing data created in step 50 and stored in the memory 13 to obtain an output image such as a bilingual form, a word book, and a idiom book (140). In Step 1000, Step 50
When the transfer paper for which this information is currently selected is used, the number of sheets to be used is calculated by using the data created in (3). The calculated number of pages is displayed on the operation unit 5 of the digital copying machine.

The first embodiment will be described. The structure of the words searched in the dictionary in step 40 is shown below. <Word information> Start X coordinate Start Y coordinate End X coordinate End Y coordinate Word number of words Word identification code Address where the corresponding recognition word data is stored

The above is the information for one word. Here, the word identification code is a code for identifying a normal word and an idiom, and the following identification is performed as an example. Common word: 0xff Jukugo: 0x01

Here, when a dictionary search is performed in step 40, an intermediate structure between lines and words (called an intermediate) is obtained.
Store the result in memory. The structure of the intermediate is shown below.

<Intermediate Information> Number of Words Included in Intermediate Address where the word information at the beginning of the intermediate is stored

Here, how to refer to the intermediate structure will be described. In step 40, the dictionary search result structures shown below are stored in the memory for each identified area as many as the number of intermediates in each area, and the result of performing the above operation for all areas is step 50. , Stored.

<Dictionary search information> Number of translated words Address where the corresponding intermediate data is stored Address where the translated word data is stored

FIG. 5 shows an explanatory view of the intermediate structure. Thick lines indicate line positions, and thin lines indicate words.

FIG. 5A shows the recognition result after OCR,
(2) shows the word (intermediate) after the dictionary search using the OCR result. As can be seen by comparing (1) and (2), it is understood that the word recognition positions of the idiom and the connected word are different.

Here, a method of storing data in the case of corresponding to a compound word in which another word intervenes will be described. In the case of the compound word "too compiled to protect", the word data of the OCR result has data stored in the order of "too compiled to protect". As the intermediate data after the dictionary search result, data is stored in the order of "to compliant", "to protect", and "complicated probe". Related idioms can be marked by referring to the intermediate information above.

FIG. 6 shows the output format. Area A (101), area B (102), area C (1
03), area D (104) and area E (105) are area A ″ (111), area B ″ (112), area C ″ (113), area D ″ (114) and area, respectively. E ''
Corresponds to (115). As described above, although the number of sheets is two in FIG. 6, the output is made according to the layout of the document shown in FIG. 10 (1).

Further, FIG. 7 shows a parallel translation output format in the case where a picture area exists. Although the area filled with black is the picture area, in this case, a margin is generated, but the layout is stored by memorizing the coordinate values in the drawing unit so that the layout of the original is preserved.

The operation of the second embodiment will be described below.
The character string recognized by the OCR unit 8 is stored as recognition result data in step 35 in association with position information on the document on a word-by-word basis. Therefore, in step 120, the recognized English word stored in step 35 is drawn between the start and end coordinates of the corresponding word in the original sentence. FIG. 8 shows an example of the parallel translation output format of the second invention. Here, "Paris" in the third line should be translated as "Paris", but it is translated as "Pannabe". So OC
Looking at the R result, we saw r and i as one character n. "P
It is "ans" and it can be understood that it was translated as Pan = pan.

The operation of the third embodiment will be described. If the word identification code of the word information is 0x01, step 1
In the drawing process of 20, the process from the start coordinate to the end coordinate is surrounded by a line to obtain an output as shown in FIG. 8 (1).

At this time, by utilizing the fact that 0x01 is not continuous in the word identification code, processing may be performed such that the word is divided and surrounded as shown in FIG. 8 (2). At this time, a line as shown in FIG. 8 (2) may be added to show that they are related.

Also, only the word identification code of 0x01 can be output as the idiom sheet.

The operation of the fourth embodiment will be described. Before the bilingual processing, the following selections can be made. Standard dictionary only Special dictionary only Both

At this time, when both are designated and translations exist for both in the same word, the translation of the standard dictionary is prioritized, or vice versa, or both translations are displayed. As a result, FIG. 9 (1),
Outputs as shown in (2) and (3) can be obtained.

The operation of the fifth embodiment will be described below.
In step 5, when an image is input, an area on the document in which blank characters and outline characters exist is designated. After being binarized in the image processing unit 3, the negative / positive inversion and the embedding processing which are known in the image processing are performed on the designated area. By inputting the data into the parallel translation device, the outline characters and outline characters can be recognized without being distinguished from the words that can be normally OCR.

At this time, since the image of the original document is cut as it is in the bilingual form and combined with the output, the image processing part is processed by the negative / positive inversion and the half-filling process until the OCR result is obtained. Remember as a temporary thing.

[0051]

According to the first aspect of the invention, the image reading means for reading the image of the original, the image storing means for storing the read image information, the image processing means for performing the image processing on the read image information, and In a digital image copying machine provided with an image recording means for recording an image signal output by an image processing means on a recording medium, an identifying means for identifying a character area from an image read by the reading means, and an identifying means for identifying the character area A recognition means for performing optical character recognition, a dictionary search means for performing a dictionary search to translate a language composed of the recognized characters into a different language, and a means for searching for words and phrases that are connected by symbols in the dictionary search. When storing the retrieved translated word, the concatenated word and the idiom are stored in the same structure as the word, and the retrieved translated word is output in an arbitrary format. Because and means for performing, amount of memory, amount of work can be reduced.

According to the second aspect of the invention, the parallel translation device according to the first aspect includes means for marking and outputting the retrieved idiom, so that both the idiom and the words that compose it are translated. Can be output. Also, by distinguishing from a word by adding a mark in the output format, it can be easily understood that it is a idiom. Further, by devising the marking method, it is possible to mark all the words related to a compound word that is not connected or to mark only the compound word.

According to the third aspect of the invention, the parallel translation apparatus according to the first aspect is provided with means for outputting the corresponding translated word below the original sentence in a layout similar to the original document in an arbitrary output format. It is possible to easily refer to and the parallel output.

According to the invention described in claim 4, in the bilingual apparatus according to claim 1, a unit for outputting the recognized character as well as the corresponding translated word under the original sentence is provided in an arbitrary output format. Therefore, it is possible to easily understand which error occurred in the process of outputting the translation.

According to the invention described in claim 5, in the bilingual apparatus according to claim 1, as a dictionary to be used, a means that uses both a dictionary that is generally widely used and a dictionary that corresponds to a technical term, Since a means for selecting the usage form of the dictionary is provided, it is possible to eliminate a translation error or a translation failure.

According to the invention described in claim 6, in the parallel translation apparatus according to claim 1, since the input range designating means is provided, it is possible to perform the translation for the outline character and the outline font.

[0057]

[Brief description of drawings]

FIG. 1 is an external view of a digital copying machine and a parallel translation apparatus according to the present invention.

FIG. 2 is a configuration diagram of a digital copying machine and a parallel translation apparatus according to the present invention.

FIG. 3 is a block diagram of a digital copying machine and a parallel translation apparatus according to the present invention.

FIG. 4 is a flowchart of a CPU according to the present invention.

FIG. 5 is an explanatory diagram showing an intermediate structure according to the present invention.

FIG. 6 is an explanatory diagram showing an output format according to the present invention.

FIG. 7 is an explanatory diagram showing an output format according to the present invention.

FIG. 8 is an explanatory diagram showing an output format according to the present invention.

FIG. 9 is an explanatory diagram showing an output format according to the present invention.

FIG. 10 is an explanatory diagram showing a conventional output format.

[Explanation of symbols]

 1: CPU 2: scanner 4: image processing unit 5: printer 6: parallel processing CPU 7: area identification unit 8: OCR unit + spell check unit 9: dictionary search unit 10: drawing unit 11: dictionary 12: font 13: Memory 400: Cable between digital copier and translation device

Claims (6)

[Claims] 1. An image reading unit for reading an image of an original, an image storage unit for storing read image information, an image processing unit for performing image processing on the read image information, and an image signal output by the image processing unit. In a digital image copying machine having an image recording means for recording on a medium, an identifying means for identifying a character area from an image read by the reading means, and a recognizing means for performing optical character recognition of the identified character area, Dictionary search means for performing a dictionary search to translate the language consisting of the recognized characters into a different language,
In the dictionary search, a means for searching for words and phrases that are connected by symbols in the dictionary search, a means for storing the connected words and phrases in the same structure as the words when storing the searched translations, and using the searched translations And a means for performing output in an arbitrary format. 2. The bilingual apparatus according to claim 1, further comprising means for marking and outputting the retrieved idiom, and separately outputting the idiom by comparing it with the translated word. 3. The parallel translation apparatus according to claim 1, further comprising means for outputting a corresponding translated word below the original sentence in a layout like a manuscript in an arbitrary output format. 4. The parallel translation apparatus according to claim 1, further comprising means for outputting a recognized character as well as a corresponding translated word under the original sentence in an arbitrary output format. . 5. The bilingual apparatus according to claim 1, wherein the dictionary to be used is provided with means for using both a widely used dictionary and a dictionary corresponding to a technical term, and a usage form of the dictionary can be selected. A parallel translation device comprising means and outputting translated words of both dictionaries to an output format. 6. The bilingual apparatus according to claim 1, further comprising an input range designating means, wherein bilingual processing is performed on the image subjected to the image processing processing means.