JPH05324709A - Bilingual image forming device - Google Patents

Abstract

PURPOSE:To retrieve translation information and improve the bilingual efficiency by removing a detected specific symbol when the specific symbol is at the tail of a line or including the specific symbol when not. CONSTITUTION:It is judged whether or not there is a hyphen as a specific symbol (S1001) and when not, a return is made. When there is the hyphen, on the other hand, it is judged whether or not a divided word flag is set (S1002). When it is judged that the flag is set, equivalent composition is performed in the center of the line below a recognized word (S1003) and a return is made. When it is judged that the divided word flag is not set, on the other hand, it is judged whether or not the number of characters before the hyphen is larger than the number of following characters (S1004). When not, equivalent composition to the line below the words following the hyphen is performed (S1006). When it is judged that the number of the characters before the hyphen is larger, on the other hand, equivalent composition to the line below the word before the hyphen is performed (S1005). Consequently, the position of the equivalent to the word is made clear.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a bilingual image forming apparatus applied to a copying machine, a facsimile machine or the like which uses digital image data having a bilingual function, and more specifically,
A bilingual image forming apparatus that optically reads a manuscript (or inputs image information via a communication line), recognizes a word in the read (or input) image information, and outputs a translated word corresponding to the word. Regarding

[0002]

2. Description of the Related Art For a person who reads a sentence written in another language, there is no problem in reading comprehension of many parts of the sentence, but it is difficult to read because only a specific word does not know an appropriate translation. There are many situations where it is not possible. In such a situation, it is a common experience to know the meaning of the specific word using a dictionary and proceed with reading.

Therefore, as a translation copying machine for automatically translating and outputting a sentence written in another language, there is, for example, a "translation copying machine" disclosed in Japanese Patent Laid-Open No. 154845/1987. According to the publication disclosed in that publication, the first
Firstly, the image reading means reads the original image, and secondly, the original image read by the identifying means is distinguished into pattern information and character information,
Third, the translation means recognizes the character information for each character,
Further, the content is translated into another language and the translated image is output together with the original image. Therefore, according to the technique disclosed in the above publication, a translated image relating to character information can be obtained together with a copy image of an original image, and therefore the labor of using a dictionary can be omitted.

[0004]

However, in the above-mentioned "translation copying apparatus" in the prior art, since a translated image relating to character information can be obtained together with a copied image of the original image, it is not necessary to use a dictionary. However, for many users who have a certain level of language knowledge, considering the current translation linguistics and implementation costs, which are not necessarily error-free, it is highly redundant and is always a desirable function. There was a problem that I could not say.

Further, in the conventional art, for a manuscript image having a plurality of columns, the position of the translated word corresponding to the word existing in the manuscript image is not clear, and the addition of the translated word causes Not only is the content changed drastically, the initial layout is broken, and the copied image becomes very difficult to read,
Completed at the end of a line or a group word in which words are connected by special symbols "-", "&", "/", etc.
In the case where one word is divided into two lines, there is a problem that the processing cannot be properly executed and the translation efficiency is reduced.

The present invention has been made in view of the above problems. The position of a translated word corresponding to a word existing in an original image is clarified, and the content of the output image is significantly changed by adding the translated word. Even if it does, the object of the present invention is to improve the translation efficiency by laying out the content of the copied image in an easy-to-read layout and outputting it, and by appropriately executing the processing of the set words connected by the special symbols.

[0007]

SUMMARY OF THE INVENTION According to the present invention, an image reading means for optically reading a document image and converting it into image information, and image information read by the image reading means or an image input through a communication line. Image storage means for storing information, image processing means for performing various image processing on the image information, image recording means for recording an image on a recording medium based on an image signal output by the image processing means, A translation image forming apparatus comprising: a translation word storage unit for storing the document image information and translation word information for the information; and a control unit for executing a translation process for the document image information based on the translation word storage unit information. The control means detects the predetermined symbol and, when judging that it is at the end of the line, removes the predetermined symbol to search the translation information, and if it is not at the end of the line. When disconnection is to provide a translation image forming apparatus for searching the translation information including the predetermined symbol.

Further, when the translation means information is retrieved by removing the predetermined symbol, the control means compares the number of characters in the portion before the predetermined symbol with the number of characters in the subsequent portion, and determines the one with the larger number of characters. It is desirable to combine the translated word information under the part of.

Further, when the control means searches the translated word information including the predetermined symbol and finds that there is no corresponding translated word in the translated word storage means, it ends (last).
It is desirable to search the translation information by removing the portion after the predetermined symbol of.

According to the present invention, when a special symbol is detected and it is determined that it is at the end of a line, the special symbol is removed to search the translated word information, and when it is determined that it is not at the end of the line, the special symbol is detected. Search for translation information including.

Further, when the translated word information is retrieved by removing the special symbol, the number of characters in the portion before the special symbol is compared with the number of characters in the subsequent portion, and the translated word information is placed below the portion having the larger number of characters. To synthesize.

When it is determined that there is no corresponding translated word in the dictionary when searching the translated word information including the special symbols, the translated word information is searched by removing the portion after the last special symbol.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of a bilingual image forming apparatus according to the present invention will be described below. (1) Schematic configuration of bilingual image forming apparatus (2) Configuration of scanner section and printer section (3) CPU processing operation (3)- Manuscript recognition processing (3) -Classification of special symbols (3) -Paragraph, line, page synthesis processing (3) -Translation word dictionary retrieval processing (3) -Translation word storage processing (3) -Fixed composite image generation processing (3)- The translation word synthesizing process will be described with reference to the drawings.

(1) Schematic Configuration of Parallel Image Forming Apparatus FIG. 1 is a block diagram showing a schematic configuration of a parallel image forming apparatus which is an embodiment of the present invention. This bilingual image forming apparatus includes a scanner unit 101 that reads an original image, a CPU 109 that controls the entire apparatus, a ROM 110 that stores a control program, a RAM 111 that is temporarily used by the control program, and an input that stores the read image. Image memory 104, output image memory 10 for storing output images
5, a recognition dictionary memory 106 in which data used for character recognition is stored, a translation dictionary memory 107 in which a word and a translation corresponding to the word are stored, and an output in which character font data used for image generation is stored Font memory 108, internal system bus 115 for exchanging data between devices, A / D converter 102 for converting analog signals into digital signals, I / Fs 103, 113 for interfacing with system bus 115, and outputs A printer unit 114 for outputting the image stored in the image memory,
Operation board 112 that gives instructions such as start and stop
Etc.

Further, this bilingual image forming apparatus has a facsimile communication function, and 116 is a network control unit (NCU) 117 for closing and opening a telephone line or sending a dial signal and detecting a ringing frequency. Is a modem (modulator / demodulator) that performs data modulation when transmitting data to a telephone line (analog line) and demodulates data when receiving.
Reference numeral 8 denotes a communication control unit for executing processing before sending the image information to the telephone line (re-compressing the data according to the facsimile function of the receiving side or reducing the image information). Descriptions of other parts necessary for the facsimile communication function are omitted.

(2) Configuration of Scanner Unit and Printer Unit FIG. 2 is an explanatory diagram showing a schematic configuration of the inside of the scanner unit 101 and the printer unit 114. The contact glass 201 for placing a document on the scanner unit 101 is
Illuminated by the light sources 202a and 202b, the reflected light from the read document is imaged on the light receiving surface of the CCD image sensor 209 via the mirrors 203 to 207 and the lens 208. Light source 202 (202a, 202b) and mirror 20
3 is mounted on a traveling body 210 which moves the lower surface of the contact glass 201 parallel to the contact glass 201 in the sub-scanning direction. Moving body 2 moving to
It is installed in 11. CC in the main scanning direction
The original image is read by the CCD image sensor 209 by the solid-state scanning of the D image sensor 209, and the entire original is scanned by moving the optical system as described above. Incidentally, reference numeral 239 in the figure is a pressure plate for pressing the document.

Next, the printer unit 114 comprises a laser writing system, an image reproducing system and a paper feeding system. The laser writing system includes a laser output unit 221, an imaging lens 22.
2 and a mirror 223. Inside the laser output unit 221, a laser diode, which is a laser light source, and a polygon mirror that rotates at high speed by a motor are provided. The laser beam output from the laser writing system is applied to the photosensitive drum 224 of the image reproducing system. Around the photosensitive drum 224, a charging charger 225, an eraser 226, a developing unit 227, a transfer charger 228, a separation charger 229, a separation claw 2 are provided.
30, a cleaning unit 231 and the like are provided. A beam sensor (not shown) for generating a main-scanning synchronization signal is arranged near the one end of the photosensitive drum 224 at a position where the laser beam is emitted.

The operation of the image reproduction process in the printer section 114 will be briefly described. The peripheral surface of the photosensitive drum 224 is uniformly charged to a high potential by the charging charger 225. When the peripheral surface is irradiated with laser light, the potential of the irradiated part is lowered. Laser light is black for recording / playback
Since the ON / OFF control is performed according to the white color, the potential distribution corresponding to the recorded image, that is, the electrostatic latent image is formed on the peripheral surface of the photosensitive drum 224 by the irradiation of the laser light. When the portion on which the electrostatic latent image is formed passes through the developing unit 227, toner adheres according to the level of the potential, and the electrostatic latent image becomes a visualized toner image. In the part where the toner image is formed,
The recording sheet 232 is fed from the paper feed cassette at a predetermined timing and overlaps the toner image. The toner image is transferred to the recording sheet 232 by the transfer charger 228, and then separated from the photoconductor drum 224 by the separation charger 229 and the separation claw 230. The separated recording sheet 232 is conveyed by a conveyor belt 234, heated by a fixing roller 235 having a heater built therein, and then ejected to an ejection tray 236.

Further, in this embodiment, as shown in FIG. 2, the printer section 114 has two paper feeding systems. One of the paper feed systems includes an upper paper feed cassette 233a and a manual paper feed tray 2
33c, and the recording sheets 232 set in the upper paper feed cassette 233a and the manual paper feed tray 233c.
The sheet a is fed by the sheet feeding roller 237a. The other paper feed system is provided with a lower paper feed cassette 233b, and the recording sheet 232b in the lower paper feed cassette 233b is fed by a paper feed roller 237b. Then, the recording sheet 232 fed from any one of the sheet feeding rollers is temporarily stopped while being in contact with the registration roller 238, and the photosensitive drum 224 is synchronized with the progress of the recording process.
Is sent to the transfer section of.

(3) Processing Operation of CPU FIGS. 3 to 10 are flowcharts showing the processing operation of the CPU 109, respectively. First, refer to the flowchart of FIG. When the power is turned on (ON), the processing mode and the like are initialized (S301). Next, it is determined whether there is a parallel translation operation instruction from the operation board 112 (S30).
2). When there is no parallel translation operation instruction, the normal copying operation (S
303) is performed and the state of waiting for the parallel translation operation instruction (S302) is resumed.

On the contrary, when the parallel translation operation instruction is given, it is judged again whether or not the start instruction is given from the operation board 112 (S304). If it is determined that the start instruction is given, the original image set on the scanner unit 101 is read (S305) and stored in the input image memory 104.
Next, the number of pages of the document is counted (S306), and document recognition (S307) is executed.

(3) -Document Recognition Processing The document recognition (S307) processing is shown in FIG. In document recognition, first, column recognition (S401) is executed. The column recognition recognizes a case where a portion (white area) of the read image data where no character exists in a main scanning direction / sub-scanning direction is a certain interval or more as a column break. Next, column position recognition (S402) is executed. In the column position recognition, coordinates are given in pixel units to the image information read by the scanner unit 101 and stored in the input image memory 104, and the coordinates occupied by the recognized column in the image are obtained. For example,
As an example of the column starting from “Chapter 1” in the manuscript shown in FIG. 3, as shown in FIG.
A rectangular area completely including a column starting with "1" is obtained, and the upper left coordinates (bsx, bsy) and the lower right coordinates (bex, bey) of the area are recognized.

Next, line recognition (S403) is executed. Line recognition is performed for each recognized column, and when a portion (white area) of the read image data where there is no character in the sub-scanning direction is present over a certain interval, it is recognized as a line delimiter. Next, line position recognition (S404) is executed. In the line position recognition, coordinates are given in pixel units to the image information read by the scanner unit 101 and stored in the input image memory 104, and the coordinates occupied by the recognized line in the image are obtained.
For example, in the manuscript shown in FIG.
Taking a line beginning with "gle" as an example, as shown in FIG. 14, a rectangular region that completely includes a line beginning with "from single" is obtained, and the upper left coordinates (ls
x, lsy) and the lower right coordinate (lex, ley) are recognized.

Next, character recognition (S405) is executed.
The character recognition algorithm uses a combination of conventional techniques such as template matching and topological matching. Next, word extraction (S406) is executed from the characters recognized by the character recognition. Words are extracted line by line, and if there are more than a certain interval (white area) in the scanned image data where there are no characters in the scanned image data, or if there are space characters, punctuation marks, or other delimiters, the words are separated. To extract. Next, the classification of special symbols (S40
Execute 7). Then, word position recognition (S408) is executed. In the word position recognition, coordinates are given in pixel units to the image data read by the scanner unit 101 and stored in the input image memory 104, and the coordinates occupied by the recognized word in the image are obtained. For example, taking the word "This" in the manuscript shown in FIG. 13 as an example, a rectangular area in which "This" is completely contained is obtained as shown in FIG. 14, and the coordinates (wsx, wsy) at the upper left of the area are obtained. ) And the lower right coordinate (wex, way).

(3) -Classification of Special Symbols FIG. 5 is a flowchart showing a classification operation of special symbols. First, the line feed position recognition is executed (S501), and then
Character recognition is executed (S502), and then it is determined whether or not there is a hyphen (S503). If it is determined that there is no hyphen, the process returns to step 502, and steps 502-503 are repeated until it is determined that there is a hyphen. On the contrary, if it is determined that there is a hyphen, then it is determined whether it is the end of the line (S504), and if it is determined that it is the end of the line, the divided word flag is reset (S50).
5). On the contrary, if it is determined that the line is not at the end of the line, the divided word flag is set (S506) and the process returns.

(3) -Paragraph, line, page combination processing Next, returning to the flowchart of FIG. 3, paragraph, line, page combination (S308) is executed. This process is shown in FIG.
In the paragraph, line, and page combination processing, first, read image copying (S601) is executed. In the read image copying, the image information read by the scanner unit 101 and stored in the input image memory 104 is copied to the output image memory 105.

Next, a column framework drawing (S602) is executed. In the column framework drawing, a rectangle completely including each column is drawn based on the coordinate data obtained by the column position recognition in step 402. That is, in each column, (bsx, bsy) and (bex, be
Draw a rectangle with y) as diagonal coordinates. FIG. 15 shows an example of drawing a column framework.

Next, stage number composition (S603) is executed. In column number composition, a column number is sequentially added to the upper left of each column for each column recognized by the column recognition in step 401. The method of adding the column number is such that font data used for the column number to be added is output to the output font memory 10
It is read out from the memory 8 and expanded in the output image memory 105.
(Bsx, bsy) is used as the reference coordinate at this time. That is, the expansion of font data is (bsx, bsy)
Execute starting from. FIG. 15 shows an example in which a step number is drawn.

Next, line number composition (S604) is executed. In the line number synthesis, a line number is sequentially added to the left side of each line recognized by the line recognition in step 403. As a method of adding a line number, the font data used for the line number to be added is read from the output font memory 108 and expanded in the output image memory 105. (Lsx, lsy) is used as the reference coordinate at this time. That is, the expansion of the font data is executed starting from (lsx, lsy). FIG. 15 shows an example in which line numbers are drawn. In this example, the output line number is limited to a multiple of 5.

Next, page number composition (S605) is executed. In the page number combining, the page number counted by the document page number counting in step 306 is added to the lower center portion of the image in the output image memory 105. The page number addition method is similar to the step number addition method, the font data used for the page number to be added is read from the output font memory 108, and the output image memory 1 is read.
Expand to 05. As the reference coordinates at this time, the coordinates of the lower central portion of the image in the output image memory 105 are determined in advance. The expansion of font data is performed with the determined coordinates as the starting point. FIG. 15 shows an example in which page numbers are drawn.

By the above processing, an image in which the read original image is combined with the paragraph number, line number and page number is created in the output image memory 105. Next, print output (S309) for recording this image by the printer unit 114 is executed (see FIG. 3).

(3) -Translation dictionary search processing Next, a translation dictionary search (S310) is performed. This process is shown in FIG. The translation dictionary search is first performed in the above step 406.
Recognized word read (S701) is executed to read out one of the words extracted and stored by the word extraction processing of.
Next, it is determined whether or not hyphen is included in the recognized words (S702). If it is determined that there is a hyphen, then it is determined whether the divided word flag is set (S703). The term “divided word” as used herein means a combination of completed words, for example, electron-
It means spin, which means the above group word. If it is determined that it is set, a dictionary search including hyphen is executed (S704). Then, it is determined whether or not the retrieved word is in the dictionary (S705). If it is determined that it is not in the dictionary, the dictionary search is executed excluding the word after the last hyphen (S706), and it is determined whether the result is in the dictionary (S707). If it is not in the dictionary, it is then determined whether the searched word has a hyphen (S708). If you determine that there is no hyphen,
When the output flag is reset (S709) and it is determined that there is a hyphen, on the contrary, the process returns to step 706, and the subsequent processing is repeated until it is determined that there is no hyphen. If it is determined in step 707 that it is in the dictionary, the translated word is stored (S710) and the output flag is set (S711).

After resetting / setting the output flags in the above steps 709/711, it is judged whether or not the retrieval of all the read recognition words is completed (S712). If it is determined that all of the recognition words have not been completed,
After reading the removed words (S713),
Returning to step 705, the subsequent processing is repeated until it is determined that the retrieval of all the read recognition words has been completed.
On the contrary, when it is determined in step 712 that the retrieval of the read recognition word is completed,
It is determined whether or not all recognized words are completed (S719), and if completed, the process returns. On the contrary, if not completed, the process returns to step 701, and the subsequent processes are repeated until it is determined that all recognized words are completed.

If it is determined in step 703 that the divided word flag is not set,
The hyphen is removed and the next recognition word is read out and connected (S714). Then, it is determined whether or not the word is in the dictionary (S715). If it is determined that the word is in the dictionary, the translated word is stored (S716) and the output flag is set (S717). On the other hand, if it is determined that it is not in the dictionary, the output flag is reset (S718). If it is determined in step 702 that there is no hyphen in the recognized word, the process proceeds to step 715. (3) -Translated word storage processing Here, the translated word storage processing (S710 / S716) will be described with reference to the flowchart of FIG. In this translated word storage process, a word and a translated word corresponding to the word are read from the translated word dictionary memory 107 and stored. In the translated word dictionary memory 107, a word and a translated word corresponding to the word are stored as a pair, and the translated word has a plurality of meanings registered for one word. Each translated word is prioritized by part of speech.

Next, it is judged in the part-of-speech identification inspection (S801) whether or not the part-of-speech is identified by the part-of-speech identification. If the part of speech is specified, the translated word of the specified part of speech is read (S803). On the contrary, when the part of speech is not specified, it is determined whether or not there are a plurality of parts of speech (S802).
When there are a plurality of parts of speech, the translated word of the first priority part of speech is read (S804). When there are no plural parts of speech, the translated word is read (S805).

For example, the word "play" is stored in the translated word dictionary memory 107 as follows. play [verb] play; to do; to play [name] to play This is the first priority part of speech is a verb, and the translated words at that time are "play", "...", "play", and second priority. Part of speech is a noun, and the translated word at that time is "play." That is, if the part of speech is specified as a verb, "play" is read, and if it is specified as a noun, "play" is read. When the part of speech is specified, the word "play" which is the translation of the first priority part of speech is read. The translated word thus read out is stored together with the word (S806).

Next, returning to the flowchart of FIG. 3, the number of output pages is counted (S311). In the present invention, since the image generated and output by the image generation unit has a larger amount of information than the read document image, a plurality of pages of output images may correspond to one page of the document image. In the output page number counting process, the number of pages of this output image is counted.

(3) -Inset Composite Image Generation Process Next, an inset composite image generation process (S312) is executed. This process is shown in FIG. The embedded composite image generation is
First, paragraph number counting (S901) is executed. In the column number counting, the columns recognized by the column recognition in step 401 are sequentially counted. Next, the column number composition (S902) is executed. Column number composition
This is executed by reading the font data corresponding to the counted column number from the output font memory 108 and expanding it in the output image memory 105. FIG. 16 shows an example of executing this column number combination.

Next, line number counting (S903) is executed. The line number counting sequentially counts the lines recognized by the line recognition in step 403. Next, line number composition (S904) is executed. In line number composition, the font data corresponding to the counted line number is read from the output font memory 108, and the output image memory 105
Run by deploying to. An example of executing this line number combination is shown in FIG. Here, only the line numbers that are multiples of 5 are expanded.

Next, read line copying (S905) is executed. In the read line copying, the coordinates (lsx, lsy) and (lex, ley) obtained by the line position recognition in step 404 are diagonally read from the image information read by the scanner unit 101 and stored in the input image memory 104. This is executed by copying the rectangular area that it has into the output image memory 105.

Next, the output flag check (S906) is executed. The output flag check is performed in the above steps 711/717.
Output flag is set, and it is determined whether the output flag for each word added by the output flag reset in step 709/718 is set or reset. When it is determined that the output flag is set, the translated word is read (S907). In the translated word reading, the translated word stored by the translated word storage in step 710/716 is read. Next, translated word synthesis (S
908) is executed.

(3) -Translated word synthesizing process The translated word synthesizing process (S908) will be described with reference to the flowchart shown in FIG. First, it is determined whether or not there is a hyphen (S1001), and if there is not, a return is made. On the contrary, if there is a hyphen, it is judged whether or not the divided word flag is set (S1002), and if it is judged that the divided word flag is set, translation word synthesis is executed at the center of the lower line of the recognized word (S1003). ), Return. On the contrary, if it is determined that the divided word flag is not set, then it is determined whether or not the number of characters before the hyphen is larger than that thereafter (S1004). When it is determined that it is not large, the translated word synthesis is executed in the lower line of the word after the hyphen (S1006). On the contrary, when it is determined that it is large, the translated word synthesis is executed in the lower line of the word before the hyphen.

Next, returning to the flowchart of FIG. 9, it is judged whether or not the processing for all the words in the line is completed (S
909), if not finished, the same process is repeated from the "output flag check" in step 906 for the next word. When it is determined that the processing has been completed for all the words in the line, it is determined whether or not all the lines in the column have been completed (S910). If not completed, the same process is repeated from the "line number count" in step 903 above for the next line. When it is determined that the processing has been completed for all the lines in the column, it is determined whether all the columns in the page have been completed (S911). If it is not completed, "column number counting" in step 901 above for the next column
Repeat the same process from.

Next, page number composition (S912) is executed. The page number combining process is step 60 shown in FIG.
The same process as "Page number composition" of 5 is executed.
As shown in 6, the number of pages of the original image and the number of pages of the generated image are combined with a number connected by a hyphen.

Returning to the flowchart of FIG. 3, print output (S313) is then executed. The print output process is the same as the "print output" in step 309, and the image generated by the "inset composite image generation" in step 312 is printed by the printer unit 114.
Next, it is determined whether or not all the inlaid composite images corresponding to the read document have been printed (S314).
If it is determined that the process is not completed yet, the process returns to step 311, and the same process is repeated from the next page number count.

When all the printing is completed, all the read originals are read and it is judged whether or not the processing is completed (S31).
5). If it has not been completed yet, the same process is repeated from the "document reading" in step 305 for the next document. When all the above processing is completed, the process returns to the "parallel translation operation instruction" in step 302 above.

As described above, in the case of the structure in which a word including a special symbol such as a divided English word is searched, the special character is searched and controlled in the OCR processing. This recognizes a special symbol or a registered character or symbol in a character string regarded as one unit for performing a dictionary search, and regards the symbol or characters before and after the character as one dictionary search unit. At this time, when a predetermined symbol (generally a hyphen) is recognized at the end of a line, the character strings before and after it are not considered as a dictionary search unit, but are connected to the next character string (the word at the beginning of the next line) to search the dictionary 1 The operation is performed in units, and the format is stored in the memory in consideration of special characters.

FIG. 11 illustrates the contents of the structural format of words and divided words. First, the number of characters that is a member of the word structure format represents the number of characters recognized as words by OCR, and the split word offset is the offset value of the number of split words in the line when the recognized word is a split word. It shows. Substitute a suitable fixed value for a word that is not a divided word. In FIG. 11, 0xff is substituted. Next, the number of characters that are members of the divided word structure format represents the number of characters of the recognized divided word, and the remaining number of divided words represents the number of divided words connected to the noticed divided word.

FIG. 12 is an explanatory diagram of processing when a hyphen arrives at the end of a line. 1) shows the state of the original when there is a hyphen at the end of the line. Where "communi-"
Compare the number of characters in the word and the number of characters in the divided word. As a result, it can be seen that the number of divided words is one less in 8 and 7 as shown in 2). Since this word has a known information of the last word existing in the line, it can be recognized that it is not treated as a divided word. Then, the divided word at the end of the line is taken out from the divided word text data, the word at the beginning of the next line is taken out from the word text, and the two are connected, and then a dictionary search is executed. Next, the number of characters at the end of the line is compared with the number of characters at the beginning of the line, and as shown in 3) and 4), the translated word searched by the dictionary search processing can be output to the one having the larger number of characters.

Further, FIG. 17 shows that when a group word composed of a plurality of words connected by a special symbol is not stored in the translated word storage means, the group word connection is decomposed and divided into words. 5 is a flowchart showing an operation for controlling so as to output a translated word for.

First, the special symbols "-", "&", "/"
A search for an English word including "etc." will be described. When an English word containing the special symbols "-", "&", and "/" is given, the dictionary is searched as appropriate for the English word. When the English word cannot be found in the dictionary, the English word is special symbol. Is divided into parts, and the search process is executed again for each of the divided English words. An output example is shown below. Input English word: electron-spin Headline English word 1: electron Translated word 1: Electronic Headword English word: spin Translated word 2: Spin In other modes, input English word: electron-spin Headword English word: electron translated word 1: [name ] Electron, electron Headword English 2: spin Translated word 2: [moving] spinning, spinning, turning,
Spinning The example shown above is a case where the contents of the translation word dictionary memory 107 are as follows. That is, electron [name] electron, electron spin [moving] spinning (threading), spinning,
Spin

Next, in FIG. 17, abc-de-f
The description will be made by taking the translation search process of gh as an example. First, ab
A translated word search is executed in the state of c-de-fgh (S17).
01). Then, in the abc-de-fgh state, it is determined whether the translated word is in the dictionary (S1702). If it is determined that it is in the dictionary, the process returns and the process described above is executed. On the contrary, if it is determined that the dictionary is not found in the dictionary, then a translated word search is executed in the abc-de state (S1703). Then, in the abc-de state, it is determined whether the translated word is in the dictionary (S1704). If it is determined that it is in the dictionary, a translation search for fgh is executed (S1709). On the contrary, if it is determined that the dictionary is not found in the dictionary, then a translation search is executed by abc (S170).
5). Next, a translated word search is executed in the de-fgh state (S1706), and then, in the de-fgh state, it is determined whether or not the translated word is in the dictionary (S1707). If it is determined that it is in the dictionary, it returns. Conversely,
If it is determined that the dictionary is not found in the dictionary, the translated word search is executed in de (S1708), the translated word search is executed in fgh (S1709), and then the process returns.

Further, in the above embodiment, various controls in the bilingual image forming apparatus are realized by software. However, in view of high speed of processing, hardware (whole or partial) may be used as necessary. It may be replaced with.

[0053]

As described above, according to the bilingual image forming apparatus of the present invention, a special symbol is detected, and when it is judged that it is at the end of a line, the special symbol is removed and the translated word information is obtained. When it is determined that the special character is not at the end of the line, the translated word information including the special symbol is searched, and when the translated word information is searched by removing the special symbol, the number of characters in the portion before the special symbol is calculated as follows. When it is determined that there is no corresponding translation word in the dictionary when comparing the translation information under the part with the larger number of characters and searching for the translation information including the special symbol , The part after the last special symbol is removed to search the translated word information, so the position of the translated word corresponding to the word existing in the manuscript image is clarified, and the content of the output image is significantly increased by adding the translated word. Changed to Also, with easy to lay outputs Read the copied image, appropriately performs processing to set words that are connected by special symbols, it is possible to improve the translation efficiency.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a bilingual image forming apparatus according to the present invention.

FIG. 2 is an explanatory diagram showing a configuration of a copying machine that embodies a bilingual image forming apparatus according to the present invention.

3 is a flowchart showing a main operation of the bilingual image forming apparatus shown in FIG.

FIG. 4 is a flowchart showing the operation of the document recognition process shown in FIG.

5 is a flowchart showing a classification operation of the special symbols shown in FIG.

FIG. 6 is a flowchart showing the operation of the paragraph / line / page combining process shown in FIG.

FIG. 7 is a flowchart showing the operation of the translation dictionary search process shown in FIG.

8 is a flowchart showing the operation of the translated word storage process shown in FIG.

9 is a flowchart showing the operation of the embedded composite image generation process shown in FIG.

10 is a flowchart showing the operation of the translated word synthesizing process shown in FIG.

FIG. 11 is an explanatory diagram showing contents of a structural format of a word and a divided word.

FIG. 12 is an explanatory diagram showing a process when a hyphen comes to the end of a line.

FIG. 13 is an explanatory diagram illustrating an example of a document image.

FIG. 14 is an explanatory diagram showing a state in which coordinates are designated for the document image shown in FIG.

15 is an explanatory diagram showing a state in which paragraph numbers are attached to the document image shown in FIG.

FIG. 16 is an explanatory diagram showing an example of a copied image output after executing a series of parallel translation adding processes according to the present invention on the original image shown in FIG.

FIG. 17 is a flowchart showing an example of division processing operation for a group of words.

[Explanation of symbols]

101 Scanner Unit 104 Input Image Memory 105 Output Image Memory 106 Recognition Dictionary Memory 107 Translation Dictionary Memory 108 Output Font Memory 109 CPU 110 ROM 111 111 RAM 114 Printer Unit 115 System Bus 116 Network Control Unit (NCU) 117 Modem 118 Communication Control Unit

Claims (3)

[Claims] 1. An image reading unit for optically reading a document image and converting it into image information, and an image storage unit for storing the image information read by the image reading unit or the image information input via a communication line. An image processing means for executing various image processing on the image information, an image recording means for recording an image on a recording medium based on an image signal output by the image processing means, the original image information and the information In a bilingual image forming apparatus having a translated word storage unit that stores translated word information and a control unit that executes a bilingual translation process on the document image information based on the information in the translated word storage unit, the control unit is configured to display a predetermined symbol. If it is detected and it is determined that it is at the end of the line, the predetermined symbol is removed to search the translation information, and if it is determined that it is not at the end of the line, the predetermined symbol is deleted. A parallel translation image forming apparatus characterized by searching for translated word information including the translation word information. 2. The control means compares the number of characters in the portion before the predetermined symbol with the number of characters in the following portion when the translation information is retrieved by removing the predetermined symbol, and the one having the larger number of characters is compared. 2. The bilingual image forming apparatus according to claim 1, wherein translated word information is synthesized under the portion of. 3. The control means, when searching the translated word information including the predetermined symbol, and judging that there is no corresponding translated word in the translated word storage means, the control means determines that the last (last) symbol or more 2. The bilingual image forming apparatus according to claim 1, wherein the translation information is retrieved by removing the portion.