JP4291870B1 - Kanji recognition program, portable terminal device, and kanji recognition method - Google Patents

Abstract

Even in the case of a portable terminal device whose processing capability is not high-performance, it is possible to display a kana reading kana included in image data acquired by an imaging unit on a display unit at any time with a simple operation with a simple operation. A kanji recognition program, a portable terminal device, and a kanji recognition method are provided.
A mobile terminal device including an imaging unit and a display unit that obtains image data by a camera function determines whether or not a character included in the image data is a kanji, and the possibility of a kanji is found. Based on the kanji determination unit 20 that outputs a certain image portion, the character recognition unit 30 that performs character recognition processing on the image portion that is output by the kanji determination unit 20 and acquires a character code, and the character code that the character recognition unit 30 acquires. A reading acquisition unit 40 that reads the reading kana information from the reading data storage unit 60 and causes the display unit 50 to display the read reading kana information.
[Selection] Figure 1

Description

  The present invention relates to a Chinese character recognition program, a mobile terminal device, and a Chinese character recognition method used in a mobile terminal device having a camera function.

In recent years, there is a method of reading a dictionary by retrieving a character to be searched as an image, recognizing the character by OCR (Optical Character Recognition), even if it is a Chinese character whose reading is unknown when drawing a dictionary. For example, as disclosed in Patent Document 1, in a mobile phone with a camera function, character data is obtained by performing character recognition processing on captured image data recorded in an internal storage memory, and the character data is used as a keyword. There is a method for searching and displaying dictionary data.
JP 2005-301914 A

However, in the technique of the above-described Patent Document 1, it takes time to acquire an image by photographing, execute an instruction for character recognition processing, character recognition processing, dictionary search processing, and display processing, or if there are many input procedures for the operating user. There is a problem in that it is difficult to understand the flow of the text read up to that point and the story because complicated processing must be performed while reading the text.
When reading newspapers, books, documents, etc., the meaning can be inferred by the reader by taking into account the context before and after, but if you simply want to know only the reading of the kanji, There is a problem that it is difficult to guess.

  The present invention has been made in consideration of such circumstances and has been made to solve the above-described problems. The purpose of the present invention is to enable easy operation, and to read kana characters in kanji included in image data acquired by the imaging unit. Is to provide a kanji recognition program, a portable terminal device, and a kanji recognition method.

In order to solve the above problem, the present invention provides a computer used in a portable terminal device that includes an imaging unit that captures a character string and acquires it as image data, and a display unit. The character included in the image data is a Chinese character. A character recognition unit that determines whether or not and outputs an image portion that may be a kanji character, performs character recognition processing on the image portion output by the kanji determination unit, and acquires corresponding character identification information; A reading data storage unit associating character identification information with reading kana information indicating kanji reading kana, reading the reading kana information from the reading data storage unit based on the character identification information acquired by the character recognition unit, and the display parts to thereby function as reading acquisition unit for displaying, the Chinese character determination unit includes a plurality of said-size a maximum percentage of the number of black pixels to the total number of pixels in the image range to be determined character Black pixels for the total number of pixels in the image range of the character to be determined based on a threshold value that is different depending on whether or not the calculated maximum value exceeds a predetermined value. It is a kanji recognition program characterized in that it is determined whether or not the character to be determined is a kanji based on whether the number ratio exceeds the threshold .

Further, the kanji determination unit of the present invention determines whether the character to be determined is a kanji based on the number of white areas that are white pixel regions surrounded by black pixels in the image range. Features.

The kanji determination unit of the present invention is characterized in that, when the number of the white areas is four or more, the determination target character is determined to be a kanji.

  In addition, the kanji determination unit according to the present invention is characterized in that a determination target character sandwiched between kanji and kanji is determined as a partial kanji of a idiom.

In addition, the present invention is a portable terminal device that includes an imaging unit that captures a character string and acquires it as image data, and a display unit, and determines whether or not the character included in the image data is a Chinese character. A character recognition unit that outputs an image portion that may be a kanji, a character recognition unit that performs character recognition processing on the image portion output by the kanji determination unit and obtains corresponding character identification information, and character identification information; Based on the reading data storage unit that associates the reading kana information indicating the reading kana of the kanji and the character identification information acquired by the character recognition unit, the reading kana information is read from the reading data storage unit and displayed on the display unit The kanji determination unit is a maximum value of the ratio of the number of black pixels to the total number of pixels in the image range of the character to be determined, and the maximum value among the plurality of characters to be determined Calculate Whether or not the ratio of the number of black pixels to the total number of pixels in the image range of the character to be determined exceeds the threshold based on a different threshold depending on whether or not the calculated maximum value exceeds a predetermined value. It is a portable terminal device characterized by determining whether the character of determination object is a Chinese character based on this.

In addition, the present invention includes an imaging unit that captures a character string and acquires it as image data, a reading data storage unit that associates character identification information and reading kana information indicating kanji reading kana, and a display unit. A method for recognizing kanji in a mobile terminal device, comprising: determining whether a character included in the image data is a kanji and outputting an image portion that may be a kanji; and an output image portion It performs character recognition processing, and the corresponding character identification information character recognition process of obtaining, based on the obtained the character identification information, read kana reading information from the read data storing unit, and reading acquisition process to be displayed on the display unit In the kanji determination process, the maximum value of the ratio of the number of black pixels to the total number of pixels in the image range of the character to be determined, and the maximum value among the plurality of characters to be determined Whether or not the ratio of the number of black pixels to the total number of pixels in the image range of the character to be determined exceeds the threshold based on a different threshold depending on whether the calculated maximum value exceeds a predetermined value It is a Chinese character recognition method characterized by determining whether the character of judgment object is a Chinese character based on this.

According to the present invention, in the mobile terminal device, the kanji determination unit outputs the image portion corresponding to the kanji in the image data acquired by the imaging unit to the character recognition unit.
As a result, when character recognition processing is performed on an image portion to be character-recognized, the character recognition processing time can be minimized as compared with the case where character recognition processing is performed on the entire range of image data. As a result, it is possible to shorten the processing time until the reading kana is displayed on the display unit.

According to the present invention, the kanji determination unit determines whether or not the character to be determined is a kanji based on the ratio of the number of black pixels.
As a result, it is possible to recognize a character having a large number of strokes and a high ratio of black pixels as a Chinese character, and to display a reading kana for the image portion on the display unit.

According to the present invention, the kanji determination unit determines whether or not the character to be determined is a kanji based on pattern matching.
Thereby, since it can be determined whether a character is a Chinese character based on the characteristic of a Chinese character, there exists an effect that it becomes possible to perform a Chinese character detection process with high precision.

  Further, according to the present invention, the kanji determining unit determines the character to be determined sandwiched between the kanji and the kanji as part of the idiom. As a result, among the characters determined to be other than kanji, some characters of the idiom can be determined as kanji, and the kana can be displayed as the idiom.

Hereinafter, a mobile terminal device 1 according to an embodiment of the present invention will be described with reference to FIG.
FIG. 1 is a block diagram showing a mobile terminal device 1 according to the present embodiment. The mobile terminal device 1 is a computer terminal such as a PDA (Personal Digital Assistants) or a mobile phone, for example, and includes an imaging unit 10, a kanji determination unit 20, a character recognition unit 30, a reading acquisition unit 40, and a display unit 50. A reading data storage unit 60 and an input unit 70 are provided. The mobile terminal device 1 performs a character recognition process on the image data acquired by the imaging unit 10 every predetermined period t seconds, and reads a kana reading of a kanji character included in a predetermined position (for example, the central portion) of the image data. The image is displayed on the display unit 50 together with the image captured by the imaging unit 10.

The imaging unit 10 of the mobile terminal device 1 acquires image data by the camera function and outputs the image data to the kanji determination unit 20 and the display unit 50.
The kanji determination unit 20 extracts only the kanji part from the image data input from the imaging unit 10, and outputs the extracted image data of the kanji part to the character recognition unit 30.
The character recognition unit 30 has a timer function, and performs character recognition processing (OCR processing) on image data every predetermined period t seconds out of the image data input from the imaging unit 10. The kanji code is read and output to the acquisition unit 40. The kanji code is a character code (character identification information) for identifying a kanji.

Based on the kanji code input from the character recognition unit 30, the reading acquisition unit 40 reads a kanji reading kana corresponding to the kanji code from the reading data storage unit 60 and indicates the kanji reading kana read on the display unit 50. Outputs the reading information.
The display unit 50 is a display such as an LCD (Liquid Crystal Display), and displays image data input from the imaging unit 10 and reading kana information input from the reading acquisition unit 40 as images.
The reading data storage unit 60 is a storage device that stores a kanji code and a reading kana in association with each other, and stores sound reading for each kanji character, kanji reading, reading of a idiom combining two or more kanji characters, and the like.
The input unit 70 is an input device such as a keyboard, a touch panel, or a button. The input unit 70 receives an input operation by a user and outputs control signals for various functions of the mobile terminal device 1.

Next, the outline | summary of operation | movement of the portable terminal device 1 by one Embodiment of this invention is demonstrated using a flowchart. FIG. 2 is a diagram showing an outline of the reading display processing flow of the mobile terminal device 1.
In the mobile terminal device 1, the input unit 70 receives an input of a reading kana display request by a user input operation, and outputs an activation request to the imaging unit 10 and the kanji determination unit 20 (step S1).

In response to the activation request from the input unit 70, the imaging unit 10 activates the camera function, outputs the acquired image data to the kanji determination unit 20 and the display unit 50, and the display unit 50 displays the input image data. Display (step S2). In response to an activation request from the input unit 70, the kanji determination unit 20 extracts a kanji part from the center of the image data as kanji recognition processing for the image data input from the imaging unit 10, and also determines a kanji determination unit. 20 starts a timer for obtaining a timing of a predetermined period t seconds.
In addition, when the kanji part cannot be extracted from the center of the image data, the kanji determination unit 20 repeats the kanji recognition process until it determines that there is no kanji and extracts the kanji part. When the Chinese character determination unit 20 extracts the Chinese character part from the central part of the image data input from the imaging unit 10, the Chinese character determination unit 20 outputs the extracted image data to the character recognition unit 30 (step S3).

The character recognition unit 30 performs an OCR process on the image data of the kanji portion input from the kanji determination unit 20 (step S4), thereby acquiring a kanji code corresponding to the input kanji character. The acquired kanji code is read and output to the acquisition unit 40 (step S5).
The reading acquisition unit 40 reads the reading kana information corresponding to the input kanji code from the reading data storage unit 60 by, for example, reverse FEP processing, and outputs the read reading kana information to the display unit 50 (step S6). .

The display unit 50 displays image data sequentially input from the image capturing unit 10 at any time. When reading kana information is input from the reading acquisition unit 40, the display unit 50 displays the image data until the next reading kana information is input. Is displayed (step S7).
The kanji determination unit 20 repeats the process of step S3 every time a predetermined period t seconds elapses (step S8).
According to the above-described embodiment, the kanji determination unit 20 extracts only the kanji part from the image data, and outputs the extracted image data to the character recognition unit 30, so that the processing takes time depending on the size of the image. In the character recognition process, the character recognition processing time can be minimized. Accordingly, there is an effect that it is possible to shorten the processing time from the acquisition of the image data of the imaging unit 10 to the display of the kanji reading on the display unit 50.

  FIG. 3 is a diagram illustrating a screen transition example of the display unit 50. FIG. 3A is a screen display example of image data by the display unit 50 in step S2 of FIG. For example, in the case of the image shown in FIG. 3A, the kanji determination unit 20 extracts the kanji portion “player” from the center of the image and outputs it to the reading acquisition unit 40 in step S3. The reading kana information “senshu” is read from the reading data storage unit 60 and output to the display unit 50. FIG. 3B is a screen example of the display unit 50 when the reading information is displayed in FIG. In addition, although the case where only "senshu" is displayed as reading kana information is illustrated here, the present invention is not limited to this, and a plurality of reading kana information corresponding to kanji codes may be displayed.

  On the other hand, FIG. 3C shows a screen example of the display unit 50 when there is no kanji in the center of the image when the kanji determination unit 20 performs kanji recognition processing after displaying the reading kana information “senshu”. It is. As shown in the figure, when there is no kanji near the center, the kanji recognition process is repeated until the kanji determination unit 20 extracts the kanji portion in step S2, and after the character recognition by the character recognition unit 30, new kana information is obtained. Until the reading acquisition unit 40 outputs, the kana information “senshu” recognized last time is displayed on the display unit 50.

  FIG. 3D shows a screen example of the display unit 50 when the reading kana information “saiyo” corresponding to the kanji part “adopted” is displayed as the reading kana information. As shown in the figure, when the kanji determination unit 20 extracts the kanji part “adopted” from the center of the image, after the character recognition by the character recognition unit 30, the reading acquisition unit 40 displays the reading kana information “saiyo”. 50, the display unit 50 displays new reading kana information “saiyo” instead of the previous reading kana information “senshu”.

Next, details of the operation of the kanji recognition process by the kanji determination unit 20 will be described with reference to the drawings. FIG. 4 is a flowchart showing an outline of the flow of kanji recognition processing in the kanji determination unit 20.
The kanji determination unit 20 has a process of pre-processing from steps S11 to S18 and a process of main processing executed as a subroutine in steps S20 to S60 as kanji recognition processing. Detect as Kanji.

<Pretreatment>
The kanji determination unit 20 performs preprocessing to improve the extraction accuracy of the kanji part. The kanji determination unit 20 removes vertical and horizontal bar lines as a process of removing frame lines that are not character elements from the input image data subject to kanji recognition processing every predetermined period t seconds. Processing is performed (step S11). The bar line removal processing is performed by, for example, scanning the upper side, left side, lower side, and right side lines of the image to be processed one pixel at a time, determining black pixels as bar pixels, and determining all black pixels adjacent to the black pixels. This is a process of replacing a pixel with a white pixel. The black pixel at the edge of the image is likely to be part of a character or part of a bar. Also, since part of the characters and bar lines are not subject to character recognition processing, by replacing the black pixels existing at the edge of the image and other black pixels adjacent to the black pixels with white pixels, It becomes possible to remove black pixels unnecessary for the character recognition process.

  Next, the kanji determination unit 20 determines whether the character string included in the image data to be processed is vertical writing or horizontal writing (step S12). This determination is performed based on, for example, the ratio of black pixels to white pixels in the horizontal direction (x-axis direction) or vertical direction (y-axis direction) of the image. Specifically, for example, when determining the horizontal direction, the kanji determination unit 20 performs image scanning one line at a time in the horizontal direction from the upper left coordinates of the image data, and when a black pixel is detected, the line to be processed is determined. A line in which black pixels exist is determined as a line in which characters exist (character line), and image scanning processing for the next line is performed. If no black pixel is detected in the image scanning up to the end of the line, it is determined that a line in which no black pixel to be processed exists is a line corresponding to a space between characters or a line. The kanji determination unit 20 performs image scanning for all lines of the image data, and calculates the ratio between the number of character lines and the number of lines that are not characters.

When the calculated ratio is less than a predetermined ratio (for example, the ratio of non-character lines is less than 20% of the whole), the interval between non-character lines is between characters, and the target character It is determined that the character string is vertical writing. In addition, when the calculated ratio is equal to or greater than a predetermined ratio, the kanji determination unit 20 determines that the interval between lines that are not characters is between lines and the target character string is horizontal writing.
If the kanji determination unit 20 can determine whether the writing is vertical or horizontal, the process proceeds to step S14. In addition, when vertical writing or horizontal writing cannot be determined due to failure to detect a line that is not a character, the image data to be processed is discarded, and the kanji recognition process for the image data is stopped, The process proceeds to kanji recognition processing for image data (step S13).

The kanji determination unit 20 detects the line spacing based on the determination result of vertical writing or horizontal writing (step S14). For example, when the character string to be detected is vertically written, the kanji determination unit 20 performs image scanning in the vertical direction one column at a time starting from the center coordinate in the x-axis direction, and sets a column in which no character exists as a predetermined column. When a number (for example, three columns) is continuously detected, these columns are detected as the line spacing with respect to the character string at the center of the image data. Similarly, the kanji determination unit 20 operates the image one column at a time in the left direction starting from the center coordinate in the x-axis direction, and detects the line spacing existing on the left and right of the character string existing in the center in the x-axis direction.
If the kanji determination unit 20 can detect the line spacing, the process proceeds to step S16. In addition, when the line spacing cannot be detected due to failure to detect a column that is not a character, the image data to be processed is discarded, and the kanji recognition processing for the image data is stopped and another image data is transferred. The process proceeds to kanji character recognition processing (step S15).

The kanji determination unit 20 detects a space between characters in the image range sandwiched between the two lines detected in step S14 (step S16). For example, when the character string to be detected is vertically written, the kanji determination unit 20 scans the image horizontally by one line downward in the range from the left line to the right line in the x-axis direction. When lines that do not exist are detected, these lines are detected as characters. The kanji determination unit 20 repeats image scanning from the upper side to the lower side of the image data, and detects a range between the characters as an image range of one character.
If the kanji determination unit 20 detects the image range of one character by detecting the space between characters, the process proceeds to step S18. In addition, when a line between characters cannot be detected because a line that is not a character cannot be detected, the image data to be processed is discarded, and the kanji recognition process for the image data is stopped, and other image data The process shifts to kanji recognition processing (step S17).

The kanji determination unit 20 binarizes the image data of the character range detected based on the line spacing detected in step S14 and the character spacing detected in step S16 (step S18). The processing of steps S11 to S18 described above is preprocessing.
Note that the specific processing content in each process of the preprocessing described above is an example, and any processing can be applied as long as it is a processing that improves the recognition accuracy of the kanji recognition processing in the main processing described later.

<Main processing>
Next, an outline of a subroutine of the main process for detecting characters with a large number of strokes as kanji will be described. The kanji determination unit 20 performs kanji recognition processing based on the character area (step S20). FIG. 5 is a diagram showing an outline of the kanji recognition process, taking as an example the case where the process is performed on the vertically written character string “kanji recognition app”.
In FIG. 5A, character ranges M1 to M7 are character ranges for each character detected by the kanji determination unit 20 based on the line spacing detected in steps S11 to S18 in FIG. 4 and the character spacing. For each character range in the figure, the kanji determining unit 20 detects the total number of pixels in the character range and the number of black pixels in the character range, and calculates the ratio of the number of black pixels to the total number of pixels.

Next, the kanji determination unit 20 determines whether the character range is a kanji for each character range based on whether the calculated ratio of the number of black pixels is equal to or greater than a predetermined threshold (for example, 50%). To do.
When comparing kanji with non-kanji characters such as hiragana, katakana, and alphanumeric characters, kanji tends to have more strokes. In addition, there is a tendency that the number of strokes is also large for the kanji for which the user checks the reading kana. Since a character range with a high proportion of black pixels is considered to be a character with a large number of strokes, it can be determined whether or not the character range is kanji based on the proportion of black pixels.

  Note that the predetermined threshold value used in step S20 may be a different value depending on whether or not the maximum value of the ratio of the number of black pixels exceeds 50%. For example, in the character range M1 to M7 of the thin character string shown in FIG. 5A, the maximum black pixel number ratio is 48% (<50%) of the black pixel number of the character range M4. In the determination unit 20, a predetermined threshold value for determining whether or not the character is kanji is set to 30%. On the other hand, in the character range m1 to m7 of the bold character string shown in FIG. 5B, the maximum black pixel number ratio is 72% (≧ 50%) of the black pixel number of the character range m4. In the determination unit 20, a predetermined threshold value for determining whether or not the character is a Chinese character is set to 50%.

Thus, for example, even in the case of using different fonts and fonts for title characters and body characters, such as newspapers, the title character or body character is determined based on the maximum ratio of the number of black pixels. It is possible to increase the accuracy of kanji recognition by making a determination and setting the threshold value for the title text in bold type to be larger than the threshold value for the text character.
The kanji determination unit 20 determines the character ranges M1, M3, and M4 as kanji characters in FIG. 5 (a), and the character ranges m1, m3, and m4 in FIG. 5 (b).

  Returning to FIG. 4, the kanji determination unit 20 performs kanji recognition processing by pattern matching (step S30). The kanji recognition processing in step S30 is processing performed based on the number of closed white areas in the rectangle of each character range. In the rectangular range of the character range, the kanji determination unit 20 scans the image from the upper left, and when a white pixel is detected, the area of the all white pixels (white area) adjacent to the detected white pixel is painted on the black pixel. Repeat the process of filling the white area until all the black pixels are filled. The number obtained by subtracting 1 from the number of times of filling processing is the number of white areas in the rectangular range of the character range. Since the maximum number of white areas in hiragana, katakana, and alphanumeric characters is 3, here, the kanji determination unit 20 determines a character range having 4 or more white areas as kanji.

  As described above, kanji tends to have a large number of strokes, and a kanji having a large number of strokes may have a plurality of closed white areas. Based on this, it is possible to determine that there are four or more white areas as kanji. For example, in Hiragana, there are very few characters that have three white areas such as “A” (depending on the font) and “Nu”. It may be detected as

  FIG. 6 is a diagram showing an outline of the kanji recognition process, taking as an example the case where the process of step S30 is performed on the character “knowledge” in the character range M3 in the vertically written character string “kanji recognition app”. is there. As shown in the figure, in the character “K”, the white area needs to be painted five times before the entire area is painted with black pixels. "Is calculated as the number of white areas of the character" Kan ". Since the number of white areas is four or more, the kanji determination unit 20 determines the character range M3 as a kanji.

  Returning to FIG. 4, the kanji determination unit 20 determines whether or not all the character ranges (for example, the character ranges M1 to M7) detected in steps S11 to S18 are kanji, and determines any one or more character ranges. If it can be recognized as a Chinese character, the process proceeds to step S50. If no Kanji character can be detected in any of the character ranges, the image data to be processed is discarded and the Kanji recognition process for the image data is stopped, and the process proceeds to Kanji recognition processing for other image data. (Step S40).

  Next, the kanji determination unit 20 determines whether or not the detected kanji is an idiom (step S50). This idiom recognition process is a process for recognizing the preceding and following characters, that is, the character ranges in which the upper and lower adjacent character ranges are kanji as kanji. For example, the character range M2 in FIG. 5 does not satisfy the condition of being in Chinese characters for both the ratio of the number of black pixels and the number of white areas, but the character range M1 and the character range M3 before and after are both Chinese characters. Therefore, it is determined that the character range M2 is also a Chinese character.

  Returning to FIG. 4, the kanji determination unit 20 calculates the cutout coordinates of the image range of the kanji that is the character recognition target, and outputs the calculated cutout coordinate range to the character recognition unit 30 as the image data of the character recognition target (step). S60). In the cut-out coordinate calculation process in step S60, the kanji determination unit 20 detects the closest kanji from the center coordinates of the image data, that is, the x-coordinate center and the y-coordinate center, and before and after centering on the detected kanji. In this character range, a range until a character that is not a Chinese character appears is defined as a cutout range. Specifically, as shown in FIG. 7, the kanji determination unit 20 detects the range of the kanji with the kanji “knowledge” in the character range m4 closest to the center of the image as the center. In the figure, the kanji determination unit 20 sets the kanji character “Kan” in the character range m1 as the start character for extraction, and ends the character range m4 before the character range m5 in which the katakana “a” appears as a character that is not a Kanji character. A range to be cut out is detected as a character.

If there is only one Kanji character with a small number of strokes due to the main processing described above, it is considered that the reading difficulty of Kana characters is likely to be low. It is possible to reduce the number of processes.
In addition, by the process of step S50, the kanji (for example, the character “character” in the character range m2 as shown in FIG. 7) included in the idiom recognized as not being a kanji has a small number of strokes in steps S20 and S30. Thus, the image data of the idiom portion can be output to the character recognition unit 30 as a continuous group.

<Step S20: Kanji recognition processing based on black pixel number ratio>
Next, details of the subroutine processing of steps S20 to S60 of the main processing of FIG. 4 described above will be described with reference to the drawings. FIG. 8 is a flowchart showing the flow of kanji recognition processing based on the black pixel number ratio in step S20 of FIG.
The kanji determination unit 20 starts image scanning from the upper left coordinates in the rectangle of the character range to be processed (step S201). The kanji determination unit 20 determines whether or not the pixel value of the pixel to be processed is 0. If the pixel value is determined to be 0, the process proceeds to step S204 (step S202). If the pixel value is not 0, The number of black pixels is incremented by 1 (step S203).

The kanji determination unit 20 adds 1 to the total number of pixels (step S204), determines whether the pixel is the last pixel to be processed, and determines that the pixel is the last pixel, the process proceeds to step S207. If it is determined that the pixel is not the last pixel, the process proceeds to step S206 (step S205). The kanji determination unit 20 moves to the next pixel and returns to the process of step S202 (step S206).
The kanji determination unit 20 calculates the black pixel number ratio based on the following equation (step S207).
Black pixel ratio = (number of black pixels / total number of pixels) x 100

The kanji determination unit 20 determines whether or not the character range subject to kanji determination processing is the last character. When it is determined that the character range is the last character, the process proceeds to step S209 and is determined not to be the last character. If so, the process returns to step S1 for the next character (step S208).
The kanji determination unit 20 compares the calculated maximum value of the number of black pixels with 50% of the determination threshold value between the title part and the body part (step S209).
If the maximum value is less than 50% as a result of the comparison, the kanji determination unit 20 sets a predetermined threshold value for determining whether or not it is a kanji as 30% (step S210-1). If the maximum value is 50% or more, the kanji is determined. The predetermined threshold value for determining whether or not is 50% (step S210-2).

The kanji determination unit 20 compares the black pixel number ratio with a predetermined threshold (step S211), and determines that the character to be determined is a kanji character when the black pixel number ratio exceeds the predetermined threshold (step S212- 1) When the ratio of the number of black pixels is equal to or less than a predetermined threshold value, it is determined that the character to be determined is not a Chinese character (step S212-2).
The kanji determination unit 20 determines whether or not the character range subject to kanji determination processing is the last character. If it is determined that the character range is the last character, the kanji determination processing based on the black pixel number ratio ends. If it is determined that the character is not, the process returns to step S201 for the next character (step S213).

<Step S30: Kanji recognition processing based on pattern matching>
Next, details of the subroutine processing of step S30 of the main processing of FIG. 4 described above will be described with reference to the drawings. FIG. 9 is a flowchart showing the flow of kanji recognition processing based on pattern matching in step S30 of FIG.
The kanji determination unit 20 starts image scanning from the upper left coordinates in the rectangle (step S301), determines whether or not the pixel value is 0, and proceeds to step S303 if the pixel value is determined to be 0 (step S303). S302). The kanji determination unit 20 executes a subroutine for painting the white area (step S303), adds 1 to the number of paintings, and returns to the process of step S301 (step S304).

If the kanji determination unit 20 determines in step S302 that the pixel value is not 0, it determines whether or not the pixel is the last pixel to be processed, and if it determines that the pixel is the last pixel, step If it is determined that the pixel is not the last pixel, the process proceeds to step S306 (step S305), moves to the next pixel, and returns to the process of step S302 (step S306).
The kanji determination unit 20 calculates the number of closed white areas by subtracting 1 from the number of times of filling (step S307), and compares the calculated number of closed white areas with 3 (step S308). .

When the number of closed white areas is 4 or more, the kanji determination unit 20 determines that the character to be determined is a kanji (step S309-1).
If the number of closed white areas is 3 or less, the kanji determination unit 20 determines that the character to be determined is not a kanji (step S309-2).
The kanji determination unit 20 determines whether or not the character range subject to kanji determination processing is the last character. If it is determined that the character range is the last character, the kanji determination processing based on pattern matching is terminated, and the last character is determined. If it is determined that it is not, the process returns to step S301 for the next character (step S310).

<Step S303: White Area Filling Process>
Next, details of the subroutine processing in step S303 of FIG. 9 described above will be described with reference to the drawings. FIG. 10 is a flowchart showing the flow of white area filling processing in step S303 of FIG.
The kanji determination unit 20 replaces the pixel value “0” of the pixel to be processed with the pixel value “1” (step S331). The kanji determination unit 20 acquires the pixel value of the pixel (X, Y-1) that is one pixel above the pixel (X, Y) to be processed (step S332).
The kanji determination unit 20 determines whether or not the acquired pixel value is 1 (step S333). If the pixel value is not 1, that is, if the pixel value is 0, the pixel (X, Y−1) is further determined. ) For the white area filling process is executed (step S334).

When the pixel value is 1, the Chinese character determination unit 20 acquires the pixel value of the pixel (X + 1, Y) that is one right of the pixel (X, Y) to be processed (step S335).
The kanji determination unit 20 determines whether or not the acquired pixel value is 1 (step S336), and if the pixel value is not 1, that is, if the pixel value is 0, the pixel value (X + 1, Y) is further increased. A subroutine for white area filling processing is executed (step S337).
When the pixel value is 1, the Chinese character determination unit 20 acquires the pixel value of the pixel (X, Y + 1) immediately below the pixel (X, Y) to be processed (step S338).

  The kanji determination unit 20 determines whether or not the acquired pixel value is 1 (step S339). When the pixel value is 1, the kanji determination unit 20 ends the currently executed white area filling subroutine. When the pixel value is not 1, that is, when the pixel value is 0, the pixel (X, A subroutine of white area filling processing for Y + 1) is executed (step S340).

<Step S50: idiom recognition processing>
Next, details of the subroutine processing of step S50 of the main processing of FIG. 4 described above will be described with reference to the drawings. FIG. 11 is a flowchart showing the idiom recognition process in step S50 of FIG.
The kanji determination unit 20 determines whether or not the character range to be determined is kanji. If it is a kanji, the process proceeds to step S505. If not, the process proceeds to step S502 (step S501). The kanji determination unit 20 determines whether or not the character range immediately before the character range to be processed is kanji. If it is kanji, the process proceeds to step S503. If not, the process proceeds to step S505. Proceed (step S502).

The kanji determination unit 20 determines whether or not the character range immediately after the character range to be processed is a kanji. If it is a kanji, the process proceeds to step S504. If not, the process proceeds to step S505. Proceed (step S503).
When determining that the preceding and following character ranges are all kanji, the kanji determining unit 20 determines that the character range to be processed is also kanji (step S504).
The kanji determination unit 20 determines whether or not the character range subject to kanji determination processing is the last character. If it is determined that the character range is the last character, the kanji determination processing ends, and it is determined that the character range is not the last character. If so, the process proceeds to step S506 (step S505). The kanji determination unit 20 moves to the next character range and returns to the process of step S501 (step S506).

<Step S60: Cutout coordinate calculation process>
Next, details of the subroutine processing of step S60 of the main processing of FIG. 4 described above will be described with reference to the drawings. FIG. 12 is a flowchart showing the cut-out coordinate calculation process in step S60 of FIG.
The kanji determination unit 20 determines whether the character range to be determined is not kanji or kanji. If it is not kanji, the process proceeds to step S604, and if it is kanji, the process proceeds to step S602 (step S602). S601).
The kanji determination unit 20 determines whether or not the character range to be processed is closer to the center of the image than the central kanji determined so far. If the character range is closer to the central kanji, the process proceeds to step S603. If it is too far, the process proceeds to step S604 (step S602). The kanji determination unit 20 determines that the character range to be processed is the central kanji (step S603).

  The kanji determination unit 20 determines whether or not the character range subject to determination processing is the last character. If it is determined that the character range is the last character, the process proceeds to step S607 and determines that the character range is not the last character. In this case, the process proceeds to step S605 (step S604). The kanji determination unit 20 moves to the next character range and returns to the process of step S601 (step S605).

The kanji determination unit 20 sets the reference character range of the processing target as the central kanji (step S606), and moves the processing target character range to the previous character range (step S607).
The kanji determination unit 20 determines whether or not the character range to be processed is the first character range. If the character range is the first character range, the process proceeds to step S611. If the character range is not the first character range, the process proceeds to step S609. (Step S608).
The kanji determination unit 20 determines whether or not the character range to be processed is kanji. If it is a kanji, the process proceeds to step S610. If not, the process proceeds to step S611 (step S609). The kanji determination unit 20 determines that the character range to be processed is a start character, and returns to the process of step S607 (step S610).

The kanji determination unit 20 sets the reference of the character range to be processed as a central Chinese character (step S611), and moves the character range to be processed to the next character range (step S612).
The kanji determination unit 20 determines whether or not the character range to be processed is the last character range. If the character range is the last character range, the cut-out coordinate calculation process ends, and if it is not the last character, the process proceeds to step S614. The process proceeds (step S613).
The kanji determination unit 20 determines whether or not the character range to be processed is a kanji. If it is a kanji, the cut coordinate calculation process ends, and if it is not a kanji, the process proceeds to the process of step S615 (step S614). . The kanji determination unit 20 determines that the character range to be processed is an end character, and returns to the process of step S612 (step S615).

In the present embodiment, a pixel value of 0 (zero) means a white pixel, and a pixel value of 1 means a black pixel.
In the above-described processing, a specific processing example for a vertically written character string has been described. However, for the horizontally written character string, the above processing is performed in the horizontal direction (x-axis direction) and the vertical direction (y-axis direction). Is a process in which right and left and up and down are replaced.
According to the embodiment described above, by determining a character with a large number of strokes as a kanji to be detected, it is easy even in the case of a portable terminal device whose processing capability is not high performance for the target kanji for which the user wants to check the reading kana. Thus, it becomes possible to detect with high accuracy by a method with a short processing time (light operation).

In addition, since the processing time is short and kanji recognition processing can be continuously performed and the reading kana information can be displayed on the display unit 50, the reading kana can be quickly grasped simply by holding the difficult-to-read kanji over the camera portion. It is possible to give the user a sense of use like a loupe, thereby reducing stress in terms of speed of response.
Further, the present invention is not limited to difficult-to-read kanji characters, and the present invention can be applied to cases where it is difficult to read a character having a larger number of strokes than hiragana, or when it is affected by presbyopia (so-called presbyopia) that makes it difficult to see nearby characters By using the portable terminal device 1, it is possible to suppress reading hindrance by displaying a reading kana of characters having a large number of strokes.

In addition, since kanji recognition processing is performed without recording an image to be processed on a recording medium such as a micro SD memory card provided in the imaging unit 10, reading kana information is displayed without saving image data that takes time. There is an effect that it is possible to display on the part 50.
In addition, kanji that is difficult to read with kanji with a small number of strokes tend to be kanji other than regular kanji, and kanji other than regular kanji is not used in newspapers. By excluding them as characters, it is possible to omit the exception process and simplify the reading display process.
As described above, according to an embodiment of the present invention, it is possible to quickly display a reading kana of a character with a large number of strokes on the display unit 50, so that it is possible to reduce the user from losing sight of the story.

  In the above-described kanji recognition processing, it has been described that the kanji recognition processing results of both the black pixel number ratio and the pattern matching are used, but only one of them may be used. This can further simplify the kanji recognition process.

In addition, the above-mentioned portable terminal device 1 has a computer system inside. The operation process of the imaging unit 10, the kanji determination unit 20, the character recognition unit 30, the reading acquisition unit 40, the display unit 50, the reading data storage unit 60, and the input unit 70 of the mobile terminal device 1 is a program format. Is stored in a computer-readable recording medium, and the above-described processing is performed by the computer system reading and executing this program. The “computer system” herein includes a CPU, various memories, an OS, and hardware such as peripheral devices.
Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.

2, 4 and 8 to 12 are recorded on a computer-readable recording medium, and the function of the processing unit in the mobile terminal device 1 shown in FIG. 1 is realized. In the image data acquired by the mobile phone terminal device using the camera function by recording the program for recording on a computer-readable recording medium, causing the computer system to read and execute the program recorded on the recording medium, A process of recognizing the part and cutting out the image data of the Chinese character part as a character recognition target may be performed.
The “computer-readable recording medium” means a flexible disk, a magneto-optical disk, a ROM, a writable nonvolatile memory such as a flash memory, a portable medium such as a CD-ROM, a hard disk built in a computer system, etc. This is a storage device.

Further, the “computer-readable recording medium” refers to a volatile memory (for example, DRAM (Dynamic DRAM)) in a computer system that becomes a server or a client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. Random Access Memory)), etc., which hold programs for a certain period of time.
The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. Here, the “transmission medium” for transmitting the program refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line.
The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, and what is called a difference file (difference program) may be sufficient.

It is a block diagram which shows the whole structure of the portable terminal device 1 by one Embodiment of this invention. It is a flowchart which shows the outline | summary of operation | movement of the portable terminal device 1 in the embodiment. It is a figure which shows the example of a display of the display part 50 in the same embodiment, and a screen transition example. It is a flowchart which shows the outline | summary of operation | movement of the Chinese character determination part 20 in the embodiment. It is a figure which shows the specific example of the Chinese character recognition process based on the black pixel number ratio in the embodiment. It is a figure which shows the specific example of the Chinese character recognition process based on the pattern matching in the embodiment. It is a figure which shows the specific example of the cut-out coordinate calculation process in the embodiment. It is a figure which shows the operation | movement flow of the Chinese character recognition process based on the black pixel number ratio in the embodiment. It is a figure which shows the operation | movement flow of the Chinese character recognition process based on the pattern matching in the same embodiment. It is a figure which shows the operation | movement flow of the white area filling process in the embodiment. It is a figure which shows the operation | movement flow of the idiom recognition process in the embodiment. It is a figure which shows the operation | movement flow of the cut-out coordinate calculation process in the embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Portable terminal device 10 Imaging part 20 Kanji determination part 30 Character recognition part 40 Reading acquisition part 50 Display part 60 Reading data storage part 70 Input part

Claims (6)

A computer used in a mobile terminal device that includes an imaging unit that captures a character string and obtains it as image data and a display unit determine whether or not the character included in the image data is kanji, and the possibility of kanji Kanji determination unit that outputs a certain image part,
A character recognition unit that performs character recognition processing on an image portion output by the kanji determination unit and acquires corresponding character identification information;
A reading data storage unit that associates character identification information with reading kana information indicating kanji reading kana,
Based on the character identification information acquired by the character recognition unit, read kana information from the reading data storage unit, function as a reading acquisition unit to be displayed on the display unit ,
The kanji determination unit calculates the maximum value of the ratio of the number of black pixels to the total number of pixels in the image range of the determination target character and calculates the maximum value among the plurality of determination target characters. The character to be determined based on whether or not the ratio of the number of black pixels to the total number of pixels in the image range of the character to be determined exceeds the threshold based on a different threshold depending on whether or not exceeds a predetermined value. A kanji recognition program characterized by determining whether or not is a kanji. The kanji determination unit
The kanji recognition according to claim 1, wherein it is determined whether or not the character to be determined is a kanji character based on the number of white areas that are white pixel regions surrounded by black pixels in the image range. program. The kanji determination unit
When the number of white areas is 4 or more, the character to be determined is determined to be kanji
The Chinese character recognition program according to claim 2, wherein: The kanji determination unit
The kanji recognition program according to any one of claims 1 to 3, wherein a determination target character sandwiched between kanji and kanji is determined as a partial kanji of the idiom. A mobile terminal device including an imaging unit that captures a character string and acquires it as image data, and a display unit,
Determining whether or not the characters included in the image data are kanji, and outputting a kanji determination unit that may be a kanji; and
A character recognition unit that performs character recognition processing on the image portion output by the kanji determination unit and acquires corresponding character identification information;
A reading data storage unit associating character identification information with reading kana information indicating kanji reading kana,
Based on the character identification information acquired by the character recognition unit, the reading acquisition unit that reads the reading kana information from the reading data storage unit and displays on the display unit ,
The kanji determination unit calculates the maximum value of the ratio of the number of black pixels to the total number of pixels in the image range of the determination target character and calculates the maximum value among the plurality of determination target characters. The character to be determined based on whether or not the ratio of the number of black pixels to the total number of pixels in the image range of the character to be determined exceeds the threshold based on a different threshold depending on whether or not exceeds a predetermined value. It is determined whether or not is a Chinese character. Kanji recognition in a portable terminal device comprising: an imaging unit that captures a character string and acquires it as image data; a reading data storage unit that associates character identification information with reading kana information indicating kanji reading kana; and a display unit A method,
Determining whether or not the character included in the image data is kanji, and outputting a portion of the image that may be kanji,
A character recognition process for performing character recognition processing on the output image portion and obtaining corresponding character identification information;
Based on the obtained the character identification information, read kana reading information from the read data storage unit, possess a reading acquisition process to be displayed on the display unit,
In the kanji determination process, the maximum value of the ratio of the number of black pixels to the total number of pixels in the image range of the character to be determined and the maximum value among the plurality of characters to be determined is calculated, and the calculated maximum Based on a threshold value that differs depending on whether the value exceeds a predetermined value, based on whether the ratio of the number of black pixels to the total number of pixels in the image range of the character to be determined exceeds the threshold value, A method for recognizing kanji, wherein it is determined whether or not the character is kanji .

Images