JP2015225559A - Image processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily detect characters from an original image even if color shift occurs in the characters in the original image.SOLUTION: An image processing part 3 acquires image data for character recognition of an original image. An OCR processing part 4 performs character recognition processing of the image data for character recognition of the original image acquired by the image processing part 3. The image processing part 3 (a) identifies a color shift amount of the original image, and (b) makes the image data for character recognition either color image data or monochrome image data in accordance with the identified color shift amount.

Description

  The present invention relates to an image processing apparatus.

  In some apparatuses, a monochrome image in a document image is separated, and a character in the document image is processed by character recognition (OCR: Optical Character Recognition) processing on a monochrome binary image obtained by binarizing the separated monochrome image. It is detected (see, for example, Patent Document 1).

  In general, since there is a possibility that characters of various sizes are included in various languages in an original image, in order to increase the character recognition rate, It is desirable to perform OCR processing.

JP 2004-104435 A

  However, in the case of the above-described apparatus, if color misregistration occurs in characters in the document image, the character portion is not accurately separated as a monochrome image, and may not be accurately detected as a character in the OCR process. .

  Further, generally, when an OCR process is performed on a color original image in which a color shift occurs in characters (especially small characters), the edge of the character portion is blurred with a chromatic color. There is a possibility that the character portion becomes close to a gradation image and the character portion is not accurately detected as a character in the OCR processing.

  Such color misregistration of characters in a document image may occur when a color document is printed or may occur when a document is color scanned.

  The present invention has been made in view of the above problems, and an object of the present invention is to obtain an image processing apparatus in which characters can be easily detected from a document image even if color misregistration occurs in the characters in the document image.

  An image processing apparatus according to the present invention includes a character recognition image acquisition unit that acquires image data for character recognition of a document image, and character recognition processing for the character recognition image data acquired by the character recognition image acquisition unit. And a character recognition processing unit. The character recognition image acquisition unit (a) specifies the color misregistration amount of the document image, and (b) converts the character recognition image data into color image data and monochrome image data according to the color misregistration amount. Either.

  According to the present invention, it is possible to obtain an image processing apparatus in which characters are easily detected from a document image even if color deviation occurs in the characters in the document image.

FIG. 1 is a block diagram showing a configuration of an image processing apparatus according to an embodiment of the present invention. FIG. 2 is a diagram for explaining the amount of color shift used in the image processing apparatus shown in FIG. FIG. 3 is a flowchart for explaining the operation of the image processing apparatus shown in FIG.

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

  FIG. 1 is a block diagram showing a configuration of an image processing apparatus according to an embodiment of the present invention.

  The image processing apparatus illustrated in FIG. 1 is a multifunction peripheral, and includes an image reading apparatus 1, an image input unit 2, an image processing unit 3, an OCR processing unit 4, a display unit 5, an operation unit 6, a print control unit 7, and a printing device. 8, a state storage unit 9, a file creation unit 10, a data communication unit 11, and a system control unit 12.

  The image reading device 1 optically reads a document image from a document and generates image data of the document image. The image reading apparatus 1 includes a plurality of line sensors respectively provided for a plurality of color components (here, RGB), scans a document with the plurality of line sensors, and outputs an image of the document image as color image data. Generate data.

  The image input unit 2 acquires image data of a document image generated by the image reading device 1 or image data of a document image received by the data communication unit 11.

  The image processing unit 3 acquires character recognition image data of the document image. Specifically, image data for character recognition is generated for each page of the document image from the image data of the document image acquired by the image input unit 2, or the image data of the document image is used as character recognition image data. Or use it.

  The OCR processing unit 4 performs OCR processing on the character recognition image data acquired by the image processing unit 3 to detect characters in the document image.

  The image processing unit 3 (a) specifies the color misregistration amount of the original image, and (b) sets the character recognition image data as either color image data or monochrome image data according to the specified color misregistration amount. As the monochrome image data, gray scale image data or monochrome binary image data is used.

  FIG. 2 is a diagram for explaining the amount of color shift used in the image processing apparatus shown in FIG.

  For example, as shown in FIG. 2, when the RGB (or CMY) color component images 31 to 33 are shifted in the main scanning direction and the sub-scanning direction, the images 31 and 33 on both sides from the central image 32 in the main scanning direction. , Dp1 and Dp2, and the distance (or the number of pixels) from the center image 32 to the images 31 and 33 on both sides in the sub-scanning direction are Ds1 and Ds2. The color misregistration amount in the direction is the larger of Dp1 and Dp2, and the color misregistration amount in the sub-scanning direction is the greater of Ds1 and Ds2. As the color misregistration amount of the document image, for example, one of the color misregistration amount in the main scanning direction, the color misregistration amount in the sub scanning direction, or the larger one of the two is used.

  For example, the sum of Dp1 and Dp2 may be used as the color misregistration amount in the main scanning direction, and the sum of Ds1 and Ds2 may be used as the color misregistration amount in the sub scanning direction.

  When only one color image is shifted, the distance (or the number of pixels) between the one color image and the remaining two color images is used as the color shift amount.

  In this embodiment, when an image obtained by scanning by the image reading device 1 is used as a document image, the image processing unit 3 displays color misregistration information indicating a scan timing difference between the image reading device 1 and the line sensor. Acquired, and specifies the color shift amount of the document image from the acquired color shift information.

  In addition, when the character authentication image data is monochrome image data, the image processing unit 3 causes the image reading apparatus 1 to scan using only one of the plurality of line sensors, and the image reading apparatus 1 performs scanning. The obtained monochrome image data may be used as image data for character authentication, or a color image obtained by causing the image reading apparatus 1 to scan using a plurality of line sensors and scanning by the image reading apparatus 1. Monochrome image data converted from data may be used as image data for character authentication.

  At that time, the image processing unit 3 and the character recognition processing unit 4 sequentially perform generation of character recognition image data and character recognition processing for a plurality of document images, and the image processing unit 3 For a page having character authentication image data as monochrome image data, the image reading device 1 is scanned using only one of the plurality of line sensors, and the monochrome image data obtained by scanning by the image reading device 1 is used. It may be used as image data for character authentication. In that case, a white light source is used as the light source used corresponding to the line sensor used.

  Alternatively, the image processing unit 3 converts the monochrome image data converted from the color image data obtained by scanning with the image reading apparatus 1 using a plurality of line sensors, for pages in which the character authentication image data is monochrome image data. Is used as character authentication image data, and each page after the next page is scanned by the image reading device 1 using only one of the plurality of line sensors, and obtained by scanning by the image reading device 1. Monochrome image data may be used as image data for character authentication.

  The display unit 5 includes, for example, a liquid crystal display, an indicator, and the like, and displays the state of the image processing apparatus, a change in image quality setting for OCR processing, and the like. The operation unit 6 includes, for example, a touch panel, a hard key, etc., receives a user operation, and outputs it as an electrical signal.

  The print control unit 7 controls the printing device 8 to cause the printing device 8 to print a document image or the like. The printing device 8 prints an original image on a printing paper by, for example, an electrophotographic method.

  The state storage unit 9 is a storage device that stores the state of the image processing apparatus, various settings, and the like.

  The file creation unit 10 creates a file including raster image data of a document image and text obtained by character recognition by the OCR processing unit 4. For example, the file creation unit 10 creates a PDF (Portable Document Format) file including a raster image data layer of a document image and a text layer obtained by character recognition of the OCR processing unit 4.

  The data communication unit 11 is a network interface or the like, for example, and performs data communication with a terminal device such as a personal computer, receives image data of an original image from the terminal device, or receives a file created by the file creation unit 10 in a predetermined manner. Send to the terminal device by protocol.

  The system control unit 12 includes an image reading device 1, an image input unit 2, an image processing unit 3, an OCR processing unit 4, a display unit 5, an operation unit 6, a print control unit 7, a printing device 8, a status storage unit 9, and a file creation The unit 10 and the data communication unit 11 are controlled to execute various jobs.

  The image input unit 2, the image processing unit 3, the OCR processing unit 4, the print control unit 7, the file creation unit 10, and the system control unit 12 are a computer, ASIC (Application Specific Integrated Circuit) that executes a predetermined program. ) Etc.

  Next, the operation of the image processing apparatus will be described. FIG. 3 is a flowchart for explaining the operation of the image processing apparatus shown in FIG.

  The user places an original on the original table of the image reading apparatus 1 or the automatic original conveying apparatus, and performs an operation (for example, pressing of a start button) for executing a predetermined job (for example, scan to send) on the operation unit 6. To do. Upon receiving an instruction to execute the predetermined job from the operation unit 6, the system control unit 12 causes the image reading unit 1 to read the original image page by page using the image input unit 2, and the image processing unit 3 and The OCR processing unit 4 is caused to execute the following processing.

  The image processing unit 3 acquires, for example, original image data (color original image data) and color misregistration information of one page of the manuscript image from the image reading device 1 via the image input unit 2 (step S1). The amount of color misregistration is specified from the information (step S2).

  Then, the image processing unit 3 determines whether or not the specified color misregistration amount is equal to or greater than a predetermined threshold (step S3).

  When the specified color misregistration amount is equal to or greater than a predetermined threshold, the image processing unit 3 converts the original image data of the document image into monochrome image data (grayscale or monochrome binary) and character recognition of the monochrome image data. Image data, or through the image input unit 2, the image scanning device 1 is caused to scan the same page by gray scanning using a single line sensor, and monochrome image data is acquired as image data for character recognition. (Step S4). In this way, by using monochrome image data as the image data for character recognition, there is no variation in the saturation and hue of the edge of the character portion, so that the character recognition rate is improved.

  On the other hand, when the specified color misregistration amount is less than a predetermined threshold, the image processing unit 3 performs color image obtained by performing predetermined image processing on the original image data of the original image or the original image data of the original image. The data is character recognition image data (step S5).

  Then, the OCR processing unit 4 performs OCR processing on the character recognition image data generated by the image processing unit 3 to detect characters in the document image (step S6).

  As described above, according to the above embodiment, the OCR processing unit 4 performs the character recognition process on the character recognition image data of the document image acquired by the image processing unit 3. The image processing unit 3 (a) specifies the color misregistration amount of the original image, and (b) sets the character recognition image data as either color image data or monochrome image data according to the specified color misregistration amount.

  As a result, even if a color shift occurs in the characters in the document image, the characters are easily detected from the document image.

  The above-described embodiments are preferred examples of the present invention, but the present invention is not limited to these, and various modifications and changes can be made without departing from the scope of the present invention. is there.

  For example, the threshold value of the color misregistration amount in the above embodiment may be set by an input value to the operation unit 6 by the user or an input value received by the data communication unit 11 from an external terminal device.

  In the above-described embodiment, the image processing unit 3 may acquire original image data of a document image and specify a color shift amount of the document image based on the original image data. In this case, for example, the image processing unit 3 detects a position where edges of a plurality of predetermined color components (for example, RGB or CMY) are parallel in the document image as a color shift position, and determines the color shift amount at the position. Identify.

  The present invention can be applied to an image forming apparatus such as a scanner or a multifunction peripheral.

3 Image processing unit (an example of a character recognition image acquisition unit)
4 OCR processing unit (an example of a character recognition processing unit)

Claims (6)

A character recognition image acquisition unit that acquires image data for character recognition of a document image;
A character recognition processing unit that performs character recognition processing on the character recognition image data acquired by the character recognition image acquisition unit;
The character recognition image acquisition unit (a) specifies a color misregistration amount of the document image, and (b) determines the character recognition image data as one of color image data and monochrome image data according to the color misregistration amount. To do,
An image processing apparatus. A plurality of line sensors respectively provided for a plurality of color components, further comprising an image reading device that scans a document with the plurality of line sensors;
The character recognition image acquisition unit acquires color misregistration information indicating scan timing misalignment between the line sensors from the image reading device, and specifies the color misregistration amount of the document image from the color misregistration information;
The image processing apparatus according to claim 1.   When the character recognition image acquisition unit uses monochrome image data as the character authentication image data, the character recognition image acquisition unit causes the image reading device to scan using only one of the plurality of line sensors. 3. The image processing apparatus according to claim 2, wherein monochrome image data obtained by scanning is used as the image data for character authentication. The character recognition image acquisition unit and the character recognition processing unit sequentially perform generation of character recognition image data and character recognition processing on the plurality of pages of the document image,
The character recognition image acquisition unit causes the image reading apparatus to re-scan the page using the character authentication image data as monochrome image data by using only one of the plurality of line sensors. Using monochrome image data obtained by scanning with an apparatus as the image data for character authentication;
The image processing apparatus according to claim 3. The character recognition image acquisition unit and the character recognition processing unit sequentially perform generation of character recognition image data and character recognition processing on the plurality of pages of the document image,
The character recognition image acquisition unit is a monochrome image converted from color image data obtained by scanning with the image reading apparatus using the plurality of line sensors for a page having the character authentication image data as monochrome image data. The data is used as image data for character authentication, and each page after the next page is scanned by the image reading device using only one of the plurality of line sensors, and scanning by the image reading device is performed. Using the obtained monochrome image data as the character authentication image data;
The image processing apparatus according to claim 3.   The image processing apparatus according to claim 1, wherein the character recognition image acquisition unit acquires original image data of the document image and specifies a color shift amount of the document image based on the original image data.

Images