JP2012205133A - Image processor and control method therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately determine a character area included in an image.SOLUTION: An image processor includes: division means for dividing an image into multiple blocks respectively having a predetermined number of pixels; derivation means for deriving the edge intensity values of the respective pixels; reference value derivation means for deriving a reference value based on the maximum value and minimum value of the edge intensity value of each pixel; threshold setting means for setting a first threshold and a second threshold based on the reference value; counting means for counting the number of pixels having the edge intensity values to be respectively included in a first range equal to or more than the first threshold value and lower than the second threshold value and a second range equal to or more than the second threshold; and determination means for determining a character area when the pixels having the edge intensity values of the second range exist and determining a non-character area when the pixels having the edge intensity values of the second range and the first range do not exist, based on the counting result.

Description

  The present invention relates to an image processing technique, and more particularly to a technique for detecting a character area included in an image.

  In general, a document read by a digital multifunction peripheral or the like is roughly classified into three types: a character document, a photo document, and a halftone print document. A halftone print original is an image in which an image with gradation is decomposed into a collection of fine dots, and the gradation is expressed in a pseudo manner by the size of the dots, that is, the density. For these original images read by the scanner, it is often performed to identify areas such as characters, halftone dots, and photographs, and to perform processing suitable for each attribute. For example, after processing with emphasis on resolution for character areas, character recognition and character vectorization processing are performed, and processing with emphasis on gradation is performed on halftone areas and photo areas. Try to improve. Therefore, it is necessary to identify characters, halftone dots, and the like with high accuracy.

  Conventionally, as a technique for identifying characters, halftone dots, and the like, a method of determining using color information or edge information has been proposed. In Patent Document 1, an input image is divided into blocks of a certain size, the sum of absolute values of density differences between adjacent pixels is obtained for each block, and each block is compared with a predetermined threshold value. There has been proposed a method for determining whether a halftone area, a character area, or a halftone area. Further, in Patent Document 2, if a continuous edge in which a boundary portion having a large pixel density difference in at least one of the main scanning direction and the sub-scanning direction continues for two or more pixels is not detected, it is determined as a character region, and if a continuous edge is detected. Judged as non-character area.

  Furthermore, in Patent Document 3, the number of edge pixels in a block is calculated, and a block in which the calculated number of edge pixels is equal to or greater than a threshold value is determined as a halftone dot region, and a block in which the calculated edge pixel number is less than the threshold value is determined as a non-halftone dot region (character region). A method has been proposed. Further, in Patent Document 4, a halftone dot or a character line contour component is extracted, a halftone dot contour component is removed from the extracted contour component, and the number of black pixels in the M × N pixel block is counted. Then, the character area and the halftone image area are identified by the number of black pixels in the block. At this time, the halftone dot area is an area having a small number of black pixels, and thus is identified as a halftone image area.

Japanese Patent Publication No. 5-50187 Japanese Patent Laid-Open No. 08-51537 JP 2002-158873 A Japanese Patent Laid-Open No. 01-286571

  However, depending on the density and composition color of the halftone dots constituting the halftone dot area, the halftone dot area and the character area may contain the same edge component or the density difference between the pixels may be equivalent. . In such a case, it becomes easy to misrecognize the dot area and the character area.

  Further, even in an image with characters on a halftone dot background as shown in FIG. 1, erroneous recognition is likely to occur for the following reason. FIG. 1A is a diagram exemplarily showing an image in which white characters are present on a halftone dot background composed of black and white. In the figure, it is assumed that a block of a predetermined size (square block) is divided and the area of each block is determined. In this example, both a block including only halftone dots (halftone dot block) and a block including halftone dots and characters (character block) are expressed in white and black, so that the density between pixels included in each block is expressed. The difference between the character block and the halftone dot block cannot be distinguished. In addition, since black and white continuous edges exist in both the character block and the halftone dot block, it is difficult to distinguish the character block from the halftone dot block based on the edge information. FIG. 1 (b) shows an image containing dark blue letters on a light blue background. In this case, since the color difference between the character color and the background color is small, it is difficult to discriminate between a character block and a non-character block in the determination using colors. In addition, since the density difference between the character and the background is small and it is difficult to recognize continuous edges around the character, it is difficult to determine a character block and a non-character block even by edge information. As a result, erroneous determination of a region occurs, and when image processing based on the erroneous determination result is performed, image quality deterioration occurs.

  The present invention has been made in view of the above-described problems, and an object of the present invention is to provide an image processing technique for suitably determining an attribute of a region in an image.

  In order to solve one or more problems described above, an image processing apparatus according to the present invention includes a dividing unit that divides an image into a plurality of blocks each including a predetermined number of pixels, and an edge of each pixel included in the target block. Deriving means for deriving an intensity value, reference value deriving means for deriving a reference value in the target block based on the maximum and minimum edge intensity values of pixels included in the target block, and the derived reference value Based on the threshold value setting means for setting a first threshold value greater than the reference value and a second threshold value greater than the first threshold, and in the target block, a first threshold value that is greater than or equal to the first threshold value and less than the second threshold value. Counting means for counting the number of pixels having edge intensity values included in each of the first range and the second range equal to or greater than the second threshold, and based on the result of counting by the counting means, the second category. If the pixel having the edge intensity value of the second range and the first range does not exist, the target block is determined to be a non-character region. Determining means.

  ADVANTAGE OF THE INVENTION According to this invention, the image processing technique which determines suitably the attribute of the area | region in an image can be provided.

It is a figure which shows the example in which determination of a character area and a non-character area is difficult. It is a block diagram which shows the function structure of the image processing apparatus which concerns on 1st Embodiment. FIG. 3 is a block diagram of a digital multi-function peripheral (MFP) in which the image processing apparatus shown in FIG. 2 is mounted. It is a flowchart explaining an edge feature-value output process. It is a figure explaining a reference value (reference threshold) derivation processing. It is a figure explaining each parameter of the edge feature-value concerning 1st Embodiment. It is a flowchart which derives | leads-out each parameter of the edge feature-value based on 1st Embodiment. It is a figure which shows the example of the edge extraction result by each threshold value. It is a flowchart of the area | region determination process which concerns on 1st Embodiment. It is a figure explaining each parameter of the edge feature-value concerning 2nd Embodiment. It is a flowchart which derives | leads-out each parameter of the edge feature-value which concerns on 2nd Embodiment. It is a flowchart of the area | region determination process which concerns on 2nd Embodiment.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings. The following embodiments are merely examples, and are not intended to limit the scope of the present invention.

(First embodiment)
A first embodiment of an image processing apparatus according to the present invention will be described below by taking a digital multi-function peripheral (MFP) as an example.
<Device configuration>
FIG. 3 is a block diagram showing a main configuration of a digital multi-function peripheral (MFP) in which the image processing apparatus according to the present invention is mounted. FIG. 2 is a block diagram showing a detailed configuration of the scanner image processing unit 2080 of the image processing apparatus according to the first embodiment. In the present embodiment, the scanner image processing unit 2080 is configured by hardware such as an electronic circuit, and the hardware of the scanner image processing unit is described as executing image processing to be described later, but is not limited thereto. . For example, the CPU 2001 may be configured to function as each processing unit of the scanner image processing unit 2080 by executing a program stored in the HDD 2004.

  As shown in FIG. 3, the MFP includes a controller unit 2000 that functions as a control unit of the image processing apparatus. The controller unit 2000 is connected to a scanner 2070 as an image input device and a printer 2095 as an image output device via a device interface (I / F). Then, the controller unit 2000 can perform control for realizing a copy function for printing out image data obtained by reading a document with the scanner 2070 by the printer 2095. The controller unit 2000 can perform control for inputting / outputting pattern images, device information, and the like to / from other apparatuses via the LAN 1006 or the public line (WAN) 1008.

  As shown in FIG. 3, the controller unit 2000 has a CPU 2001. The CPU 2001 starts up an operation system (OS) by a boot program stored in the ROM 2003. Various processes are executed by executing application programs stored in an HDD (Hard Disk Drive) 2004 on the OS. A RAM 2002 is used as a work area for the CPU 2001. The RAM 2002 is used not only as a work area for the CPU 2001 but also as an image memory area for temporarily storing image data. The HDD 2004 stores the application program and image data.

  A ROM 2003 and a RAM 2002, an operation unit I / F 2006, a network I / F 2010, a modem 2050, and an image bus I / F 2005 are connected to the CPU 2001 via a system bus 2007.

  An operation unit I / F (interface) 2006 is an interface with an operation unit 2012 having a touch panel, and outputs image data to be displayed on the operation unit 2012 to the operation unit 2012. Further, the operation unit I / F 2006 sends information input by the user through the operation unit 2012 to the CPU 2001.

  A network I / F (interface) 2010 is connected to the LAN 1006 and inputs / outputs information to / from each device connected to the LAN 1006 via the LAN 1006. The modem 2050 is connected to the public line 1008 and inputs / outputs information to / from other devices via the public line 1008.

  An image bus I / F (interface) 2005 is a bus bridge for connecting a system bus 2007 and an image bus 2008 for transferring image data at high speed and converting a data structure. The image bus 2008 includes a PCI bus or IEEE1394. A raster image processor (RIP) 2060, a device I / F 2020, a scanner image processing unit 2080, a printer image processing unit 2090, an image rotation unit 2030, and an image compression unit 2040 are connected to the image bus 2008.

  The RIP 2060 is a processor that develops a PDL code into a bitmap image. A scanner 2070 and a printer 2095 are connected to the device I / F 2020, and perform synchronous / asynchronous conversion processing of image data transfer. A scanner image processing unit 2080 corrects, processes, and edits input image data input from the scanner. A printer image processing unit 2090 generates print data by performing print correction, resolution conversion, and the like on the print output image data, and outputs the print data to the printer. The image rotation unit 2030 rotates image data. The image compression unit 2040 compresses multi-value image data into JPEG data and binary image data into data such as JBIG, MMR, and MH, and also performs decompression processing.

  Next, the configuration of each processing unit used when the scanner image processing unit 2080 determines the attribute of each region in the image and performs predetermined image processing according to each attribute will be described with reference to FIG. For example, the block input unit 11 converts a monochrome or color image obtained by reading a paper document using the scanner 2070 into a plurality of block images each including a predetermined number of pixels (hereinafter also simply referred to as “blocks”). To divide. That is, the input image is divided into blocks each having a predetermined size. Then, the divided block images are input to the edge extraction unit 12 and the smoothing processing unit 15.

  The edge extraction unit 12 extracts edge strength from the block input from the block input unit 11. The edge intensity extracted by this processing can use differential data obtained by applying a known Prewitt filter or Sobel filter to each pixel in the block. It is. The edge feature extraction unit 13 calculates a reference value (hereinafter also referred to as a reference threshold) for the edge strength for each block from the edge strength (edge strength value) of each pixel calculated by the edge extraction unit 12. An edge feature acquisition parameter is set from the reference threshold value to obtain an edge feature amount. Details of the edge feature extraction processing will be described later with reference to FIGS. The region determination unit 14 determines whether the block of interest is a character region or a non-character region based on the edge feature amount determined by the edge feature extraction unit 13. Details of this area determination processing will be described later with reference to FIG.

  The smoothing processing unit 15 performs a smoothing process on the block image input by the block input unit 11. The smoothing process is for reducing fine changes for each pixel included in the block image, squashing the halftone dot pattern, and facilitating the analysis of the color information. For the smoothing process, a known low-pass filter, moving average filter, edge holding type smoothing filter, or the like can be used.

  The representative color selection unit 16 analyzes the distribution of the appearance colors of the pixels in the block image smoothed by the smoothing processing unit 15 and refers to the region determination result of the block of the region determination unit 14 to represent the representative color (representative color). Select (Candidate). If the result of area determination is a character area, a plurality of colors (character color, background color) are selected as representative colors. If the result of area determination is a non-character area (for example, a monochrome background), a representative color is selected. Decide on one color. The representative color selection process can use a method of selecting a color having a high frequency of appearance colors of each pixel in the block.

  The color quantization unit 17 assigns one of the representative colors selected by the representative color selection unit 16 to each pixel in the block of interest. In this color quantization process, a representative color having the shortest color distance is assigned to each pixel in a block in which a plurality of representative colors are selected, and for a pixel in a block having one representative color. It is possible to use a method of assigning the same one color.

  Further, the image processing may be switched on a block basis. For example, edge enhancement may be performed on character pixels in a character block, or pixels in a non-character block may be smoothed. Also, using block processing results and information between blocks, connecting character parts and separating them into character units, then performing character recognition processing and vectorization processing that extracts character outlines and converts them into vector data It may be configured to do.

<Edge feature amount>
FIG. 4 is a flowchart showing details of the edge feature amount extraction processing executed by the edge feature extraction unit 13.

  In step S131, the edge strength of each pixel in the block of interest acquired by the edge extraction process is input. In step S132, a reference value (reference threshold) in the target block is derived based on the edge strength of each pixel in the target block (reference value deriving unit).

  FIG. 5 is a diagram for explaining a reference threshold value derivation process according to the first embodiment. The solid line exemplarily shows the distribution of the edge intensity of each pixel in the block acquired from the edge extraction unit 12. In a plurality of blocks obtained by dividing a document image, the difference between the maximum value and the minimum value of the edge strength of pixels (that is, the range of edge strength) can vary greatly. Therefore, in order to perform the same processing for all of a plurality of blocks, it is considered that the edge feature amount is determined after normalizing the range of the edge strength of the pixel to a value of 0 to 255 (8-bit value). It is done. (Dotted lines indicate the distribution of edge strength when normalized.) However, if the edge strength of each pixel in the block of interest is normalized, the processing load increases. Therefore, in the present embodiment, instead of normalizing the edge intensity of each pixel, the threshold T in the actual edge intensity range is set as follows based on the threshold F used in normalization. That is, the threshold value F (empirically set value (eg, “120”)) set in advance in the edge strength range of 0 to 255 is used to set the reference threshold T in the actually acquired edge strength range as follows: Obtained by Equation 1.

T = (Emax−Emin) × F / 255 + Emin (Formula 1)
Here, Emin and Emax are the minimum value and the maximum value of the edge intensity of the pixel in the block of interest obtained by the extraction process, respectively. That is, the value of the threshold F (predetermined constant) is common to each block, but the values of Emin and Emax are different for each block. Therefore, the value of T can be different for each block.

  In step S133, an edge feature amount derivation parameter is set using the derived reference threshold value T (threshold setting means), and in step S134, an edge feature amount is derived using the set parameter. In step S135, the derived edge feature amount is output.

FIG. 6 is a diagram for explaining parameters used when deriving the edge feature amount according to the first embodiment. The edge feature derivation parameters include a reference threshold T (third threshold), a strong edge determination threshold ST (second threshold) obtained based on the reference threshold T, and a weak edge determination threshold WT (first threshold). There are three thresholds. here,
Reference threshold T
Strong edge determination threshold ST = T + B
Weak edge determination threshold WT = T + C
B and C are constants specified in advance set so as to satisfy ST>WT> T. For example, B = 70 and C = 50.

In addition, the edge feature is derived based on the edge strength of each pixel in the block of interest and the above parameters. Specifically,
Number of strong edge pixels: number of pixels having edge strength of ST or more (second range of second threshold or more) weak edge pixels: first of WT or more and less than ST (first threshold or more and less than second threshold) Number of pixels having edge strength of (range) Reference edge pixel number: obtained by counting the number of pixels having edge strength of T or more and less than WT (third range that is greater than or equal to the third threshold and less than the first threshold). It is

  FIG. 7 is a flowchart for deriving each parameter of the edge feature amount according to the first embodiment.

  In step S13301, the edge strength of each pixel in the block of interest and the reference threshold T of the block of interest are input. In step S 13302, parameters for edge feature calculation are set from the reference threshold T. For example, the strong edge determination threshold ST and the weak edge determination threshold WT are set to T + 70 and T + 50, respectively.

  In steps S13303 to S13311, the edge strength is determined for each pixel in the block of interest (counting means). In step S13303, it is determined whether the edge intensity of the pixel to be processed is equal to or greater than the strong edge determination threshold ST. If the edge strength of the pixel is greater than or equal to the strong edge determination threshold ST, the process proceeds to step S13304, where the pixel is determined to be a strong edge pixel, and the count value SN of the strong edge pixel is incremented by one. On the other hand, if the edge strength of the pixel is less than the strong edge determination threshold ST, the process proceeds to step S13305.

  In step S13305, it is determined whether the edge intensity of the pixel is equal to or greater than the weak edge determination threshold WT. If the edge strength of the pixel is equal to or greater than the weak edge determination threshold WT, the process proceeds to step S13306, where the pixel is determined to be a weak edge pixel, and the count value WN of the weak edge pixel is increased by one. If the edge strength of the pixel is less than the weak edge determination threshold WT, the process proceeds to S13307.

  In step S13307, it is determined whether the edge intensity of the pixel is greater than or equal to the reference threshold T. If the edge strength of the pixel is greater than or equal to the reference threshold T, the process proceeds to step S13308, where the pixel is determined to be an edge pixel of the reference threshold, and the count value TN is incremented by one. If the edge strength of the pixel is less than the reference threshold T, the process proceeds to S13311.

  In step S13311, it is determined whether or not the processing has been completed for all the pixels included in the block of interest. If there is an unprocessed pixel, the process returns to step S13303, and the above determination is repeated for the next pixel. If there is no unprocessed pixel, the process advances to step S13312 to output edge feature amounts (SN, WN, TN).

<Example of edge extraction result by each threshold>
FIG. 8 is a diagram illustrating an example of an edge feature extraction result based on each threshold value. Here, the edges for images of “Easy-to-read character area” with clear characters, “Unreadable character area” with unclear characters due to the same color of foreground and background, and “Non-character area” The result of feature extraction is shown. In the diagram showing the edge extraction result, white pixels are pixels determined to have each edge feature. As can be seen from the figure, in the “easy-to-read character region”, the outline of the character is suitably extracted with all three threshold values. On the other hand, in the “hard-to-read character region”, the outline of the character is extracted at the lowest threshold (reference threshold), but only a few outlines of the character are extracted at the other two thresholds. Also. With respect to the “non-character region”, a noise component is extracted at the lowest threshold (reference threshold).

  In other words, the following can be read as characteristics when an edge pixel is a pixel having an edge strength equal to or greater than each threshold value.

  Strong edge threshold: Edge pixels (strong edge pixels) are extracted more than a predetermined value only in the “readable character region”.

  Weak edge threshold: In the “easy-to-read character region”, edge pixels (weak edge pixels) or more are extracted, and in the “hard-to-read character region”, edge pixels (weak edge pixels) are extracted in some places. In the “non-character region”, edge pixels (weak edge pixels) may be extracted in some places, but edge pixels (weak edge pixels) may not be extracted.

  Reference threshold: In the “easy-to-read character region” and “unreadable character region”, the edge pixels of the character are extracted. In the “non-character region”, a large number of edge pixels are observed when there is noise such as halftone dots. On the other hand, in the case of a non-character region that is not a halftone dot, there are very few edge pixels.

<Character area determination processing>
FIG. 9 is a flowchart of region determination (detection) processing according to the first embodiment.

  In step S1401, the edge feature values (SN, WN, TN) of the block of interest are input.

  In steps S1402 to S1407, it is determined whether the target block is a character area or a non-character area. Specifically, in step S1402, attention is paid to the strong edge pixel number SN, and if there is a strong edge pixel, the process proceeds to step S1403 to determine the target block as a character area. At this time, in consideration of noise components, a predetermined threshold value may be provided to determine whether or not a strong edge pixel exists. On the other hand, if there is no strong edge pixel, the process proceeds to step S1404, and the determination is made by paying attention to the number WN of weak edge pixels.

  If there is no weak edge pixel, the process proceeds to step S1405 and the target block is determined to be a non-character area. On the other hand, if there is a weak edge pixel, the process proceeds to step S1406, and determination is made by paying attention to the number of edge pixels of the reference threshold value. In other words, if the reference threshold edge pixels are very large (first number or more) or very small (less than the second number), the process proceeds to S1405, and the block of interest is determined as a non-character area. When the number of edge pixels of the reference threshold is in the range greater than or equal to the second number and less than the first number, it may be considered as a “hard character” or a “non-character region”. Therefore, in consideration of correction in post-processing, in Step S1407, indeterminate information indicating an uncertain area (“UNKNOWN”) may be given to the block of interest.

  As described above, according to the first embodiment, a plurality of threshold values are set for the block of interest, and the character region / the number of pixels having edge strength in each range set based on the plurality of threshold values is set. Non-character area / indeterminate area is determined. That is, the region is determined based on the distribution of the edge intensity of the pixels included in the block of interest. Moreover, what is necessary is just to determine an area | region by post-processing about the area | region determined as the uncertain area | region. With this configuration, it is possible to determine the character area more accurately than in the past. In particular, an area including a character having a color close to a halftone dot background is determined as an indeterminate area, not a non-character area, so if it is configured to correct the attribute as a character area in post-processing, It becomes possible to distinguish more accurately. As a result, for example, it is possible to further improve the accuracy of character recognition in the subsequent process. Note that by setting information on an indeterminate area for an area that could not be determined, for example, it is possible to selectively accept designation of area attributes from the user (reception means) in a later step.

(Second Embodiment)
In the second embodiment, an example will be described in which, in each of the character area and the non-character area, whether or not the area includes a halftone dot is identified together with the determination of whether or not the area is a character area. . That is, in the second embodiment, any one of “halftone background only”, “non-halftone background only”, “halftone background + character”, and “non-halftone background + character” is identified. Below, a different part from 1st Embodiment is mainly demonstrated.

<Edge feature amount>
FIG. 10 is a diagram for explaining each parameter of the edge feature amount according to the second embodiment. The edge feature derivation parameters include a reference threshold T (reference value), a strong edge determination threshold ST (second threshold) and a weak edge determination threshold WT (first threshold) obtained based on the reference threshold T, And four threshold values of a low threshold value LT (fourth threshold value). here,
Reference threshold T
Strong edge determination threshold ST = T + B
Weak edge determination threshold WT = T + C
Low threshold LT = D
And B, C, and D are predetermined constants set to satisfy ST>WT>T> D.

In addition, the edge feature is derived based on the edge strength of each pixel in the block of interest and the above parameters. Specifically,
Number of strong edge pixels: number of pixels having edge strength of ST or more (second range) Number of weak edge pixels: number of pixels having edge strength of WT or more and less than ST (first range) Reference edge pixel number: T or more The number of pixels having edge strengths less than WT (third range) Low edge pixel count: Number of pixels having edge strengths greater than or equal to LT and less than T (fourth range greater than or equal to the fourth threshold and less than the third threshold) It is expressed as

  FIG. 11 is a flowchart for deriving each parameter of the edge feature value according to the second embodiment. Steps S13301 to S13308 are substantially the same as those in the first embodiment. However, the difference is that in step S13307, if the edge strength of the pixel is less than the reference threshold T, the process proceeds to S13309.

  In step S13309, it is determined whether the edge intensity of the pixel is equal to or higher than a low threshold LT (for example, 40). If the edge strength of the pixel is equal to or higher than the low threshold LT, the process advances to step S13310 to determine that the pixel is a low threshold edge pixel, and increment the count value LN by one. On the other hand, if it is less than the low threshold LT, the process proceeds to step S13311.

  In step S13311, it is determined whether or not the processing has been completed for all the pixels included in the block of interest. If there is an unprocessed pixel, the process returns to step S13303, and the above determination is repeated for the next pixel. If there is no unprocessed pixel, the process advances to step S13312 to output edge feature amounts (SN, WN, TN, LN).

<Character area determination processing>
When the block of interest is a character on a non-halftone dot (background), no edge pixel is generated on a uniform color background, so the number of low threshold edge pixels is reduced. On the other hand, the character on the halftone dot has a very large number of low threshold edge pixels because the halftone dot pixel in the halftone dot background becomes an edge pixel. A non-halftone dot background (background) without characters has few edge pixels at a low threshold. Also, in a halftone dot background without characters, there are many edge pixels at a low threshold. Therefore, it is possible to identify one of the four types of “halftone background only”, “non-halftone background only”, “halftone background + character”, and “non-halftone background + character” as follows. I can do it.

  FIG. 12 is a flowchart of region determination processing according to the second embodiment. Note that steps S1411 to S1417 are substantially the same as steps S1401 to S1407 in the first embodiment. However, the processing after the “character area” or “non-character area” is determined is different from the first embodiment.

  After determining that the block of interest is a “character area” (S1413), in step S1418, it is determined whether or not the number LN of low-threshold edge pixels is a predetermined number or more. If the number LN of low threshold edge pixels is greater than or equal to the predetermined number, it is determined that the character is on the halftone dot background (character region on the halftone dot) (S1419). On the other hand, when the number LN of the edge pixels having the low threshold value is less than the predetermined number, it is determined that the character area is an overlying character area (a non-halftone character area with characters on a non-halftone background (background)) (S1420).

  Similarly, after it is determined that the target block is a “non-character area” (S1415), in step S1421, it is determined whether or not the number LN of low-threshold edge pixels is a predetermined number or more. When the number LN of low-threshold edge pixels is a predetermined number or more, it is determined that the pixel is a halftone dot (halftone dot background region) (S1422). On the other hand, when the number LN of edge pixels having a low threshold value is less than the predetermined number, it is determined that the background is a background (non-halftone background area) (S1423).

  As described above, according to the second embodiment, in addition to the determination of the character area and the non-character area, it is possible to perform the determination of the halftone dot background and the non-dot background.

(Other examples)
The present invention can also be realized by executing the following processing. That is, software (program) that realizes the functions of the above-described embodiments is supplied to a system or apparatus via a network or various storage media, and a computer (or CPU, MPU, or the like) of the system or apparatus reads the program. It is a process to be executed.

Claims (8)

Dividing means for dividing the image into a plurality of blocks each including a predetermined number of pixels;
Derivation means for deriving the edge intensity value of each pixel included in the block of interest;
Reference value deriving means for deriving a reference value in the target block based on the maximum value and the minimum value of the edge intensity values of the pixels included in the target block;
Threshold setting means for setting a first threshold value greater than the reference value and a second threshold value greater than the first threshold based on the derived reference value;
A count for counting the number of pixels having edge intensity values included in each of the first range that is greater than or equal to the first threshold and less than the second threshold and the second range that is greater than or equal to the second threshold in the block of interest. Means,
Based on the result of counting by the counting means, if there is a pixel having an edge intensity value in the second range, the block of interest is determined as a character area, and the edge intensity values in the second range and the first range Determining means for determining that the block of interest is a non-character area when there is no pixel having
An image processing apparatus comprising: The threshold value setting means further sets a third threshold value equal to the reference value,
The counting means further counts the number of pixels having edge intensity values included in a third range that is greater than or equal to the third threshold and less than the first threshold,
The determination means further includes no pixel having an edge intensity value in the second range, and a pixel having an edge intensity value in the first range, and an edge intensity value in the third range. 2. The image processing according to claim 1, wherein the block of interest is determined to be a non-character region when the number of pixels having a pixel number is greater than or equal to the first number or less than a second number less than the first number. apparatus.   3. The determination means further provides determination impossible information indicating that determination is impossible for a block that has not been determined as the character area or the non-character area. Image processing apparatus. The threshold value setting means further sets a fourth threshold value smaller than the reference value,
The counting means further counts the number of pixels having edge intensity values included in a fourth range that is greater than or equal to the fourth threshold and less than the third threshold,
The determination means further includes:
When it is determined as the character area and the number of pixels having the edge intensity value in the fourth range is a predetermined number or more, it is determined as a character area on a halftone dot,
When it is determined as the character area and the number of pixels having the edge intensity value in the fourth range is less than a predetermined number, it is determined as a non-half-dot character area;
When it is determined as the non-character region and the number of pixels having the edge intensity value in the fourth range is equal to or greater than a predetermined number, it is determined as a halftone dot background region,
4. A non-halftone background region is determined when the number of pixels determined to be the non-character region and having the edge intensity value in the fourth range is less than a predetermined number. The image processing apparatus according to any one of the above.   5. The apparatus according to claim 1, further comprising: a receiving unit configured to receive, from a user, whether the block is the character area or the non-character area with respect to the block to which the undecidable information is given. An image processing apparatus according to claim 1. The reference value deriving means includes
A reference value in the target block is derived based on a predetermined constant set in common for the plurality of blocks and a maximum value and a minimum value of edge intensity values of pixels included in the target block. The image processing apparatus according to any one of claims 1 to 5. A dividing step of dividing the image into a plurality of blocks each including a predetermined number of pixels;
A derivation step in which the derivation means derives an edge intensity value of each pixel included in the block of interest;
A reference value deriving step for deriving a reference value in the target block based on the maximum value and the minimum value of the edge intensity values of the pixels included in the target block;
A threshold value setting step in which a threshold value setting means sets a first threshold value larger than the reference value and a second threshold value larger than the first threshold based on the derived reference value;
The number of pixels having edge intensity values included in each of the first range that is greater than or equal to the first threshold and less than the second threshold and the second range that is greater than or equal to the second threshold in the target block. A counting step for counting
The determination unit determines that the block of interest is a character area when there is a pixel having an edge intensity value in the second range based on the result of the counting, and determines the edge intensity in the second range and the first range. A determination step of determining the block of interest as a non-character region when there is no pixel having a value;
A control method for an image processing apparatus.   The program for functioning a computer as each means of the image processing apparatus as described in any one of Claims 1 thru | or 6.