JP5221401B2 - Image reading apparatus, image characteristic determination method and program in the apparatus - Google Patents

Description

  The present invention relates to an image characteristic determination technique (original determination technique) for determining whether image data read from an original by an image reading apparatus is color, monochrome, or blank.

  Conventionally, an image reading apparatus that reads a large number of originals by an automatic original feeder (ADF) has been put into practical use. In such an image reading apparatus, it is desirable to output a color image when the read original is a color original and a black and white image when the original is a black and white original. This is because the color image data has a larger data size than the monochrome image data, and the storage capacity of the hard disk or the like can be saved by outputting the image in a mode corresponding to the original appropriately. However, the operation of designating the reading mode for each of a large number of documents set on the ADF by the user is very troublesome and the work efficiency is poor.

  Therefore, a document type determination function (image characteristic determination function) that determines whether the document type is color or black and white and automatically sets a document reading mode has been put into practical use. According to this function, it is not necessary for the user to distinguish between a color original and a black and white original, and if a large number of originals are simply set on the ADF and the start key is pressed, the color image is converted to black and white. If the document is a manuscript, a monochrome image is output. Thus, by outputting an image in an appropriate reading mode corresponding to the document type, it is possible to contribute to cost reduction without wasting storage capacity such as a hard disk. In addition, it is determined whether or not the original type is blank, and if it is blank, the storage capacity can be saved without saving.

  In addition, when image processing or the like is performed on the read image, the black and white document can be processed with the data size of the black and white image, so that an improvement in reading speed in a series of reading processes can be expected. Furthermore, since the user does not have to worry about whether the document is a color document or a monochrome document, more comfortable operability can be realized.

  With respect to this document type determination function, for example, in Patent Document 1, a color pixel / monochrome pixel is determined for each pixel, and based on the determination result, the color pixels are counted in the main scanning direction and the sub-scanning direction, whereby the document Describes a technique for determining the type of each. However, with this technique, there is a risk of erroneously determining the document type when the document includes wrinkles.

  FIG. 1 is a cross-sectional view of a sensor periphery during image reading by a three-line image sensor that can individually read R (red), G (green), and B (blue) components of reflected light from a document. . In FIG. 1, SR is an R (red) component reading sensor. SG is a G (green) component reading sensor. SB is a B (blue) component reading sensor. D is a document being conveyed. In the image reading apparatus having the configuration as shown in FIG. 1, in order to correctly read the image information of the document D, the distances DR, DG, and DB between the sensors of the three-line image sensor and the document D must be equal to each other. Become.

  However, as shown in FIG. 1, when the document D includes wrinkles, the distances DR, DG, and DB between the sensors of the three-line image sensor and the document D are not equal. For this reason, the output level of each of the R component reading sensor SR, the G component reading sensor SG, and the B component reading sensor SB varies, and image data (hereinafter referred to as noise) having a different hue and color from the original image. ) May be formed. This results in erroneous determination of the document type.

Therefore, in order to reduce the influence of wrinkles contained in the document, for example, in Patent Document 2, the document is pressed with a reference surface extending in the width direction of the document conveyance path and a pressing member facing the reference surface, and the document A technique for correcting wrinkles formed on a surface is described. By reading a document surface in which wrinkles are corrected by this technique, it is possible to realize document type determination with reduced wrinkle effects.
JP 2002-199239 A Japanese Patent No. 3912917

  However, since the technique described in Patent Document 2 requires a restraining member for pressing the document, the cost is increased in exchange for realizing high accuracy of the document type determination.

  An object of the present invention is to realize a technique capable of determining image characteristics with high accuracy for image data including a wrinkle region.

In order to solve the above-described problems, an image reading apparatus according to the present invention is an image reading apparatus that reads an image of a document that is being transported, a reading unit that reads an image from a document that is being transported, and an image that is read by the reading unit. Calculation means for calculating color space information of each pixel of data, wrinkle area determination means for determining whether each pixel is included in the wrinkle area using the color space information of each pixel calculated by the calculation means, using the result of the determination by the wrinkle area determination means have a, and determining characteristic judgment means a characteristic of the image data, the color space information includes a chroma value of each pixel, the characteristic determination unit, the Based on the number of pixels of the image data that are determined not to be included in the wrinkle region by the wrinkle region determination unit and whose saturation value does not satisfy the first threshold value, the image data is color image data. Determines whether the black-and-white image data whether it is data.
The image reading apparatus of the present invention includes a reading unit that reads an image from a document, a calculation unit that calculates color space information of each pixel of image data read by the reading unit, and each pixel calculated by the calculation unit. Wrinkle area determination means for determining whether each pixel is included in the wrinkle area using the color space information, and characteristic determination means for determining the characteristics of the image data using the determination result by the wrinkle area determination means, The color space information includes a saturation value of each pixel, and the characteristic determination means determines that the pixel is in the image data and is not included in the wrinkle area by the wrinkle area determination means. And the number of pixels whose saturation value is less than the second threshold and the number of pixels that are determined to be included in the wrinkle region and whose saturation value is less than the third threshold. The image data is a color image. Determines whether the data or black-and-white image data.

The image characteristic determination method of the present invention is a method for determining image characteristics in image data obtained by using an image reading apparatus having a reading unit for reading an image from a document being conveyed, the reading being performed by the reading unit. A calculation step of calculating color space information of each pixel of the image data; a wrinkle region determination step of determining whether each pixel is included in the wrinkle region using the color space information of each pixel calculated by the calculation step; have a, and determining characteristic determination step the characteristics of the image data by using the result of the determination by the wrinkle area determination step, the color space information includes a chroma value of each pixel, the characteristic determination step, Based on the number of pixels of the image data that are determined not to be included in the wrinkle region by the wrinkle region determination means and whose saturation value does not satisfy the first threshold value, the image data is color-coded. Determines whether the monochrome image data whether the image data.
The image characteristic determination method according to the present invention is a method for determining image characteristics in image data obtained by using an image reading apparatus having a reading unit for reading an image from a document, the image characteristic being read by the reading unit. A calculation step of calculating color space information of each pixel, a wrinkle region determination step of determining whether each pixel is included in the wrinkle region using the color space information of each pixel calculated by the calculation step, and the wrinkle region A characteristic determination step of determining a characteristic of the image data using a determination result of the determination step, wherein the color space information includes a saturation value of each pixel, and the characteristic determination step includes the image data The number of pixels that are determined not to be included in the wrinkle region in the wrinkle region determination step and the saturation value is less than the second threshold, and are included in the wrinkle region. Judgment Is, the image data by using the number of pixels chroma value is less than the third threshold value determines whether the color image data or monochrome image data.

  According to the present invention, it is possible to perform highly accurate image characteristic determination that is not easily affected by a wrinkle region included in image data of a document. Further, since it can be realized with an inexpensive configuration, cost reduction can be expected.

  The best mode for carrying out the present invention will be described below in detail with reference to the accompanying drawings. The following embodiments are examples as means for realizing the present invention, and should be appropriately modified or changed depending on the configuration of the apparatus to which the present invention is applied and various conditions without departing from the spirit of the present invention. The present invention is not limited to the following embodiments.

  Also, in the following, the present invention is applied to an image reading apparatus such as a scanner, and color / monochrome determination performed when a document composed of black text on a white background is read as characteristic determination of color image data or monochrome image data. However, the present invention can also be applied to an image forming apparatus such as a copying machine or a printer that executes a similar image processing method. Furthermore, the dimensions, materials, shapes, relative arrangements, and the like of the components exemplified as the present embodiment should be changed as appropriate according to the configuration of the apparatus to which the present invention is applied and various conditions. It is not intended that the scope of the present invention be limited to the following embodiments.

[Embodiment 1]
FIG. 2 is a side sectional view showing a schematic configuration of the image reading apparatus according to the embodiment of the present invention.

  As shown in FIG. 2, the image reading apparatus 1 includes a pickup roller 2 that picks up a large number of documents D stacked thereon, and a feeding roller 3 that feeds the documents D picked up by the pickup roller 2 into the apparatus. . The image reading apparatus 1 also includes a separation roller 4 that separates the picked-up documents D one by one and a document detection sensor 10 that detects whether or not the documents D are stacked.

  Further, the image reading apparatus 1 includes a registration roller pair 5 for conveying the document D, a first image reading unit 33 for reading the front surface of the document D, and a second image reading unit 34 for reading the back surface of the document D. . The first image reading unit 33 includes a line image sensor 7a that reads an image on the surface of the document D through the contact glass 6a. The second image reading unit 34 includes a line image sensor 7b that reads an image on the back side of the document D through the contact glass 6b, and a white plate 9 that is a member facing the line image sensor 7b. The operations of the first and second image reading units 33 and 34 are controlled by the scanner control unit 301. Further, the image reading apparatus 1 includes a pair of conveying rollers 8 disposed on the downstream side of the first and second image reading units 33 and 34.

  The image reading apparatus 1 illuminates the document D with a light source and reads the document image with the line image sensors 7a and 7b. However, since the light amount of the light source is uneven and the sensitivity of the line image sensors 7a and 7b is uneven, a white color reference is used. It is necessary to read and correct the member. For this reason, the shading correction data is acquired when executing the light amount adjustment process for optimizing the light emission amount of the light source and the gain adjustment process for optimizing the amplification factor for the image signals output from the line image sensors 7a and 7b. Done. The shading correction data is used for shading correction processing for correcting the light amount unevenness of the light source and the sensitivity unevenness of the line image sensors 7a and 7b in units of pixels of the line image sensors 7a and 7b.

  In the present embodiment, the correction process for the line image sensors 7a and 7b to uniformly read the document image including the light amount adjustment process and the gain adjustment process is referred to as “shading correction”. Here, the color of the facing member 9 arranged facing the line image sensors 7a and 7b and the light source so as to sandwich the document D is white, and the facing member 9 is read to acquire shading correction data.

  The image reading apparatus 1 introduces the document D into the apparatus by the pickup roller 2 and the feeding roller 3 and separates the original D one by one by the separation roller 4. The document D is nipped and conveyed by the registration roller pair 5 and the conveyance roller pair 8, and the image on the front surface of the document D in the first image reading unit 33 and the image on the back surface of the document D in the second image reading unit 34 are respectively in the main scanning direction. Scanning is performed in a direction (perpendicular to the document conveyance direction). The shading correction is performed on the image data generated from the image signals of the line image sensors 7a and 7b with reference to the shading correction data described above. After reading the image, the document D is nipped and conveyed by the conveyance roller pair 8 and is discharged to the outside of the apparatus.

  The image reading timing starts when the leading edge of the document D passes the document detection sensor 10 and time 1a has elapsed, and the trailing edge of the document D passes the document detection sensor 10 and time 2a has elapsed. When set to finish reading. Time 1a is the time from when the leading edge of the document D passes through the document detection sensor 10 until it reaches the line image sensors 7a and 7b. Time2a is the time from when the trailing edge of the document D has passed through the document detection sensor 10 to when it has passed through the line image sensors 7a and 7b. These values are obtained from time = L / V where V is the document conveyance speed and L is the distance between the document detection sensor 10 and the line image sensors 7a and 7b.

  FIG. 3 is a block diagram illustrating a schematic configuration of the scanner control unit 301 and the external control unit 302. In FIG. 3, a scanner control unit 301 includes an image reading unit 11, an A / D conversion unit 12, an image processing unit 13, and a CPU 14 functioning as a control center corresponding to the first and second image reading units 33 and 34 in FIG. A memory 15 for temporarily storing image data, and a communication unit 16. The communication unit 16 is a general-purpose USB, SCSI, or the like, and is communicably connected to an external control unit 302 as the host computer 17.

  An image signal read and output by the image reading unit 11 of the scanner control unit 301 is input to the A / D conversion unit 12. The A / D converter 12 converts the analog electric signal into a digital signal. Image data generated by the A / D conversion process in the A / D conversion unit 12 is output to the image processing unit 13. Image data that has been subjected to various types of image processing by the image processing unit 13 and is in a suitable state as an output image is transferred to the host computer 17 of the external control unit 302 via the communication unit 16.

  The memory 15 stores a control program to be executed by the CPU 14, and the CPU 14 controls the apparatus while reading program instructions from the memory 15. The memory 15 also has a function of temporarily storing data processed by the image processing unit 13.

  FIG. 4 is a block diagram showing a schematic configuration of the image processing unit 13 of FIG. 4, the image processing unit 13 includes a shading correction unit 2 that corrects the sensitivity unevenness of the line image sensors 7a and 7b in units of pixels. The image processing unit 13 includes a wrinkle region detection unit 18 that detects a wrinkle region from the read image, and a document type determination unit 19 that determines whether the document D is a color document or a monochrome document. These units are connected via a bus line 21. The image processing unit 13 also includes a correction processing unit (not shown) that performs various correction processes for making the read image suitable for an output image.

  Next, an image reading operation by the image reading apparatus 1 will be described with reference to the flow of FIG. In FIG. 5, when an image reading start instruction is received, it is first determined by the document detection sensor 10 whether or not the document is stacked on a document table (not shown) (step S501). If it is determined that there is no document, this processing is terminated. If it is determined that the document is stacked on the document table, the image reading unit 11 shown in FIG. 3 starts reading the document (step S502). Thereafter, the image processing unit 13 stores the image data subjected to A / D conversion processing by the A / D conversion unit 12 in the memory 15 (step S503). Further, the wrinkle area detection process described later is performed on the image data by the wrinkle area detection unit 18 of the image processing unit 13, and the position information of the wrinkle area as a detection result is stored in the memory 15 (step S504). Thereafter, a document type determination process is executed by the document type determination unit 19 of the image processing unit 13 (step S505), and image data corresponding to the document type is created by a correction processing unit (not shown) of the image processing unit 13. (Step S506). In the document type determination process in step S505, the wrinkle region position information detected in step S504 is used as will be described later. Various correction processes are performed on the image data created in step S506 by a correction processing unit (not shown) of the image processing unit 13 (step S507). Thereafter, the data is output to the external control unit 302 via the communication unit 16 (step S508) and stored in the hard disk of the host computer 17. The operations from step S501 to step S508 are repeatedly executed until there are no stacked originals.

  Next, the wrinkle region detection process in step S504 of FIG. 5 will be described with reference to the flow of FIG. In FIG. 6, the wrinkle region detection unit 18 first calculates the background brightness value of the document D (step S601), and based on the calculated background brightness value, the brightness value of pixels that form wrinkles in the document (hereinafter referred to as the wrinkle pixel brightness). ) Is determined (step S602). Further, it is determined whether or not each pixel in the document is a pixel corresponding to a ridge portion and a valley portion of the wrinkle of the document (hereinafter referred to as a ridge pixel) (step S603). Based on the information thus obtained, it is determined whether or not each pixel in the document is included in the wrinkle region (step S604). Here, when a certain pixel is included in the wrinkle region, the position information of the pixel is stored in the memory 15.

  Next, the background lightness value detection processing of the document D in step S601 in FIG. 6 will be described with reference to FIGS. First, histogram information is created based on image data stored in the memory 15 by histogram creation processing. 7 and 8 show the results of creating a histogram from the document image stored in the memory 15. Reference numeral 511 denotes a read image. The coordinate system of the image has an origin (x, y) = (0, 0) at the upper left, and has an x axis in the width direction and a y axis in the height direction. Reference numeral 512 represents the histogram information created from the image data as a graph. The histogram is represented by the lightness value p of the pixel and the number of appearances t of the pixel having the lightness value p in the image. The lightness value p is calculated by Equation 1 exemplified below.

p = MAX (R, G, B) (1)
However, each of R, G, and B is a three-color separation signal obtained by separating each pixel into an R component, a G component, and a B component. In the present embodiment, the maximum value of R, G, and B is the maximum that can be represented by 8-bit data. 255 and the minimum value are 0, and p can take the same maximum and minimum values.

  In the histogram creation process, the pixels in the image are scanned, and the brightness value p of each pixel and the number of appearances t thereof are stored in the memory 15 to create a histogram. In the histogram of FIG. 7, the part surrounded by the ellipse 520 represents the number of appearances of the lightness value of the pixels constituting the text character, and the part enclosed by the ellipse 521 represents the number of appearances of the lightness value of the pixels constituting the background. Represents. In FIG. 7, the document is composed of two colors, a text character and a background, and the number of appearances of the black portion of the text character and the white portion of the background occupies most of the time, so that the histogram is as shown.

  In FIG. 8, reference numeral 513 represents a histogram created from the original 513a by the histogram creation process, and is represented by a solid line. The functional expression of the graph 513 is t = F1 (p). Reference numeral 514 represents a histogram of the document 514a created by the histogram creation process, which is represented by a dotted line, and the function expression of the graph 514 is t = F2 (p). The document 514a is a state in which wrinkles are formed on the document 513a. The part enclosed by the ellipse 525 of the graph 513 and the graph 514 represents the brightness of a text character as mentioned above. Even when the wrinkle overlaps with the character portion, the character portion is black enough to absorb the light from the light source, and thus is not affected by the wrinkle.

  In the background lightness value detection process, first, a point (Pmax2, tmax2) having a maximum value of t = F2 (p) in a right-side region that is a high lightness portion in the graph 514 is detected. FIG. 9 is a flowchart showing a process for detecting a point (Pmax2, tmax2) having a maximum value. In FIG. 9, first, Pmin is substituted into a variable P (step S901). Next, the appearance count t = F2 (p) is substituted into the variable Max_t, and the value P is substituted into Buffer_P (step S902). The variable P, the variable Max_t, and the variable Buffer_P are stored in the memory 15. Next, the value of P is incremented by 1 (step S903), and the value of P is compared with the magnitude of 255 (step S904). If P is greater than 255, Buffer_P is substituted for Pmax2 and Max_t is substituted for tmax2 (step S905), and the process ends. When P is 255 or less, it is determined whether or not the number of appearances t = F2 (p) is larger than Max_t (step S906). If larger, t = F2 (p) is substituted into Max_t, P is substituted into Buffer_P (step S907), and the process returns to step S903. With the above processing, the lightness and the number of appearances (Pmax2, tmax2) at which the number of appearances reaches the maximum value in the right-side region where the lightness is Pmin or more in the histogram of the graph 14 are obtained. In this embodiment, Pmin is set to 128 on the assumption that the brightness value is represented by 8 bits, but the present invention is not limited to this. Pmax2 is the brightness at which the number of appearances is the maximum value, but a value in the vicinity thereof may be used and is substantially equivalent to the present embodiment.

  Next, the detection process of the brightness value of the wrinkle pixel in step S602 in FIG. 6 will be described with reference to FIG.

First, each position of P1, P2, P3, and P4 is obtained by the following formulas 2 to 5 with the detected Pmax2 as a reference. P2 = Pmax2 × 0.95 (2)
P3 = Pmax2 × 1.05 (3)
P1 = P2− (P2−Pmin) × 0.7 (4)
P4 = P3 + (255−P3) × 0.8 (5)
However, Pmin is a value representing an intermediate brightness value between the maximum brightness value of the pixels constituting the text character and the maximum brightness value of the pixels constituting the background, and is set to 128 in the present embodiment as described above. Equation 4 is a case where the brightness value is represented by 8 bits, and 255 is the maximum value of the detectable brightness value. For example, when the brightness value is represented by 10 bits, the maximum value of the brightness value that can be detected is set to a value such as 1023. In addition, the range determined by the above formula is the widest range, which can be appropriately reduced and is included in the present invention.

  A wrinkle pixel is a pixel having a brightness value in the range of P1 to P2 and P3 to P4. The reason for this will be described with reference to FIGS. 10 and 11 show a state where the wrinkle region of the document is read by the line image sensor 7a, (a) shows a state where a wrinkle-free portion is read, and (b) and (c) show a state where the wrinkle portion is read. Each state is shown. Further, in FIG. 10B, the light irradiated from the light source and reflected from the document surface is weakened due to the influence of the inclined portion formed from the wrinkle ridge. In FIG. 10C, the reflected light is stronger than the background portion without wrinkles. In FIG. 11B, the reflected light is strengthened due to the influence of the inclined portion formed by the wrinkle valley. In FIG. 11C, the reflected light is weakened due to the influence of the inclined portion formed by the wrinkle valley. As a result, in the graph 514 of FIG. 8, the number of appearances of the dark lightness portion and the bright lightness portion increases with the background lightness value Pmax2 interposed therebetween, and the appearance frequency of the background lightness value Pmax2 decreases. That is, a dark lightness portion and a bright lightness portion across the background lightness value Pmax2 are lightness values generated due to the influence of wrinkles. Therefore, the lightness value in the range from P1 to P2 and the lightness value in the range from P3 to P4 are determined as the lightness value of the wrinkle pixel.

  Next, the ridge pixel determination process in step S604 of FIG. 6 will be described.

  FIG. 12A is an enlarged view of a ridge wrinkle, and FIG. 12B is an enlarged view of a valley wrinkle. In the present embodiment, the vertices 518 and 519 of the wrinkle 522 shown in FIG. 12 are referred to as a wrinkle ridge and a wrinkle valley, respectively, and the pixels constituting the wrinkle ridge portion and the wrinkle valley portion are ridge pixels. The ridge pixel preferably corresponds to both the ridge of the wrinkle and the valley of the wrinkle, but may correspond to either one, and an effect according to the effect described later can be obtained. Wrinkle ridges and valleys are sandwiched between a bright part 524 and a dark part 523. In the present embodiment, when two pixels (A and B) separated by S pixels across the pixel of interest (x, y) satisfy the following Expression 6 or 7, it is determined that the pixel of interest is a ridge pixel. In the present embodiment, S is 3.

P1 <A brightness value <P2 and P3 <B brightness value <P4 (6)
Or
P1 <B brightness value <P2 and P3 <A brightness value <P4 (7)

FIG. 14A is a diagram illustrating a pixel position that is separated by S pixels from the target pixel (x, y) on the image. In this embodiment, the combination of two pixels (A and B) that are separated by S pixels across the pixel (x, y) of interest is
Pix1: (x-3, y-3) and Pix13: (x + 3, y + 3)
Pix2: (x-2, y-3) and Pix14: (x + 2, y + 3)
Pix3: (x-1, y-3) and Pix15: (x + 1, y + 3)
Pix4: (x, y-3) and Pix16: (x, y + 3)
Pix5: (x + 1, y-3) and Pix17: (x-1, y + 3)
Pix6: (x + 2, y-3) and Pix18: (x-2, y + 3)
Pix7: (x + 3, y-3) and Pix19: (x-3, y + 3)
Pix8: (x + 3, y-2) and Pix20: (x-3, y + 2)
Pix9: (x + 3, y-1) and Pix21: (x-3, y + 1)
Pix10: (x + 3, y) and Pix22: (x-3, y)
Pix11: (x + 3, y + 1) and Pix23: (x-3, y-1)
Pix12: (x + 3, y + 2) and Pix24: (x-3, y-2)
It is. However, the parentheses indicate the coordinates of Pix1 to Pix24.

If any one of these combinations meets the condition of Expression 6 or Expression 7, the pixel (x, y) is determined to be a ridge pixel. For example, in the combination [Pix1: (x-3, y-3) and Pix13: (x + 3, y + 3)], when the brightness values of the pixels of Pix1 and Pix13 are f1 and f13, respectively,
P1 <f1 <P2 and P3 <f13 <P4
Or P1 <f13 <P2 and P3 <f1 <P4
It becomes. When this condition is satisfied, the pixel (x, y) is determined as a ridge pixel. For pixels in which any one of Pix1 to Pix24 is outside the range of the read image (for example, 516 in FIG. 14B), the portion outside the range is not used in the determination process, and is within the range of the read image. Only the portion is used (for example, 517 in FIG. 14B).

  Next, processing for detecting the position of the ridge pixel from the entire read image will be described with reference to the flow of FIG. The coordinate system of the image is a coordinate system in which the upper left corner of the image is the origin (x, y) = (0, 0), the x axis is in the width direction, and the y axis is in the height direction. The height and width of the image are represented by height and width, respectively. Further, the brightness value of the pixel at the position (x, y) is represented by f (x, y).

  In FIG. 13, first, scanning is started from the upper left end (x, y) = (0, 0) of the image (step S1301). Next, it is determined whether or not the pixel of interest (x, y) is a ridge pixel (step S1302). If the pixel of interest (x, y) is a ridge pixel, the coordinates (x, y) of the pixel (x, y) are stored in the memory 15 (step S1303). Thereafter, the x coordinate of the pixel of interest (x, y) is incremented by one (step S1304). When the x coordinate of the pixel of interest (x, y) exceeds the width width, the y coordinate of the pixel of interest (x, y) is incremented by 1, and the x coordinate is reset to 0 (step S1306). Next, it is determined whether or not the y coordinate of the pixel (x, y) of interest exceeds height (step S1307). If it exceeds height, the process is terminated, and if it is not more than height, the process returns to step S1302.

  Next, the wrinkle region determination process in step S604 of FIG. 6 will be described with reference to the flow of FIG. As described above, the coordinate system of the image is a coordinate system in which the upper left corner of the image is the origin (x, y) = (0, 0), the x axis is in the width direction, and the y axis is in the height direction. The height and width of the image are represented by height and width, respectively. Further, the brightness value of the pixel at the position (x, y) is represented by f (x, y).

  In FIG. 15, first, scanning is started from the left end (x, y) = (0, 0) of the image (step S1501). If the pixel of interest (x, y) is a pixel within the distance r (r is a constant) with the ridge pixel as the center (r is a constant) (step S1502), the pixel is determined as a wrinkle region candidate, and the process proceeds to step S1503. In step S1503, it is determined whether or not the pixel is a wrinkle region candidate and has a lightness value in the range of P1 to P2 described above, or a pixel having a lightness value in the range of P3 to P4. To do. If the brightness value of the pixel is within one of the above ranges, the pixel is determined to be a pixel in the wrinkle region, and the position of the pixel is stored in the memory 15 as wrinkle region data (step S1508). Thereafter, the x coordinate of the pixel of interest (x, y) is incremented by one (step S1504). When the x coordinate of the pixel of interest (x, y) exceeds the width width, the y coordinate of the pixel of interest (x, y) is incremented by 1, and the x coordinate is reset to 0 (step S1506). Next, it is determined whether or not the y coordinate of the pixel of interest (x, y) exceeds height (step S1507). If it exceeds height, the process is terminated, and if it is not more than height, the process returns to step S1502.

  Next, the document type determination process in step S505 in FIG. 5 will be described with reference to the flow in FIG. The coordinate system of the image is a coordinate system in which the upper left corner of the image is the origin (x, y) = (0, 0), the x axis is in the width direction, and the y axis is in the height direction. The width of the image is represented by height and width, respectively.

  In FIG. 16, first, the color counter SC held in the memory 15 is reset to 0, and scanning is started from the upper left end (x, y) = (0, 0) of the image (step S1601).

  Next, referring to the wrinkle area data stored in the memory 15 by the wrinkle area detection unit 18, it is determined whether or not the pixel of interest (x, y) is a wrinkle pixel (step S 1602). If the pixel of interest (x, y) is a wrinkle pixel, the process proceeds to step S1606, where the x coordinate of the pixel of interest (x, y) is incremented by 1, and the next pixel is processed. If the pixel of interest (x, y) is not a wrinkle pixel, the process proceeds to step S1603. In step S1603, a saturation value S and a lightness value p are calculated as color space information of the pixel of the image data.

  The saturation value S and the lightness value p are calculated using a three-color separation signal obtained by separating each pixel into an R component, a G component, and a B component. Here, the R component, the G component, and the B component are pixel values obtained by decomposing each pixel data into three color components of red (Red), green (Green), and blue (Blue). Here, the saturation value S is calculated by Equation 8 exemplified below.

Saturation value S = {MAX (R, G, B) −MIN (R, G, B)} / MAX (R, G, B) (8)
However, when MAX (R, G, B) is 0, the saturation value S is processed as 0. The lightness value p is calculated according to the above-described exemplary equation 1.

  In the case of the present embodiment, whether a pixel is a color pixel or a monochrome pixel is determined based on the saturation value S and the brightness value p obtained by the above equations 8 and 1, whether or not the pixel having the brightness value p is a color pixel. It is determined by whether or not it is larger than the function J (p) that gives the first threshold value for In the present embodiment, the first threshold value is a curve J as shown in FIG. In FIG. 17, the vertical axis represents the saturation value S, the horizontal axis represents the lightness value p, and the curve J is determined by various factors such as the optical characteristics of the image reading apparatus 1.

  The color determination using the curve J is performed using the following formulas 9 and 10 based on the saturation value S and the lightness value p of the determination target pixel (step S1604).

J (p) <S: Color pixel (9)
J (p) ≧ S: monochrome pixel (10)
(J (p) is the saturation value at the lightness value p in the curve J)
Therefore, when the combination of the saturation value S and the lightness value p of the pixel to be determined is located in the shaded portion in FIG. 17, the pixel to be determined is determined to be a color pixel, and is located on the opposite side across the curve J Are determined to be monochrome pixels. Hereinafter, the threshold for determining the color information of each pixel such as the curve J is referred to as a determination curve. The determination curve J is stored in advance in the memory 15, and the determination by the above formulas 9 and 10 is performed with reference to the memory 15. If S is larger than J (p), the count number of the color counter SC is incremented assuming that the pixel (x, y) is a color pixel (step S1605). Subsequently, the x coordinate of the pixel (x, y) is incremented by one (step S1606).

  Next, it is determined whether or not x exceeds the width width (step S1607). If it exceeds, the y coordinate of the pixel (x, y) is incremented by 1 and x is reset to 0 (step S1608). . Next, it is determined whether or not y exceeds the height height (step S1609). If not, the process returns to step S1602, and if it exceeds, the process proceeds to step S1610. In step S1610, it is determined whether the read document is a color document or a monochrome document at a ratio of the count number of the color counter SC to the total pixel number A. If the ratio of SC to the total number of pixels A excluding the wrinkle area is larger than a predetermined threshold value K (%), the document is determined as a color document, and the process proceeds to step S506 in FIG. A color image is created by the correction processing unit. If the document is small, the document is determined as a monochrome document, and the process advances to step S506 to generate a monochrome image by the correction processing unit of the image processing unit 13.

  As described above, when determining whether the read original is a color original or a black-and-white original and performing subsequent image processing, the pixels included in the wrinkled area are not used for the original type determination. be able to.

[Embodiment 2]
In the second embodiment, the document type determination process in step S505 in FIG. 5 is different, and the document type determination process executed by the document type determination unit 19 in the second embodiment will be described with reference to the flow in FIG. Other configurations and processes are the same as those in the first embodiment, and thus description thereof is omitted. Further, in the following description, the description proceeds with reference to the drawings described in the first embodiment as appropriate. The coordinate system of the image is a coordinate system in which the upper left corner of the image is the origin (x, y) = (0, 0), the x axis is in the width direction, and the y axis is in the height direction. The width of the image is represented by height and width, respectively.

  In FIG. 18, first, the count number of the color counter SC held in the memory 15 is reset to 0, and scanning is started from the upper left end (x, y) = (0, 0) of the image (step S1801). In step S1802, the saturation value S and the lightness value p of the pixel (x, y) are calculated. As the calculation method of the saturation value S and the lightness value p of the pixel (x, y), Expressions 1 and 3 exemplified in the same manner as Step S1603 in FIG. 16 are used.

  Next, it is determined whether or not the pixel (x, y) is a wrinkle region (step S1803). If it is not a wrinkle region, the process proceeds to step S1804. If it is a wrinkle region, the process proceeds to step S1805. In step S1804, a function JA (p) that gives a determination curve as the second threshold is read (see FIG. 19). In step S1805, a function JB (p) that gives a determination curve as a third threshold different from the determination curve JA (p) is read (see FIG. 20). The way of judging the judgment curves JA and JB is the same as that shown in FIG.

  In step S1806, it is determined whether the pixel (x, y) is a color pixel or a monochrome pixel based on the determination curves JA and JB. The determination method here is in accordance with the method shown in step S1604 of FIG. That is, it determines with the following formulas 11-14.

JA (p) <S: Color pixel (11)
JA (p) ≧ S: monochrome pixel (12)
JA (p) is the saturation value S at the lightness value p in the curve JA.
JB (p) <S: Color pixel (13)
JB (p) ≧ S: monochrome pixel (14)
JB (p) is the saturation value S at the lightness value p in the curve JB.
If the pixel (x, y) is not a wrinkle region, it is determined whether the pixel (x, y) is a color pixel or a black-and-white pixel according to equations 11 and 12. On the other hand, when the pixel (x, y) is a wrinkle region, the determination is made using equations 13 and 14. If it is determined in step S1806 that the saturation S is greater than JA (p) or the saturation S is greater than JB (p), the process proceeds to step S1807. Otherwise, the process proceeds to step S1808.

  In step S1807, the count number of the color counter SC is incremented. Thereafter, the x coordinate of the pixel (x, y) is incremented by one (step S1808). Next, it is determined whether or not x exceeds the width width (step S1809). If it exceeds, the y coordinate of the pixel (x, y) is incremented by 1 and x is reset to 0 (step S1810). . Next, it is determined whether or not y exceeds the height height (step S1811). If not, the process returns to step S1802, and if it exceeds, the process proceeds to step S1812. In step S1812, it is determined based on the count number of the color counter SC whether the read document is a color document or a monochrome document. The determination method is the same as step S1610 in FIG. Thereafter, the document determination unit 19 ends the series of document type determination processing, and proceeds to step S506 in FIG.

  As described above, according to the present embodiment, the determination of whether the target pixel is a color pixel or a monochrome pixel is performed using two determination curves. The determination curve JB in the wrinkle region is a curve that is less likely to be erroneously recognized as a color document than the determination curve JA in consideration of the noise caused by the wrinkle region described above. It can be performed.

[Embodiment 3]
In the third embodiment, instead of the document type determination process in the first and second embodiments as the image characteristic determination process, a blank document determination process (blank determination process) for determining whether or not the image data read from the document is blank image data. ). Other configurations and processes are the same as those in the first embodiment, and thus description thereof is omitted. Further, in the following description, the description proceeds with reference to the drawings described in the first embodiment as appropriate.

  Conventionally, there are models that have a blank document determination function for determining whether or not a document is blank, and when a scanned document is a blank document, do not save the image data of the scanned document, This is useful for saving storage capacity such as hard disks. This blank page original determination function creates binary image data obtained by binarizing image data of a read original, scans the created binary image data from left to right, and from top to bottom, and white pixels are black pixels. And the number of black pixels changed to white pixels are counted, and when the count value is equal to or smaller than a predetermined threshold value, it is determined that the document is blank (Japanese Patent Laid-Open No. 10-229484). ) And a method (Japanese Patent Laid-Open No. 06-141126) that determines that the black pixel in the created binary image data is blank when the ratio is equal to or less than a predetermined ratio.

  Here, in the above-described conventional blank document determination method, when a wrinkle is included in a document, it may be erroneously determined that even a blank document is not blank. Further, in the blank document determination method using the binary image data, the shadow of the wrinkled portion in the document is determined to be a black pixel. This is because it is determined that a document including many wrinkles even if it is blank is not blank. Further, as in Patent Document 1, an apparatus for correcting wrinkles generated in a document is expensive because a member for pressing the document is required.

  Accordingly, the present embodiment is intended to perform blank document determination as a highly accurate image characteristic determination with an inexpensive configuration that is less affected by the wrinkle region included in the image data of the read document.

  FIG. 21 is a block diagram illustrating a schematic configuration of an image processing unit according to the third embodiment. In FIG. 21, the image processing unit 13 includes a blank document determination unit 219 instead of the document type determination unit 19 as means for performing an image characteristic determination process. After the wrinkle pixel detection by the wrinkle area detection unit 18 is completed, the blank document determination unit 219 determines whether or not the document D is blank. The blank document determination unit 219 determines the blank document while referring to the wrinkle pixel position information stored in the memory 15, and the image data determined to be blank is discarded from the memory 15. The image data of the document determined not to be blank is transferred to the host computer 17 via the communication unit 16. Further, the wrinkle pixel position information stored in the memory 15 is discarded after the blank page original determination is completed.

  Next, an image reading operation performed by the image reading apparatus according to the third embodiment will be described with reference to the flowchart of FIG. In FIG. 22, when an image reading start instruction is received, the document detection sensor 10 first determines whether or not the document is stacked on a document table (not shown) (step S2201). If it is determined that there is no document, the process ends. If it is determined that the document is loaded on the document table, the image reading unit 11 in FIG. 3 starts reading the document (step S2202). Thereafter, the image processing unit 13 stores the image data subjected to A / D conversion processing by the A / D conversion unit 12 in the memory 15 (step S2203). Further, the wrinkle area detection processing similar to that in step S 504 of FIG. 5 is performed on the image data by the wrinkle area detection unit 18 of the image processing unit 13, and the position information of the wrinkle area as a detection result is stored in the memory 15. (Step S2204). Thereafter, the blank document determination unit 219 of the image processing unit 13 executes blank document determination processing (step S2205), and the image data of the document determined to be blank is discarded from the memory 15 (step S2206).

  On the other hand, the image data when the document is not blank is processed by the image processing unit 13 based on another image processing command stored in the memory 15 (for example, correction of image data distortion caused by the document being conveyed obliquely). And the like are executed (step S2207). Thereafter, the data is output to the external control unit 302 via the communication unit 16 (step S2208) and stored in the hard disk of the host computer 17. The operations from step S2201 to step S2208 are repeated until there are no stacked documents.

  Next, the blank document determination process in step S2205 in FIG. 22 will be described with reference to the flow in FIG. The blank document determination unit 219 of the present embodiment scans a binary image from left to right and from top to bottom, and counts the number of white pixels changed to black pixels and the number of white pixels changed to black pixels. Perform blank page original determination.

  In FIG. 23, each pixel of the image data of the document read from the left to the right and from the top to the bottom is first converted to a white pixel if the brightness value of the pixel is equal to or greater than the threshold value, and to a black pixel if the pixel brightness value is less than the threshold value. A binary image is created from the image data read from (Step S2301). In the binary image created in step S2301, the target pixel N is shifted from left to right and from top to bottom, and blank document determination is performed while scanning all pixels. It is assumed that the target pixel N is at the coordinates (x, y) on the image data. If N is the leftmost pixel on the image data, the subsequent processing is not performed and the process proceeds to step S2308 (step S2302). In step S2308, the number of unused pixels H is incremented as the number of pixels that are not used for blank page original determination. In step S2303, it is determined that the target pixel N is a wrinkle pixel in the wrinkle pixel determination process, and if it matches the position information stored in the memory 15, the process proceeds to step S2308. Otherwise, the process proceeds to step S2304. Proceed to

  If N is a black pixel and an adjacent pixel N-1 immediately before N (showing a pixel whose coordinates are (x-1, y)) is a wrinkle pixel (YES in step S2304), the pixel Is determined to have changed, and the change number B is incremented (step S2307). Further, when N is a black pixel and the adjacent pixel N-1 is a white pixel (YES in step S2305) or when N is a white pixel and N-1 is a black pixel (YES in step S2306). It is determined that the color of the pixel has changed, and the change number B is incremented (step S2307). Thereafter, the process from step S2302 to S2307 is repeated via any of steps S2310 to S2312 until the target pixel N becomes the last pixel (N (xmax, ymax)) in step S2309, and all the pixels are When the scanning is completed, the process proceeds to step S2313.

  In step S2313, it is determined whether the number of pixels obtained by subtracting the number of unused pixels H that has not been used for blank sheet document determination from the total number of pixels is zero. If it is 0, it is considered that there are only wrinkle pixels in the document excluding the leftmost pixel, so the document is determined to be blank. Otherwise, the process proceeds to step S2314. In step S2314, it is determined whether the change number B is equal to or less than a blank page determination threshold value J that is predetermined with respect to the number of pixels used for the determination (the number obtained by subtracting the number of unused pixels H from the number of all pixels). If it is less than or equal to the threshold value J, the document is determined to be blank and the process proceeds to step S2206 in FIG. 22 to discard the image data. Otherwise, it is determined that the document is not blank and the process proceeds to step S2207 in FIG.

  The wrinkle area detection process and the blank document determination process may be executed by the external control unit 302 instead of the scanner control unit 301.

  According to the present embodiment, since the wrinkle area is detected before blank document determination is performed and the detected wrinkle portion is not used for blank document determination, erroneous determination due to the wrinkle portion in the document can be reduced, and it is inexpensive and accurate. High blank paper document determination can be realized.

  In each of the above-described embodiments, the image data created by the scanner control unit 301 is described as being sent to the host computer 17 via the communication unit 16 and stored in the hard disk of the host computer 17. Is not limited to hard disks. For example, a storage medium device such as a CD-ROM, a CD-R, a CD-RW, a DVD-ROM, a DVD-RAM, a DVD-RW, a DVD + RW, or another optical disk, a magnetic tape, a nonvolatile memory card, a ROM, or the like may be used. Good.

  Note that the present invention can also be applied to a program for causing a computer to execute the processing according to each flow described above and a computer-readable storage medium storing the program. In this case, a storage medium storing a computer program for executing the processing according to each flow described above is supplied to the computer of the image reading apparatus, and the computer reads the program code stored in the storage medium and executes the above processing. To do. In addition, a part or all of the processing according to each flow of the present invention may be replaced with processing by hardware or software. Further, some or all of the processing of the present invention may be performed in a host device such as the host computer 17 connected to the image reading device. In this case, an image reading system including the image reading device and the host device corresponds to the image reading device of the present invention.

It is sectional drawing of the sensor periphery at the time of the image reading by a 3 line image sensor. 1 is a side sectional view showing a schematic configuration of an image reading apparatus according to an embodiment of the present invention. It is a block diagram which shows schematic structure of a scanner control part and an external control part. It is a block diagram which shows schematic structure of an image process part. 3 is a flowchart illustrating an image reading operation according to the first embodiment. It is a flowchart which shows a wrinkle area | region detection process in step S504 of FIG. FIG. 7 is a diagram illustrating a background brightness value detection process of a document in step S601 in FIG. FIG. 7 is a diagram illustrating a background brightness value detection process of a document in step S601 in FIG. It is a flowchart which shows the process which detects the point (Pmax2, tmax2) used as the maximum value in the detection process of a background lightness value. It is a figure explaining the brightness nonuniformity by the influence of a wrinkle. It is a figure explaining the brightness nonuniformity by the influence of a wrinkle. It is a figure explaining the determination process of the ridge pixel in step S604 of FIG. It is a flowchart which shows the process which detects the position of a ridge pixel. It is a figure explaining the pixel position away from the attention pixel (x, y) on the image by S pixel. It is a flowchart which shows a wrinkle area | region determination process in step S604 of FIG. 6 is a flowchart illustrating document type determination processing according to the first exemplary embodiment. FIG. 6 is a diagram illustrating a determination curve for determining a color pixel or a monochrome pixel in the document type determination process of the first embodiment. 10 is a flowchart illustrating document type determination processing according to the second embodiment. FIG. 10 is a diagram illustrating a determination curve for determining a color pixel or a monochrome pixel in the document type determination process of the second embodiment. FIG. 10 is a diagram illustrating a determination curve for determining a color pixel or a monochrome pixel in the document type determination process of the second embodiment. FIG. 10 is a block diagram illustrating a schematic configuration of an image processing unit according to a third embodiment. 10 is a flowchart illustrating an image reading operation according to the third embodiment. FIG. 23 is a flowchart showing blank document determination processing in step S2205 of FIG.

DESCRIPTION OF SYMBOLS 1 Image reading apparatus 2 Pickup roller 3 Feeding roller 4 Separation roller 5 Registration roller 6a, 6b Contact glass 7a, 7b Line image sensor 8 Conveyance roller 9 White plate 10 Document detection sensor 11 Image reading part 12 A / D conversion part 13 Image processing Part 14 CPU
15 Memory 16 Communication Unit 17 Host Computer 18 Wrinkle Region Detection Unit 19 Document Type Determination Unit 20 Shading Correction Unit 21 Bus Line 33 First Image Reading Unit 34 Second Image Reading Unit 219 Blank Document Determination Unit 301 Scanner Control Unit 302 External Control Unit D Document SR R component reading sensor SG G component reading sensor SB B component reading sensor

Claims (8)

An image reading apparatus for reading an image of a document being conveyed,
Reading means for reading an image from a document being conveyed ;
Calculating means for calculating color space information of each pixel of the image data read by the reading means;
Wrinkle area determination means for determining whether each pixel is included in the wrinkle area using the color space information of each pixel calculated by the calculation means;
Have a, and determining characteristic judgment means a characteristic of the image data by using the result of the determination by the wrinkle area determination means,
The color space information includes a saturation value of each pixel,
The characteristic determination means is based on the number of pixels that are pixels of the image data and are determined not to be included in the wrinkle area by the wrinkle area determination means, and whose saturation value does not satisfy the first threshold value. An image reading apparatus for determining whether the image data is color image data or monochrome image data . Reading means for reading an image from a document;
Calculating means for calculating color space information of each pixel of the image data read by the reading means;
Wrinkle area determination means for determining whether each pixel is included in the wrinkle area using the color space information of each pixel calculated by the calculation means;
Characteristic determination means for determining the characteristic of the image data using the result of determination by the wrinkle area determination means,
The color space information includes a saturation value of each pixel,
The characteristic determination means determines the number of pixels in the image data that are determined not to be included in the wrinkle area by the wrinkle area determination means, and whose saturation value does not satisfy the second threshold value, And determining whether the image data is color image data or black-and-white image data using a number of pixels that are determined to be included in the wrinkle region and whose saturation value is less than a third threshold value. It is that images reader. The characteristic determination unit is a blank page determination unit that determines whether the image data is blank page image data based on the brightness of a pixel that is not determined to be included in the wrinkle region by the wrinkle region determination unit. The image reading apparatus according to claim 1 or 2 . The blank page determination unit creates binary image data obtained by binarizing the image data, and the binary image data is a pixel that is not included in the wrinkle region and changes from black to white or from white to black. 4. The image reading apparatus according to claim 3 , wherein it is determined whether or not the image data is blank sheet image data based on the number of pixels obtained. The color space information includes the brightness value of each pixel,
The wrinkle area determination means creates a histogram from the brightness value of each pixel, determines a brightness value of a pixel included in the wrinkle area from the created histogram, and a pixel included in the wrinkle area from the determined brightness value of the pixel the image reading apparatus according to any one of claims 1 to 4, characterized in that determining whether a. A method for determining image characteristics in image data obtained by using an image reading apparatus having a reading means for reading an image from a document being conveyed ,
A calculation step of calculating color space information of each pixel of the image data read by the reading unit;
A wrinkle region determination step for determining whether each pixel is included in the wrinkle region using the color space information of each pixel calculated by the calculation step;
Using the result of the determination by the wrinkle area determining step have a, and determining characteristic determination step the characteristics of the image data,
The color space information includes a saturation value of each pixel,
The characteristic determination step is based on the number of pixels of the image data that are determined not to be included in the wrinkle region by the wrinkle region determination unit and whose saturation value is less than the first threshold value. A method of determining image characteristics, characterized in that it is determined whether the image data is color image data or monochrome image data .   A method for determining image characteristics in image data obtained using an image reading apparatus having a reading means for reading an image from a document,
  A calculation step of calculating color space information of each pixel of the image data read by the reading unit;
  A wrinkle region determination step for determining whether each pixel is included in the wrinkle region using the color space information of each pixel calculated by the calculation step;
  A characteristic determination step of determining a characteristic of the image data using a result of determination by the wrinkle region determination step,
  The color space information includes a saturation value of each pixel,
  The characteristic determination step is a number of pixels in the image data that are determined not to be included in the wrinkle region by the wrinkle region determination step, and whose saturation value is less than the second threshold; And determining whether the image data is color image data or black-and-white image data using a number of pixels that are determined to be included in the wrinkle region and whose saturation value is less than a third threshold value. Method of determining image characteristics. Program for executing a method of determining the image characteristics according to computer to claim 6 or 7.

Images