JP2015023468A - Image processing apparatus and image processing method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image reading apparatus capable of correctly detecting blank paper even if a stripe is generated in an image due to an adhered dust.SOLUTION: In an image reading apparatus, a document fed by an automatic document feeder and passing a reading position is read and it is determined whether the read document is blank paper or not by using a read pixel signal and a first blank paper determination parameter (506, 2301). While referring to an outputted blank paper determination result, a second blank paper determination parameter for removing influence of a foreign substance, such as dust and dirt, appearing in an image is calculated (S1605). Before reading the document, image data of a white member is acquired, a second blank paper determination parameter for removing influence of a stripe from a blank paper determination result for the image data is calculated, then blank paper detection processing for the document data is executed by using the calculated second blank paper determination parameter.

Description

  The present invention relates to an image processing apparatus for determining whether read image data is image data having print information or blank image data having no print information.

  In an image reading apparatus such as a digital copying machine, it is possible to suppress wasteful printing and consumption of paper and toner by deleting image data determined to be blank. Various blank sheet detection methods have been devised so far. . For example, Patent Document 1 proposes a blank page determination method for detecting an edge portion in image data from image data read from a document, and determining whether the detected edge amount is a blank page from the ratio of the total number of pixels. Yes.

  On the other hand, in a flow reading method in which the position of the optical system is fixed and an image is read while an original is conveyed by an automatic document feeder (ADF), dust or dust is present at the original reading position on the platen glass. In the case of adhesion or when there is a scratch or dirt (hereinafter referred to as attached dust), the same attached dust may continue to be read. As a result, a straight line along the conveyance direction may occur in the read image. Due to the effect of this streak, the blank sheet detection method of Patent Document 1 may not be able to detect a blank sheet correctly.

  To solve this problem, conventionally, a reference member of a predetermined color is provided facing the surface of the platen glass through which the document passes, and the presence / absence of attached dust is determined based on the reading result of the reference member. There has been proposed a technique for correcting an image by interpolating or replacing a region corresponding to attached dust in a document image with pixel values around the region (see Patent Document 2).

  Also, before reading the document image, the blank sample document is read, and the image data of the read blank sample document and all white image data having the same size as the image data of the read document are generated, and two code data amounts are generated. A technique for generating a threshold parameter for blank sheet detection based on a relative ratio has been proposed (Patent Document 3). According to this prior art, the blank sheet detection parameters are generated based on the scanned blank sample document image data, and therefore it is possible to determine the image data of the blank document with streaks as the blank document. Become.

JP 2010-178377 A JP 2012-99946 A JP 2006-333284 A

  However, in the invention described in Patent Document 2, streaks are corrected by performing an interpolation process from pixel values around the detected streak area. For this reason, when the detected streak area is thick, a wide range is interpolated from surrounding pixels, and an image existing in the document may be lost. When the correction target is up to a certain thickness in order to prevent the disappearance of the original image, there is a problem that a thick streak cannot be corrected.

  Further, in the invention described in Patent Document 3, since it is necessary to read a blank sample document before reading the document, the load on the user, such as labor and time, increases. Further, when the amount of attached dust increases or decreases, it is necessary for the user to recognize the situation and read a blank sample document each time, which causes a problem that the load increases significantly.

  The present invention has been made in view of the above-described conventional example, and an object thereof is to provide an image processing apparatus and an image processing method for correctly detecting a blank sheet even when streaks occur in an image due to attached dust.

The image processing apparatus of the present invention includes a blank page determination unit that executes blank page determination for determining whether image data obtained by reading an image is blank or not at a plurality of determination levels with different levels of determination as blank page,
The determination level corresponding to each of the plurality of determination levels is performed with reference to a reference image, and the determination result corresponding to each of the plurality of determination levels is determined. Adjusting means for specifying a change point determination level or adjusting the plurality of determination levels until there is no room for adjustment;
Blank sheet determination control means for performing blank sheet determination by the blank sheet determination section for image data obtained by reading a document image at one of the plurality of determination levels adjusted by the adjustment section. It is characterized by.

According to the present invention, even if a streak occurs in an image due to adhering dust and the streak is not corrected, it is possible to reduce the influence of the streak on the blank sheet detection process without reading the blank sample document. It becomes.
Therefore, it is possible to improve blank sheet detection performance and prevent erroneous determination.

1 is a diagram illustrating an example of an overview of a copying machine. It is a figure which shows the structural example at the time of the flow-reading operation | movement of a scanner part. It is a block diagram which shows the structural example of a controller. It is a figure which shows the structural example of an operation unit. FIG. 3 is a block diagram illustrating a configuration example of a scanner IF image processing unit in the first embodiment. It is a block diagram which shows the structural example of a stripe detection part. It is a figure which illustrates the residence of the dust on the platen glass. FIG. 6 is a diagram illustrating an example of a read image when dust adheres to a platen glass. It is a flowchart which shows an example of the flow of a stripe detection process. It is a flowchart which shows an example of the flow of a stripe correction process. It is a figure for demonstrating an example of the interpolation calculation for correct | amending a streak. 3 is a block diagram illustrating a configuration example of a blank paper detection unit. FIG. It is a block diagram which shows an example of a histogram generation part and an edge information generation part. It is a block diagram which shows an example of a histogram analysis part. It is a block diagram which shows an example of an edge information analysis part. 3 is a flowchart illustrating a flow of a flow reading mode according to the first embodiment. 6 is a flowchart illustrating a flow of blank sheet determination parameter change processing according to the first exemplary embodiment. 6 is a diagram illustrating a relationship between a blank page determination parameter and a blank page determination result according to the first exemplary embodiment. FIG. It is a figure which shows an example of the parameter change of Embodiment 1. FIG. 6 is a diagram illustrating an example of parameter change amount adjustment according to the first embodiment. 6 is a table illustrating an example of a plurality of blank sheet determination results according to the first exemplary embodiment. It is a figure which shows the example of a display of a liquid crystal operation panel. FIG. 10 is a block diagram illustrating a configuration example of a scanner IF image processing unit according to a second embodiment. 10 is a flowchart illustrating a flow in a flow reading mode according to the second embodiment. It is a figure which shows an example of a blank paper detection parameter.

  Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings.

[Embodiment 1]
FIG. 1 is a diagram illustrating an example of the appearance of a copying machine according to the first embodiment. The scanner unit 110 serving as an image reading unit converts the image information into an electrical signal by inputting reflected light obtained by exposing and scanning the image on the document by light emitted from the illumination lamp to a linear image sensor (CCD sensor). To do. The scanner unit 110 further converts the electrical signal into a luminance signal composed of R, G, and B colors, and outputs the luminance signal as image data to a controller 300 (FIG. 3) described later.

  The document is set on the tray 112 of the document feeder 111. When the user gives an instruction to start reading from the operation unit 120, the controller 300 sends an original reading instruction to the scanner unit 110. Upon receiving this instruction, the scanner unit 110 feeds documents one by one from the tray 112 of the document feeder 111 and performs a document reading operation (hereinafter, this operation mode is referred to as a flow reading mode). The original can also be read by placing it on an original platen glass described later.

The printer 100 is an image forming device that forms image data received from the controller 300 on a sheet. The image forming method in this embodiment is an electrophotographic method using a photosensitive drum or a photosensitive belt. The printer 100 also includes a plurality of paper cassettes 101, 102, and 103 that can handle different paper sizes or different paper orientations. The printed paper is discharged to the paper output tray 104 <Copier-Scanner Unit>
FIG. 2 is a schematic diagram showing the main configuration and reading operation of the scanner unit 110 using the linear image sensor in the present embodiment. In particular, FIG. 2 shows an outline of a main configuration and a reading operation in the case of “flow reading” in which a document is read by operating the document feeder 111.

  In FIG. 2, a document bundle 202 to be read is placed on a tray 112. A delivery roller 221, a separation / conveyance roller 222, and a registration roller 223 are disposed at the lower part in the document conveyance direction. The feed roller 221 is rotated by a driving source (not shown) and feeds the document bundle 202 placed on the tray 112. Next, a separation / conveying roller 222 disposed downstream of the feeding roller 221 separates and conveys the uppermost document 201 from the conveyed document bundle 202. The start of rotation of the registration roller 223 disposed downstream of the separation conveyance roller 222 is a reference for subsequent conveyance timing and image reading timing of the original 201.

  A drive source for driving the feed roller 221, separation transport roller 222, and registration roller 223 is, for example, a stepping motor.

  The document 201 discharged from the registration roller 223 travels along the guide plate 227, is nipped by the rotating large-diameter conveying roller 224 and the driven rollers 225a, 225b, and 225c, and is conveyed along the outer periphery of the conveying roller 224. Is done. At this time, the document 201 is once conveyed at a constant speed in the direction of the arrow in FIG. 2 through the surface of the document table glass 211.

  Image reading of the original 201 by an image reading unit described later is performed when the original 201 passes through the surface of the original table glass 211.

  After the image reading, the image is continuously conveyed along the outer periphery of the conveyance drum 224 and is discharged onto the document feeder (original supply unit) 111 by the discharge roller 226.

  In this flow-reading mode, since it is only necessary to move the document in a certain direction, a large number of documents can be continuously read at a high speed.

  Next, an image reading unit in this embodiment will be described. In the flow-reading mode, the document 201 passes through the surface of the document table glass 211 as described above. At this time, the first mirror unit 219 and the second mirror unit 220 are moved by the motor 218 and fixedly arranged at the positions shown in the drawing. Therefore, the original 201 is irradiated by the illumination lamp 212 in the first mirror unit 219 when opposed to the surface of the original table glass 211, and the reflected light passes through the mirrors 213, 214, and 215 and onto the CCD sensor 217 by the lens 216. Imaged. The reflected light input to the CCD sensor 217 is converted into an electrical signal by the sensor, and the electrical signal of the pixel is converted to digital data by an A / D converter (not shown) and input to the controller 300 as a pixel signal Din. The image data is data in units of original pages.

  In this method, a rod-shaped light source is used, a reading line is set parallel to the longitudinal direction, and the document is conveyed in a direction perpendicular to the reading line. A direction parallel to the reading line is defined as a main scanning direction, and a direction perpendicular to the reading line (original transport direction) is defined as a sub-scanning direction. The CCD sensor 217 may be color or monochrome, but in the present embodiment, it is assumed to be a monochrome sensor for the sake of simplicity.

  In addition to the flow-reading mode, there is a method in which an image is read by placing a document to be read on the platen glass 211. In the case of this method, the first mirror unit 219 including the mirror 213 and the illumination lamp 212 moves under the platen glass 211 on which the document is placed at a speed v. Further, the second mirror unit 220 including the mirrors 214 and 215 moves in the same direction as the first mirror unit 219 at a speed of 1/2 v, thereby scanning the front surface of the document 201. The first mirror unit 219 and the second mirror unit 220 are driven by a motor 218.

  Note that the streak detection process, the streak correction process, and the blank sheet detection process, which will be described later in this embodiment, are executed by reading the white surface of the transport roller 224 facing the sensor at the start of the flow reading mode or between sheets. The interval between sheets is a section where the document 201 does not pass through the surface of the document table glass 211 in the flow reading mode, and is a section between the document to be read and the document. The streak correction process and the blank sheet detection process are also executed when the document is read.

<Details of waste accumulation>
FIG. 7 is a diagram illustrating the stagnation of dust on the platen glass. In a copying machine equipped with an automatic document feeder as in the present embodiment, when a document is read in the above-described flow reading mode, dust fixedly attached on the platen glass 211 is read, and a streak appears in the read image. Occurs. This occurs when dust stays on the original platen glass 211 between the first mirror unit 219 and the transport roller 224 (the intersection of the original platen glass 211 and the dotted line in the figure). This streak is generated due to stagnation of dust and is not an image written on the original. Therefore, there is no correlation with surrounding pixels in the document.

  Next, FIG. 8 is a diagram showing an example of a read image when dust adheres to the platen glass. First, FIG. 8A shows a read image when dust does not adhere. Next, FIG. 8B shows an image obtained when dust adheres to the platen glass by reading the document as in FIG. In this case, it can be understood from the figure that black streaks occur in the acquired image. FIG. 8C shows an image obtained when dust adheres to the original platen glass when a blank original is read. In this case, it can be understood from the drawing that it has an adverse effect on the blank sheet detection process described later.

<Copier-Controller>
FIG. 3 is a block diagram showing in detail a hardware configuration of the copying machine used in the present embodiment, particularly a configuration example of the controller.

  The controller 300 is connected to a scanner unit 110 as an image input device, a printer 100 as an image output device, a LAN 313, and a public line (WAN) 317, and controls the operation of the copier as well as image information and Performs device information input / output control.

  The CPU 301 is a processor that controls the entire copying machine, and comprehensively controls access to various connected devices based on a control program stored in the ROM 303. Further, the CPU 301 comprehensively controls various processes performed in the controller 300.

  A RAM 302 is a system work memory for the CPU 301 to operate, and is also an image memory for temporarily storing image data and the like.

  A ROM 303 is a boot ROM and stores a system boot program.

  The HDD 304 is a hard disk drive, and mainly stores information (system software) and image data necessary for starting up and operating the computer. These data may be stored not only in the HDD 304 but also in a recording medium that can be stored even when the power is turned off.

  A LANC (LAN controller) 305 is connected to the LAN 313 and inputs / outputs output image data and information related to device control with the user PC 314.

  A local IF (local interface) 306 is an interface such as a USB, and is connected to a user PC 316 or a printer via a cable 315 to input / output data.

  The MODEM 307 is connected to the public line 317 and inputs / outputs data.

  A printer IF image processing unit 308 is connected to the printer 100 and communicates with a CPU mounted on the printer 100. The printer IF image processing unit 308 performs image processing for synchronous / asynchronous conversion of image data and print output.

  A scanner IF image processing unit 309 is connected to the scanner unit 110 including the document feeder 111 and communicates with a CPU mounted on the scanner unit 110. The scanner IF image processing unit 309 performs image processing for image reading including synchronous / asynchronous conversion of image data, streak detection processing described later, and the like.

  The image rotation unit 310 performs rotation processing on the input image data based on the processing conditions set by the user via the operation unit 120 and the orientation of the document.

  The image compression / decompression unit 311 performs a process of compressing multi-value image data to JPEG, binary image data to JBIG, MMR, MH, and the like, and further expanding the compressed image data as necessary.

  The operation unit IF 312 is an interface for outputting image data to be displayed on the operation unit 120 from the controller 300 to the operation unit 120 and outputting information input from the operation unit 120 by the user of the copier to the controller 300. .

<Copier-operation unit>
FIG. 4 is a diagram illustrating a configuration example of the operation unit 120. The liquid crystal operation panel 401 is a combination of a liquid crystal and a touch panel, displays an operation screen, and sends information to the controller 300 when a display key is pressed by the user. A start key 402 is used to start an operation for reading and printing a document image, or to start other functions. The start key incorporates LEDs of two colors, green and red, indicating that start is possible when the green is lit, and that start is not possible when the red is lit. A stop key 403 functions to stop an operation in progress. The hard key group 404 includes a numeric keypad, a clear key, a reset key, a guide key, and a user mode key.

<Scanner IF image processing unit>
FIG. 5 is a block diagram illustrating a configuration example of the scanner IF image processing unit 309. As described above, in the reading mode, fixed adhering dust, dust, scratches, dirt, etc. present at the original reading position of the optical system fixing position are streaks. In the following description, a case in which the streak generation factor is dust is taken as an example, but the present embodiment can be applied even to a streak generation factor other than dust.

  The scanner IF image processing unit 309 is connected to a register (not shown), and operates based on a control parameter set in the register. It is assumed that writing to the register is performed by the CPU 301 and the scanner IF image processing unit 309.

  The shading correction unit 501 performs a correction process on the image data Din so as to obtain an image with uniform brightness with respect to luminance unevenness due to the characteristics of the optical system and the imaging system using a known technique. The pixel signal Dsh subjected to the shading correction process is output to the subsequent stage. Note that the image data Din is a luminance signal, and the whiter the pixel value is, and the blacker the pixel value is.

  The streak correction unit 502 acquires streak position information G from a streak detection unit 504 described later when dust that causes streaks adheres to the document reading position, and the streak correction unit 502 streaks in the scanned image generated due to the dust. Correction processing is performed to make the image less noticeable. Specifically, the streak correction unit 502 performs correction processing on streaks using normal pixels existing around the streaks to reduce the influence of dust. Hereinafter, the pixel signal in the streak is referred to as an abnormal pixel. A detailed correction method of the streak correction process will be described later. The pixel signal Dh subjected to the streak correction process is output to the subsequent stage.

  The gamma correction unit 503 corrects a difference in color characteristics between the reading element and the device using a known technique. The pixel signal Dg subjected to the gamma correction process is output to the subsequent stage.

  The DMAC 505 is a DMA controller. This is a DMA controller 505 that directly plays the role of writing the pixel signal Dg output from the gamma correction unit 503 as data Dout to a specified area of the image memory (RAM 302) without using a CPU.

  The streak detection unit 504 detects the streak position due to the attached dust when fixed dust that causes streaks adheres to the document reading position. The detected streak position information G is passed to the streak correction unit 502, and is used to specify a correction range (streaks) when an image having streaks is actually input.

The blank sheet detection processing unit 506 performs blank sheet detection processing, that is, determines whether or not the target image data is blank, for the image data that has undergone the line correction and the gamma correction.
Here, the blank paper refers to a document determined to have no print information (that is, content). If there is no print information, a colored document such as colored paper and recycled paper are also handled as blank paper. That is, blank sheet detection is determination of the presence or absence of content printed on a document. In addition, image data when these are read, and image data only of show-through when they are read are also referred to as blank sheets below. On the other hand, an original document described with a small amount of characters or thin characters printed with halftone dots is not a blank sheet. Image data obtained by reading a sheet containing print information handwritten or printed by a printer is called content data.

<Streak detection unit>
FIG. 6 is an exemplary block diagram of the streak detection unit 504 according to the present embodiment. As described above, the streak detection unit 504 operates only when the flow reading mode using the document feeder 111 is started or when the white surface on the conveyance roller 224 is read between sheets. This is controlled by the CPU 301 setting a control parameter indicating ON / OFF of the operation of the streak detection unit 504 in a register (not shown), and the streak detection unit 504 reading the register set value.

  In this embodiment, X pixels are read in the main scanning direction (that is, the line sensor arrangement direction), and Y lines are read in the sub-scanning direction (that is, the document conveyance direction). The image data obtained by reading the white surface is hereinafter referred to as white plate data.

  The binarization circuit 601 binarizes the pixel signal by comparing the image data Dsh after shading correction with the binarized slice level, and inverts it. If dust adheres to the white surface and the pixel signal is affected, the binarization result is almost “1” for each line with respect to the same position in the main scanning direction (less than the binarized slice level). It becomes. The binarized slice level is a parameter given by the CPU 301 and can be changed. For example, the binarization circuit 601 compares the pixel values serially input in raster order with the slice level for each pixel and outputs an inverted binarized signal.

  The accumulation counter 602 counts the number of 1 of the binary signal output from the binarization circuit 601. In order to specify the dust position, it is necessary to execute the cumulative count for a plurality of lines in the sub-scanning direction. Therefore, the cumulative counter 602 has a configuration in which counters corresponding to pixels in one main scanning line are arranged in parallel, for example. The output signal of the binarization circuit 601 is supplied to a counter corresponding to the pixel position in the main scanning direction of the pixel, and the pixels at the same position in the main scanning direction and less than the binarized slice level arranged in the sub scanning direction. Numbers are accumulated. Hereinafter, the arrangement of pixels in the sub-scanning direction at the same position in the main scanning direction is referred to as a column, and the arrangement of pixels in the main scanning direction at the same position in the sub-scanning direction is referred to as a line.

  The comparator 603 compares the cumulative count value for each column output from the cumulative counter 602 with the streak determination level. Here, the main scanning position of the column having the cumulative count value exceeding the streak determination level is determined as the streak position. If the comparator 603 determines that the position is a streak position, the comparator 603 transmits the streak position information G to the streak correction unit 502. The streak position information can be, for example, a binary flag signal (streaky flag) of “1” and “0” for each position in the main scanning direction. That is, the streak position information is, for example, a line of streak flags for one line, and the streak flag “1” represents a streak. The streak determination level is a parameter given by the CPU 301 and can be changed.

<Streak detection procedure>
FIG. 9 is an exemplary flowchart of the streak detection process according to the embodiment. This streak detection process is executed by the streak detection unit 504. The counter used here is initialized before reading the document image, and the procedure of FIG. 9 is executed for the target pixel.

  Here, it is assumed that the streak detection unit 504 is connected to a register (not shown), and operates based on a control parameter set in the register, for example, a binarized slice level or a streak determination level. It is assumed that writing to the register is performed by the CPU 301 and the streak detection unit 504.

  In step S901, the binarization circuit 601 sequentially sets image data input in synchronization with the pixel synchronization signal as a target pixel, and determines whether the target pixel exceeds the binarized slice level. If the binarized slice level is exceeded, the process proceeds to step S902. On the other hand, if it does not exceed the binarized slice level, the process proceeds to step S903. In S902, the cumulative counter 602 increments the cumulative count value corresponding to the main scanning position of the pixel that has not reached the binarized slice level by 1, that is, increases it by one.

  In step S903, the comparator 603 determines whether the accumulated count value for each column exceeds the streak determination level at each main scanning position. If the streak determination level is exceeded, the process proceeds to S904. On the other hand, if it does not exceed the streak determination level, the process proceeds to step S905.

  In step S904, the comparator 603 determines that there is a streak at the main scanning position where the accumulated count value exceeds the streak determination level, and sets the detection result relating to the main scanning position to “1”. In step S905, the comparator 603 determines that there is no streak in the main scanning position where the accumulated count value is equal to or lower than the streak determination level, and sets the detection result relating to the main scanning position to “0”.

  In step S <b> 906, the comparator 603 transmits the line detection result to the line correction unit 502.

  In step S907, the detection result holding unit 604 holds the above-described streak information. For example, when a pixel determined to be a new streak is generated, the number information of the retained streak is incremented by one.

<Streak correction section>
Next, streak image correction processing will be described. In the present embodiment, it is assumed that the streak correction unit 502 is used to correct the streak image. FIG. 10 is a flowchart illustrating an example of the streak correction process according to the embodiment. This streak correction process is executed by the streak correction unit 502. Here, it is assumed that the streak correction unit 502 is connected to a register (not shown), and operates based on a control parameter set in the register. It is assumed that writing to the register is performed by the CPU 301 and the line correction unit 502. It is assumed that the detection result signal G (that is, the streak flag) input to the streak correction unit 502 is a detection result for a position in the main scanning direction (that is, the same column) that is synchronized with the input image data Dsh. Further, the streak correction unit 502 is the number (Tw + 1) of detection results (streaks flag) G input from the streak detection unit 501 synchronized with the input pixel and the position in the main scanning direction plus one threshold value Tw referred to in S1002. A buffer for buffering pixels is provided. Accordingly, if the stripe has a width equal to or smaller than the threshold value Tw, the width can be determined. In order to synchronize with the detection information, the pixel signal Dsh has a buffer that holds the same amount. These buffers can access information therein in units of pixels. For example, referring to FIG. 11, if the target pixel is D2 and the threshold value Tw is a value w, the buffer includes w + 1 pixels from the pixel D1 to the pixel immediately before the target pixel D2, and The corresponding detection signal G is stored. This buffer is initialized every time processing is completed for one main scan so as not to extend between main scans. Here, in order to simplify the configuration. The line correction unit 502 may be configured to share a memory that holds the detection result for one line of the line detection unit 501. In this case, the buffer described above is unnecessary. Alternatively, every time the line correction unit 502 finishes processing for one line, the detection result for one line may be read from the line detection unit 501 in a batch. In this case, the streak correction unit 502 still needs a buffer for holding the detection result for one line, but there is no restriction on the settable value of the threshold Tw.

  In step S <b> 1001, the line correction unit 502 detects the width in the main scanning direction of the line detected by the line detection unit 504. For example, since the streak detection unit 504 outputs “1” for the main scanning position constituting the streak, the number of “1” s continuously arranged in the main scanning direction is detected. For example, if two “1” s are consecutive, the width of the stripe is “2”. However, the width of the streak determined here is a streak whose width is fixed by sandwiching pixels on both sides of the streak. For example, when the detection result of the main scanning position corresponding to the target pixel is “1”, the width cannot be determined even if it is a streak. Accordingly, in S1001, first, if the determination result signal G corresponding to the target pixel input from the streak detection unit 2440 is 0, it is determined whether the detection signal G corresponding to the pixel immediately before the buffered target pixel is 1 or not. If it is not 1, it is determined that there is no streak, and the process is terminated for the target pixel. Since this branch does not exist in FIG. 10, for example, a value exceeding the threshold value Tw, for example, Tw + 1 is set as the width value. Thereby, S203 to S204 are skipped. If the signal G corresponding to the pixel immediately before the buffered pixel of interest is 1, 0 is searched from the position toward the head of the buffer (ie, the pixel that is no longer buffered). If 0 is found in the buffer, the number of 1s following the 0 becomes the stripe width. Therefore, the value is stored as the line width of the stripe. Also, the found 0 and the target pixel correspond to pixels that sandwich the stripe. If there is no 0 in the buffer, the width of the stripe exceeds the threshold value Tw, so a value exceeding the threshold value Tw, for example, Tw + 1 is set as the stripe width. This procedure determines the streak width. When the detection result is stored in a line unit buffer, the line width can be easily determined by scanning the buffer. In this case, the processing of S1001-S1002 may be performed once per line, and S1003-S1004 is executed for the target pixel belonging to the streak.

  In S1002, the streak correction unit 502 determines whether or not the detected width is equal to or less than a predetermined threshold (Tw or less). When the predetermined threshold value Tw is exceeded, the correction process is skipped. On the other hand, if it is equal to or smaller than the predetermined threshold value, the process proceeds to step S1003. This threshold is, for example, 10 pixels at 600 dpi. This is a threshold value that is conscious that characters and the like appear to be crushed when an image is confirmed with human eyes. For example, a streak having a width of 11 pixels or more is not corrected.

In step S <b> 1003, the streak correction unit 502 calculates a correction value for correcting the data of the pixels constituting the streaks using the data of other pixels adjacent to the pixels constituting the streaks. For example, a linear interpolation value is calculated from data of adjacent pixels located on the left and right of the correction target pixel. One of the two pixels sandwiching the stripe is the pixel of interest. The other is a pixel corresponding to the pixel position where the value 0 searched for in the buffer of the signal G in the description of S201 is stored. In S1004, the streak correction unit 502 uses the correction target pixel data as a correction value. Correct by In this correction, for example, correction may be made from the pixel on the sub-scanning line existing in the place that goes back Th line from the target pixel to the pixel. As described above, the reason for performing the correction from the location that goes back the Th line is that when the correlation is detected continuously for the Th line in the streak image detection process (step S904), the target pixel at that time is detected as a streak. Due to being. For this purpose, for example, a line buffer for Th lines may be provided and the pixel data in the line buffer may be corrected.

Next, FIG. 11 shows details of the line correction processing. In the figure, Y represents a correction result pixel, D1 represents a left end reference adjacent pixel, D2 represents a right end reference adjacent pixel, K represents a pixel position from the streak correction processing start position, and W represents a streak pixel width. At this time, when the dust correction process is executed by linear interpolation, the correction result, that is, the pixel value Yk at the position K is calculated by the following equation.
Yk = D1 + K / W × (D2-D1)
The pixels determined to be streaks are replaced by Yk thus obtained. Further, the streak correction process may be executed by simply performing a replacement process using surrounding normal pixels (reference adjacent pixels) without using an operation.

  Here, it will be described that this streak correction process does not necessarily correct all streaks. For example, if the correction is not started from the place where the streak is traced back by the Th line, the streak image remains for the Th line. Also, no correction is made when the streak exceeds the predetermined width in step S1002. There are also cases where floating dust that causes streaks moves as the document is conveyed, streaks become diagonal lines or curves, or the density (that is, the reflected light of the dust) changes and the correlation cannot be determined correctly. Such streaks that cannot be corrected greatly affect the blank sheet detection process described later.

  Note that the streak information used for blank sheet determination result selection processing described later may be all the detected streak information or only the streak information that cannot be corrected as described above.

<Blank detection processing section>
FIG. 12 is a diagram showing an internal configuration of the blank sheet detection processing unit 506 in the present embodiment. A register (not shown) is connected to the blank sheet detection processing unit 506, and a blank sheet determination parameter (referred to as a control parameter or simply a parameter) and a processing result are held in the register. Writing to the register is performed by the CPU 301 and the blank page detection processing unit 506, and the blank page detection processing unit 506 operates by reading the control parameters set in the register.

  In the present embodiment, it is assumed that the RAM 302 holds control parameters that are equal to or greater than the number of parameter sets that can be set in the register. This will be described in detail with reference to FIG. FIG. 18 shows ten sets of control parameters (determination criteria) P [0] to P [9] stored in the RAM 302. Here, P [0] is a parameter that is most difficult to determine as blank paper (it is difficult to determine that print information is not included). On the other hand, P [9] is a parameter that is most easily determined to be blank (easily determined that print information is not included).

  In this embodiment, the CPU 301 selects five control parameters RP [0] to RP [4] to be set in the register from P [0] to P [9]. In this embodiment, P [0] to P [4] are selected as initial values of RP [0] to RP [4], but another parameter may be selected as the initial value. The CPU 301 sets the initial value of the selected parameter, in this example, RP [0] to RP [4] in the register. Also, the parameters RP [0] to RP [4] set in the register are changed by blank sheet detection parameter changing processing described later. The blank sheet detection parameter changing process will be described later with reference to FIG.

  The blank sheet detection processing unit 506 holds the determination result as to whether or not the read document is finally blank, and as a feature in the present embodiment, holds a plurality of the results in association with the parameters (in FIG. 18). R [0] to R [4]). That is, it is possible to have a plurality of levels of control parameters and hold blank sheet determination results at a plurality of detection levels corresponding to each parameter. Details such as what control parameters are provided will be described later.

  The description returns to the blank sheet detection processing unit 506 shown in FIG. A region control unit 1201 controls a region for generating histograms and edge information from input image data. When the original is read by the original feeder 111, the leading end, the rear end, the left end, and the right end of the original depend on the original conveyance configuration and the CCD sensor 217 light source configuration. The area control unit 1201 determines whether the pixel position of the currently input pixel in the document is an effective area or an invalid area, and generates a signal indicating whether the pixel is an effective area or an invalid area. Further, the area control unit 1201 divides the main scanning effective area and the sub-scanning effective area of the document into a plurality of areas. That is, the region control unit 1201 outputs the valid / invalid region signal 1301 and the region signal 1302 to the subsequent processing unit (histogram generation unit 1202 and edge information generation unit 1204) in addition to the pixel signal Dg.

  The histogram generation unit 1202 generates a histogram of each region from the image data Dg, the valid / invalid region signal 1301, and the region signal 1302.

  A histogram analysis unit 1203 determines whether or not the document is blank from the first to ninth histograms generated by the histogram generation unit 1202.

  The edge information generation unit 1204 generates edge information of each region from the image data Dg, the valid / invalid region signal 1301, and the region signal 1302.

  The edge information analysis unit 1205 determines whether the document is blank from the first to ninth edge information generated by the edge information generation unit 1204.

  A blank page determination unit 1206 finally determines whether or not the document is blank from the determination signals of the histogram analysis unit 1203 and the edge information analysis unit 1205. Here, if the determination signal from the histogram analysis unit 1203 is a blank sheet and the determination signal from the edge information analysis unit 1205 is a blank sheet, a signal for determining that the read document image is a blank sheet is acquired. If any one of the determination signals from the histogram analysis unit 1203 and the edge information analysis unit 1205 is content, a signal for determining that the content is content is acquired. In addition, the CPU 301 is notified that the blank sheet detection process is completed.

<Internal Configuration of Histogram Generation Unit 1202>
The internal configuration of the histogram generator 1202 is shown in FIG. The data distribution unit 1303 is a data distribution unit that reflects the frequency of the pixel value in the histogram of each subsequent divided region according to the pixel signal Dg, the valid / invalid region signal 1301, and the region signal 1302. In each histogram, the frequency corresponding to the pixel value of the distributed image data Dg is added. Since the image data Dg, the valid / invalid area signal 1301 and the area signal 1302 need to be synchronized, the image data Dg is also delayed in accordance with the signal delay in the area control unit or the like, but is omitted in the description. . If the valid / invalid area signal 1301 indicates an invalid area, no output is performed to the subsequent stage. In addition, in the case of generating a histogram with 32 gradations (5 bits) for the bit precision (for example, 8 bits) of the image data Dg, a function of outputting 5 bits obtained by removing the lower 3 bits of the pixel value to the subsequent stage Also have. In other words, each pixel value included in the input image data Dg is quantized and reflected in the frequency distribution. For the sake of simplicity, the second to eighth histograms are omitted, and only the first histogram 702 and the ninth histogram 703 are shown in FIG.

<Internal Configuration of Edge Information Generation Unit 1204>
FIG. 13B shows an internal configuration of the edge information generation unit 1204. An edge extraction unit 1306 extracts edges from the image data Dg. Here, for example, a convolution operation is performed using a 7 × 7 matrix, and an edge signal indicating that the output is an edge portion when the output is equal to or greater than the threshold value and a non-edge portion when the output is less than the threshold value is output to the subsequent stage. It is assumed that the 7 × 7 matrix coefficient and the threshold value used here are read from a register (not shown). The data distribution unit 1307 reflects the number of edges in the subsequent stage according to the edge signal, the valid / invalid region signal 1301 and the region signal 1302 output from the edge extraction unit 1306. That is, if the edge signal indicates an edge, a signal is output to the number of edges in the divided area specified by the valid / invalid area signal 1301 and the area signal 1302, and thereby the number of edges in the corresponding divided area is, for example, Add one. Here, when the valid / invalid area signal indicates an invalid area, output to the subsequent stage is not performed. For simplicity of explanation, the second to eighth edge numbers are not shown, and only the first edge number 1308 and the ninth edge number 1309 are shown in FIG.

  Note that the edge extraction unit 1306 has a plurality of threshold parameters to be used in order to obtain a plurality of levels of blank page determination results. Thereby, for example, the number of edges according to each threshold parameter is held in the edge number holding unit such as the first edge number 1308 and the ninth edge number 1309.

<Internal Configuration of Histogram Analysis Unit 1203>
Details of the histogram analysis unit 1203 will be described below with reference to FIG. The average value calculation unit 1402 calculates first to ninth average values 808 from the first to ninth histograms 1401 generated by the histogram generation unit 1202.

  The variance value calculation unit 1403 calculates the first to ninth variance values from the first to ninth histograms generated by the histogram generation unit 1202 and the first to ninth average values calculated by the average value calculation unit 1402. To do. The variance value calculation unit 1403 receives the average value, the input luminance value, and the histogram 1401 output from the average value calculation unit 1402, and calculates a difference from the average value. Here, the difference value is calculated from (average value−luminance value) × (average value−luminance value) × frequency. Then, the difference value is cumulatively added to all the luminance values, and further, the variance value is calculated by dividing the cumulative added value by the total frequency. In other words, the difference value here is the sum of squares of deviations from the average luminance with respect to the pixel of the focused luminance. For example, when there is some printed matter, the variance value is high, and when there is only the background color, the variance value is small.

  The average value determination unit 1404 compares the average value calculated by the average value calculation unit 1402 with a threshold value, and determines whether or not the divided area is covered with a dark printed matter (for example, a dark portion of a photograph). Here, when the average value is equal to or greater than the threshold value, it indicates that the print information is not included, that is, blank paper, and when it is less than the threshold value, a determination signal indicating that the print information is included (not blank) is output. The average value determination unit 1404 performs determination for each of the first to ninth regions.

  A variance value determination unit 1405 compares the variance value calculated by the variance value calculation unit 1403 with a threshold value, and determines a variation in luminance values of the divided areas. Here, when the variance value is equal to or greater than the threshold, it is determined that the printed matter is present because the luminance variation is large, i.e., includes print information. That is, it is determined that the sheet does not include print information, and a determination signal indicating the determination result is output. The variance value determination unit 1405 performs determination for each of the first to ninth regions.

  A histogram determination unit 1406 determines whether or not the document image 203 is blank from the determination signal of the average value determination unit 1404 and the determination signal of the variance value determination unit 1405. Here, if the determination signal of the average value determination unit 1404 and the determination signal of the variance value determination unit 1405 in the first to ninth regions are all blank page candidates, a blank page candidate signal is output. When there is a signal indicating the content even in one area, a determination signal 1407 for determining that it has print information, that is, is not blank is output.

  Here, if there is a signal indicating the content even in one area, it is determined that it has print information. For example, threshold processing is performed on the number of areas determined to have print information, and printing is performed for a predetermined number of areas or more. When it is determined that information is included, a signal that determines that the image data has print information may be output.

  Note that a set of threshold parameters used in the average value determination unit 1404 and a threshold parameter used in the variance value determination unit 1405 in order to obtain a plurality of levels of blank page determination results is held in a plurality of registers or the like according to the determination level selected in advance. Each threshold parameter is used for determination to obtain a determination result corresponding to the determination level. Accordingly, the determination signal 1407 output from the histogram determination unit 1406 becomes a plurality of determination results based on a plurality of levels (a plurality of parameters).

<Internal Configuration of Edge Information Analysis Unit 1205>
Hereinafter, the details of the edge information analysis unit 1205 will be described with reference to FIG.

  The maximum value calculation unit 1501 obtains the maximum number of edges from the first to ninth number of edges generated by the edge information generation unit 1205. The minimum value calculation unit 1502 obtains the minimum number of edges from the first to ninth number of edges generated by the edge information generation unit 1205.

  An upper limit determination unit 1503 performs threshold processing on the maximum number of edges obtained by the maximum value calculation unit 1501 and outputs a determination signal as to whether or not the image data has print information. Here, if it is equal to or greater than the threshold, there is print information, and if it is less than the threshold, a blank candidate signal is output. For example, in the case of a digital multi-function peripheral or the like, security dots or the like may be printed for the purpose of restricting copying of printed printed matter. This printing may be performed in front of the image, and when comparing the edge distribution between the subsequent areas, the same number of edges may be counted in all the areas, and may be determined to be blank. That is, when the number of edges exceeds a predetermined number, it is necessary to determine that there is print information.

  The lower limit determination unit 1504 performs threshold processing on the maximum number of edges obtained by the maximum value calculation unit 1501 and outputs a determination signal as to whether or not the sheet is blank. Here, when the value is equal to or greater than the threshold, there is print information, and when the value is less than the threshold, a determination signal for determining that the sheet is blank is output. For example, in the case of good quality paper such as coated paper, edges may not be extracted. In other words, if the number of edges is 10 in one area and the number of edges is 0 in another area, 0/10 = 0 when compared with the relative value between the areas, and the correlation value is minimized, and it is determined that the print information is included. May be. Here, a low correlation value indicates that the difference in the number of edges between the regions is large. For example, in the case of general white paper, when the maximum number of edges is 320 and the minimum number of edges is 300, 300/320 = 0.93, and the correlation value is high. That is, when the number of edges is less than a predetermined number in each area, it is necessary to determine that the sheet is blank with no print information.

  A division unit 1505 performs a division process using the maximum number of edges calculated by the maximum value calculation unit 1501 and the minimum number of edges calculated by the minimum value calculation unit 1502 to calculate a correlation value between regions. Here, the correlation value is calculated by the minimum edge number / maximum edge number.

  The minimum value determination unit 1506 outputs a determination signal as to whether or not the sheet is blank from the correlation value calculated by the division unit 1505. Here, the correlation value calculated by the dividing unit 1505 is compared with a threshold value, and if it is equal to or greater than the threshold value, it is determined that the blank paper has no print information, and if it is less than the threshold value, it has print information. In other words, when the correlation value is high, the difference between the maximum number of edges and the minimum number of edges is small, so it is determined to be blank. It is determined that it is (original).

  The edge determination unit 1507 determines whether or not the sheet is blank from the determination signal from the upper limit determination unit 1503, the determination signal from the lower limit determination unit 1504, and the determination signal from the minimum value determination unit 1506. Here, when the determination signal of the upper limit determination unit 1503 indicates that there is print information, the determination signals of the lower limit determination unit 1504 and the minimum value determination unit 1506 are not referred to, and a determination signal for determining that the print information is included is output. Further, when the determination signal of the upper limit determination unit 1503 is a blank sheet candidate and the determination signal of the lower limit determination unit 1504 is a blank sheet, the determination signal of the minimum value determination unit 1506 is not referred to and a determination signal for determining the blank sheet is output. When the determination signal of the upper limit determination unit 1503 is a blank candidate and the determination signal of the lower limit determination unit 1504 is an image data candidate having print information, the determination signal of the minimum value determination unit 1506 is output.

  Note that the threshold parameter used in the upper limit determination unit 1503 to obtain a plurality of levels of blank page determination results is held in a plurality of registers or the like according to the determination level selected in advance, and each threshold parameter is used for blank page determination. The determination result according to the determination level is obtained. As a result, the edge determination unit 1507 outputs a plurality of edge determination results based on each threshold parameter.

  In addition, the blank sheet determination unit 1206 obtains a determination result of a plurality of levels with reference to the result of the plurality of threshold parameters used in the histogram analysis unit 1203 and the edge information analysis unit 1205 in order to obtain a plurality of levels of blank sheet determination result. That is, a plurality of determination results for determining whether the image data finally input to the blank sheet detection processing unit 506 is a blank sheet having no print information is acquired and held.

  In addition, although it was set as the structure which has multiple threshold parameters of each part (1306, 1404, 1405, 1503) as a suitable embodiment, it is not restricted to this but finally it has multiple threshold parameters and various control parameters of another process part. It is good also as a structure which acquires and hold | maintains several determination results.

  Further, if it is attempted to comprehensively acquire the results of each threshold parameter, a result with an excessive calculation load “number of threshold parameters × number of types of threshold parameters” is calculated. In order to avoid this, the threshold parameter may be provided with levels (for example, five levels) and the threshold parameter for each level may be set. As a result, various calculations are limited to combinations of parameters fv for each level (for example, five patterns and five determination results), so that the calculation load is reduced.

<Blank judgment level setting>
FIG. 22 shows a screen for setting a blank sheet determination level, which includes a minus button 2201, a plus button 2202, a cancel button 2203, an OK button 2204, and an adjustment bar 2205 indicating the current determination level.

In FIG. 22, the adjustment bar 2205 indicates that the blank page determination level is set to 3. The user can press the minus button 2201 to lower the blank page determination level by 1, that is, make it difficult to determine that the page is blank. When adjustment is performed to lower the blank sheet determination level by 1 using the adjustment bar 2205, the threshold value used for blank sheet determination is changed to a value that is difficult to determine blank sheet.
On the other hand, when the plus button 2202 is pressed, the blank page determination level can be increased by 1, that is, it can be easily determined that the page is blank. . When adjustment is performed to increase the blank page determination level by 1 using the adjustment bar 2205, the threshold used for blank page determination is changed to a value that is easily determined to be blank.
FIG. 22 shows that the user can set the blank page determination level in five stages.

  When the CPU 301 detects that the OK button 2204 has been pressed, the CPU 301 holds the blank sheet determination level setting content that was displayed immediately before and closes this operation screen. When the pressing of the cancel button 2203 is detected, the setting content of the blank sheet determination level immediately before opening the operation screen is held, and the operation screen is closed. The blank page determination level is set before the start key 402 of the operation unit 120 is pressed in the follow-up chart of the scanning mode to be described later. However, the present invention is not limited to this. You may do it.

<Flowchart of the scanning mode>
FIG. 16 is a flowchart showing the flow of the non-stop reading mode in the present embodiment. The flowchart shown in FIG. 16 is executed by the CPU 301 in accordance with a program stored in the HDD 304.

  In step S <b> 1601, it is detected that a plurality of documents to be read are set on the tray 112 of the document feeder 111. In S1602, the flow-reading mode is specifically started by detecting the pressing of the start key 402 of the operation unit 120.

  In S1603, white plate data is read. The white plate data is image data obtained by reading the white surface on the transport roller 224 as described above. This is obtained by fixing the optical system shown in FIG. 3 and capturing the reflected light on the white surface before starting the conveyance of the original 201 by the CCD sensor 217. At this time, the conveyance roller 224 is driven to read the white surface for a plurality of lines. In step S1604, the scanner IF image processing unit 309 performs various types of image processing on the read white plate data, and executes processing for storing the white plate data in the RAM 302. Here, the above-described streak detection process in the streak detection unit 504, the streak correction process by the streak correction unit 502, and the blank page detection process in the blank page detection processing unit 506 are executed.

  In step S <b> 1605, using the plurality of blank sheet detection results held in the blank sheet detection processing unit 506, processing for changing the blank sheet detection parameter to a parameter that is not affected by streaks is performed. Details will be described later with reference to FIG.

  In S1606, it is determined whether or not the change of the blank sheet detection parameter is completed. If the change of the blank page detection parameter is not completed, the changed blank page detection parameter is set in a register (not shown), and the process returns to S1603. If the change is completed, the changed blank page detection parameter is set in a register (not shown), and the process proceeds to S1607. Steps after S1607 are blank page determination control processing for the scanned original image instead of the reference image.

  In step S1607, the original is actually read using the original feeder 111. Specifically, as described with reference to FIG. 2, the original is read one by one from the tray 112 of the original feeder 111, and the original is read. In step S <b> 1608, the scanner IF image processing unit 309 performs various kinds of image processing on the read document data, and executes processing for storing the document data in the RAM 302. Here, the above-described streak correction process by the streak correction unit 502 and the blank page detection process by the blank page detection unit 506 are executed. Here, the line correction processing unit 502 corrects the line using the line position information G generated by the line detection processing unit 504 in step S1604. Further, the blank sheet detection processing unit 506 performs blank sheet detection processing using the blank sheet detection parameter acquired by the blank sheet detection parameter changing process in S1605.

  In step S1609, blank page processing is executed using the blank page determination result generated in step S1608. Blank page skip processing refers to control so that a document surface determined to be blank is not subjected to print processing. In addition, this indicates all useful processing using the blank sheet determination result.

  In step S1610, it is determined whether image reading is to be performed. This is determined by whether or not a document is still left on the tray 112 on the document feeder 111. If there remains a document to be read on the tray 112, the process returns to S1603. If there are no remaining documents to be read, the process ends.

<Blank Detection Parameter Change Processing Flowchart>
FIG. 17 is a flowchart showing the flow of blank sheet detection parameter change processing in the present embodiment. This flowchart describes step S1605 in FIG. 16 in detail, and the flowchart shown in FIG. 17 is executed by the CPU 301 in accordance with a program stored in the HDD 304. In the present embodiment, the blank sheet detection parameter changing process is performed only when reading white plate data (also referred to as a reference white plate or a reference image), and is not performed during an operation of reading a normal document. As described above, the blank sheet detection parameter includes, for example, a threshold parameter used in the average value determination unit 1404, a threshold parameter used in the variance value determination unit 1405, a threshold parameter used in the upper limit determination unit 1503, and the like in this embodiment. A parameter set in which the values of the respective parameters are combined according to the level is configured. In the following description, the threshold parameter refers to a parameter set combined according to the level.

  In S1701, a blank page determination result is acquired. In this case, the CPU 301 accesses the blank page determination unit 1206 and acquires the stored blank page determination result. The blank page determination result is a signal indicating whether or not the read original is blank, for example, 1-bit information (1 is blank (no printing information), 0 is printing information (not blank)). In the present embodiment, the blank page detection (blank page determination) process is executed using each of the five threshold parameters set in the register, and five determination results (blank page detection results) corresponding to each set of parameters. ) Has been acquired. Therefore, in S1701, information indicating the 5-bit blank page determination result generated by the blank page determination unit 1206 is acquired.

  In S1702, using the blank page determination result acquired in S1701, a parameter for switching the blank page determination result is searched. This operation will be described with reference to FIG.

  FIG. 19 shows blank sheet determination results R [0] to R [corresponding to each parameter when blank sheet detection processing is executed on blank sheet data with the blank sheet detection parameters RP [0] to RP [4] registered in the register. 4]. Here, the search is sequentially performed from the parameter results that are difficult to determine as blank pages, and the parameters for which the blank page determination results change from 0 to 1 are specified. Here, if the blank sheet determination result for the blank board data is 0 (print information is present (not blank sheet)), it can be determined that an erroneous determination is made due to the influence of streaks. On the other hand, when the blank sheet determination result for the white board data is 1 (no printing information (blank sheet)), it can be determined that a correct blank sheet determination result can be obtained without being affected by streaks. That is, the blank sheet determination result for the white board data is scanned from the lower detection level, and the parameter for switching from 0 to 1, that is, the parameter for which the detection result is 1 for the first time is specified. This specified parameter can be determined to be the parameter that is most difficult to be determined as a blank sheet without being affected by streaks when blank image data is read. In FIG. 19, since the blank page determination result R [4] is determined to be 1 (blank page), the control parameter RP [4] corresponding to the blank page determination result R [4] is specified as the change point. Note that the currently set detection level corresponds to the middle of five stages, that is, R [2].

  In S1703, it is determined whether a change point exists as a result of the search in S1702. If it is determined that there is a change point (YES in S1703), the process proceeds to S1704. On the other hand, if it is determined that there is no change point (NO in S1703), the process proceeds to S1709.

  In S1704, the parameter change amount up to the change point specified in S1702 is calculated. The amount of parameter change is increased based on the current detection level, for example, so that the parameter corresponding to the detection level at the change point is the parameter corresponding to the lowest detection level among the parameter groups registered in the register. The number of detection levels that should be detected. Therefore, when the detection level is lowered, the amount of change becomes a negative value. In the example of FIG. 19, since the parameter registered in RP [0] is changed to the parameter registered in RP [4], the change amount is 4. The change referred to here is a change by shifting or shifting the parameter. The change is not only the change point, but also other parameters are shifted by the same amount in conjunction with the movement of the parameter at the change point.

  In step S1705, it is determined whether to adjust the change amount calculated in step S1704. If it is determined that the change amount is adjusted, the process advances to step S1706. On the other hand, if it is determined that there is no adjustment of the change amount, the process proceeds to S1707. Specifically, there is a parameter to be set in the register among the parameters P [0] to P [9] shown in FIG. It is determined whether or not to perform the change amount adjustment process. FIG. 20 shows an example where the amount of change is adjusted. If the current parameter set (the parameter currently set in the register is called the current parameter) is P [2] to P [6], and the amount of change calculated in S1704 is 4, The parameter sets are P [6] to P [10]. However, since the upper limit parameter is P [9] and P [10] does not exist, there is no parameter to be set in RP [4]. Accordingly, it can be determined that the amount of change needs to be adjusted.

  In step S1706, the change amount is adjusted. Here, the amount of change is adjusted so that the corresponding parameter does not exist. That is, when there is no parameter to be set on the high level side of the register, the number of parameters that do not exist is subtracted from the change amount determined in S1702. Conversely, if there is no parameter to be set on the low level side of the register, the number of parameters that do not exist is added to the change amount determined in S1702. In the example of FIG. 20, when shifting to the high level side by four levels, there is no parameter to be set in the high level register RP [4]. Subtract and adjust the change amount to 3.

  In S1707, the parameters are changed based on the change amounts calculated in S1704 and S1706. In the example of FIG. 20, parameters P [5] to P [9] are set in registers PR [0] to PR [4].

  In step S1708, the CPU 301 is notified that the parameter change has been completed, and the blank sheet detection parameter change process is terminated.

On the other hand, if a change point cannot be found in S1703, the process branches to S1709. In step S1709, a change amount is set. In this embodiment, five parameters are set in the register. Since it is determined in S1709 that no change point has been found in S1703, the five blank sheet determination results R [0] to R [4] are all 1 or 0. Here, if R [0] to R [4] are all 1, the change amount is set to -4, and if all are 0, 4 is set.
If the blank page determination result R is all 0 (with print information), it is considered that all of the parameters are difficult to determine as blank pages. Therefore, the change amount is set to 4, and control is performed so that the parameters are easily determined as blank pages. . On the other hand, if all the blank page determination results R are 1 (no print information), it is considered that all the parameters are easy to determine as blank pages. Therefore, the change amount is set to -4 and the parameter is changed to a parameter that is not easily determined as blank pages. Control as follows.

  The same processing as S1707 is performed in S1705, S1706 in S1711, and S1712 and S1715 in S1710.

  In S1713, it is determined whether there is a change amount. If it is determined that there is no change amount, the process advances to step S1714. In this case, either the highest detection level is not determined to be blank, or the lowest detection level is determined to be blank. Therefore, there is no room for changing the parameters. On the other hand, if it is determined that there is a change amount, the process of S1714 is not performed, and the blank sheet detection parameter change process ends.

  In step S1714, the CPU 301 is notified that the parameter change has been completed, and the blank page detection parameter change process is terminated. This flowchart is now complete.

If the parameters are changed in S1712, S1715, it is not determined whether there is a change point, so the change is not completed and the process returns to S1606. Therefore, in these cases, the process branches from S1606 to S1603, the reference white plate is read again, and the blank sheet detection process is retried.
Thus, the more print information (image detected as a streak) obtained when the reference image is read, the easier it is for the determination result to be determined as content. Therefore, the more print information obtained when the reference image is read, the more the parameter needs to be changed.

<Details of blank paper detection parameters>
Next, a detailed example of the parameters P [0] to P [9] described above will be described with reference to FIG. As a preferred embodiment, each parameter individually has a threshold value used in the four processes of the average value determination unit 1404, the variance value determination unit 1405, the edge extraction unit 1306, and the upper limit determination unit 1503. In the present embodiment, each threshold parameter is indicated by a level.

  The threshold parameter used in the average value determination unit 1404 is a histogram average luminance threshold in the table. In this embodiment, the lower the level notation, the higher the luminance threshold. Levels that can be set are four levels from level 1 to level 4. The lower the level notation, the higher the histogram variance threshold. That is, the higher the histogram dispersion threshold value, the easier it is to determine that the page is blank.

  The threshold parameter used in the variance value determination unit 1405 is a histogram variance threshold in the table. In this embodiment, the variance threshold is smaller when the level notation is lower. There are 10 levels that can be set, from level 1 to level 10. The lower the level notation, the smaller the dispersion threshold. That is, the smaller the dispersion threshold, the easier it is to determine that the page is blank.

  The threshold parameter used in the edge extraction unit 1306 is the edge detection luminance threshold in the table, and in this embodiment, the lower the level notation, the lower the edge detection threshold. Levels that can be set are four levels from level 1 to level 4. The lower the level notation, the lower the edge detection threshold. That is, the lower the edge detection threshold, the easier it is to determine that the page is blank.

  The threshold parameter used in the upper limit determination unit 1504 is the edge number threshold in the table. In this embodiment, the lower the level notation, the smaller the edge number threshold. There are 10 levels that can be set, from level 1 to level 10. The lower the level notation, the smaller the edge detection value. That is, the smaller the edge detection value, the easier it is to determine that the page is blank.

  In the present embodiment, a suitable combination of these threshold parameters is prepared, and parameters P [0] to P [9] are formed. Here, the parameters P [0] to P [9] indicate the blank page determination level as described above, and the parameter No. 1, that is, P [0] is the parameter that is most difficult to determine as blank page. For example, since the threshold value for the number of edges is level 1, the threshold value is exceeded only by detecting a small number of edges, and it is easy to determine that the sheet is not blank.

  Parameter No. 10, that is, P [9] is the parameter that is most easily determined to be blank. For example, since the threshold value for the number of edges is level 10, even if a small number of edges are detected, the threshold value is rarely exceeded, and it is easy to determine “blank paper”. The same applies to other parameters.

  In the present embodiment, parameters are formed by gradually increasing the level of each threshold parameter. Here, in order to make it easy to determine the density of the content, which is the print information included in the image data, as blank paper according to the thinness, the levels of the histogram average luminance value and the edge detection luminance threshold are increased.

  On the other hand, if it is desired to easily determine that the page is blank according to the print amount of the content, the levels of the histogram dispersion value and the edge number threshold may be increased.

Therefore, in the present embodiment, a parameter that can minimize the influence of streaks is automatically acquired from a blank page determination result using a plurality of parameters, and a blank sheet is obtained according to the content desired by the user from the state where the influence of streaks is excluded. Detection is possible.
The above is an example of the parameter set, but the parameter set is selected by selecting a combination of threshold values for each determination process so that the degree of ease with which it is determined to be blank is changed stepwise.

    Here, FIGS. 21A to 21C show examples of blank sheet detection results in the present embodiment. FIG. 21A shows an example of the result of blank page determination for the white plate data before reading the read image shown in FIG. At this time, streaks are generated in the white plate data before reading the originals shown in FIGS. 8B and 8C, and there is a result that the blank page determination result for the white plate data is determined to be 0 (not a blank page). . In FIGS. 8B and 8C, the blank page determination result is 1 only for the parameter of RP [4].

  FIG. 21B shows a result of parameter change from the result shown in FIG. At this time, since there is a change point in the blank sheet determination result before reading the original in FIGS. 8B and 8C, the parameter is changed. On the other hand, since all the blank page determination results before reading the original in FIG. 21 (a) are 1, the change amount should be −4. However, since the adjustment is +4 in S1711, it is understood that there is no change in the parameters. This can also be understood from the fact that the blank sheet is determined even at the lowest detection level, that is, the level where it is difficult to determine the blank sheet.

  FIG. 21C shows the result of blank page detection performed on the document data shown in FIGS. 8A to 8C using the parameters calculated in FIG. 21B. In FIG. 21C, the originals shown in FIGS. 8A and 8B have the same determination result, and the original shown in FIG. 8C is determined to have a blank sheet detection result of 1 (blank sheet) for all parameters. . The user can select a blank page determination result from the result of FIG. 21C, and it is feasible to select an appropriate blank page detection result without the influence of streaks. That is, for example, according to the procedure of FIG. 17, the blank page detection parameters P [4] to P [8] are set in the registers RP [0] to RP [4] as shown in (b) of FIG. To do. As shown in the user interface of FIG. 22, the user selects one of the five levels of blank paper detection levels set in the register, that is, one of the parameters set in RP [0] to RP [4]. Can be selected. The opportunity for the user to change the detection level, that is, the opportunity to operate the user interface, for example, does not return to S1606 in FIG. 16 immediately after the setting change is completed in S1708 and S1714 in FIG. 17, but the user interface 401 in FIG. It can be given by displaying. In the user interface 401, the user moves the slide bar 2205 to a desired detection level and completes the setting change when the OK button 2204 is touched. In the subsequent blank sheet detection process, the detection result at the set detection level is adopted. In this example, detection results using the five parameter sets RP [0] to RP [4] set in the register are obtained, respectively, of which the parameter set RP [i] corresponding to the detection level selected by the user. The detection result R [i] using is used as the detection result of the blank paper detection process. Before the change by the user, the lowest detection level determined as blank paper (that is, the detection level of RP [0] after parameter resetting) may be set as a default value, or intermediate RP [2] may be set as a default value. It is good. In order to let the operator know whether the detection level set as a result of blank page determination using the image data of the reference white board (that is, the parameter corresponding to the detection level) is that level, the detection level is clearly indicated on the user interface. May be displayed automatically. In the user interface 401 of FIG. 22, it is displayed by the slide bar 2205, but a number indicating the detection level or a word indicating the degree (for example, “strong”, “weak”, etc.) may be displayed. With the above processing, even if there is a situation where streaks can occur in the image due to adhering dust, it is possible to automatically select blank page detection parameters that can exclude the influence of streaks from the blank page judgment result for white board data, and the influence of streaks in the image This makes it possible to execute a blank sheet detection process that reduces the above.

[Modification]
In the above embodiment, the lowest detection level (that is, the detection level that is most difficult to be determined to be blank) among the change points, that is, the detection level determined to be blank for the image data of the reference white board is set as the blank sheet detection parameter. The blank sheet detection parameter is reset so as to be as low as possible in the register group. However, when a change point is found in a plurality of determination results (five in this example), the detection level of the change point is set as a default value of the blank paper detection level for the original, and in particular, a parameter The resetting may not be performed. In this case, the parameter is shifted only when there is no change point in the detection result group using the parameter currently set in the register, and blank page detection is performed until the change point is found or there is no room for parameter change. Try.

[Embodiment 2]
In the first embodiment, when there is a streak on the whiteboard data, the whiteboard data is repeatedly read, and the blank page determination parameter is automatically adjusted from the blank page determination result for the white board data, thereby obtaining a blank page determination result excluding the influence of the streak. Described how.

  In the second embodiment, a description will be given of a case where the load is reduced by automatically adjusting the blank page determination parameter without reading the white board data a plurality of times when the white board data has a streak.

In this embodiment, the processing flow of the scanner IF image processing unit 309 and the flow scanning mode is different from that of the first embodiment. Other configurations and processing flows are the same as those in the first embodiment.
Hereinafter, the present embodiment will be described in detail with respect to differences from the first embodiment.

<Scanner IF image processing unit>
FIG. 23 is a block diagram illustrating a configuration example of the scanner IF image processing unit 309 in the present embodiment. Note that FIG. 23 is partly in common with FIG. 5 and description of common parts is omitted.

  The DMAC 2302 is a DMA controller that plays a role of reading data Din ′ from a specified area of the image memory (RAM 302) and outputting the data Din ′ to the blank sheet detection processing unit 2301.

  The blank sheet detection processing unit 2301 has the same function as the blank sheet detection processing unit 506 shown in the first embodiment. However, the blank sheet detection processing unit 2301 has a selector function for switching data Dg and data Dg ′ as input data based on an input switching setting value read from a register (not shown). When the blank page detection process is executed again after changing the blank page detection parameter, the input switching setting value is set to a value using the pixel signal Dg ′ as input data, and the blank page detection processing unit 2301 uses the pixel signal Dg ′ as input data. And In other cases, the input switching setting value is set to a value using the pixel signal Dg as input data, and the blank sheet detection processing unit 2301 uses the pixel signal Dg as input data.

  With this configuration, it is possible to read the white board data output to the RAM 302 by the DMAC 2302 and execute the blank paper detection processing unit 2301, thereby reducing the processing load for changing the blank paper detection parameter.

<Flowchart of the scanning mode>
FIG. 24 is a flowchart showing the flow of the flow reading mode in the present embodiment. The flowchart shown in FIG. 24 is executed by the CPU 301 in accordance with a program stored in the HDD 304. S2401 to S2405 shown in FIG. 24 are the same processes as S1601 to S1605 in FIG. Also, S2408 to S511 shown in FIG. 24 are the same processing as S1607 to S310 in FIG. Here, the processing flow of S2406 and S2407 shown in FIG. 24 will be described.

  In S2406, it is determined whether or not the change of the blank sheet detection parameter is completed. If the change of the blank paper detection parameter is not completed, the process proceeds to S2407. On the other hand, if the change has been completed, the process proceeds to S2408.

  In step S <b> 2407, the blank sheet detection processing unit 2301 executes the blank sheet data stored in the RAM 302. That is, the input switching setting value is set so that Dg ′ becomes the input image signal of the blank paper detection processing unit 2301. At the same time, the DMAC 2302 is operated to read the data Din ′ from the designated area of the image memory (RAM 302). The designated area is, for example, an area corresponding to a reference white plate in the image data.

This is the end of the flowchart in the flow reading mode in the present embodiment.
With the above processing, it is possible to calculate the blank page detection parameter without repeatedly reading the white board data, and it is possible to reduce the processing load as compared with the first embodiment.

  In the present embodiment, for simplification of explanation, it is assumed that the image data read by the scanner is a monochrome signal having only luminance as a component. However, the present embodiment can also be applied to color signals such as RGB. For example, in the binarization circuit 601 in FIG. 6, for example, RGB is once converted into luminance and color difference signals such as L * a * b *, the saturation is within a predetermined value, and the luminance is also less than the predetermined value. In this case, “1” is output, and “0” is output otherwise. The subsequent blank sheet detection configuration may be the same as that in the above embodiment. As a result, the streak can be determined for the color image data. The streak correction unit 502 performs color image interpolation processing. That is, interpolation is performed for each color component.

(Other embodiments)
The present invention is also 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.

  Further, although the electrophotographic apparatus has been described as an example of the above embodiment, an inkjet printer, a thermal printer, or the like may be used, and the gist of the present invention is not limited to the type of printer. Further, the toner in electrophotographic printing has been described as an example of the recording agent. However, the recording agent used for printing is not limited to the toner, and may be other recording agents such as ink. The gist of the present invention is the kind of the recording agent. It is not limited to.

Claims (19)

Determination means for determining whether or not the image data includes print information according to a plurality of determination criteria;
Obtaining means for obtaining print information included in the image data;
Adjusting means for adjusting the plurality of determination criteria according to the amount of print information acquired by the acquiring means;
Determination control means for performing determination by the determination means on image data obtained by reading a document image using a plurality of determination criteria adjusted by the adjustment means;
An image processing apparatus comprising:   The image processing according to claim 1, wherein the adjustment unit shifts the plurality of determination criteria in a direction that makes it easier to determine that the print information is not included as the print information acquired by the acquisition unit increases. apparatus. Determination means for determining whether or not the image data includes print information according to a plurality of determination criteria;
A determination by the determination unit is performed on a reference image, and a determination result corresponding to each of the plurality of determination criteria is referred to, and the print information is not included among the determination criteria determined not to include print information. An adjustment unit that identifies a first determination criterion that is a determination criterion that is difficult to determine, and adjusts a plurality of determination criteria including the first determination criterion from the plurality of determination criteria;
A determination control unit that performs determination by the determination unit on image data obtained by reading a document image using a plurality of determination criteria including the first determination criterion adjusted by the adjustment unit;
An image processing apparatus comprising:   The adjustment unit determines that all the results of determination by the determination unit according to a plurality of determination criteria with respect to the reference image do not include print information, and when the first determination criterion cannot be specified, A plurality of determination criteria are shifted in a direction that makes it difficult to determine that print information is not included, and if it is determined that all the determination results include print information and the first determination criterion cannot be specified, the plurality of determination criteria The image processing apparatus according to claim 3, wherein the determination criterion is shifted in a direction in which it is easy to determine that print information is not included and the plurality of determination criteria are selected.   The adjustment means includes a determination criterion that is most difficult to determine that the plurality of determination criteria do not include the print information, and all the results of the determination performed on the reference image according to the plurality of determination criteria are print information. If the first determination criterion cannot be specified, or if the plurality of determination criteria include a determination criterion that makes it easy to determine that print information is not included most, and the plurality of the determination criteria are not included. The determination result according to the determination criterion is a determination result that all the print information is included, and when the first determination criterion cannot be specified, it is determined that there is no room for adjustment. An image processing apparatus according to 1.   In the case where the first determination criterion is included in a result of the determination performed on the reference image according to the plurality of determination criteria, the adjustment unit further includes the plurality of first determination criteria. 6. The image processing according to claim 3, wherein the plurality of determination criteria are adjusted so as to be a determination criterion that is most difficult to determine that the print information is not included among the determination criteria. apparatus. In accordance with an input by the user, further comprising a selection means for selecting one determination level from the adjusted determination criteria,
The image processing apparatus according to claim 3, wherein the determination control unit makes a determination based on a determination criterion selected by the selection unit.   The image processing apparatus according to claim 7, wherein the selection unit further includes a unit that displays the plurality of determination criteria and a determination criterion used for determination by the determination control unit.   The image processing apparatus according to claim 3, further comprising a unit that displays a determination criterion used for determination by the determination control unit.   10. The determination unit according to claim 3, wherein the determination unit holds a plurality of determination parameters corresponding to the plurality of determination criteria, and performs determination using each of the plurality of determination parameters. Image processing apparatus. The determination means compares the variance and average of the brightness of the image data with a threshold value to determine whether the image data includes print information,
The image processing apparatus according to claim 10, wherein the determination parameter includes a threshold for the variance and the average. The determination means determines whether the image data includes print information based on the number of edges included in the image data,
The image processing apparatus according to claim 10, wherein the determination parameter includes a threshold for edge detection and a threshold for the number of edges.   The determination means compares the brightness variance and average of the image data with threshold values to determine whether the image data has print information, and based on the number of edges included in the image data The image processing apparatus according to claim 3, wherein the image data is determined to be blank when it is determined that no print information is included in the image data. Manuscript supply means;
A reading unit that reads a document conveyed by the document feeding unit at a predetermined position and generates image data for each page;
The reference image is a white surface of a conveyance roller facing a reading position by the reading unit,
The image processing apparatus according to claim 3, wherein the determination control unit determines whether or not print information is included in the image data for each page.   The image processing apparatus according to claim 14, further comprising a processing unit configured to process a page determined not to include print information from the image data for each page read by the reading unit. An image processing apparatus according to claim 14,
An image forming apparatus comprising: an output unit configured to output the image data for each page read by the reading unit except for a page determined not to include print information. A determination step in which a determination unit of the image processing apparatus determines whether or not print information is included in the image data according to a plurality of determination criteria;
An acquisition step of an image processing apparatus for acquiring print information included in the image data;
An adjustment step in which the adjustment unit of the image processing apparatus adjusts the plurality of determination criteria according to the amount of print information acquired in the acquisition step;
The determination control means of the image processing apparatus has a determination control step for performing determination by the determination step on image data obtained by reading a document image using a plurality of determination criteria adjusted by the adjustment step. An image processing method. A determination step in which a determination unit of the image processing apparatus determines whether or not print information is included in the image data according to a plurality of determination criteria;
Of the determination criteria determined by the adjustment unit of the image processing apparatus to determine the reference image as a target, refer to the determination results corresponding to each of the plurality of determination criteria, and not include print information. The first determination criterion that is the determination criterion that is most difficult to determine that the print information is not included is identified, and adjustment is performed so that a plurality of determination criteria including the first determination criterion are selected from the plurality of determination criteria. An adjustment process to
The determination control means of the image processing apparatus uses the plurality of determination criteria including the first determination criterion adjusted by the adjustment step to perform determination by the determination step for image data obtained by reading a document image. And a determination control step to be performed.   A program for causing a computer to execute the image processing method according to claim 17 or 18.

Images