IMAGE READING DEVICE, IMAGE READING SYSTEM, CONTROL DEVICE, AND DOCUMENT AREA CUTTING METHOD

Abstract

To provide a technique capable of performing a document region extraction process in which an effect of dirt is suppressed, even if a reading section has dirt including, for example, scratches and adhering substances such as dust.SOLUTION: A controller of an image reading device is configured to perform: a dirt detection process of detecting a dirt position of a reading section in a pixel column direction intersecting with a relative movement direction between a document and the reading section; and an extraction process of extracting a document region from read image data on the basis of, analysis image data to discriminate the document region or a background region. The controller searches the analysis image data for an edge candidate position where the background region changes to the document region inward from an edge portion in the pixel column direction, in the extraction process, and continues the search when the retrieved edge candidate position corresponds to the dirt position and extracts the document region from the read image data at the retrieved edge candidate position when the retrieved edge candidate position does not correspond to the dirt position.SELECTED DRAWING: Figure 8

Description

[Technical field]

[0001] The present invention relates to an image reading device for reading an original, an image reading system, a control device, and an original area extraction method. [Background technology]

[0002] Image reading devices use image sensors such as those based on the Contact Image Sensor method abbreviated as CIS and the Charge Coupled Devices method abbreviated as CCD. If there is dirt on the document reading surface of the image sensor, such as fingerprints, paper dust, pencil dust, correction fluid, glue, condensation, scratches, or other stains, an abnormality will appear in the reading result. For example, if the image sensor does not move and the document moves in the relative movement direction, the above-mentioned stains will appear in the reading result as vertical streaks along the relative movement direction. The image reading device disclosed in Patent Document 1 issues a stain alert when it detects stains on the document reading surface. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2020-17856 A Summary of the Invention [Problem to be solved by the invention]

[0004] When an image reading device performs document area cutout processing to remove background areas outside the document area from scanned image data, if the position of a stain on the document reading surface is outside the document area, the image reading device cuts out the document area including the stain position. [Means for solving the problem]

[0005] The image reading device of the present invention comprises: A reading unit that reads an original and outputs a reading result of the original; a drive unit that moves the document and the reading unit relative to each other in a relative movement direction; a storage unit for storing read image data based on the read result and analysis image data based on the read result, the analysis image data being used to determine whether the image data is a document area or a background area outside the document area; a control unit that performs a dirt detection process to detect a dirt position of the reading unit in a pixel row direction intersecting with the relative movement direction, and a cut-out process to cut out the document area from the read image data based on the analysis image data, The control unit, in the cut-out process, A search is performed for a candidate edge position where the background region changes to the document region in an inward direction from an edge portion in the pixel column direction in the analysis image data; If the edge candidate position found corresponds to the stain position, the search is continued; If the searched edge candidate position does not correspond to the stain position, the document area is cut out from the read image data at the searched edge candidate position.

[0006] Further, an image reading system of the present invention is an image reading system including the image reading device and a control device, the image reading device further includes a transmission unit configured to transmit output image data corresponding to the document region cut out from the read image data and the stain position to the control device; The control device includes: a receiving unit that receives the output image data and the stain position; and an image processing unit that reduces streaks along the relative movement direction in the output image data based on the stain position.

[0007] Furthermore, the control device of the present invention is a control device connected to an image reading device, the image reading device moves the document and a reading unit that reads the document relative to each other in a relative movement direction, and transmits output image data based on a result of reading the document and a dirt position of the reading unit in a pixel row direction that intersects with the relative movement direction to the control device; The control device includes: a receiving unit that receives the output image data and the stain position; and an image processing unit that reduces streaks along the relative movement direction in the output image data based on the stain position.

[0008] Furthermore, the document area extraction method of the present invention includes: A reading unit that reads an original and outputs a reading result of the original; a drive unit that moves the document and the reading unit relative to each other in a relative movement direction; A method for extracting an original region in an image reading device, the method comprising: storing read image data based on the read result; and analysis image data based on the read result, the analysis image data being used to determine whether the original region is an original region or a background region outside the original region, the method comprising: a stain detection step of detecting a stain position of the reading unit in a pixel row direction intersecting with the direction of the relative movement; a cutting step of cutting out the document area from the read image data based on the analysis image data, In the cutting step, A search is performed for a candidate edge position where the background region changes to the document region in an inward direction from an edge portion in the pixel column direction in the analysis image data; If the edge candidate position found corresponds to the stain position, the search is continued; If the searched edge candidate position does not correspond to the stain position, the document area is cut out from the read image data at the searched edge candidate position. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is a perspective view illustrating an example of the appearance of an image reading apparatus. [Diagram 2] FIG. 2 is a vertical cross-sectional view showing an example of a document transport path of the image reading device. [Diagram 3] FIG. 1 is a block diagram illustrating a configuration example of an image reading system. [Figure 4] FIG. 2 is a diagram illustrating an example of the configuration of a reading unit. [Diagram 5] 4A and 4B are diagrams showing examples of first read data Ri and second read data Li of a read reference surface; [Figure 6] 6 is a flowchart illustrating an example of a stain detection process. [Figure 7] FIG. 2 is a diagram illustrating a schematic example of how a document is read by a reading unit; [Figure 8] 11A and 11B are diagrams illustrating an example of a document region extraction process according to a stain position. [Figure 9] 6A and 6B are diagrams illustrating an example in which read image data and analysis image data are generated from a reading result. [Figure 10] FIG. 13 is a diagram showing an example of analysis image data. [Figure 11] 11 is a flowchart illustrating an example of a document region cutout process. [Figure 12] 4 is a diagram illustrating a schematic example of a state in which a carrier sheet holding a document is read by a reading section; FIG. [Figure 13] 10 is a flowchart illustrating an example of a threshold setting process. [Figure 14] 10A to 10C are diagrams illustrating an example of a process for reducing streaks in output image data by image processing. [Figure 15] 10 is a flowchart illustrating an example of a process for reducing streaks included in output image data. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following embodiments of the present invention will be described. Of course, the following embodiments are merely illustrative of the present invention, and not all of the features shown in the embodiments are necessarily essential to the solution of the invention.

[0011] (1) Overview of the technology included in this invention: First, the outline of the technology included in the present invention will be described with reference to the examples shown in Figures 1 to 15. The figures in this application are diagrams showing examples in a schematic manner, and the magnification in each direction shown in these figures may differ, and the figures may not be consistent with each other. Of course, each element of the present technology is not limited to the specific example indicated by the symbol. In the "outline of the technology included in the present invention," the words in parentheses mean supplementary explanations of the immediately preceding words.

[0012] [Aspect 1] As illustrated in FIGS. 1 to 3, 7, etc., an image reading device (e.g., scanner 1A) according to one aspect of the present technology includes a reading unit 20, a driving unit UD, a storage unit (e.g., RAM 38), and a control unit 40. The reading unit 20 reads an original P0 and outputs a reading result RS of the original P0. The driving unit UD moves the original P0 and the reading unit 20 relatively in a relative movement direction D1. The storage unit (38) stores read image data DA1 based on the reading result RS, and analysis image data DA3 based on the reading result RS, which is for determining whether the original area A1 is a document area A1 or a background area A2 outside the document area A1. As illustrated in FIGS. 6 and 11, the control unit 40 performs a dirt detection process for detecting a dirt position P1 of the reading unit 20 in a pixel column direction D2 intersecting with the relative movement direction D1, and a cut-out process for cutting out the original area A1 from the read image data DA1 based on the analysis image data DA3.

[0013] The above-mentioned control unit 40, in the cut-out process, A search is performed for an edge candidate position P2 where the background area A2 changes to the document area A1 from an edge portion (e.g., a left side DA1c or a right side DA1d) in the pixel column direction D2 toward the inside in the analysis image data DA3, If the edge candidate position P2 found corresponds to the stain position P1, the search is continued; If the detected edge candidate position P2 does not correspond to the stain position P1, the document area A1 is cut out from the read image data DA1 at the detected edge candidate position P2.

[0014] In the above-mentioned aspect 1, if the position of dirt such as dust or scratches on the reading unit 20 is outside the actual document portion, that position does not necessarily become the cut-out position of the document portion. If the dirt position P1 is inside the actual document portion, the actual document portion becomes the cut-out position. Therefore, the above-mentioned aspect can provide an image reading device that can perform document area cut-out processing while suppressing the influence of dirt even if dirt such as dust or scratches is present on the reading unit.

[0015] Here, the image reading device includes a scanner, a facsimile, a multifunction machine having a function of reading an original and a function of outputting image data, and the like. The relative movement between the document and the reading unit includes the document moving without the reading unit moving, the reading unit moving without the document moving, and both the document and the reading unit moving. The above remarks also apply to the following aspects.

[0016] [Aspect 2] 6 and 7, the control unit 40 may search for the edge candidate position P2 in the cut-out process from both edges (e.g., the left side DA1c and the right side DA1d) in the pixel column direction D2 in the analysis image data DA3 toward the inside. This allows the document area cut-out process to be performed while suppressing the influence of dirt on both sides in the pixel column direction D2. Therefore, this embodiment can provide an image reading device suitable for performing the document area cut-out process while suppressing the influence of dirt. Although not included in the above-mentioned second aspect, the present technology also includes a case in which the edge candidate position P2 is searched for from one edge in the pixel column direction D2 but not from the other edge.

[0017] [Aspect 3] As illustrated in FIG. 4, the reading unit 20 may have a reading reference surface 22a. As illustrated in FIG. 3, the image reading device (1A) may further include a storage unit 43 that stores the first read data Ri of the reading reference surface 22a. As illustrated in FIG. 6, the control unit 40 may obtain the second read data Li of the reading reference surface 22a after the first read data Ri is stored in the dirt detection process, and may detect the dirt position P1 based on the second read data Li and the first read data Ri. This suppresses the influence of dirt on the reading unit 20 caused by the use of the image reading device (1A). Therefore, this embodiment can provide an image reading device suitable for performing a document area cut-out process in which the influence of dirt is suppressed.

[0018] [Aspect 4] The background area A2 may be an area of ​​a color within a first set range from the color of the reading reference surface 22a. In FIG. 9, Rb≦Ra≦Rc, Gb≦Ga≦Gc, and Bb≦Ba≦Bc are shown as examples of the first set range. As shown in FIG. 9 and FIG. 11, the control unit 40 may generate the analysis image data DA3 based on the reading result RS and a first threshold value TH1 that separates the color of the document area A1 from the color of the background area A2. This embodiment can provide an image reading device that is more suitable for performing document area cutout processing while suppressing the influence of dirt.

[0019] [Aspect 5] The background area A2 may be an area of ​​a color within a second set range from the color of the portion of the carrier sheet SH1 (see FIG. 12) holding the document P0 that is superimposed on the reading reference surface 22a. FIG. 9 shows Re≦Rd≦Rf, Ge≦Gd≦Gf, and Be≦Bd≦Bf as examples of the second set range. As exemplified in FIGS. 9, 11, and 13, the control unit 40 may generate the analysis image data DA3 based on the reading result RS and a second threshold value TH2 that separates the color of the document area A1 from the color of the background area A2. This embodiment also provides an image reading device that is more suitable for performing document area cutout processing while suppressing the influence of dirt.

[0020] [Aspect 6] As illustrated in FIGS. 1 and 2, the image reading device (1A) may further include a cover (e.g., upper unit 4) that opens and closes the transport path 10 of the document P0. As illustrated in FIG. 6, the control unit 40 may perform the dirt detection process at least some of the following times: when the image reading device (1A) is turned on, when the reading of the document P0 is completed, and when the cover unit (4) is closed. This suppresses the influence of dirt detected at least some of the following times: when the image reading device (1A) is turned on, when the reading of the document P0 is completed, and when the cover unit (4) is closed. Therefore, this embodiment can provide an image reading device suitable for performing document area cut-out processing with suppressed influence of dirt.

[0021] [Aspect 7] 9, the storage unit (38) may store the analysis image data DA3 having a lower resolution than the scanned image data DA1. This embodiment can reduce the data capacity of the storage unit (38) and can speed up the analysis image data DA3 generation process and the document area cut-out process. Although not included in the seventh aspect, the case where the resolution of the analysis image data DA3 is the same as the resolution of the read image data DA1 is also included in the present technology.

[0022] [Aspect 8] As illustrated in Fig. 11, in the cut-out process, after the edge candidate position P2, which is the stain position P1, is searched for, if the edge candidate position P2 is not searched for up to the middle in the pixel column direction D2 in the analysis image data DA3, the control unit 40 may cut out the document area A1 from the read image data DA1 at the searched edge candidate position P2. In the above case, the stain position P1 overlaps with the edge of the actual document area A1 in the pixel column direction D2, so that the document area A1 is cut out from the read image data DA1 at the searched edge candidate position P2, thereby preventing the edge of the actual document area A1 from being cut out in the pixel column direction D2. Therefore, this embodiment can provide an image reading device capable of more appropriately performing the document area cut-out process.

[0023] [Aspect 9] 3, 14, and 15, an image reading system SY1 according to an embodiment of the present technology includes an image reading device (1A) according to any one of embodiments 1 to 8, and a control device 60. The image reading device (1A) further includes a transmitting unit (e.g., I / F 47) that transmits output image data DA4 corresponding to the document area A1 cut out from the read image data DA1, and the stain position P1 to the control device 60. The control device 60 includes a receiving unit (e.g., I / F 67) that receives the output image data DA4 and the stain position P1, and an image processing unit 68 that reduces a streak L1 along the relative movement direction D1 in the output image data DA4 based on the stain position P1. In this embodiment, the stain position P1 detected by the image reading device (1A) is sent to the control device 60, and the control device 60 performs processing to reduce the streak L1 along the relative movement direction D1 in the output image data DA4 based on the stain position P1. Therefore, in this embodiment, in an image reading system equipped with an image reading device, it is possible to perform image processing that suppresses the influence of stains even if there is dirt such as dust or scratches on the reading unit.

[0024] [Aspect 10] Furthermore, the above-mentioned control device 60 also has technical features. Therefore, the control device 60 according to one aspect of the present technology includes a receiving unit (67) that receives the output image data DA4 and the stain position P1, and an image processing unit 68 that reduces the streak L1 along the relative movement direction D1 in the output image data DA4 based on the stain position P1. Therefore, this aspect can provide a control device that can perform image processing that suppresses the influence of dirt even if there is dirt such as dust or scratches on the reading unit.

[0025] [Aspect 11] Furthermore, a document area cutting method relating to one embodiment of the present technology is a document area cutting method in an image reading device (1A) equipped with the reading unit 20, the drive unit UD, and the storage unit (38), and includes the following steps (A) and (B). (A) A stain detection step ST1 of detecting a stain position P1 of the reading unit 20 in a pixel column direction D2 intersecting with the relative movement direction D1. (B) A cutting step ST2 of cutting out the document area A1 from the read image data DA1 based on the analysis image data DA3. In the document region cut-out method, in the cut-out step ST2, A search is performed for an edge candidate position P2 where the background area A2 changes to the document area A1 from an edge portion (e.g., a left side DA1c or a right side DA1d) in the pixel column direction D2 toward the inside in the analysis image data DA3, If the edge candidate position P2 found corresponds to the stain position P1, the search is continued; If the detected edge candidate position P2 does not correspond to the stain position P1, the document area A1 is cut out from the read image data DA1 at the detected edge candidate position P2. Therefore, this aspect can provide a document area extraction method that can extract a document area while suppressing the influence of dirt, such as dust, or dirt such as scratches, even if the reading unit has dirt or other attachments. Incidentally, it is also possible to add the above-mentioned aspects 2 to 8 to the above-mentioned aspect 11.

[0026] Furthermore, the present technology is applicable to a multifunction device including the above-mentioned image reading device, a multifunction system including the above-mentioned image reading system, an image reading method implemented in the above-mentioned image reading device, an image generation method implemented in the above-mentioned image reading system, an image reading program that causes a computer to realize functions corresponding to the above-mentioned control unit, an image processing program that causes a computer to realize functions corresponding to the above-mentioned image processing unit, a computer-readable medium having the aforementioned program recorded thereon, and the like.

[0027] (2) Specific examples of the configuration of an image reading system including an image reading device: Fig. 1 is a schematic diagram showing the appearance of a scanner 1A as a specific example of an image reading device. Fig. 2 is a schematic vertical cross-sectional view showing an example of a document transport path 10 of the scanner 1A. Fig. 3 is a schematic diagram showing the configuration of an image reading system SY1. As shown in Fig. 3, the image reading system SY1 includes a scanner 1A and a control device 60. The scanner 1A shown in Figs. 1 and 2 is a document scanner capable of reading both sides of a document P0, which is an example of a medium.

[0028] In the XYZ coordinate system shown in Figs. 1 and 2, the X direction is the width direction of the scanner 1A and also the width direction of the document P0, the Y direction is the direction perpendicular to the X direction and also the direction of transport of the document P0, and the Z direction is the direction perpendicular to the X direction and the Y direction and also generally perpendicular to the surface of the document P0. The Y direction is also the relative movement direction D1 between the document P0 and the reading unit 20, and the X direction is also the pixel row direction D2 that intersects with the relative movement direction D1. Here, the X direction and the Y direction do not have to be perpendicular as long as they intersect with each other, the X direction and the Z direction do not have to be perpendicular as long as they intersect with each other, and the Y direction and the Z direction do not have to be perpendicular as long as they intersect with each other. In addition, the +Y direction side, which is the direction of the Y direction arrow, is the front side of the device, and the -Y direction side opposite to the +Y direction side is the rear side of the device. The -Y direction side where the document P0 is fed is sometimes called the upstream side, and the +Y direction side where the document P0 is transported is sometimes called the downstream side.

[0029] The scanner 1A shown in FIGS. 1 and 2 includes a device body 2 having therein a reading unit 20 that reads an image of a document P0. The device body 2 includes a lower unit 3 and an upper unit 4. The upper unit 4 is provided so as to be openable and closable around a pivot point on the downstream side of the document transport direction with respect to the lower unit 3. The pivot point is configured by a rotation shaft arranged along the X direction near the front edge of the scanner 1A. A user can open the closed upper unit 4 by rotating it toward the front side of the device, and by exposing the transport path 10 of the document P0, it is possible to easily handle jams of the document P0, etc. A user can close the open upper unit 4 by rotating it to match the lower unit 3, and the transport path 10 can be closed. The upper unit 4 is an example of a cover part that opens and closes the transport path 10.

[0030] An original placement section 11 having a placement surface 11a on which the original P0 to be fed is placed is provided on the rear side of the device body 2. The original placement section 11 is provided so as to be detachable from the device body 2. The original placement section 11 is provided with a pair of left and right edge guides 12, 12 that guide the side edges in the X direction, which is the width direction intersecting with the Y direction, which is the feeding direction of the original P0. The original placement section 11 also includes a first paper support 8 and a second paper support 9. The first paper support 8 and the second paper support 9 can be stored inside the original placement section 11 and can be pulled out from the original placement section 11 as shown in FIG. 1, and the length of the placement surface 11a can be adjusted.

[0031] The device main body 2 is provided with an operation panel 7 on the front side of the upper unit 4, which is a UI for performing various reading settings and reading execution operations and displaying the reading setting contents, etc., that is, which realizes a user interface. In this embodiment, the operation panel 7 is a so-called touch panel that can perform both display and input, and serves both as an operation section for performing various operations and a display section for displaying various information. An inlet 6 leading to the inside of the device body 2 is provided at the top of the upper unit 4. The document P0 placed on the document placement section 11 is sent from the inlet 6 towards the reading section 20 inside the device body 2. An output tray 5 for receiving the document P0 to be discharged is provided at the front side of the lower unit 3.

[0032] 2 is a space sandwiched between the lower unit 3 and the upper unit 4, and shows the path along which the original P0 passes. On the transport path 10 shown in Fig. 2, a document placement unit 11, rollers 14 and 15, a first document detection unit 31, a double feed detection unit 30, a transport roller pair 16, a second document detection unit 32, a reading unit 20, and a discharge roller pair 17 are arranged in this order in the transport direction D1. The document placement unit 11 located at the most upstream side of the transport path 10 is provided with a sensor (not shown) for detecting whether or not the document P0 is present on the document placement unit 11.

[0033] In the transport path 10, a feed roller 14 that sends the document P0 placed on the placement surface 11a of the document placement unit 11 toward the reading unit 20, and a separation roller 15 that nips and separates the document P0 between the feed roller 14 and the document placement unit 11 are provided downstream of the document placement unit 11. The feed roller 14 is driven to rotate by a motor 45 for the feed roller shown in FIG. 3. The feed roller 14 obtains a rotation torque from the motor 45 for the feed roller and rotates in a counterclockwise direction in FIG. 2. The feed roller 14 contacts the lowest document P0 placed on the placement surface 11a of the document placement unit 11. Therefore, when a plurality of documents P0 are set on the document placement unit 11 in the scanner 1A, the documents P0 are fed in order from the document placement surface 11a side toward the downstream side. A rotation torque is transmitted to the separation roller 15 from a motor 46 for the transport roller shown in FIG. 3 via a torque limiter (not shown).

[0034] When no document P0 is interposed between the feed roller 14 and the separation roller 15, or when only one document P0 is interposed, the separation roller 15 rotates due to slippage occurring at a torque limiter (not shown) regardless of the rotational torque received from the transport roller motor 46. In this case, the rotational direction of the separation roller 15 is clockwise in FIG. 2. When a second or subsequent document P0, in addition to the document P0 to be fed, enters between the feed roller 14 and the separation roller 15, slippage occurs between the documents. As a result, the separation roller 15 rotates counterclockwise in FIG. 2 due to the rotational torque received from the transport roller motor 46, preventing double feeding of documents P0.

[0035] In the transport path 10, a transport roller pair 16, a reading unit 20 that reads an image of the document P0, and a discharge roller pair 17 are provided downstream of the feed roller 14. The transport roller pair 16 includes a transport drive roller 16a that is driven to rotate by a transport roller motor 46 shown in FIG. 3, and a transport driven roller 16b that rotates in response to the drive roller 16a. The document P0 nipped between the feed roller 14 and the separation roller 15 and fed downstream is nipped by the transport roller pair 16 and transported to the reading unit 20 located downstream of the transport roller pair 16.

[0036] In the transport path 10, a first original detection unit 31 is provided between the feed roller 14 and the transport roller pair 16. An optical sensor or the like can be used for the first original detection unit 31. The first original detection unit 31 shown in FIG. 2 is an optical sensor including a light emitting unit 31a and a light receiving unit 31b arranged opposite to each other across the transport path 10. The light receiving unit 31b detects light emitted from the light emitting unit 31a and travels over the transport path 10, and transmits an electrical signal indicating the intensity of the detected light to the control unit 40 shown in FIG. 3. When the transported original P0 blocks the light emitted from the light emitting unit 31a, the electrical signal indicating the intensity of the detected light from the light receiving unit 31b changes. This allows the control unit 40 to detect the passage of the leading and trailing ends of the original P0.

[0037] In the transport path 10, a multi-feed detection unit 30 that detects multi-feeding of the document P0 is disposed between the first document detection unit 31 and the pair of transport rollers 16. The multi-feed detection unit 30 includes an ultrasonic transmitter 30a and an ultrasonic receiver 30b that are disposed opposite each other across the transport path 10. The ultrasonic receiver 30b detects ultrasonic waves that have passed over the transport path 10 from the ultrasonic transmitter 30a, and transmits an electrical signal indicating the intensity of the detected ultrasonic waves to the control unit 40. When a multi-feed of the document P0 occurs, the electrical signal indicating the intensity of the ultrasonic waves from the ultrasonic receiver 30b changes. This allows the control unit 40 to detect the multi-feeding of the document P0.

[0038] In the transport path 10, a second original detection unit 32 is provided downstream of the pair of transport rollers 16. A contact sensor or the like can be used for the second original detection unit 32. The second original detection unit 32 shown in Fig. 2 is a contact sensor having a lever, and when the lever rotates as the leading edge or trailing edge of the original P0 passes through, the electrical signal sent to the control unit 40 changes. The change in the electrical signal sent from the second original detection unit 32 to the control unit 40 enables the control unit 40 to detect the passage of the leading edge or trailing edge of the original P0. The control unit 40 can grasp the position of the original P0 on the transport path 10 by the first original detection unit 31 and the second original detection unit 32 described above.

[0039] The reading section 20, which is provided downstream of the second document detection section 32 on the transport path 10, includes an upper reading sensor 20A provided on the upper unit 4 side and a lower reading sensor 20B provided on the lower unit 3 side. The upper reading sensor 20A reads the upper surface of the document P0 and outputs a reading result RS of the upper surface. The lower reading sensor 20B reads the lower surface of the document P0 and outputs a reading result RS of the lower surface. The upper reading sensor 20A and the lower reading sensor 20B can be a CISM, i.e., a contact type image sensor module. The image of at least one of the upper and lower surfaces of the document P0 is read by the reading section 20. Details of the reading section 20 will be described later.

[0040] The document P0 that has passed through the reading unit 20 on the transport path 10 is nipped by a discharge roller pair 17 located downstream of the reading unit 20, and is discharged from a discharge outlet 18 provided on the front side of the lower unit 3. The discharge roller pair 17 includes a discharge drive roller 17a that is driven to rotate by the transport roller motor 46 shown in FIG. 3, and a discharge driven roller 17b that is driven to rotate. In this specific example, the rollers 14 and 15, the roller pairs 16 and 17, and the motors 45 and 46 are an example of a drive unit UD that moves the document P0 and the reading unit 20 relatively in the relative movement direction D1.

[0041] As shown in FIG. 3, the scanner 1A includes a control unit 40 and a RAM 38, which is a semiconductor memory, as a control system. Here, RAM is an abbreviation for Random Access Memory. The RAM 38 is an example of a storage unit. The control unit 40 performs various controls of the scanner 1A, including feed control and reading control of the document P0. A signal from the operation panel 7 is input to the control unit 40, and a signal for realizing the display of the operation panel 7, particularly the UI, is sent from the control unit 40 to the operation panel 7. The control unit 40 controls a feed roller motor 45, which is a drive source for the feed roller 14, and a transport roller motor 46, which is a drive source for the separation roller 15 and the roller pairs 16 and 17. Detection signals from the detection units 30, 31, and 32 are input to the control unit 40, and the control unit 40 performs necessary control based on these detection signals. The control unit 40 transmits a signal for controlling the reading unit 20 to the reading unit 20, and controls the reading image data DA1 based on the reading result RS from the reading unit 20 to be stored in the RAM 38, which serves as a buffer.

[0042] The control unit 40 includes a CPU 41 which is a processor, a ROM 42 which is a semiconductor memory, a storage unit 43, an I / F 47 which is an example of a transmission unit, and the like. Here, CPU is an abbreviation for Central Processing Unit, ROM is an abbreviation for Read Only Memory, and I / F is an abbreviation for interface. The ROM 42 stores a program 44 which realizes various functions of the scanner 1A. The ROM 42 may be a semiconductor memory in which data including the program 44 can be rewritten. The program 44 may be stored in the storage unit 43, or may be read from the storage unit 43 to a RAM, for example, the RAM 38, and executed. The program 44 includes a UI program which controls a UI to be displayed on the operation panel 7, various control programs required for transporting and reading a document, a dirt detection program for executing a reading unit dirt detection process illustrated in FIG. 6, a cut-out program for executing a document area cut-out process illustrated in FIG. 11, a threshold setting program for executing a threshold setting process illustrated in FIG. 13, an image transmission program for executing an image transmission process illustrated in FIG. 15, and the like. The storage unit 43 is a non-volatile memory capable of reading and writing data, and stores the first read data Ri of the reading reference surface 22a shown in Fig. 4. The storage unit 43 may be a non-volatile semiconductor memory such as a flash memory, a magnetic storage device such as a hard disk, or the like. The CPU 41 performs various arithmetic processing according to a program 44, and controls the operation of the entire scanner 1A. The I / F 47 is connected to the I / F 67 of the control device 60, and transmits various information to the control device 60 and receives various information from the control device 60 by communicating with the I / F 67 according to a predetermined communication protocol.

[0043] The control unit 40 in this specific example controls the storage of thinned image data DA2 based on the reading result RS from the reading unit 20 in RAM 38, the generation of analysis image data DA3 from the thinned image data DA2 in RAM 38, the detection of a dirt position P1 on the reading unit 20, and the generation of output image data DA4 from the read image data DA1 based on the analysis image data DA3 and the dirt position P1.

[0044] A control device 60 connected to the scanner 1A includes a CPU 61 which is a processor, a ROM 62 which is a semiconductor memory, a RAM 63 which is a semiconductor memory, a storage device 64, an input device 65, a display device 66, an I / F 67 which is an example of a receiving unit, etc. These elements 61 to 67, etc. are electrically connected to each other, so that they can input and output information to each other.

[0045] The storage device 64 stores an operating system (not shown), a read control program PR1 for causing the control device 60 to realize the image processing function FU1, and the like. The information stored in the storage device 64 is read out to the RAM 63 as appropriate and used for image processing. The storage device 64 may be a magnetic storage device such as a hard disk, a non-volatile semiconductor memory such as a flash memory, or the like. The input device 65 may be a pointing device, a hard key including a keyboard, a touch panel attached to the surface of a display panel, or the like. The display device 66 may be a liquid crystal display panel or the like. The input device 65 and the display device 66 constitute a UI in the control device 60. The I / F 67 is connected to the I / F 47 of the scanner 1A, and communicates with the I / F 47 according to a predetermined communication protocol to transmit various information to the scanner 1A and receive various information from the scanner 1A.

[0046] The CPU 61 performs image processing corresponding to the image processing function FU1 by executing the reading control program PR1 read from the storage device 64 to the RAM 63. The reading control program PR1 causes the control device 60, which is a computer, to function as an image processing unit 68 corresponding to the image processing function FU1. The computer-readable medium storing the reading control program PR1 is not limited to the storage device 64, and may be a recording medium external to the control device 60.

[0047] Fig. 4 is a schematic diagram illustrating the configuration of the reading unit 20. The upper reading sensor 20A and the lower reading sensor 20B have point-symmetrical structures in Fig. 4 and have the same basic configuration. Therefore, the following description will focus on the configuration of the lower reading sensor 20B. The lower reading sensor 20B includes a transmissive plate 21, a reference plate 22, a light guide 23, a lens 24, and a line sensor 25. The lower reading sensor 20B is elongated in the pixel column direction D2, and the transmissive plate 21, the reference plate 22, the light guide 23, the lens 24, and the line sensor 25 are also elongated in the pixel column direction D2.

[0048] The transparent plate 21 can be made of colorless and transparent glass, a resin plate such as a colorless and transparent acrylic plate, etc. The outer surface of the transparent plate 21, i.e., the surface facing the upper reading sensor 20A, is a reading surface that reads the original P0 and forms the transport path 10 for the original P0. For shading correction, the reference plate 22 is read by the line sensor 25 of the opposing upper reading sensor 20A. The surface of the reference plate 22 facing the upper reading sensor 20A is a reading reference surface 22a. For example, the reference plate 22 may be a resin plate in a color such as white, gray, or black, or a metal plate painted in a color such as white, gray, or black.

[0049] The light guide 23 is a light guide member for irradiating light toward the document P0 or the reference plate 22, and guides light emitted from a light source (not shown) in the pixel row direction D2 toward the opposing upper reading sensor 20A. In this case, for example, an LED, i.e., a light emitting diode, can be used as the light source. Also, a fluorescent lamp or the like can be used as the light source. The light irradiated from the light guide 23 to the upper reading sensor 20A is reflected by the document P0 or the reading reference surface 22a of the upper reading sensor 20A, reaches the line sensor 25 through the lens 24, and is photometered by the line sensor 25. The line sensor 25 includes a large number of photoelectric conversion elements arranged along the pixel row direction D2. The arrangement density of the photoelectric conversion elements is not particularly limited, but can be, for example, equivalent to 300 to 600 dpi. For example, a photodiode or the like can be used as the photoelectric conversion element. One photoelectric conversion element corresponds to one pixel, and the line sensor 25 outputs the reading result RS obtained by measuring the light of each photoelectric conversion element to the RAM 38. The line sensor 25 in this specific example is a CMOS image sensor, but a solid-state imaging element such as a line sensor or area sensor made up of a CCD may also be used as the light measuring sensor. Here, CMOS is an abbreviation for Complementary Metal-Oxide Semiconductor.

[0050] FIG. 5 shows a schematic example of the first read data Ri and the second read data Li of the read reference surface 22a. The first read data Ri is reference data obtained by reading the reference plate 22 in a manufacturing process of the scanner 1A, i.e., before the product is shipped, when the reference plate 22 is not dirty. When the control unit 40 receives an instruction to generate the first read data Ri from the operation panel 7, the control unit 40 acquires the first read data Ri from the reading unit 20 by having the reading unit 20 read the reference plate 22, and stores the first read data Ri in the storage unit 43. The second read data Li is data obtained by reading the reference plate 22 in a user's usage environment, i.e., after the product is shipped, when there is a possibility that the transparent plate 21 is dirty, and is so-called data for shading correction. In FIG. 5, the horizontal axis is pixels, with 1 pixel at the left end and n pixels at the right end. For example, when the short side of an A4 size document is read at 600 dpi, n=5100. It should be noted that the first pixel and the nth pixel are not necessarily pixels at either end of the line sensor 25. It is preferable that the number of pixels n is a value that takes into consideration some margin compared to the document size. In Fig. 5, the vertical axis indicates the pixel read value, which represents the received light intensity.

[0051] When the scanner 1A is used, foreign matter such as fingerprints, paper dust, pencil dust, correction fluid, glue, condensation, etc. may adhere to the reading reference surface 22a of the transparent plate 21. If there is any adhesion on the reading reference surface 22a, an abnormality will appear in the reading result. Also, if the reading reference surface 22a is scratched, an abnormality will appear in the reading result. When dirt such as dust or scratches occurs on the transparent plate 21, the second read data Li will have, for example, singular points Y1, Y2, Y3, and Y4. The first read data Ri, which is the reference data, does not have any singular points. Therefore, as a basic concept, for a pixel Xi (i=1 to n) shown in FIG. 6, if the difference between the read data Ri and Li exceeds a set threshold Shi (i=1 to n), the pixel Xi can be treated as a dirt position P1. In this specific example, the difference is an absolute value and is a value equal to or greater than 0, and the threshold Shi is a positive value. Of course, the difference and threshold Shi may be positive or negative.

[0052] However, a difference in level may occur in the light irradiated from the light guide 23 when the first read data Ri is acquired and when the second read data Li is acquired. In the example shown in FIG. 5, it is shown that the level of the second read data Li is increased from that of the first read data Ri. In this case, in order to compare the read data Ri and Li, for example, it is necessary to increase the level of the first read data Ri to the level of the second read data Li. In this specific example, the arithmetic mean value of the first read data Ri for each pixel Xi is set as AR, the arithmetic mean value of the second read data Li for each pixel Xi is set as AL, and a threshold value Shi is applied to the difference between the adjusted data RRi obtained by multiplying the first read data Ri by the ratio AL / AR for each pixel Xi and the second read data Li.

[0053] FIG. 6 shows a schematic diagram of a dirt detection process for detecting a dirt position P1 of the reading unit 20 in the pixel column direction D2. The control unit 40 that performs the dirt detection process implements a dirt detection step ST1 that detects a dirt position P1 of the reading unit 20. The control unit 40 starts the dirt detection process when (a) the power of the scanner 1A is turned on, (b) the reading of the document P0 is completed, and (c) the upper unit 4 is closed. The dirt detection process may be performed only in some of the above-mentioned cases (a), (b), and (c). The dirt detection process includes the processes of steps S102 to S120. Hereinafter, the description of "step" will be omitted, and the reference numerals of each step will be shown in parentheses.

[0054] When the stain detection process starts, the control unit 40 reads out the first read data Ri from the storage unit 43 (S102). The control unit 40 also acquires the second read data Li from the reading unit 20 by having the reading unit 20 read the reference plate 22 (S104). After acquiring the second read data Li, the control unit 40 generates adjusted data RRi by adjusting the level of the first read data Ri (S106). The adjusted data RRi can be generated by multiplying the first read data Ri for each pixel Xi by the ratio AL / AR of the average value AR of the first read data Ri for each pixel Xi to the average value AR of the second read data Li for each pixel Xi.

[0055] After generating the adjustment data RRi, the control unit 40 substitutes 1 for the variable i that identifies the pixel Xi (S108), and calculates the difference Di between the adjustment data RRi and the second read data Li (S110). Di = |RRi - Li| After calculating the difference Di, the control unit 40 branches the process according to whether the difference Di is greater than or equal to the threshold value Shi (S112). If Di ≥ Shi, the control unit 40 sets the pixel Xi to the dirt position P1 (S114). The setting process of S114 can be, for example, a process of storing the value of the variable i in the dirt position data table prepared in the RAM 38. If Di < Shi, the control unit 40 determines that there is no abnormality in the pixel Xi (S116) and does not set the pixel Xi to the dirt position P1. After the process of S114 or S116, the control unit 40 determines whether the variable i has reached the number of pixels n (S118). If the variable i is less than the number of pixels n, the control unit 40 increments the value of the variable i by 1 (S120) and returns the process to S110. When the variable i reaches the number of pixels n, the control unit 40 ends the dirt detection process. Therefore, the process of detecting the dirt position P1 based on the second read data Li and the first read data Ri is performed from the first pixel to the nth pixel.

[0056] FIG. 7 schematically illustrates a state in which the document P0 is read by the reading unit 20. The document P0 conveyed in the relative movement direction D1 along the conveyance path 10 is read by the reading unit 20. The reading unit 20 sequentially outputs the reading result RS of the document P0 to the RAM 38. Each reading result RS is reading data having reading values of R, G, and B for a plurality of pixels PX1 included in the pixel column PXL along the pixel column direction D2. Here, R means red, G means green, and B means blue. The resolution of the reading result RS is not particularly limited, but can be, for example, 300 to 600 dpi. As a result of the reading results RS being sequentially accumulated in the RAM 38, reading image data DA1 is generated in the RAM 38. The resolution of the reading image data DA1 in the relative movement direction D1 is not particularly limited, but can be, for example, 300 to 600 dpi.

[0057] Although details will be described later, the control unit 40 performs document area cut-out processing to cut out document area A1 from the scanned image data DA1 at edge areas where the background portion changes to the document portion from the four sides DA1a, DA1b, DA1c, and DA1d toward the center of the scanned image. Here, the edge area is searched downward from the upper side DA1a, the edge area is searched upward from the lower side DA1b, the edge area is searched toward the right from the left side DA1c, and the edge area is searched toward the left from the right side DA1d.

[0058] If a part of the reading unit 20 is dirty, the color of the document P0 is not read correctly at the dirty part, and a streak L1 occurs along the relative movement direction D1. Hereinafter, the streak along the relative movement direction D1 is also called a vertical streak. FIG. 7 shows the read image data DA1 with a vertical streak L1 at the dirty position P1 of the reading unit 20. The vertical streak may occur outside the actual document part in the read image as shown in FIG. 7, or may occur inside the actual document part in the read image. When the vertical streak L1 exists outside the actual document part, if the document part is cut out from the read image at the boundary B9 where the color of the background changes from the four sides DA1a, DA1b, DA1c, and DA1d of the read image toward the center, a wider area than the actual document part is cut out. In this way, when the reading unit 20 is dirty, the cut-out result changes compared to when the reading unit 20 is not dirty, and correct output image data may not be obtained.

[0059] This specific example has a feature that when a vertical line L1 is present in the scanned image outside the document portion, the position of the vertical line L1 is not used as the cut-out position, and the document area A1 is cut out with the boundary B1 between the document portion and the background portion. In the document area cut-out process, when the edge candidate position P2 coincides with the stain position P1 in the pixel column direction D2, the edge candidate position P2 is not used as the cut-out position, and when the edge candidate position P2 does not coincide with the stain position P1 in the pixel column direction D2, the edge candidate position P2 is used as the cut-out position. This allows the background area A2 outside the document area A1 to be appropriately deleted. Hereinafter, various cases C1 to C7 in which the document area A1 is cut out from the read image data DA1 and the output image data DA4 is generated will be described with reference to FIG.

[0060] 8 is a schematic diagram illustrating the document region extraction process according to the stain position P1. For ease of understanding, the vertical stripe L1 is exaggerated. The vertical stripe L1 in case C1 is present outside the document portion at a distance. When the edge candidate position P2 is searched for leftward from the right side DA1d shown in Fig. 7, the first edge candidate position P2 coincides with the stain position P1 and is not an edge position. As a result, a rectangular document area A1 not including the vertical stripe L1 is cut out, and output image data DA4 representing an image of the document area A1 is generated. In case C2, the vertical stripe L1 is inside the document portion. When the edge candidate position P2 is searched for leftward from the right side DA1d, the first edge candidate position P2 does not match the stain position P1, so it becomes the edge position. As a result, the document area A1 is cut out with the boundary B1 between the document portion and the background portion.

[0061] In case C3, the right edge of the vertical stripe L1 coincides with the right edge of the document portion. When edge candidate position P2 is searched for leftward from the right side DA1d, the first edge candidate position P2 coincides with the stain position P1, but the next edge candidate position is not until the middle of the scanned image, so it becomes the edge position. As a result, document area A1 is cut out with boundary B1 between the document portion and the background portion. The left edge of the vertical line L1 in case C4 coincides with the right edge of the document portion. When the edge candidate position P2 is searched for leftward from the right side DA1d, the first edge candidate position P2 coincides with the stain position P1, but the next edge candidate position is the edge position because it is not until the middle of the scanned image. As a result, the document area A1 is cut out with the right edge of the vertical line L1 connected to the document portion as the boundary B1. In case C4, the document area A1 is cut out with the vertical line L1 added to the document portion, but since the vertical line L1 is narrow, this does not pose a problem.

[0062] The ideas of cases C1 to C4 are applicable even if the original P0 is skewed, that is, oblique. In case C5, the document portion is slanted and has a vertical stripe L1 spaced from the outside of the document portion. When edge candidate position P2 is searched for leftward from the right side DA1d, the first edge candidate position P2 coincides with the stain position P1 and is therefore not an edge position. As a result, a document area A1 that does not include the vertical stripe L1 is cut out. Note that from a scanned image having a slanted document portion, a rectangular document area A1 circumscribing the slanted document portion is cut out. In case C6, the document portion is slanted and has a vertical stripe L1 on the inside of the document portion. When the edge candidate position P2 is searched for leftward from the right side DA1d, the first edge candidate position P2 does not match the stain position P1 and is therefore the edge position. As a result, a rectangular document area A1 circumscribing the slanted document portion is cut out. In case C7, the document portion is slanted, and a vertical line L1 is connected to the rightmost corner of the document portion. When edge candidate position P2 is searched leftward from the right side DA1d, the first edge candidate position P2 coincides with the stain position P1, but the next edge candidate position is the edge position because it is not until the middle position of the scanned image. As a result, a rectangular document area A1 is cut out with the right edge of the vertical line L1 connected to the document portion as the boundary B1. In case C7, the document area A1 with the vertical line L1 added to the document portion is cut out, but since the vertical line L1 is narrow, this does not pose a problem.

[0063] The control unit 40 of this specific example searches for edge candidate positions P2 based on analysis image data DA3 generated from the reading result RS, as illustrated in Figs. 9 and 10. Fig. 9 shows an example of generating read image data DA1 and analysis image data DA3 from the reading result RS. Fig. 10 shows an example of analysis image data DA3. The analysis image data DA3 shown in Figs. 9 and 10 is binary data for determining whether the area is the document area A1 or the background area A2.

[0064] As described above, the reading unit 20 sequentially outputs the reading results RS having R, G, and B read values ​​for a plurality of pixels PX1 included in the pixel column PXL along the pixel column direction D2 to the RAM 38. The control unit 40 controls the storage of the read image data DA1 based on the reading results RS from the reading unit 20 in the RAM 38. The read image data DA1 is image data having R, G, and B read values ​​for a plurality of pixels PX1 arranged in the relative movement direction D1 and pixel column direction D2. The resolution of the read image data DA1 shown in FIG. 9 in the relative movement direction D1 and pixel column direction D2 is 300 dpi or 600 dpi.

[0065] The control unit 40 also controls storing thinned image data DA2 based on the read result RS from the reading unit 20 in the RAM 38. The thinned image data DA2 is also image data having R, G, and B read values ​​for a plurality of pixels PX1 arranged in the relative movement direction D1 and pixel column direction D2. The resolution of the thinned image data DA2 shown in FIG. 9 in the relative movement direction D1 and pixel column direction D2 is 75 dpi. The control unit 40 generates the thinned image data DA2 by sampling 1 pixel out of 4 pixels in the relative movement direction D1 and pixel column direction D2 when the read result RS is 300 dpi, and sampling 1 pixel out of 8 pixels in the relative movement direction D1 and pixel column direction D2 when the read result RS is 600 dpi.

[0066] The control unit 40 performs control to store binary analysis image data DA3 in the RAM 38 by applying a first threshold value TH1 that separates the color of the document area A1 from the color of the background area A2 to each pixel PX1 of the thinned image data DA2. Here, the R, G, and B colors of the reading reference surface 22a shown in FIG. 4 are (Ra, Ga, Ba), and the first setting range that is a setting range from the color (Ra, Ga, Ba) is Rb≦Ra≦Rc, Gb≦Ga≦Gc, and Bb≦Ba≦Bc. The first threshold value TH1 collectively refers to the boundary values ​​Rb, Rc, Gb, Gc, Bb, and Bc. The control unit 40 assigns black to pixels whose R value is Rb or more and Rc or less, whose G value is Gb or more and Gc or less, and whose B value is Bb or more and Bc or less, among the multiple pixels PX1 included in the thinned image data DA2, and assigns white to the remaining pixels, thereby generating analysis image data DA3. Therefore, analysis image data DA3 is generated based on the reading result RS with the background region A2 as a region of color within a first set range from the color of the reading reference surface 22a. The control unit 40 generates analysis image data DA3 based on the first threshold value TH1 and the reading result RS. In the analysis image data DA3, the value indicating black may be 1 and the value indicating white may be 0, or the value indicating black may be 0 and the value indicating white may be 1. In the analysis image data DA3 shown in FIG. 10, the document portion and the vertical streaks corresponding to the three stain positions P1 are displayed in white, and the remaining portion is displayed in black.

[0067] 11 is a schematic diagram illustrating an example of a document region cut-out process for cutting out a document region A1 from the read image data DA1 based on the analysis image data DA3. The control unit 40 performs the cut-out process by carrying out a cut-out step ST2 for cutting out the document region A1 from the read image data DA1 based on the analysis image data DA3. The control unit 40 starts the cut-out process when the reading unit 20 starts reading the document P0. When the cut-out process starts, the control unit 40 generates the thinned image data DA2 based on the read result RS from the reading unit 20 (S202). After generating the thinned image data DA2, the control unit 40 generates analysis image data DA3 by applying a first threshold value TH1 to the thinned image data DA2 (S204). Here, the control unit 40 branches the process depending on whether or not a setting for automatically cutting out the document area A1 has been accepted from the operation panel 7 or the control device 60 shown in FIG. 3 (S206). When the control unit 40 accepts a setting of a standard size such as A4 size or B5 size from the operation panel 7 or the control device 60, the control unit 40 cuts out the read image data DA1 at the set standard size (S224) and ends the cut-out process.

[0068] When the control unit 40 receives a setting to automatically cut out the document area A1 from the operation panel 7 or the control device 60, it first sets the direction in which to search for edge candidates from the analysis image data DA3 (S208). Explaining this with reference to Figs. 7 and 10, there are four directions for searching for edge candidates: downward from the upper side DA1a, upward from the lower side DA1b, rightward from the left side DA1c, and leftward from the right side DA1d. The control unit 40 sets one of these four directions.

[0069] After setting the search direction, the control unit 40 searches for the edge candidate position P2 where the background area A2 changes to the document area A1 from the edge (DA1a, DA1b, DA1c, or DA1d) of the analysis image data DA3 toward the inside (S210). The control unit 40 searches for the edge candidate position P2 in the search direction for all the pixel arrays along the search direction. For example, when the search direction is leftward along the pixel column direction D2, the edge candidate position P2 is searched for in all the pixel columns PXL along the pixel column direction D2, from the pixel on the right side DA1d toward the left in sequence. In the analysis image data DA3, the first location where the pixel PX1 changes from black to white in any of the pixel columns PXL is the edge candidate position P2 searched for in S210. Focusing on the pixel column direction D2, the control unit 40 searches for the edge candidate position P2 in the analysis image data DA3 from both edges toward the inside in the pixel column direction D2.

[0070] When the search direction is upward or downward along the relative movement direction D1, the vertical streak L1 does not become the edge candidate position P2, so theoretically the upper and lower ends of the document portion become the edge candidate position P2. When the search direction is right or left along the pixel column direction D2, the location where the vertical streak L1 appears becomes the edge candidate position P2 in cases C1, C3, C4, C5, and C7 shown in Fig. 8, and the left or right end of the document portion becomes the edge candidate position P2 in cases C2 and C6 shown in Fig. 8.

[0071] When the search direction is the right or left along the pixel row direction D2, the control unit 40 branches the process depending on whether the searched edge candidate position P2 corresponds to the dirt position P1 of the reading unit 20 (S212). When the dirt position P1 has a higher resolution than the analysis image data DA3, the control unit 40 may, for example, convert the dirt position P1 to the resolution of the analysis image data DA3 and determine whether the edge candidate position P2 matches the dirt position P1 after conversion. In the example shown in FIG. 8, the dirt position P1 is 300 dpi or 600 dpi, and the analysis image data DA3 is 75 dpi. In this case, the control unit 40 converts the dirt position P1 to a position of 75 dpi and determines whether the edge candidate position P2 matches the dirt position P1 equivalent to 75 dpi. If the search direction is upward or downward along the relative movement direction D1, the vertical stripe L1 is not a candidate edge position P2. In this case, the control unit 40 may always advance the process from S212 to S214.

[0072] When the edge candidate position P2 does not correspond to the stain position P1 as in cases C2 and C6 shown in FIG. 8, the control unit 40 cuts out the document area A1 from the read image data DA1 at the edge candidate position P2 in the search direction (S214), and proceeds to the process of S222. When the edge candidate position P2 corresponds to the stain position P1 as in cases C1, C3, C4, C5, and C7 shown in FIG. 8, the control unit 40 continues searching for the edge candidate position P2 up to the middle in the search direction in the analysis image data DA3, and searches for the next edge candidate position P2 (S216). After the process of S216, the control unit 40 branches the process depending on whether the next edge candidate position P2 is found in S216 (S218). When the next edge candidate position P2 is found as in cases C1 and C5 shown in FIG. 8, the control unit 40 returns the process to S212. When the next edge candidate position P2 also corresponds to the stain position P1, the next edge candidate position P2 is further searched for (S216), and the determination process of S218 is performed. As a result, even if multiple vertical lines L1 are present outside the document portion in the read image at a distance, these vertical lines L1 are not determined as edge positions. Furthermore, if there is a next edge candidate position P2, but this edge candidate position P2 does not correspond to the stain position P1, in S214, the document area A1 is cut out from the read image data DA1 at this edge candidate position P2.

[0073] 8, if the next edge candidate position P2 is not found, the control unit 40 cuts out the document area A1 from the read image data DA1 at the first edge candidate position P2 found in the search direction in S210 (S220), and proceeds to the process at S222. Therefore, if the edge candidate position P2, which is the stain position P1, is found, and then the edge candidate position P2 is not found until the middle of the analysis image data DA3 in the pixel column direction D2, the control unit 40 cuts out the document area A1 from the read image data DA1 at the found edge candidate position P2.

[0074] After the process of S214 or S220, the control unit 40 judges whether or not all search directions have been set (S222). If there are unset search directions, the control unit 40 returns the process to S208. If all search directions have been set, the control unit 40 ends the document area cut-out process. By using the above-mentioned document area extraction process, it is possible to determine whether each part of the scanned image is a document area A1 or a background area A2 based on the analysis image data DA3, and output image data DA4 can be generated by extracting the document area A1 from the scanned image data DA1.

[0075] As described above, even if the vertical line L1 is located outside the document portion in the read image, the position of the vertical line L1 is not the cut-out position, and the document area A1 is cut out with the boundary B1 between the document portion and the background portion. If the document portion and the vertical line L1 are connected in the read image, the document area A1 is cut out including the vertical line L1. Even if the vertical line L1 is located at the edge of the document area A1 that is cut out because the vertical line L1 is located at the edge of the document portion, it is not a problem because the vertical line L1 is narrow. Therefore, in this specific example, even if the reading unit 20 has dirt such as dust or scratches, it is possible to perform a document area cut-out process that suppresses the influence of the dirt.

[0076] (3) Application examples: As shown in FIG. 12, the scanner 1A may read the document P0 held on the carrier sheet SH1. In this case, the background of the document P0 placed on the carrier sheet SH1 will be the color of the carrier sheet SH1 at the portion overlapping the reading reference surface 22a shown in FIG. 4. Even if the carrier sheet SH1 is colorless and transparent, the color of the portion of the carrier sheet SH1 overlapping the reading reference surface 22a will actually be different from the color of the reading reference surface 22a itself. Therefore, when the scanner 1A reads the document P0 held on the carrier sheet SH1, it is only necessary to change the threshold value applied to the thinned image data DA2 in the document area cut-out process shown in FIG. 11.

[0077] Fig. 12 is a schematic diagram illustrating how the carrier sheet SH1 holding the document P0 is read by the reading unit 20. The carrier sheet SH1 shown in Fig. 12 holds two documents P0. The carrier sheet SH1 on which the document P0 is placed is transported in the relative movement direction D1 along the transport path 10 and is read by the reading unit 20. The reading unit 20 sequentially outputs the reading results RS of the document P0 and the carrier sheet SH1 to the RAM 38. FIG. 12 shows the read image data DA1 having a vertical streak L1 at the stain position P1 of the reading unit 20. The control unit 40 performs a document area cut-out process to cut out a document area A1 from the read image data DA1 at an edge area where the background part changes to the document part from the four sides DA1a, DA1b, DA1c, and DA1d toward the center of the read image. When multiple documents P0 are held on the carrier sheet SH1, a rectangular document area A1 circumscribing the multiple documents P0 is cut out so as to include all the documents P0.

[0078] Explaining with reference to FIG. 9, the control unit 40 generates analysis image data DA3 by applying a second threshold value TH2, which separates the color of the document area A1 from the color of the background area A2, to the thinned image data DA2 when the carrier sheet SH1 is used. Here, the R, G, and B colors of the part of the carrier sheet SH1 that is superimposed on the reading reference surface 22d are (Rd, Gd, Bd), and the second setting range, which is a setting range from the color (Rd, Gd, Bd), is Re≦Rd≦Rf, Ge≦Gd≦Gf, and Be≦Bd≦Bf. The second threshold value TH2 collectively refers to the boundary values ​​Re, Rf, Ge, Gf, Be, and Bf. The control unit 40 assigns black to pixels whose R value is Re or more and Rf or less, whose G value is Ge or more and Gf or less, and whose B value is Be or more and Bf or less, among the multiple pixels PX1 included in the thinned image data DA2, and assigns white to the remaining pixels, thereby generating analysis image data DA3. Therefore, the analysis image data DA3 is generated based on the reading result RS, with the background region A2 being an area of ​​color within the second set range from the color of the portion of the carrier sheet SH1 that is overlapped with the reading reference surface 22a. The control unit 40 generates the analysis image data DA3 based on the second threshold value TH2 and the reading result RS.

[0079] Fig. 13 illustrates a threshold setting process for applying the second threshold TH2 to the document area cut-out process shown in Fig. 11. The threshold setting process starts when an input for setting a threshold for generating analysis image data is received from the operation panel 7 or the control device 60 shown in Fig. 3. When the threshold setting process starts, the control unit 40 judges whether or not to use the carrier sheet SH1 based on the received input (S302). When the setting is such that the carrier sheet SH1 is not used, the control unit 40 sets the threshold of S204 shown in FIG. 11 to the first threshold TH1 (S304) and ends the threshold setting process. In this case, when the document area cutout process shown in FIG. 11 is performed, the first threshold TH1 is applied to the thinned image data DA2 to generate analysis image data DA3. On the other hand, when the setting is such that the carrier sheet SH1 is used, the control unit 40 sets the threshold of S204 shown in FIG. 11 to the second threshold TH2 (S306) and ends the threshold setting process. In this case, when the document area cutout process shown in FIG. 11 is performed, the second threshold TH2 is applied to the thinned image data DA2 to generate analysis image data DA3.

[0080] As described above, even when the original P0 held on the carrier sheet SH1 is read by the scanner 1A, the original area cut-out process suppresses the influence of dirt. In the example shown in Figures 12 and 13, since the control device 60 does not need to perform the cut-out process for deleting the carrier sheet portion outside the original area A1 from the output image data DA4, it is possible to directly store the output image data DA4 from the scanner 1A in a specified folder or directly attach it to an e-mail, etc.

[0081] As shown in Fig. 14, it is preferable to make the vertical streaks L1 remaining in the output image data DA4 less noticeable by image processing. When the control device 60 performs the above-mentioned image processing, if the position of the vertical streaks L1 is unknown, the following processing must be performed. First, the control device 60 checks the values ​​of each pixel PX1 of the output image data DA4 from end to end in the pixel column direction D2, and performs vertical streak extraction processing to extract anything that appears to be a vertical streak after checking all pixels of the output image data DA4. If a vertical streak is extracted, the control device 60 performs averaging processing using pixel values ​​to the left and right of the vertical streak to make the vertical streak less noticeable. However, since it is necessary to check all pixels PX1 of the output image data DA4 to extract the vertical streaks, the vertical streak extraction process takes time. Therefore, as shown in the example of Figure 15, the scanner 1A notifies the control device 60 of the dirt position P1 of the reading unit 20, thereby shortening the time required for the vertical streak extraction process.

[0082] FIG. 14 is a schematic diagram illustrating a process for reducing the streaks L1 in the output image data DA4 by image processing. The output image data DA4 is image data having R, G, and B read values ​​for a plurality of pixels PX1 arranged in the relative movement direction D1 and pixel row direction D2. In FIG. 14, the upper side DA4a, the lower side DA4b, the left side DA4c, and the right side DA4d of the output image data DA4 are shown. Here, as shown in the enlarged view in the lower part of FIG. 14, the R, G, and B values ​​of one or more pixels 70 included in the vertical stripe L1 are (r0, g0, b0), the R, G, and B values ​​of one or more pixels 71 located near the right side of the vertical stripe L1 are (r1, g1, b1), and the R, G, and B values ​​of one or more pixels 72 located near the left side of the vertical stripe L1 are (r2, g2, b2). Note that when the reference symbol 71 indicates a plurality of pixels, the pixel value (r1, g1, b1) is the arithmetic mean value of the pixel values ​​of each pixel. When the reference symbol 72 indicates a plurality of pixels, the pixel value (r2, g2, b2) is the arithmetic mean of the pixel values ​​of the individual pixels. The control device 60 replaces the values ​​of one or more pixels 70 included in the vertical stripe L1 with, for example, the arithmetic mean (r3, g3, b3) of the pixel values ​​(r1, g1, b1) and (r2, g2, b2).

[0083] 15 shows a schematic example of a process for reducing streaks L1 included in output image data DA4. This process includes an image transmission process performed by the scanner 1A and an image adjustment process performed by the control device 60. Here, S414 corresponds to the image processing unit 68 shown in FIG. 11 is completed, the scanner 1A starts an image transmission process, and transmits output image data DA4 corresponding to the document area A1 cut out of the scanned image data DA1, and the stain position P1 of the reading unit 20 to the control device 60 (S402). After the transmission process of S402, the scanner 1A ends the image transmission process.

[0084] The control device 60, which has received a request from the scanner 1A to receive the output image data DA4 and the stain position P1, starts the image adjustment process and receives the output image data DA4 and the stain position P1 (S412). Thereafter, as described above, the control device 60 performs a process to reduce the streaks L1 in the output image data DA4 along the relative movement direction D1 based on the stain position P1 (S414), and ends the image adjustment process. As a result, the control device 60 can make the vertical streaks L1 remaining in the output image data DA4 due to the presence of dirt such as dust or stains such as scratches on the reading unit 20 less noticeable through rapid image processing. The scanner 1A may itself perform the vertical streak reduction process shown in FIG.

[0085] (4) Variation: The present invention contemplates various modifications. For example, the image reading device may be a one-sided reading only device that does not have either the upper reading sensor 20A or the lower reading sensor 20B and reads only one side of the document P0. The above-described processes can be appropriately changed, such as by changing the order of the processes. For example, in the stain detection process shown in FIG. 6, the process of S102 for reading the first read data Ri and the process of S104 for acquiring the second read data Li can be changed.

[0086] In the above-mentioned specific example, the analysis image data DA3 is generated from the thinned image data DA2, but the analysis image data DA3 may be generated by applying a threshold value to the scanned image data DA1. Therefore, the analysis image data DA3 may have the same resolution as the scanned image data DA1.

[0087] (5) Conclusion: As described above, according to the present invention, it is possible to provide a technique capable of performing document region extraction processing while suppressing the influence of dirt, such as dust, or dirt, such as scratches, on the reading section, through various aspects. Of course, even with a technique consisting only of the constituent elements of the independent claims, the basic actions and effects described above can be obtained. In addition, configurations in which the configurations disclosed in the above examples are substituted with each other or the combination is changed, configurations in which the configurations disclosed in the publicly known techniques and the above examples are substituted with each other or the combination is changed, etc. The present invention also includes these configurations. [Explanation of symbols]

[0088] 1A...scanner (example of image reading device), 2...device body, 3...lower unit, 4...upper unit (example of cover section), 7...operation panel, 10...transport path, 11...original placement section, 14...feed roller, 15...separation roller, 16...pair of transport rollers, 17...pair of discharge rollers, 20...reading section, 20A...upper reading sensor, 20B...lower reading sensor, 21...transparent plate, 22...reference plate, 22a...reading reference surface, 23...light guide, 24...lens, 25...line sensor, 30...multiple feed detection section, 31...first original detection section, 32...second original detection section, 38...RAM (example of storage section), 40...control section, 41...CPU, 42...ROM, 43...storage section, 44...program, 45...motor for feed roller, 46...motor for transport roller, 47...I / F (transmission section an example of a scanning unit for scanning an image of a document, 60... a control device, 66... ​​a display device, 67... an I / F (an example of a receiving unit), 68... an image processing unit, A1... an original area, A2... a background area, B1, B9... a boundary, D1... a relative movement direction, D2... a pixel row direction, DA1... a scanned image data, DA1a... an upper edge, DA1b... a lower edge, DA1c... a left edge, DA1d... a right edge, DA2... a thinned image data, DA3... an analyzed image data, DA4... an output image data, L1... a streak, Li... a second scanned data, P0... an original, P1... a stain position, P2... an edge candidate position, PR1... a scanning control program, PX1... a pixel, PXL... a pixel row, Ri... a first scanned data, RS... a scanning result, SH1... a carrier sheet, Shi, TH1, TH2... a threshold value, ST1... a stain detection process, ST2... a cutting process, SY1... an image scanning system, UD... a driving unit.

Claims

1. A reading unit that reads an original and outputs a reading result of the original; a drive unit that moves the document and the reading unit relative to each other in a relative movement direction; Read image data based on the read result, and analysis image data based on the read result. The analysis for determining whether the data is in a document area or a background area outside the document area. A storage unit for storing image data; A stain detection unit detects a stain position of the reading unit in a pixel row direction intersecting with the relative movement direction. A detection process is performed to extract the document area from the read image data based on the analysis image data. A control unit that performs a cut-out process, The control unit, in the cut-out process, In the analysis image data, A search is performed for a candidate edge position where the background region changes to the document region, If the edge candidate position found corresponds to the stain position, the search is continued; If the searched edge candidate position does not correspond to the stain position, the searched edge candidate position is The image reading device further comprises: an image reading unit that reads out the document area from the read image data at the edge candidate position.

2. The reading unit has a reading reference surface, The image reading device further includes a storage unit that stores the first read data of the reading reference surface. 、 The control unit, in the stain detection process, after the first read data is stored, Second read data of the read reference surface is obtained, and based on the second read data and the first read data, The image reading device according to claim 1 , wherein the position of the stain is detected by detecting the position of the stain.

3. the background area is an area having a color within a first set range from the color of the reading reference surface, The control unit determines a first threshold value for separating a color of the document region from a color of the background region and the read result. The image reading apparatus according to claim 2 , further comprising: a sensor for detecting a position of the image sensor;

4. The background area is a region overlapping the reading reference surface of a carrier sheet that holds the document. a color range within a second range from the color of the selected portion, The control unit determines a second threshold value that separates a color of the document region from a color of the background region and the read result. The image reading apparatus according to claim 2 , further comprising: a sensor for detecting a position of the image sensor;

5. A reading unit that reads an original and outputs a reading result of the original; a drive unit that moves the document and the reading unit relative to each other in a relative movement direction; Read image data based on the read result, and analysis image data based on the read result. The analysis for determining whether the data is in a document area or a background area outside the document area. A storage unit for storing image data; the reading unit has a reading reference surface and a transparent plate covering the reading reference surface, The reference data is obtained by reading the reading reference surface when the transmission plate is free of dirt. A storage unit that stores the first read data; A stain detection unit detects a stain position of the reading unit in a pixel row direction intersecting with the relative movement direction. A detection process is performed to extract the document area from the read image data based on the analysis image data. A control unit that performs a cut-out process, The control unit, in the stain detection process, after the first read data is stored, obtaining second read data which is data for shading correction obtained by reading the read reference surface; Detecting the stain position based on the second read data and the first read data; The control unit, in the cut-out process, In the analysis image data, the back A search is performed for a candidate edge position where the background region changes to the document region; If the edge candidate position found corresponds to the stain position, the search is continued; If the searched edge candidate position does not correspond to the stain position, the searched edge candidate position is The image reading device further comprises: an image reading unit that reads out the document area from the read image data at the edge candidate position.

6. the background area is an area having a color within a first set range from the color of the reading reference surface, The control unit determines a first threshold value for separating a color of the document region from a color of the background region and the read result. The image reading apparatus according to claim 5 , wherein the analysis image data is generated based on the result of the analysis.

7. The background area is a region overlapping the reading reference surface of a carrier sheet that holds the document. a color range within a second range from the color of the selected portion, The control unit determines a second threshold value that separates a color of the document region from a color of the background region and the read result. The image reading apparatus according to claim 5 , wherein the analysis image data is generated based on the result of the analysis.

8. a cover portion for opening and closing the document transport path, The control unit is configured to control the image reading device when the image reading device is turned on and when the reading of the document is completed. The dirt detection process is performed at least partly when the cleaning is completed and when the cover is closed. The image reading device according to any one of claims 1 to 7, wherein the image reading device performs the above-mentioned steps.

9. The storage unit stores the analysis image data having a resolution lower than that of the read image data. The image reading device according to any one of claims 1 to 8.

10. The control unit determines, in the extraction process, that the edge candidate position, which is the stain position, After the search, the edge is detected up to the middle in the pixel column direction in the analysis image data. If the candidate position is not found, the edge candidate position is extracted from the read image data.

10. The image reading apparatus according to claim 1, further comprising: an image processing unit for processing the document area.

11. The image reading device according to any one of claims 1 to 10, and a control device are provided. An image reading system, comprising: The image reading device reads the scanned image data corresponding to the document region that is cut out. A transmission unit that transmits the output image data and the stain position to the control device, The control device includes: a receiving unit that receives the output image data and the stain position; In the output image data, a streak along the relative movement direction is detected based on the stain position. and an image processing unit for reducing Image reading system.

12. A reading unit that reads an original and outputs a reading result of the original; a drive unit that moves the document and the reading unit relative to each other in a relative movement direction; Read image data based on the read result, and analysis image data based on the read result. The analysis for determining whether the data is in a document area or a background area outside the document area. A method for extracting an original area in an image reading device having a storage unit for storing image data There was, A stain detection unit detects a stain position of the reading unit in a pixel row direction intersecting with the relative movement direction. A detection step; Extracting the document region from the read image data based on the analysis image data. and In the cutting step, In the analysis image data, A search is performed for a candidate edge position where the background region changes to the document region, If the edge candidate position found corresponds to the stain position, the search is continued; If the searched edge candidate position does not correspond to the stain position, the searched edge candidate position is A document area extraction method for extracting the document area from the read image data at the edge candidate position 。

13. A reading unit that reads an original and outputs a reading result of the original; a drive unit that moves the document and the reading unit relative to each other in a relative movement direction; Read image data based on the read result, and analysis image data based on the read result. The analysis for determining whether the data is in a document area or a background area outside the document area. A storage unit for storing image data, and the reading unit includes a reading reference surface and a transparent sheet covering the reading reference surface. and a reference plate, the reference surface being read when the reference plate is free of dirt. a storage unit for storing first read data, which is data of the original document area of ​​an image reading device; A region extraction method, comprising the steps of: A stain detection unit detects a stain position of the reading unit in a pixel row direction intersecting with the relative movement direction. A detection step; Extracting the document region from the read image data based on the analysis image data. and In the stain detection step, A shading correction process is performed by reading the reading reference surface after the first read data is stored. The second read data, which is the correct data, is obtained, and the second read data and the first read data are compared. The position of the dirt is detected based on the In the cutting step, In the analysis image data, the back A search is performed for a candidate edge position where the background region changes to the document region; If the edge candidate position found corresponds to the stain position, the search is continued; If the searched edge candidate position does not correspond to the stain position, the searched edge candidate position is A document area extraction method for extracting the document area from the read image data at the edge candidate position 。

Images