JP4862331B2 - Image processing apparatus, input image correction method, and computer program - Google Patents

Description

  The present invention relates to an image processing apparatus and a correction method for correcting an input image input by a scanner or the like.

  When an original image is input by a scanner, a noise image may be included in the input image. In order to obtain a cleaner input image, it is desirable to remove noise from the input image. Therefore, for example, a method for removing noise from an input image as described in Patent Document 1 has been proposed.

According to such a method, the image processing apparatus extracts an input image read from the image input device as a set of circumscribed rectangles in which continuous black pixels are grouped. Next, it is determined whether or not each rectangle is noise. Noise is removed by converting black pixels in the rectangle determined to be noise into white pixels. The input image from which noise has been removed is displayed on the display. Here, the user gives an instruction to correct the erroneously determined noise while confirming the displayed input image. That is, for noise that has not been removed, it is instructed to be noise, and an instruction is given that a part of the removed image is not noise. Then, the instructed content is executed and the image is corrected. When the correction work is finished, the noise removal process is completed.
JP 2004-110128 A

  However, in the conventional method as described in Patent Document 1, it is difficult to instruct correction. This is because noise is usually scattered in a wide range of the input image. Therefore, since the noise that could not be removed at the time of the first removal process is scattered in a wide range, when the user designates these one by one, how far the work has progressed gradually It becomes difficult to understand and the mental burden increases. Work mistakes are also likely to occur.

  Also, it is more difficult to specify the part that has been removed even though it is not noise. This is because such a portion has already been removed and is not displayed on the display, so that it must be specified while comparing the original document with the image displayed on the display.

  SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and an object of the present invention is to enable a user to accurately remove a noise portion from an input image more easily than in the past.

The image processing apparatus according to the present invention, a noise image detection means for detecting an image of the noise included in the input image, the magnification rate receiving unit from the user accepts the magnification of the input image, depending on the magnification accepted wherein determining the size for dividing an input image Te, an image dividing means for dividing the input image based on the size of the determined split image, among divided the divided image by the image dividing unit, the user A first one or a plurality of designated ones indicating the positions of those included in the designated divided image among the detected noise images, which are noise images detected by the noise image detecting means, at the received enlargement ratio . display with one of the indicia, and dividing the image display means, disposed cells corresponding to the respective divided image in accordance with the relative positional relationship between the respective divided image The cell group thus formed is displayed side by side with the divided image currently displayed by the divided image display means, and corresponds to the divided image currently displayed by the divided image display means in the cell group. A cell is displayed in a first display form different from other cells, and the divided image cell display means and the noise image detection means of the detected noise images included in the displayed divided image are erroneously displayed. generating a false detection noise selecting means for selecting those detected as an image noise to the user, the noise correction image than those selected are-option is an image obtained by removing from the input image of the detected noise image Noise correction image generation means for performing the above-described processing.

Preferably, the enlargement ratio accepting unit is capable of accepting the change of the enlargement ratio, and the image dividing means is changed in response to the enlargement ratio accepting unit accepting the change of the enlargement ratio. The size is re-determined according to the enlargement ratio, the input image is re-divided into the divided images based on the re-determined size, and the divided image display means is configured to re-divide the divided images. One or more specified by the user is displayed together with the first mark at the changed enlargement ratio, and the divided image cell display means corresponds to the re-divided divided image. The cell group is displayed.
Alternatively, the divided image cell display means may select a cell corresponding to the divided image in the cell group, which has been confirmed by the user for noise detection errors, from a second display mode different from the first display form. A cell corresponding to the divided image that is not currently displayed by the divided image display means and that has not yet been confirmed by the user for noise detection errors. Display is performed so that a third display form different from the display form and the second display form is obtained.

Or, the entering force included in the image, detects a false noise image of a suspicious part When it is noise, the pseudo noise image detection means, further comprising, the divided image display means, before Symbol first A second mark is displayed at the position of the suspicious noise image included in the input image together with the mark , and the erroneous detection noise selection means is erroneously detected by the noise image detection means of the detected noise image. user to select what is image noise of the detected ones, and the pseudo noise image as the noise correction image generation means as the noise correction image was selected among the detected noise image and those selected from among the pseudo noise image and non also of generating the images is removed from the input image.

  In the present invention, the mark of the position of the detected noise image and the mark of the position of the suspicious noise image are applied, for example, by giving the detected noise image and the suspicious noise image a color different from the original image. Alternatively, the detection noise image and the suspicious noise image may be marked by surrounding them with a dotted line or a one-dot chain line. Alternatively, an icon may be displayed as a mark on the detected noise image and the suspicious noise image.

  According to the present invention, the user can more accurately remove the noise portion from the input image than before.

  FIG. 1 is a diagram illustrating an example of the overall configuration of a system provided with the image processing apparatus 1, FIG. 2 is a diagram illustrating an example of a hardware configuration of the image processing apparatus 1, and FIG. 3 is a functional diagram of the image processing apparatus 1. It is a figure which shows the example of a structure.

  An image processing apparatus 1 and an MFP (Multi Function Peripherals) 2 according to the present invention are connected to each other via a communication line 3 as shown in FIG. As the communication line 3, the Internet, a LAN, a public line, a dedicated line, or the like is used. You may connect via interfaces, such as USB or RS-232C.

  As shown in FIG. 2, the image processing apparatus 1 includes an image processing apparatus main body 10, a display 11, an input device 12, and the like. The input device 12 is a device such as a mouse or a keyboard for the user to input data or commands to the image processing apparatus main body 10.

  The image processing apparatus main body 10 includes a CPU 10a, a RAM 10b, a ROM 10c, a hard disk 10d, a communication interface 10e, a video interface 10f, an input interface 10g, and the like.

  The communication interface 10 e is a network interface card (NIC) or a modem, and is used for communicating with the MFP 2 via the communication line 3. The video interface 10f and the input interface 10g are used to connect to the display 11 and the input device 12, respectively.

  The hard disk 10d includes a document image receiving unit 101, a screen display processing unit 102, a noise image detecting unit 103, a suspicious noise image detecting unit 104, a noise-removed image generating unit 105, an image dividing unit 106, an input image, as shown in FIG. A program and data for realizing the functions of the correction unit 107, the image storage unit 1M1, the corrected noise position storage unit 1M2, and the like are stored. These programs and data are called to the RAM 10b as necessary, and the programs are executed by the CPU 10a.

  As the image processing apparatus 1, for example, a personal computer or a workstation is used.

  The MFP 2 is an image forming apparatus having functions such as copy, network printing, scanner, fax, document server, and file transfer. Sometimes called a "multifunction machine".

  Hereinafter, functions and processing contents of each unit of the image processing apparatus 1 and the MFP 2 illustrated in FIG. 3 will be described in detail.

  FIG. 4 is a view showing a display example of the document image window WN1, FIG. 5 is a view showing a display example of the noise editing window WN2 and the area confirmation window WN3, and FIG. 6 is a flowchart for explaining an example of the flow of the corrected image generation process. .

  In FIG. 3, the MFP 2 optically reads images such as photographs, characters, pictures, charts and the like drawn on a document sheet set on a document table by a scanner, and generates an image file FG. Then, the image file FG is transferred to the image processing apparatus 1 by FTP (File Transfer Protocol) or the like. Hereinafter, the document image read by the MFP 2 is referred to as “document image GA0”. In the present embodiment, a case where the document image GA0 is a monochrome binary image will be described as an example.

  The document image receiving unit 101 of the image processing apparatus 1 receives the image file FG of the document image GA0 from the MFP 2. The received image file FG of the document image GA0 is stored in the image storage unit 1M1. In addition, the image storage unit 1M1 stores image data of images such as a noise image Gz, a suspicious noise image Gd, a divided image GAb, and a corrected image GA2, which will be described later.

  After receiving the image file FG from the MFP 2, the screen display processing unit 102 displays a document image window WN1 as shown in FIG. 4 on the desktop DP. The document image window WN1 includes an image display area RY1, an enlargement ratio designation menu KM1, a division execution button BN1, and the like. In the image display area RY1, the document image GA0 is displayed based on the received image file FG.

  If the original image GA0 cannot be displayed in the image display area RY1 due to the relationship between the size or shape of the original image GA0 and the size or shape of the image display area RY1, only a part of the original image GA0 is displayed in the image display area RY1. Let In this case, the user can operate the scroll button on the right side of the image display area RY1 to scroll the document image GA0 and display a portion that is not displayed.

  The screen display processing unit 102 also performs processing for displaying windows such as a noise editing window WN2 and a region confirmation window WN3 (see FIG. 5), which will be described later, on the desktop DP.

  By the way, the document image GA0 may contain an image of noise generated when the MFP 2 reads the original document. According to the image processing apparatus 1 according to the present embodiment, the user performs image processing so as to remove noise accurately while enlarging and displaying the original image GA0 in the image display area RY1 of the original image window WN1 and checking the noise. An instruction can be given to the device 1.

  The noise image detection unit 103 detects an image of noise generated in the document image GA0. Hereinafter, the noise image detected by the noise image detection unit 103 is referred to as a “noise image Gz”. The noise image Gz may be detected by a known method, but in the present embodiment, for example, it is detected by the following method.

  “Αx × αy” is set in advance as the predetermined size α. A block consisting of one pixel having a value indicating black or a plurality of continuous pixels is searched from the document image GA0. That is, a block of pixels surrounded by pixels having a value indicating white is searched. Then, when the size of the retrieved chunk is less than the size α, that is, when the horizontal length of the chunk is less than αx and the vertical length is less than αy, an image of the chunk portion is It is determined that the noise image is Gz.

  The suspicious noise image detection unit 104 detects an image suspected of being noise from the document image GA0. Hereinafter, the image detected by the suspicious noise image detection unit 104 is referred to as “suspect noise image Gd”. The suspicious noise image Gd may be detected by the following method.

  “Βx × βy” is set in advance as the predetermined size β. A plurality of continuous pixel clusters having a value indicating black are searched from the document image GA0. When the size of the retrieved chunk is not less than size α and less than size β, that is, the length in the horizontal direction of the chunk is not less than αx and less than βx, and the length in the vertical direction is not less than αy and less than βy. If it is, it is determined that the image of the block is the suspicious noise image Gd. However, “βx ≧ αx”, “βy ≧ αy”, and “(βx × βy)> (αx × αy)”.

  The image data of the noise image Gz detected by the noise image detection unit 103 is stored in the image storage unit 1M1 together with position information Pz indicating the position of each pixel of the noise image Gz in the document image GA0. Similarly, the image data of the suspicious noise image Gd detected by the suspicious noise image detection unit 104 is stored in the image storage unit 1M1 together with the position information Pd indicating the position of each pixel of the suspicious noise image Gd in the document image GA0. Is done.

  The noise-removed image generation unit 105 generates the noise-removed image GA1 by removing the noise image Gz detected by the noise image detection unit 103 and the suspicious noise image Gd detected by the suspicious noise image detection unit 104 from the document image GA0. To do. The noise-removed image GA1 may be generated by a known method. For example, the noise-removed image GA1 is generated by changing the value of each pixel of the detected noise image Gz and the value of each pixel of the suspicious noise image Gd in the document image GA0 to a value indicating white. Can do. The generated noise-removed image GA1 is stored in the image storage unit 1M1.

  The image dividing unit 106 performs processing for dividing the noise-removed image GA1 into images of a plurality of rectangular areas. Such processing is performed based on the enlargement ratio designated by the user in the document image window WN1 of FIG.

  That is, the user operates the enlargement factor designation menu KM1 in the manuscript image window WN1 displayed on the desktop DP to designate the enlargement factor (hereinafter referred to as “enlargement factor Kr”) of the manuscript image GA0. When the split execution button BN1 is clicked, the image dividing unit 106 starts a process of dividing the noise-removed image GA1 of the document image GA0.

  First, the size when the original image GA0 is enlarged according to the enlargement ratio Kr is calculated. That is, “Lx × Kr” and “Ly × Kr” are calculated assuming that the horizontal length and the vertical length of the document image GA0 are Lx and Ly, respectively. For example, when “Kr = 500%”, Lx and Ly are each multiplied by 5. Hereinafter, the calculated horizontal length and vertical length at the time of enlargement will be referred to as Lx ′ and Ly ′, respectively.

  “Lx ′ / RYx” and “Ly ′ / RYy” are calculated, and when the division is not divisible, the first decimal place of the calculated quotient is raised. However, RYx and RYy are a horizontal length and a vertical length of an image display area RY2 (see FIG. 5) described later, respectively. Hereinafter, the values thus obtained are described as DVx and DVy, respectively.

  Then, the noise-removed image GA1 is equally divided into DVx × DVy rectangular area images. Thereby, the process of dividing the noise-removed image GA1 is completed. Hereinafter, each divided image is referred to as “divided image GAb”.

  After the noise-removed image GA1 is divided into a plurality of divided images GAb by the image dividing unit 106, the screen display processing unit 102 closes the document image window WN1. Then, as shown in FIG. 5, a noise editing window WN2 and a region confirmation window WN3 are displayed on the desktop DP.

  The noise editing window WN2 includes an image display area RY2, an enlargement ratio designation menu KM2, a division execution button BN2, a check box CB2, an up button MB21, a down button MB22, a right button MB23, and a left button MB24. ing.

  The screen display processing unit 102 enlarges and displays any one of the divided images GAb of the noise-removed image GA1 in the image display area RY2 with the enlargement factor Kr specified by the user. Further, among the noise image Gz and the suspicious noise image Gd included in the original document image GA0 of the divided image GAb, an image located in the area of the divided image GAb is displayed on the divided image GAb. Are superimposed and displayed like a layer.

  The position where the noise image Gz is superimposed on the divided image GAb is set to the position where the noise image Gz is based on the position information Pz of the noise image Gz. That is, if the noise image Gz is not deleted from the divided image GAb, the noise image Gz is superimposed at a position that appeared in the divided image GAb. The same applies to the noisy noise image Gd.

  Further, the display forms of the noise image Gz, the suspicious noise image Gd, and the divided image GAb are made different so that the user can easily distinguish them. For example, the content part of the divided image GAb is displayed in black, the content part of the noise image Gz is displayed in red, and the content part of the suspicious noise image Gd is displayed in yellow. Alternatively, the noise image Gz may be surrounded by a dotted line, and the pseudo noise image Gd may be surrounded by a one-dot chain line.

  By such processing, the divided image of the document image GA0 includes the noise image Gz portion, the noisy noise image Gd portion, and the other image portion (divided image GAb) included in the image display area RY2. They are displayed in different display forms.

  When the noise editing window WN2 is displayed, as a default, the divided image GAb positioned at the upper left of the divided images GAb of the noise-removed image GA1 is displayed in the image display area RY2. Then, when the right move button MB23 is clicked, the display of the currently displayed divided image GAb is stopped, and the divided image GAb adjacent to the right is displayed. Also at this time, the noise image Gz and the suspicious noise image Gd that were within the area of the newly displayed divided image GAb are superimposed and displayed. Similarly, when the downward movement button MB22 is clicked, the divided image GAb, the noise image Gz, and the suspicious noise image Gd adjacent to the lower side are displayed, and when the left movement button MB24 is clicked, the divided image GAb, the noise image adjacent to the left side is displayed. Gz and the suspicious noise image Gd are displayed, and when the up button MB21 is clicked, the adjacent divided image GAb, noise image Gz, and suspicious noise image Gd are displayed.

  The area confirmation window WN3 includes a display position confirmation table TL3, a correction execution button BN3, and the like. The display position confirmation table TL3 includes DVx × DVy cells, that is, cells corresponding to the number of divided divided images GAb. Each cell corresponds to a divided image GAb having the same relative positional relationship. That is, the Xth cell from the left and the Yth cell from the top corresponds to the divided image GAb that is the Xth from the left and the Yth from the top among the divided image GAb.

  The cell corresponding to the divided image GAb currently displayed in the image display area RY2 of the noise editing window WN2 is displayed in a different display form (for example, black) from the other cells. Therefore, each time the divided image GAb displayed in the image display area RY2 changes, the display form of the cell changes as if the black cursor moves.

  In the noise editing window WN2, when the user wants to enlarge the image, the user designates an enlargement ratio Kr larger than the enlargement ratio Kr designated previously by operating the enlargement ratio designation menu KM2. Alternatively, when it is desired to reduce the size, an enlargement factor Kr smaller than the enlargement factor Kr designated previously is designated. Then, the division execution button BN2 is clicked.

  Then, the image division unit 106 performs the division process of the noise-removed image GA1 again based on the newly specified enlargement ratio Kr. The screen display processing unit 102 displays the re-divided divided image GAb in the image display area RY2.

  In addition, the user can confirm and specify the noise image Gz detected by the noise image detection unit 103 in spite of not being noise in the image display area RY2 of the noise editing window WN2. In addition, it is possible to confirm and designate a noise that is a noise among the suspicious noise images Gd detected by the suspicious noise image detection unit 104. These designations are made by, for example, selecting the mouse by placing the mouse pointer on the noise image Gz or the noisy noise image Gd and double-clicking the mouse button.

  Then, the corrected noise position storage unit 1M2 in FIG. 3 indicates the position indicated by the position information Pz of the designated noise image Gz (that is, the position of the noise image Gz that is not actually noise in the original noise-removed image GA1). Stored as non-noise position information Nz. Further, the position indicated by the position information Pz of the designated suspicious noise image Gd (that is, the position of the suspicious noise image Gd that is actually noise in the original noise-removed image GA1) is stored as the noise position information Nd.

  The display form of the designated noise image Gz is different from the display form of the non-designated noise image Gz so that the user can easily distinguish it. For example, the color of the designated noise image Gz is displayed darker than the color of the noise image Gz that is not designated. The same applies to the noisy noise image Gd.

  When the user designates the noise image Gz or the suspicious noise image Gd by mistake, the user selects the noise image Gz or the suspicious noise image Gd by double-clicking again. Then, the non-noise position information Nz of the noise image Gz or the noise position information Nd of the suspicious noise image Gd is deleted from the corrected noise position etc. storage unit 1M2, and the display form of the noise image Gz or the suspicious noise image Gd is the original. Return to. As a result, the designation is canceled.

  When the necessary specification for the noise image Gz and the suspicious noise image Gd currently displayed in the image display area RY2 is completed, the user can select the up button MB21, the down button MB22, the right button MB23, and the left button MB24. Click one of them.

  Then, after temporarily clearing the display content of the image display area RY2, the screen display processing unit 102 displays the adjacent divided image GAb corresponding to the movement button in the image display area RY2, and further, the range of the divided image GAb. The noise image Gz and the suspicious noise image Gd existing inside are superimposed on the divided image GAb like a layer and displayed. As a result, adjacent divided images GAb are sequentially displayed in the image display area RY2 like frame advance.

  At this time, the user can recognize that the cell corresponding to the divided image GAb displayed in the image display area RY2 immediately before in the display position confirmation table TL3 of the area confirmation window WN3 has been confirmed. Highlight with a color such as gray. Further, the cell corresponding to the newly displayed divided image GAb is changed to black. Here, the user also designates the noise image Gz and the suspicious noise image Gd that are newly displayed in a superimposed manner and those that are erroneously detected and noise.

  When the divided image GAb that has already been confirmed is displayed again in the image display area RY2, the screen display processing unit 102 keeps the check box CB2 in the “on” state, and the entire image display area RY2 is light gray or the like. To be masked. However, the divided image GAb, the noise image Gz, and the suspicious noise image Gd under the mask are made to show through. Such a display form means that designation of the noise image Gz and the suspicious noise image Gd within the range of the divided image GAb is prohibited (that is, editing is not possible). Therefore, the user clicks the check box CB2 to turn “off” when he / she wants to specify again the noise image Gz and the suspicious noise image Gd. As a result, the prohibition of mask display and designation is released, and re-designation is possible.

  When the correction execution button BN3 in the area confirmation window WN3 is clicked, the input image correction unit 107 corrects the document image GA0 by a procedure as shown in FIG. 6A, for example, and generates a corrected image GA2. I do.

  The noise removal image GA1 of the original image GA0 to be corrected and the noise image Gz at the position indicated by the non-noise position information Nz stored in the correction noise position etc. storage unit 1M2 (that is, erroneously detected by the noise image detection unit 103) Image data of the noise image Gz) that is detected but not originally noise and the position information Pz thereof, and the suspicious noise image Gd in a position not indicated in any noise position information Nd (that is, there is a suspicion of noise) Of the suspicious noise image Gd detected by the suspicious noise image detection unit 104, image data of the suspicious noise image Gd that is not actually noise) and its position information Pd are called from the image storage unit 1M1 (FIG. 6A). # 201).

  The called noise image Gz is superimposed on the called noise-removed image GA1 (# 202). However, the position where the noise image Gz is superimposed is the position indicated by the position information Pz of the noise image Gz.

  Further, the called noise image Gd is superimposed on the noise-removed image GA1 (# 203). In this case as well, as in step # 202, the position where the suspicious noise image Gd is superimposed is the position indicated by the position information Pd of the suspicious noise image Gd.

  In this way, by superimposing an image that is not actually noise designated by the user on the noise-removed image GA1 from which the image of the part that has been mechanically determined to be noise or suspected is removed, It is possible to generate a corrected image GA2 from which noise has been more accurately removed from the image GA0.

  Alternatively, the generation process of the noise-removed image GA1 previously executed by the noise-removed image generation unit 105 is canceled (Undo) by the procedure as shown in FIG. 6B, and the contents specified by the user in the noise editing window WN2 etc. The corrected image GA2 may be generated by regenerating the noise-removed image GA1 based on the above.

  First, the position of noise generated in the document image GA0 is confirmed based on the information regarding the positions stored in the image storage unit 1M1 and the corrected noise position storage unit 1M2. Specifically, none of the positions indicated by the non-noise position information Nz stored in the corrected noise position storage unit 1M2 is indicated from the position information Pz stored in the image storage unit 1M1. The position information Pz is called (# 211 in FIG. 6B). That is, the position information Pz of the noise image Gz detected without error by the noise image detection unit 103 is called. Further, the noise position information Nd stored in the corrected noise position etc. storage unit 1M2 is called (# 212). That is, the noise position information Nd of the suspicious noise image Gd that is actually noise is called.

  Then, by changing the pixels at the positions indicated in the position information Pz and the noise position information Nd called in steps # 211 and # 212 among the pixels constituting the document image GA0 to indicate white values. Then, the noise-removed image GA1 is generated again (# 213). The noise-removed image GA1 generated in this way becomes the corrected image GA2.

  Next, an overall processing flow when the image processing apparatus 1 generates the corrected image GA2 by removing noise from the document image GA0 will be described with reference to a flowchart. FIG. 7 is a flowchart illustrating an example of the overall processing flow of the image processing apparatus 1.

  In FIG. 7, when the image processing apparatus 1 acquires the data of the document image GA0 from the MFP 2 (# 1), the image processing apparatus 1 detects the noise image Gz and its position information Pz from the document image GA0 (# 2). The noise image Gd and its position information Pd are detected (# 3). Then, the noise image G1 is generated by removing the detected noise image Gz and the suspected noise image Gd from the document image GA0 (# 4), and the document image window WN1 (see FIG. 4) is displayed on the desktop DP (#). 5). In the image display area RY1 of the document image window WN1, the noise-removed image GA1 generated in step # 4 is displayed.

  When the enlargement factor Kr is designated by the user (# 6), the image processing apparatus 1 divides the noise-removed image GA1 into a plurality of divided images GAb according to the enlargement factor Kr (# 7), and the document. After the image window WN1 is closed, the noise editing window WN2 and the area confirmation window WN3 (see FIG. 5) are displayed on the desktop DP (# 8).

  In the image display area RY2 of the noise editing window WN2, any one of the divided images GAb divided in step # 7 is enlarged and displayed according to the enlargement ratio Kr. Further, among the divided image GAb, the noise image Gz detected in step # 2 and the suspicious noise image Gd detected in step # 3, which is located in the region of the divided image GAb, The divided image GAb is superimposed and displayed like a layer. The noise image Gz and the suspicious noise image Gd are also enlarged and displayed according to the enlargement ratio Kr.

  Here, the user designates the noise image Gz erroneously detected as noise among the noise images Gz displayed in the image display area RY2, for example, by double clicking with the mouse. Further, the suspicious noise image Gd, which is noise, among the suspicious noise images Gd displayed in the image display area RY2 is designated.

  Upon receiving these designations (# 9), the image processing apparatus 1 acquires the non-noise position information Nz of the designated noise image Gz and the noise position information Nd of the suspicious noise image Gd (# 10).

  The user can switch the divided image GAb to be displayed in the image display region RY2 by operating the up button MB21 or the left button MB24 in the noise editing window WN2. Also in the switched divided image GAb, the noise image Gz erroneously detected as noise and the suspicious noise image Gd as noise are designated.

  When the user completes confirmation and necessary specification of the noise image Gz and the noisy noise image Gd for all the divided images GAb of the noise-removed image GA1 (Yes in # 11), the position information Pz and position information acquired in step # 2 Correction that is an image obtained by removing noise from the original image GA0 more accurately than the noise-removed image GA1 generated in step # 4 based on Pd and the non-noise position information Nz and noise position information Nd acquired in step # 10 The image GA2 is generated (# 12). The procedure for generating the corrected image GA2 is as described above with reference to FIGS.

  After generating the corrected image GA2, the noise editing window WN2 and the area confirmation window WN3 may be closed, a new window may be opened, and the corrected image GA2 may be displayed on the window. Alternatively, the image data of the corrected image GA2 may be converted into a predetermined format and stored in the hard disk 10d as an image file, or may be output to another device connected to the image processing device 1.

  According to this embodiment, the noise detection result (noise image Gz) is displayed superimposed on the original image (noise-removed image GA1) from which noise has been removed. Further, an image of a part that is suspected of being noise (suspected noise image Gd) is also displayed superimposed on the noise-removed image GA1. In addition, black, red, and yellow are used as display colors of the noise-removed image GA1, the noise image Gz, and the suspicious noise image Gd, respectively. Thereby, the user can remove the noise portion included in the document image more easily and accurately than in the past.

  Moreover, the noise-removed image GA1, the noise image Gz, and the suspicious noise image Gd are displayed in units of the divided image GAb of the noise-removed image GA1. In the display position confirmation table TL3 of the area confirmation window WN3, the currently displayed range, the confirmed range, and the unconfirmed range are indicated by cells of different colors. Therefore, the user can easily grasp to what extent the error of noise detection has been confirmed, and the mental burden is reduced as compared with the conventional case.

  FIG. 8 is a diagram showing an example of the comparison window WN4.

  In the present embodiment, the noise image Gz and the noisy noise image Gd are detected from the document image GA0, and after the noise-removed image GA1 is generated, the image is divided into the divided images GAb. For example, the document image GA0 may be divided into divided images first, and then the noise image Gz and the suspicious noise image Gd may be detected for each divided image to generate a noise-removed image. In this case, for each divided image, processing is performed to superimpose the noise image Gz and the noisy noise image Gd, which are designated by the user and where noise actually occurs, on the noise-removed image. Then, the corrected image GA2 of the document image GA0 may be generated by arranging and integrating the superimposed noise-removed images at the original positions.

  When the noise image Gz detected by the noise image detection unit 103 is often erroneous, the size α, which is a threshold value for determining noise, is changed to a small value, and the noise image Gz is changed to the changed size α. This detection process may be redone. The value to be changed may be designated by the user, or may be automatically determined by the image processing apparatus 1. Alternatively, when there are many detection omissions of the noise image Gz, the size α may be changed to a large value and the detection process of the noise image Gz may be performed again. Similarly, the size β, which is one of threshold values when detecting the suspicious noise image Gd, may be changed according to the detection accuracy, and the detection process of the suspicious noise image Gd may be performed again.

  In addition, by simultaneously displaying the detection result of the noise image Gz and the suspicious noise image Gd before the change of the size α or the size β and the detection result of the noise image Gz and the suspicious noise image Gd after the change, the threshold value before the change ( The user may be able to select whether the size α and the size β) should be adopted or the changed threshold value should be adopted. For example, as in the comparison window WN4 in FIG. 8, the detection result based on the threshold value before the change and the detection result based on the threshold value after the change may be displayed simultaneously. And what is necessary is just to employ | adopt the direction which the user selected from them. Alternatively, the detection result based on the threshold value before the change and the detection result based on the threshold value after the change may be overlapped and displayed in one area. In this case, the display form of the noise image Gz may be different depending on what is detected only before the change, what is detected only after the change, and what is detected both before and after the change. desirable. For example, the noise image Gz detected only before the change is displayed in dark red, the noise image Gz detected only after the change is displayed in light red, and the noise image Gz detected before and after the change is displayed. It is displayed in the middle dark red. The same applies to the noisy noise image Gd.

  In the present embodiment, the document image GA0 input by the MFP 2 is transmitted to the image processing apparatus 1 and noise removal processing is performed by the image processing apparatus 1. However, the MFP 2 may perform all processing alone. Good. In this case, the function of each unit shown in FIG.

  In the present embodiment, only one divided image GAb is displayed in the noise editing window WN2. However, a plurality of image display areas RY2 may be provided so that a plurality of divided images GAb can be displayed simultaneously.

  In the present embodiment, the case where the document image GA0 is a monochrome binary image has been described as an example. However, the present invention can also be applied to the case where the document image GA0 is a color image. In this case, the original image GA0 is color-separated into, for example, C (cyan), M (magenta), Y (yellow), and K (black) plates, and the noise image Gz and the noisy noise image Gd for each plate image. And the like. The present invention can also be applied to correct line errors automatically detected by the image processing apparatus during line emphasis processing.

  In addition, the configuration of the entire image processing apparatus 1 or each unit, processing contents, processing order, and the like can be changed as appropriate in accordance with the spirit of the present invention.

  The present invention is particularly suitable for removing noise from an input image of a document printed on a sheet having a color or pattern.

It is a figure which shows the example of the whole structure of the system provided with the image processing apparatus. It is a figure which shows the example of the hardware constitutions of an image processing apparatus. It is a figure which shows the example of a functional structure of an image processing apparatus. FIG. 6 is a diagram illustrating a display example of a document image window. It is a figure which shows the example of a display of a noise edit window and an area | region confirmation window. It is a flowchart explaining the example of the flow of a correction | amendment image generation process. It is a flowchart explaining the example of the flow of the whole process of an image processing apparatus. It is a figure which shows the example of a comparison window.

Explanation of symbols

1 Image processing device 2 MFP (image input device)
12 Input device (false detection noise selection means, selection means)
102 Screen display processing unit (divided image display means, divided image cell display means, input image display means)
103 noise image detection unit (noise image detection means)
104 Suspicious noise image detection unit (suspicious noise image detection means)
107 Input image correction unit (noise correction image generation means)
GA0 Original image (input image)
GA2 corrected image (noise corrected image)
Gd suspicious noise image Gz noise image (detection noise image)
RY2 image display area (divided image display means, erroneous detection noise selection means, selection means)
TL3 display position confirmation table (cell group)
WN3 area confirmation window (divided image cell display means)
α size (first size)
β size (second size)

Claims (11)

A noise image detection means for detecting an image of the noise included in the input image,
An enlargement rate receiving means for receiving an enlargement rate of the input image from a user;
An image dividing means for determining a size for dividing the input image according to the received enlargement ratio, and for dividing the input image into divided images based on the determined size;
A detected noise image, which is an image of noise detected by the noise image detecting means at the received enlargement ratio , of one or more specified by the user among the divided images divided by the image dividing means. A divided image display means for displaying together with a first mark indicating the position of the one included in the designated divided image;
A cell group in which cells corresponding to the respective divided images are arranged according to the relative positional relationship between the respective divided images is displayed side by side with the divided images currently displayed by the divided image display means, and the cells A divided image cell display means for displaying a cell corresponding to the divided image currently displayed by the divided image display means in a group in a first display form different from other cells;
An erroneous detection noise selection means for causing a user to select one of the detected noise images included in the displayed divided image that is erroneously detected as a noise image by the noise image detection means;
A noise corrected image generating means for generating a noise correction image is an image obtained by removing other than those selected among the detected noise image from the input image,
An image processing apparatus comprising: The enlargement ratio accepting unit can accept the change of the enlargement ratio,
In response to the enlargement rate accepting unit accepting the change of the enlargement rate,
The image dividing means re-determines the size according to the changed enlargement ratio, re-divides the input image into the divided images based on the re-determined size,
The divided image display means displays one or more designated by the user among the divided images that have been re-divided together with the first mark at the changed magnification.
The divided image cell display means displays the cell group corresponding to the divided image that has been divided again.
The image processing apparatus according to claim 1. The divided image cell display means, among the cell groups, the cell corresponding to the divided image that has been confirmed the presence of an error in the detection of noise by the user, a second display different from the first display form so as to form, and the currently displayed yet not and yet user cells corresponding to the divided image which is not check for error detection of noise by said divided image display unit, the first display form And a third display form different from the second display form .
The image processing apparatus according to claim 1 or 2 wherein. Contained in the entering force image, detects a false noise image is an image of a the suspicious part noise, a pseudo noise image detection means further includes,
The divided image display means, with the previous SL first indicia, and displaying the second mark at the position of the pseudo noise image included in the input image,
The erroneous detection noise selection means allows the user to select one of the detected noise images that is erroneously detected as a noise image by the noise image detection means and one that is a noise image from the suspicious noise image. ,
The noise correction image generation means as the noise correction image, the detected noise images, wherein the other than those selected are-option false and those selected among the noise image is removed from the input image of the image to generate,
The image processing apparatus according to claim 1. The noise image detection means detects the block as a noise image when the block of pixels indicating continuous black included in the input image is less than the first size,
The suspicious noise image detection means, when a block of pixels indicating continuous black included in the input image is equal to or larger than the first size and smaller than the second size, Detect as,
The image processing apparatus according to claim 4. An input image correction method for correcting the input image,
Detecting an image of noise included in the input image;
Receiving an enlargement ratio of the input image from the user;
Determining a size for dividing the input image according to the received enlargement ratio, and dividing the input image into divided images based on the determined size;
Of the divided the divided image, the one or more designated by the user, with the magnification accepted, are included in the designated divided image of the detected noise image is an image of the detected noise With a first mark indicating the location of the object,
A cell group in which cells corresponding to the respective divided images are arranged according to the relative positional relationship between the respective divided images is displayed side by side with the divided images currently displayed by the divided image display means, and the cells Displaying a cell corresponding to the divided image currently displayed by the divided image display means in a first display form different from other cells in the group;
Let the user select what was detected as an image of noise by mistake among the detected noise images included in the displayed divided image,
To generate a noise correction image is an image obtained by removing other than those selected among the detected noise image from the input image,
An input image correction method characterized by the above. When accepting the change of the enlargement ratio,
Re-determine the size according to the changed magnification, and re-divide the input image into the divided images based on the re-determined size,
One or more specified by the user among the re-divided divided images is displayed with the first mark at the changed magnification.
Displaying the cell group corresponding to the re-divided divided image;
The input image correction method according to claim 6. Detecting a suspicious noise image that is an image of a suspicious portion of noise included in the input image,
Together with the first indicia, and displaying the second mark at the position of the pseudo noise image included in the input image,
Is selected to an image noise of the previous SL detecting noise erroneous ones were detected as image noise and the suspected noise image ones of images to a user,
Generating a noise compensated image the other than those selected are-option false and those selected among the noise image is an image obtained by removing from the input image of the detected noise image,
The input image correction method according to claim 6 or 7 . A computer program used in a computer for correcting the input image,
A process of detecting an image of noise included in the input image;
A process of accepting an enlargement ratio of the input image from a user;
Determining a size for dividing the input image according to the received enlargement ratio, and dividing the input image into divided images based on the determined size;
Of divided by said divided image, the one or more designated by the user, with the magnification accepted, are included in the designated divided image of the detected noise image is an image of the detected noise Processing with a first mark indicating the position of the object,
A cell group in which cells corresponding to the respective divided images are arranged according to the relative positional relationship between the respective divided images is displayed side by side with the currently displayed divided image, and the current group of the cell groups is displayed. A process of displaying a cell corresponding to the divided image being displayed in a first display form different from other cells;
A process for causing the user to select an image detected as an image of noise by mistake among the detected noise images included in the displayed divided image;
And generating a noise correction image is an image obtained by removing other than those selected among the detected noise image from the input image,
A computer program for causing a computer to execute. When accepting the change of the enlargement ratio,
Re-determining the size according to the changed magnification, and re-dividing the input image into the divided images based on the re-determined size;
Processing of displaying one or more designated by the user among the divided images that have been re-divided together with the first mark at the changed magnification rate;
Processing to display the cell group corresponding to the divided image that has been re-divided;
Causing the computer to execute
The computer program according to claim 9. Contained in the entering force image, and processing for detecting a noise image false is an image of a the suspicious part noise,
Together with the first indicia, and table Shimesuru processing a second mark at the position of the pseudo noise image included in the input image,
A process of selecting what is the image noise of the previous SL detecting noise erroneous ones were detected as image noise and the suspected noise image ones of images to a user,
And generating a noise corrected image wherein a other than those selected are-option false and those selected among the noise image is an image obtained by removing from the input image of the detected noise image,
Causing the computer to execute
The computer program according to claim 9 or 10 .