JP4136819B2 - Image processing apparatus and program - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention , Painting Image processing device And programs About.
[0002]
[Prior art]
Conventionally, when a document or form image is processed by a character recognition device, if the image is inclined, the recognition rate is adversely affected, and image formation is performed by a digital composite device having functions such as image reading, facsimile reception, and copying. When there is a noise image other than the image, the image formation is adversely affected. In view of this, the inclination of an image is detected, character recognition is performed in consideration of the detected inclination, and a noise image is detected and erased to form an image.
[0003]
As an example of a technique for detecting the inclination of a document image, a corner portion of a document image is selected as a portion where a character / graphic block and a margin block are likely to form a substantially linear boundary, and then this corner corner is selected. In the portion, a circumscribed rectangle circumscribing the character graphic block is detected, and the scanning direction of the horizontal scanning line (one side of the character graphic block) is rotated around the intersection of one side of the detected circumscribed rectangle and one side of the character graphic block. . Then, the maximum rotation angle at which no character / graphic element is detected on the scanning line is detected as an inclination candidate of the document image. There has been an invention in which this is performed on images at four corners of a document image, and the average value is detected as the inclination of the document image (see Patent Document 1).
[0004]
[Patent Document 1]
JP-A-8-320916 (abstract, paragraphs (0027) to (0031), FIG. 7)
[0005]
Also, as an example of a method for eliminating image noise, in a digital copying machine, an original is read by prescanning with a scanner, an analog signal of the read image is converted into a digital signal, and an RGB color component of the digital signal is converted into a CMY color component. Convert to component. From this color component signal, a histogram relating to the gradation distribution is created for each color component, and background color data representing the background color of the document and a threshold value used for region determination processing are determined based on the histogram. When the original is read by the main scan, the pixel data of the pixel belonging to the image area to be copied is not subjected to any processing, and the pixel data of the pixel belonging to the background area is the background color data obtained by the pre-scan and the threshold value. Based on the background color conversion processing, that is, erase. For this reason, the invention that the background dirt and the show-through can be prevented is disclosed (see Patent Document 2).
[0006]
[Patent Document 2]
JP 2001-45297 A (paragraphs (0031) to (0035))
[0007]
[Problems to be solved by the invention]
However, since the invention disclosed in Patent Document 1 detects the inclination of an image using a character graphic block and a margin block, the inclination detection is performed on an original in which the information area covers the entire edge of the original. There were times when it was not possible.
[0008]
In the invention disclosed in Patent Document 2, the image area to be copied and the background area are distinguished by prescanning when copying a document, and the background area is the background color data obtained by prescanning and the threshold value. Since color conversion processing is performed based on the value, when the image reading apparatus reads an original placed under a black pressure plate, for example, if the size of the original is smaller than the size of the reading area, An image area of black pixels corresponding to other than the document may occur, and this image area is not erased.
[0009]
The present invention provides the Affection It was made in view of That My goal is, Correct the skew of original image data even for originals with missing corners That is.
[0010]
[Means for Solving the Problems]
According to the first aspect of the present invention, in the image processing apparatus for correcting the inclination of the document image data included in the image data, end point position information indicating a position where the continuation of the same color pixel values in the image data ends. From the side of the image data Means for obtaining, and change in increase / decrease tendency of coordinate values for the obtained plurality of end point position information Position And detecting the boundary information of the document image data based on the estimated position information of each corner and only a part of the end point position information based on the estimated position information of each corner. And obtaining the boundary information of the original image data from the obtained end point position information, and detecting the inclination of the original image data using the acquired boundary information. Means for correcting the inclination, and the means for acquiring the boundary information detects the boundary information of the document image data only for the end point position information separated from the estimated corner by a predetermined range. It is an image processing apparatus characterized in that it is obtained as end point position information for the purpose.
A second aspect of the present invention is the image processing apparatus according to the first aspect, wherein means for extracting an image area including the original image data with the inclination corrected, and in the extracted image area, An image processing apparatus comprising: means for determining a noise area in a peripheral portion of document image data; and means for replacing a color of the determined noise area.
The invention of claim 3 is the image processing apparatus according to claim 2, further comprising means for acquiring a boundary portion between the document image data and other image data in the extracted image region, The means for determining the noise region is an image processing device that determines a noise region from an end portion of the extracted image region to the boundary portion acquired.
According to a fourth aspect of the present invention, in order to correct the inclination of the document image data included in the image data, the computer is provided with end point position information indicating a position where the continuation of the same color pixel value ends in the image data. From the side of the image data Means for obtaining, and change in increase / decrease tendency of coordinate values for the obtained plurality of end point position information Position And means for estimating the position of the corner of the document image data, and detecting only the end point position information separated from the estimated position of each corner by a predetermined range from the boundary information of the document image data. Obtained as end point position information, means for acquiring boundary information of the original image data from the obtained end point position information, and detecting the inclination of the original image data using the acquired boundary information, and detecting the detected This is a program for functioning as a means for correcting inclination.
The invention according to claim 5 is the program according to claim 4, further comprising: a computer for extracting an image area including the original image data whose inclination is corrected; and the extracted image area The program for causing the noise region in the peripheral portion of the document image data to function as a means for color-replacement of the determined noise region.
The invention according to claim 6 is the program according to claim 5, further causing the computer to function as means for acquiring a boundary portion between the document image data and other image data in the extracted image region. In addition, the means for determining the noise region is a program for determining from the end of the extracted image region to the acquired boundary as a noise region.
[0018]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, extraction of a document image in a digital multi-function peripheral having a FAX function and a copy function according to an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram illustrating a schematic configuration of an image processing unit of a multifunction digital multifunction peripheral according to an embodiment of the present invention. In FIG. 1, 1 is a general-purpose image reading device such as a scanner, and 2 is an image. Central processing unit (CPU) for controlling the extraction process, 3 ROM and NVRAM for storing a program for starting the digital multifunction peripheral and a program for controlling the image extraction process, 4 for reading the image extraction process A RAM for temporarily storing images, 5 is a large-capacity memory (HDD) for storing read images and processed images, and 6 is a CD-ROM or DVD- that is loaded with a program for controlling image extraction. A driver for driving a ROM, etc. 7 is a display device for displaying image extraction processing commands and image extraction statuses, and 8 is for exchanging image data with other personal computers and digital multifunction peripherals via an interface (not shown). That network, 9 is exchanged for bus commands and data between these devices.
[0019]
(Embodiment 1) In this embodiment, the multi-function digital multi-function peripheral is such that a document is a binary image, and a document image input from the image reading device 1 has an image area other than the document at its peripheral edge. When this is done, the boundary noise due to the image area other than the document is eliminated, and a document image that apparently does not include the boundary noise is extracted.
FIG. 2 is a diagram showing a relationship between an image area of a document and an image area other than the document generated in the image area. In FIG. 2, 11 is a document, 12 is a character, figure, photograph, graph recorded on the document 11 Etc. (FIG. 2A). 14 schematically shows an input image read by the image reading device 1 and developed in the RAM 4, 15 is an image area of the document 11, 16 is coloring of a pressure plate (not shown) of the image reading device 1, and the like. This is an image area other than the document generated at the peripheral edge of the document image (FIG. 2B).
[0020]
FIG. 3 is a diagram illustrating a form for extracting a document image. The CPU 3 monitors the generation of the image area 16 on the RAM 4 and detects the inclination of the image area of the document when the image area 16 is generated (FIG. 3A). Then, the image area 15 is rotated in the direction without inclination by the detected inclination (FIG. 3B). From the input image whose inclination is corrected, the CPU 3 cuts out, that is, extracts only the image area of the document (FIG. 3C). At this time, the extracted document image area 15 includes an image area (boundary noise) 17 other than the document included in the document image area. Therefore, the boundary noise 17 is replaced with an appropriate color (FIG. 2D), and an image from which the boundary noise is apparently erased is obtained.
[0021]
Note that specific forms of inclination detection, boundary detection, image extraction, boundary noise detection, and color replacement will be described later.
[0022]
FIG. 4 is a flowchart of processing for extracting a document image from an input image. The processing procedure will be described with reference to the flowchart. In the following description, S indicates a process step. When the image area 16 is generated, the CPU 3 detects the inclination of the image area of the document (S1), and when the inclination occurs, the CPU 3 rotates the image area 16 in the direction without the inclination (S2). An image area of the document is extracted from the input image whose inclination is corrected (S3). Then, an image area other than the document at the document boundary is detected (S4). Then, color replacement of an image area (boundary noise) other than the original included in the image area of the original is performed (S5) and output as an output image.
[0023]
According to the present embodiment, when an image area (boundary noise) 17 other than the document is included around the document image 15, the boundary noise 17 is erased and only the image area 15 that apparently corresponds to the document 11 is extracted. Can do.
[0024]
(Embodiment 2) In this embodiment, the multi-function digital multi-function peripheral is such that the original is a grayscale or color image, and the original image input from the image reading apparatus 1 is an image area other than the original at the peripheral edge. , The boundary noise due to the image area other than the original is erased, and the original image that apparently does not include the boundary noise is extracted. When the original is a multi-value image such as a gray scale or color image, the multi-value image is converted into a binary image by a general-purpose conversion means. After conversion to a binary image, a document image that does not include an image area other than the document is extracted by the same method as that described in the first embodiment.
[0025]
According to the present embodiment, even if the document is a multi-valued image, the image area other than the document is extracted as a binary image, so that extraction can be performed at high speed. At this time, the capacity of the RAM 4 can be small.
[0026]
(Embodiment 3) In the present embodiment, when the multifunction digital multi-function peripheral has a large document image input from the image reading apparatus 1 and has an image area other than the document at the periphery of the image. Then, the boundary noise due to the image area other than the original is deleted, and an original image that apparently does not include the boundary noise is extracted. The CPU 3 monitors the size of the input image developed in the RAM 4, and when determining that the input image exceeds the predetermined size, the CPU 3 reduces the size of the input image to the predetermined size. A document image that does not include an image area other than the document is extracted from the reduced input image by the same method as that described in the first embodiment. At this time, if the document is a binary image, the image conversion is performed without performing image conversion, and if the document is a multi-value image, the image conversion described in the second embodiment is performed. A document image that does not include an image area other than the document is extracted by the same method as that described in the first embodiment.
[0027]
According to the present embodiment, even when the size of the input image 14 is large, it is possible to extract an image area other than a document at high speed without increasing the capacity of the RAM 4.
[0028]
(Embodiment 4) In the present embodiment, the multi-function digital multifunction peripheral determines the inclination of the original image as the boundary information of the image, that is, the position information of the boundary between the image area of the original and the image area other than the original, Detection is performed from position information at the four corners of the image area.
[0029]
FIG. 5 is a diagram showing an input image for explaining a form in which the inclination angle of the document image is detected from the boundary information. In FIG. 5, when the input image 14 is developed in the RAM 4, the CPU 3 determines the side e of the boundary between the image area 15 other than the document on the top and bottom and the left and right of the document image area 15. Position information (coordinate information) is acquired for, f, g, and h. Alternatively, position information (coordinate information) is acquired for the sides connecting the corners a, b, c, and d. Then, the inclination angle of the document image is determined from the gradient in the Y-axis direction of the upper side e and the lower side f of this position information or the gradient in the X-axis direction of the left side g and the right side h.
[0030]
In addition, the inclination angle of the original image is obtained from an inclination histogram created from the character string rectangle disclosed in Japanese Patent No. 3281469, which is filed by the applicant of the present application, or in the read image similarly disclosed in Japanese Patent No. 3338537. A technique can be used in which the coordinates of the black-and-white boundary point at the upper end of the character are a point sequence, a regression line is obtained from the coordinates, and the regression coefficient is obtained from the regression coefficient.
[0031]
Therefore, you may obtain | require comprehensively from the angle of the inclination obtained by each above-mentioned method. That is, the angle is determined from the average value, median value, and probable value of each angle. Each side and the side connecting the four corners are divided into upper and lower and left and right pairs, and the obtained angle is determined as the angle with the smaller difference.
[0032]
According to the present embodiment, the inclination of the original image is determined based on the position information (coordinate information) of e, f, g, h on the upper and lower sides and the left and right sides of the original image area 15, and the four corners a, b , c, d from the position information, it can be obtained easily. Further, since the position information is comprehensively calculated from the average value, the median value, etc., it can be obtained accurately.
[0033]
(Embodiment 5) In this embodiment, the multi-function digital multifunction peripheral sets a search range for obtaining a boundary between an image area of a document and an image area other than the document in advance.
[0034]
FIG. 6 is a diagram illustrating a search range set in advance in the input image. In FIG. 6, in the input image 14, a rectangular range configured by a predetermined width from the upper side to the lower side of the input image (upper search range U), and a rectangle configured by a predetermined width from the lower side of the input image to the upper side. Range (lower search range D), rectangular range configured with a predetermined width from the left side of the input image toward the right side (left search range L), rectangular range configured with a predetermined width from the right side of the input image toward the left side ( The right search range R) is set as a search range for boundary detection. These search ranges may be set by a fixed value corresponding to the document size, or a value dynamically determined corresponding to the document size may be used. When the boundary between the image area of the document and the image area other than the document is obtained, the boundary is not searched for the entire input image but is detected within the set search range.
[0035]
According to the present embodiment, the boundary detection process may be performed within the limited ranges U, D, L, and R, so that the boundary can be detected at high speed without performing useless detection operation.
[0036]
(Embodiment 6) In this embodiment, the multifunction digital multi-function peripheral performs boundary detection based on end point information of continuous colors.
FIG. 7 is a diagram illustrating an input image for explaining a form in which boundary detection is performed. In FIG. 7, the boundary search ranges U, D, L, and R described in the sixth embodiment are set. For example, in the case of a binary image, since the image area other than the original is black, an area where black pixels are continuous from the upper and lower sides and the left and right sides is obtained (thick arrow in FIG. 7A). For example, in the case of obtaining within the left search range L, an area where black pixels continue from the left side of the input image in the right direction of the drawing is obtained, and the information (end point position information) is obtained with the point where the continuity is interrupted as the end point. This is scanned at predetermined intervals from the upper side to the lower side, and boundary detection is performed by continuing the acquired end point position information (see FIG. 7B). Similarly, an area where black pixels continue from the lower side to the upper side, from the left side to the right side, and from the right side to the left side is obtained, and end point position information is obtained. At this time, since white pixels may be generated as noise in an image area other than the original, if the number of white pixels is within a predetermined value, for example, 3 or less, this is ignored and a continuous area of black pixels ( (See FIG. 7C).
[0037]
According to this embodiment, boundary detection can be easily performed.
[0038]
(Embodiment 7) In this embodiment, the multifunction digital multi-function peripheral detects a boundary from position information (corner corner position information) of four corners of an image area of a document.
FIG. 8 is a diagram for explaining a form in which the boundary is detected from the corner position information of the image area of the document. Here, each corner position information of the four corners is obtained from a position where the end point position information greatly changes. For example, the position information of the upper left corner is a location where the increase or decrease of the coordinate values in the X and Y directions change. That is, when the image area of the document is inclined to the lower right, when the end point position information from the lower side is scanned in the right side direction, the coordinate value in the Y direction continues to decrease and starts to increase suddenly at the point Sa corresponding to the upper left corner. To change. Therefore, the CPU 2 obtains the position information with the changing position as the upper left corner. Similarly, other corner corner position information can be acquired. The boundary is detected by the side connecting the corner position information.
[0039]
Also, if the image area of the document is tilted to the lower left, scanning the end point position information from the left side in the upper side direction, the coordinate value in the X direction continues to decrease, and suddenly increases from the point Sa that hits the upper left corner. Change to start. Therefore, the CPU 2 obtains the position information with the changing position as the upper left corner. Similarly, other corner corner position information can be acquired, and the boundary can be detected.
[0040]
Whether the image area of the document is lowering to the right or lowering is detected in advance as a change in coordinate value, for example, +1 when increasing compared to the previous X-axis coordinate value or Y-axis coordinate value, and -1 when decreasing. You just have to.
[0041]
According to this embodiment, the boundary can be easily detected.
[0042]
(Embodiment 8) In this embodiment, the multi-function digital multi-function peripheral is characterized in that the corners of the document may be folded, torn, or missing due to stapling or the like, and the end point near the corners. Boundary detection is performed based on end point position information excluding position information.
[0043]
FIG. 9 is a diagram for explaining a form in which boundary detection is performed excluding end point position information in the vicinity of the corner, where 14a is an input image of the image area 15a of the document lacking the upper left corner. The position information of the missing upper left corner is roughly obtained by obtaining the intersection of the left side and the upper side. Then, end point position information of ranges p, q, r, and s excluding a range of about a quarter from both corners connecting each corner is obtained and detected as a boundary.
[0044]
According to the present embodiment, it is possible to increase the detection accuracy when detecting the boundary of the image area of the document having a missing corner.
[0045]
(Embodiment 9) In this embodiment, the multifunction digital multi-function peripheral extracts an image area corresponding to a document from the image area of the document for which the boundary is determined. This extraction is performed by extracting circumscribed rectangles focusing on the background color and integrating the extracted circumscribed rectangles. The background color in this case is the background color of the image area of the document. In the binary image, since the background color is white, a circumscribed rectangle focusing on white pixels is obtained. Integration is performed so as to include the obtained circumscribed rectangle, and finally the maximum integrated rectangle is determined as an image area of the document, and this image area is extracted. At this time, since a noise region is also assumed, it is possible to perform processing such as rectangles having a predetermined size or less are not integrated as noise.
[0046]
According to the present embodiment, it is possible to easily determine an image region to be extracted.
[0047]
(Embodiment 10) In this embodiment, the multifunction digital multi-function peripheral determines a boundary noise search range in order to detect and delete boundary noise from the extracted image area of the document, and the boundary within the search range. Boundary noise is detected by detecting the circumscribed rectangle in contact with the.
[0048]
When the image area of the original is inclined, the image area other than the original is included in the boundary portion of the image area of the original. Therefore, it is necessary to erase this, but instead of setting the entire image area of the original document extracted for this erasure process as the boundary noise detection target area, the boundary noise search range is determined, and within the search range. The boundary noise is eliminated for the circumscribed rectangle that touches the boundary. The boundary noise search range is a range having a predetermined width from the upper and lower and left and right boundaries of the extracted document image area. The search range is determined in advance by a fixed value corresponding to the image size or a value dynamically determined according to the image information, for example, a value obtained by calculation based on the width, height, and the like.
[0049]
According to the present embodiment, since a circumscribed rectangle existing within a predetermined range is detected, boundary noise can be detected and eliminated at high speed.
[0050]
(Embodiment 11) In this embodiment, the multi-function digital multi-function peripheral checks the change in the run length of the black pixel again without determining the extracted circumscribed rectangle in contact with the boundary as the boundary noise. Distinguish between boundary noise and information area at large points.
[0051]
FIG. 10 is a diagram illustrating a relationship between a document for explaining why a circumscribed rectangle in contact with a boundary is not directly determined as boundary noise and a circumscribed rectangle in contact with the boundary. Recently, some magazines or the like do not have a margin around the document, and some documents have a photo laid out to the edge of the document sheet. FIG. 10A shows a document in which an information area 30 such as characters, figures, photographs, tables, and graphs is laid out to the upper left end of the document sheet and the left end. FIG. 10B shows a state where the boundary noise 17 and the information area 30 are in contact with each other when such a document is detected as a circumscribed rectangle, and the information area 30 and the boundary noise 17 must be separated forcibly. The information area 30 may be erased as the boundary noise 17. Therefore, the change in the run length of the black pixel is examined, and the pixel is forcibly separated at the place where the change is large.
[0052]
FIG. 11 is a diagram for explaining a mode for distinguishing between the information area and the boundary noise. In FIG. 11, the CPU 3 determines whether or not the circumscribed rectangular information area 30 is in contact with the boundary noise 17. When the CPU 3 determines that the information is in contact with the boundary noise 17, Start investigating changes in the continuous area (run length). This check checks the number of consecutive black pixels to scan from the upper left corner to the right and further down. At this time, since the run length of the m-1 line and the run length of the m line change abruptly, the position of this change point is obtained, and the information area 30 and the boundary noise 17 are obtained at this position. Isolate.
[0053]
According to the present embodiment, even if the boundary noise is left unerased, it is considerably less likely that the information area is erased by mistake. Therefore, even in the case of a document having no margin around the document, the document image can be extracted without erasing the information area.
[0054]
(Embodiment 12) In this embodiment, the multifunction digital multi-function peripheral performs color replacement for pixels detected as boundary noise. This color may be white, black or other color specified by the user. Alternatively, it can be output as it is without color replacement. Of course, when the user does not specify a color, white can be used.
[0055]
According to the present embodiment, since the replacement color can be arbitrarily determined, the degree of boundary noise erasure can be controlled.
[0056]
(Embodiment 13) In this embodiment, when performing color replacement on a pixel detected as boundary noise, the multifunction digital multi-function peripheral replaces it with the nearest background color existing around the pixel. As the background color, there is a method of obtaining a color histogram and setting the most colors as the background color. At this time, a color closest to the background color may be found from surrounding pixels and replaced with that color. Since the color is slightly different depending on the position of the image, it is replaced with a surrounding color in order to replace it with the most natural background color. Of course, the user may be allowed to decide whether to execute such processing.
[0057]
According to this embodiment, the boundary noise is replaced with a color closest to the background color of the surrounding pixels, so that the image area of the document can be extracted as it is.
[0058]
(Embodiment 14) In order for a computer of a general multifunction digital multi-function peripheral to function as means for performing the image processing described in Embodiments 1 to 13, the procedure is described as a computer program in a general-purpose program language. In addition, the program is recorded on an arbitrary recording medium such as a flexible disk, CD-ROM, DVD-ROM, MO, etc., and the image processing according to the present invention is performed by causing a computer of a general digital multi-function peripheral to read the program. It can be easily implemented. In addition to recording the program on a recording medium, the program can be directly read by the computer of the digital multi-function peripheral via an arbitrary network such as the Internet or an intranet.
[0059]
The embodiments described above are based on the premise that the present invention is implemented in a multi-function digital multifunction peripheral having a FAX function and a copy function, but a digital multifunction peripheral and a scanner for processing image information received from a personal computer. You may implement in the image processing apparatus which consists of a combination of a personal computer.
[0060]
【The invention's effect】
Of the present invention effect: Corrected skew of original image data even for originals with missing corners can do.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a schematic configuration of an image processing unit of a multifunction digital multifunction peripheral according to an embodiment of the present invention.
FIG. 2 is a diagram illustrating a relationship between an image area of a document and an image area other than the document generated in the image area.
FIG. 3 is a diagram illustrating a form for extracting a document image.
FIG. 4 is a flowchart of processing for extracting a document image from an input image.
FIG. 5 is a diagram illustrating an input image for explaining a form in which a tilt angle of a document image is detected from boundary information.
FIG. 6 is a diagram illustrating a search range set in advance in an input image.
FIG. 7 is a diagram illustrating an input image for explaining a form for performing boundary detection;
FIG. 8 is a diagram illustrating a form in which a boundary is detected from corner position information of an image area of a document.
FIG. 9 is a diagram illustrating a form in which boundary detection is performed except for end point position information in the vicinity of a corner.
FIG. 10 is a diagram illustrating a relationship between a document for explaining why a circumscribed rectangle in contact with a boundary is not determined as boundary noise as it is, and a circumscribed rectangle in contact with the boundary.
FIG. 11 is a diagram illustrating a form for distinguishing between an information area and boundary noise.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Scanner, 2 ... CPU3, 3 ... ROM, 4 ... RAM, 5 ... Large scale memory, 6 ... CD-ROM driver, 7 ... Display apparatus, 8 * ..Network connection line, 9 ... bus control line

Claims (6)

In an image processing apparatus for correcting the inclination of document image data included in image data,
Means for obtaining end point position information indicating a position where the continuation of pixel values of the same color in the image data ends , from the side of the image data ;
Means for estimating position information of corners of the document image data from positions at which the tendency of increase / decrease in the coordinate values of the plurality of end point position information obtained is changed;
Based on the estimated position information of each corner, only a part of the end point position information is obtained as end point position information for detecting boundary information of the document image data, and the original is determined from the obtained end point position information. Means for obtaining boundary information of image data;
Means for detecting the inclination of the document image data using the acquired boundary information and correcting the detected inclination;
With
The means for acquiring the boundary information obtains only the end point position information separated from the estimated corner by a predetermined range as end point position information for detecting the boundary information of the document image data. Image processing device. The image processing apparatus according to claim 1,
Means for extracting an image area including the original image data with the inclination corrected;
Means for determining a noise area in a peripheral portion of the document image data in the extracted image area;
Means for color replacement of the determined noise region;
An image processing apparatus comprising: The image processing apparatus according to claim 2,
Means for obtaining a boundary between the original image data and other image data in the extracted image region;
With
The image processing apparatus characterized in that the means for determining the noise region determines a noise region from an end of the extracted image region to the acquired boundary portion. In order to correct the inclination of the document image data included in the image data,
Means for obtaining end point position information indicating a position where the continuation of pixel values of the same color in the image data ends , from the side of the image data ;
From changing the position of the increasing or decreasing trend of coordinate values according to the plurality of the terminal position information obtained, and means to estimate the angular corner of the original image data,
Only the end point position information that is separated from the estimated position of each corner by a predetermined range is obtained as end point position information for detecting boundary information of the document image data, and the document image is obtained from the obtained end point position information. Means for obtaining boundary information of the data;
A program for detecting an inclination of the document image data using the acquired boundary information and causing the detected inclination to be corrected. The program according to claim 4,
In addition, the computer
Means for extracting an image area including the original image data with the inclination corrected;
Means for determining a noise area in a peripheral portion of the document image data in the extracted image area;
A program for causing a determined noise region to function as a color replacement unit. The program according to claim 5,
Computer
In the extracted image area, while functioning as a means for obtaining a boundary portion between the document image data and other image data,
The program for determining the noise area is a means for determining a noise area from an end of the extracted image area to the acquired boundary.

Images