JP2012019474A - Image reading device and image correction program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image reading device which can correct a distorted image with an appropriate correction amount, and to provide an image correction program.SOLUTION: The image reading device comprises: a correction preprocessing unit 61 which generates distortion correction image data by performing first correction on medium image data generated when the surface image of a medium having a predetermined shape is read in distorted shape, extracts the contour line from the distortion correction image data, and then corrects the contour line with a plurality of correction factors having different values; a display processing unit 62 which generates the images of correction factor outside contour line corrected with respective correction factors, and displays a list of the images on a display unit 7 so that the images can be selected for each correction factor; and an image correction unit 63 which performs second correction on the distortion correction image data so that the contour line extracted from the distortion correction image data matches the correction factor outside contour line which the user inputs selectively to an operation input unit 8.

Description

  The present invention relates to an image reading apparatus, and more particularly to an image reading apparatus and an image correction program that read a medium having a predetermined shape, generate image data, and perform distortion correction on a distorted image having a distorted shape. .

When reading a spread page of a booklet-like document with a scanner (for example, a flat bed scanner), there is a problem that characters and drawings are distorted in the page binding portion of the read image. there were.
Therefore, an image reading apparatus that extracts the outer shape (outline) of a document from the read image, obtains the degree of distortion, and expands the image based on this degree to correct the distortion of the page binding portion. It is disclosed.
For example, using the distortion correction rate (expansion rate) calculated according to the difference in the distance between the outline of the document (contour line), a reference line such as a character string or ruled line written on the document, and the center line of the image, There is an image reading device that expands an image in a certain region and corrects image distortion (Patent Document 1).
As described above, the conventional image reading apparatus calculates one value of the degree of distortion (distortion correction rate) by a predetermined method, and corrects the distortion of the page binding portion based on the distortion correction rate.

JP 2008-252945 A

However, since the conventional image reading apparatus corrects the distortion of the page binding portion based on the calculated one distortion correction rate, each user's request to be used is not reflected.
Therefore, in the correction performed by the conventional image reading apparatus, there is a problem that the correction amount is insufficient or the correction amount is excessive depending on the user to use.

  SUMMARY An advantage of some aspects of the invention is that it provides an image reading apparatus and an image correction program capable of correcting a distorted image at an appropriate correction rate.

  In order to solve the above problems, an image reading apparatus according to the present invention includes an input unit that receives a selection input from a user and a display unit, and reads a surface image of a medium having a predetermined shape to generate medium image data. An image reading apparatus that corrects distortion of the medium image data generated in a shape in which the medium surface is distorted when read, and includes a pre-correction processing unit, a display processing unit, and an image correction unit. did.

According to the present invention, a distorted image can be corrected with an appropriate correction rate. That is, by generating image data for the corrected contour line for each correction rate, making each selectable and displaying the list on the display unit, the right to select the correction rate can be given to the user.
In addition, by displaying the corrected contour line performed at each correction rate on the display unit, the user can easily determine the desired corrected contour line.
Further, by correcting the medium image data so that the contour line of the medium extracted from the medium image data matches the contour line of the correction factor selected by the user, the medium is determined based on the correction factor selected by the user. The distorted shape can be corrected.
In this way, the user can select an appropriate correction amount with a simple operation.

1 is an external view of an image reading apparatus according to the present invention. 1 is a block diagram of an overall configuration of an image reading apparatus according to the present invention. 1 is a block diagram of an image processing system included in an image reading apparatus according to a first embodiment. It is a figure which shows (a) read image data D1, (b) distortion correction image data D2. (A) It is a figure which shows the straight line contained in the read external shape data S1, and the straight line contained in (b) distortion correction external shape data S2. It is a figure which shows the curve which approximates an edge pixel. It is a figure which shows matching with the structural point P1 and the structural point P2 on a straight line. FIG. 8 is a continuation of FIG. 7 and shows a correspondence between the configuration point P1 and the configuration point P2 on the curve. FIG. 9 is a continuation of FIG. 8 showing a correspondence between the configuration point P1 and the configuration point P2 on the curve. It is a figure which shows an example in which the partial image was displayed according to the correction factor on the display panel. It is a figure which shows an example of the outline (outer shape) of correction factor outer shape data S3. It is a figure which shows an example of the relationship between the position of the virtual point U3, and the position of the estimation point U2. 3 is a flowchart illustrating an overall processing operation of the image reading apparatus according to the first embodiment. It is a flowchart which shows the correction preparation process operation | movement of a correction preparation part. It is a flowchart which shows the external shape extraction process operation | movement of an external shape extraction part. It is a flowchart which shows the component point arrangement | positioning process operation | movement of a component point arrangement | positioning part. It is a flowchart which shows the correspondence construction processing operation | movement of a correspondence construction part. It is a flowchart which shows the secondary correction process operation | movement of a secondary correction | amendment part. It is a flowchart which shows the correction image generation process operation | movement of a correction image generation part. It is a block diagram of the image processing system which the image reading apparatus which concerns on 2nd Embodiment has. It is a figure which shows an example in which the partial image was displayed according to the correction factor on the display panel. It is a flowchart which shows the secondary correction process operation | movement of a secondary correction | amendment part. It is a flowchart which shows the correction | amendment shape evaluation process operation | movement of a correction | amendment shape evaluation part. It is a figure which shows an example of the correction factor external shape data S3.

Hereinafter, embodiments of the present invention will be described with reference to the drawings as appropriate, taking as an example the case where the present invention is applied to a scanner and a book or booklet is applied to a medium as a document.
First, an image reading apparatus according to the present invention will be described. FIG. 1 shows an external view of the image reading apparatus 1, and FIG. 2 shows a block diagram of the overall configuration of the image reading apparatus 1.
As shown in FIG. 1, the image reading apparatus 1 opens a document 91 such as a book or a booklet, covers a medium placing table 81 placed so as to be in contact with the reading surface of the document 91, and the medium placing table 81. A mounting table cover 82, an image sensor 83 (image reading unit 2) for reading the reading surface of the document 91, an operation button 84 (operation input unit 8) and a start button 85 (operation input unit 8) for a user to input instructions. And a display panel 86 (display unit 7) for displaying information to the user.

<< First Embodiment >>
As shown in FIG. 2, the image reading apparatus 1 according to the first embodiment generates an image data from the read data, and an image reading unit 2 that reads the original 91 by the image sensor 83 and generates read data. An image data generation unit 3, a CPU (Central Processing Unit) 4 that controls each component included in the image reading apparatus 1, and a program ROM (Read Only Memory) 5 that stores a program for the CPU 4 to control A RAM (Random Access Memory) 6 for storing data, a display unit 7 having a display panel 86, an operation input unit 8 having an operation button 84 and a start button 85, and generated image data are transferred to a telephone line (not shown). An I / F (interface) unit 9 connected to an external device (PC (Personal Computer), server, etc.) via a network such as a LAN or a Local Area Network (LAN) Is provided.

Next, an image processing system included in the image reading apparatus according to the first embodiment of the present invention will be described. FIG. 3 is a block diagram of the image processing system.
The image data generation unit 3 generates image data (read image data D1) from the read data obtained by the image reading unit 2 reading the document 91. Then, the image data generation unit 3 stores the read image data D1 in the RAM 6 (image data storage unit 11).
The image processing unit 10 functioning by the CPU 4 acquires the read image data D1 stored in the image data storage unit 11, executes the image processing, displays the image data after the image processing on the display unit 7, and from the user. Is input from the operation input unit 8, and image data corrected based on the instruction is output.

[Image processing unit 10]
The image processing unit 10 includes a primary correction unit 12 (pre-correction processing unit 61), a correction preparation unit 13 (pre-correction processing unit 61), and a secondary correction unit 14 (display processing unit 62 and image correction unit 63). Prepare.

[Image data storage unit 11]
The image data storage unit 11 is provided in the RAM 6 and is read image data D1 (FIG. 4A) from the image data generation unit 3 and distortion corrected image data D2 (FIG. 4B) from the primary correction unit 12 described later. )).

[Primary correction unit 12]
The primary correction unit 12 acquires the read image data D1 from the image data storage unit 11, performs a conventional distortion correction process (first correction) on the read image data D1, and generates distortion correction image data D2. It is a correction unit. The distortion corrected image data D2 is stored in the image data storage unit 11.
The conventional distortion correction process is, for example, the distortion correction process described in Patent Document 1.

[Correction preparation unit 13]
The correction preparation unit 13 extracts the outer shape (outline) of the document 91 from the distortion corrected image data D2 by executing correction preparation processing on the read image data D1 and the distortion corrected image data D2 (distortion corrected outer shape data). S2) is a component that generates correction factor outline data S3.
The correction preparation unit 13 includes an outer shape extraction unit 21, a component point arrangement unit 22, a correspondence relationship construction unit 23, a correction rate setting unit 24, and a shape correction unit 25.

(Outer shape extraction unit 21)
The outer shape extraction unit 21 acquires the read image data D1 and the distortion-corrected image data D2 from the image data storage unit 11, executes the outer shape extraction process, and reads the read outer shape data S1 (read) from the read image data D1. (Outline outline) is extracted, and the distortion correction outline data S2 (distortion correction outline outline) is extracted from the distortion correction image data D2.
Then, the outer shape extraction unit 21 outputs the read outer shape data S1 and the distortion correction outer shape data S2 to both the component point arrangement unit 22 and the correction factor setting unit 24.

Here, the outer shape extraction process will be described.
The outer shape extraction unit 21 extracts pixels (outer shape pixels Mz ₁ ) that form the outer shape (outline) of the document 91 from the read image data D1, and reads all the extracted outer shape pixels Mz ₁ as read outer shape data S1. (Reading outline). Then, the outer shape extraction unit 21 outputs the read outer shape data S1 to both the component point arrangement unit 22 and the correction factor setting unit 24.
Similarly, profile-shape extractor 21, distortion correction also extracts pixels that form the outer shape of the original 91 (the contour lines) (external shape pixel Mz ₂₎ from the image data D2, all extracted contour pixels Mz ₂ Is output to both the component point arrangement unit 22 and the correction factor setting unit 24.

(Composition point arrangement part 22)
The component point arrangement unit 22 acquires the read outer shape data S1 and the distortion correction outer shape data S2 from the outer shape extraction unit 21, and executes the component point arrangement process on the read outer shape data S1 and the distortion correction outer shape data S2. It is.
By this component point arrangement processing, the component point arrangement unit 22 performs coordinate conversion of each pixel (outer shape pixel Mz ₁ ) that forms the outer shape (outline) of the document 91 from the read outer shape data S1, and then on the plane coordinates. placing the configuration point P1 to the position of the outer shape pixels Mz _1. The constituent points P1 may be arranged at a predetermined interval. Accordingly, as shown in FIG. 7, the configuration points P1 are arranged as P ₁₁ , P ₁₂ ,..., P _1a , P _1b,.
Then, the configuration point arrangement unit 22 assigns an identification number to each configuration point P1, and associates the identification number of the configuration point with the coordinate value of the configuration point and stores it in the outer shape storage unit 26 described later. At this time, the identification number assigned by the component point arrangement unit 22 is a number that does not overlap each other and determines one component point.

Similarly, the component point arrangement unit 22 performs coordinate conversion of each pixel (outer shape pixel Mz ₂ ) that forms the outer shape (outline) of the document 91 from the distortion corrected outer shape data S2 to obtain an outer shape pixel on a plane coordinate. placing the configuration point P2 to the position of the Mz _2. Accordingly, as shown in FIG. 7, the configuration points P1 are arranged as P ₂₁ , P ₂₂ ,..., P _2a , P _2b,.
Then, the configuration point placement unit 22 assigns an identification number to each configuration point P2, and associates the identification number of the configuration point with the coordinate value of the configuration point and stores them in the outer shape storage unit 26 described later. This association is performed using a database.
Hereinafter, unless otherwise specified, the association is performed using a database.

(Correspondence relationship construction unit 23)
The correspondence construction unit 23 acquires the configuration point P1 and the configuration point P2 from the outer shape storage unit 26 described later, and executes the correspondence construction process.
By this correspondence construction processing, the correspondence construction unit 23 converts all the constituent points P1 constituting the contour of the read contour data S1 (S1L) to the distortion-corrected contour data S2 (S2L) as shown in FIG. Each of the constituent points P2 constituting the outline is associated with each of the constituent points P2.
Here, the identification number assigned to the configuration point P1 and the identification number assigned to the configuration point P2 corresponding to the configuration point P1 are associated with each other and stored in the outer shape storage unit 26 to be described later. Points are associated.

(Correction rate setting unit 24)
The correction rate setting unit 24 sets the state of the read outer shape data S1 to the correction rate R = 0.0 and the state of the distortion correction outer shape data S2 to the correction rate R = 1.0, and sets the correction rate R at a predetermined interval. It is a component that sets a plurality of correction factors R whose values are changed.
For example, if the predetermined interval is “0.2”, the correction rate setting unit 24 sets the correction rate R as “..., 0.6, 0.8, 1.0, 1.2, 1.4, 1.6, 1.8,... ”Are set.

(Shape correction unit 25)
The shape correction unit 25 is a component that corrects the distortion correction outer shape data S2 with each value obtained by changing the correction rate R and generates a plurality of correction rate outer shape data S3 (for each correction rate R).
Then, the shape correcting unit 25 stores the read outer shape data S1, the distortion corrected outer shape data S2, and the correction factor outer shape data S3 in the RAM 6 (outer shape storage unit 26).
Here, the correction performed by the shape correction unit 25 is performed using field morphing, which is known two-dimensional morphing.

[Outer shape storage unit 26]
The external shape storage unit 26 is a storage unit that is provided in the RAM 6 (FIG. 2) and stores data generated by the correction preparation unit 13. According to the first embodiment, the following data (1) to (3) are stored.
(1) Identification Number and Coordinate Value of Configuration Point P1 (P2) In the external shape storage unit 26, the configuration point placement unit 22 associates each identification number of the configuration point P1 with the coordinate value of the configuration point P1. Is memorized. Further, the identification number of each constituent point P2 and the coordinate value of the constituent point P2 are stored in association with each other.
These identification numbers are numbers that do not overlap each other and determine one component point.
(2) Data correlating the configuration point P1 and the configuration point P2 Further, in the external shape storage unit 26, the identification number given to the configuration point P1 by the correspondence construction unit 23 and the configuration point corresponding to the configuration point P1 The identification number assigned to P2 is stored in association with it.
(3) Reading outline data S1, distortion correction outline data S2, and correction factor outline data S3
The outer shape storage unit 26 stores read outer shape data S1, distortion corrected outer shape data S2, and correction factor outer shape data S3 by the shape correcting unit 25.
Here, in the outer shape storage unit 26, the outline of the read outer shape data S1 is formed in the read image data D1 of the image data storage unit 11 as the read outer shape data S1, after the correspondence construction processing (S206 in FIG. 14). The identification numbers of all the constituent points P1 are stored.
Also, in the outer shape storage unit 26, the shape correction unit 25 forms a contour line of the distortion corrected outer shape data S2 in the distortion corrected image data D2 of the image data storage unit 11 as the distortion corrected outer shape data S2. The identification numbers of all the constituent points P2 later (S206 in FIG. 14) are stored.
Further, in the outer shape storage unit 26, the shape correction unit 25 corrects the coordinate value of the configuration point P2 forming the contour line of the distortion corrected outer shape data S2 as the correction rate outer shape data S3 with the correction rate R. The coordinate value of the point P2 is stored.

[Secondary correction unit 14]
The secondary correction unit 14 is a component that performs secondary correction processing, and includes a feature region extraction unit 27 (display processing unit 62), a display control unit 28 (display processing unit 62), and an input processing unit 29 (image correction). Unit 63) and a corrected image generation unit 30 (image correction unit 63).

(Feature region extraction unit 27)
The feature region extraction unit 27 acquires the read outer shape data S1, the distortion correction outer shape data S2, the correction factor outer shape data S3, the read image data D1, and the distortion correction image data D2, and executes the feature region extraction process. Then, a region (feature region) having a predetermined size with a high rate of change between the contour line of the read outer shape data S1 and the contour line of the distortion corrected outer shape data S2 is extracted. And it is a component which extracts the image (partial image) which falls in a characteristic area from all the correction factor external shape data S3.

(Display control unit 28)
The display control unit 28 is a component that displays a plurality of partial images on the display unit 7.
FIG. 10 shows an example in which the display control unit 28 causes the display panel 86 (display unit 7) to display partial images for each correction rate R. J1 is the correction rate R = 1.0, and J2 is the correction rate R =. 1.2, J3 indicates a partial image with a correction factor R = 1.4, and J4 indicates a correction factor R = 1.6.

(Input processing unit 29)
The input processing unit 29 is a configuration unit that acquires selection information input from the operation input unit 8.
For example, when the user operates the operation button 84 (FIG. 1, operation input unit 8) and selects one partial image displayed on the display panel 86 (FIG. 1, display unit 7), the input processing unit 29 is selected. Acquires selection information of the partial image.

(Correction image generation unit 30)
Based on the selection information, the corrected image generation unit 30 acquires correction rate outer shape data S3 of the correction rate R corresponding to the selected partial image (feature region) from the outer shape storage unit 26, and executes correction image generation processing Part.
First, the correction image generation unit 30 rasterizes the correction factor outer shape data S3 and acquires the contour line of the correction factor outer shape data S3 (FIG. 11). This rasterization process is performed using a known algorithm. Here, the contour correction factor profile data S3L, a line surrounding through _{D3 a ~D3 b ~D3 d ~D3 e} ~D3 a shown in FIG. 11.

Then, the corrected image generation unit 30 corrects the nearest neighbor from the virtual point U3 located in the region surrounded by the outline of the correction factor outline data S3 (S3L) by the corrected image generation process as shown in FIG. From the component point P3 (P _3n : n is an integer) of the rate contour data S3 and the component point P2 (P _2n : n is an integer) of the contour line of the distortion-corrected contour data S2 associated with the component point P3, An estimated point U2 (u _2n : n is an integer) that is estimated to be in the distortion corrected image data D2 (D2L) is calculated before the virtual point U3 is corrected. Then, the pixel value of the pixel corresponding to the position of the virtual point U3 (u _3n : n is an integer) is corrected with the pixel value of the pixel of the distortion corrected image data D2 corresponding to the position of the estimated point U2.

  The correction image generation unit 30 performs the above correction image generation processing on all points (virtual points U3) located in the region surrounded by the outline of the correction factor outer shape data S3, so that the correction image generation unit 30 corrects the correction factor outer shape data S3. The pixel value of the pixel (coordinates) in the region surrounded by the contour line can be corrected (second correction) with the pixel value of the pixel of the distortion corrected image data D2.

<Overall Processing of Image Reading Apparatus>
Next, the overall processing operation of the image reading apparatus 1 will be described with reference to FIG. 13 (refer to FIGS. 1 to 12 as appropriate).
First, the image reading unit 2 reads the surface image of the document 91 placed on the medium placement table (step S101). The image data generation unit 3 generates read image data D1 (FIG. 4A) from the read data (step S102), and stores the read image data D1 in the image data storage unit 11 (step S103).

  Next, the primary correction unit 12 included in the image processing unit 10 acquires the read image data D1 from the image data storage unit 11, executes a conventional distortion correction process (step S104), and generates distortion correction image data D2. . The distortion corrected image data D2 (FIG. 4B) is stored in the image data storage unit 11 (step S105).

  The correction preparation unit 13 acquires the read image data D1 and the distortion corrected image data D2 from the image data storage unit 11, and executes correction preparation processing (step S106). By this correction preparation process, the distortion correction outline data S2 obtained by extracting the outline shape (outline) of the document 91 from the distortion correction image data D2 and the generated correction factor outline data S3 are stored in the outline storage unit 26.

The secondary correction unit 14 acquires the distortion correction external shape data S2 and the correction factor external shape data S3 from the external shape storage unit 26, further acquires the distortion correction image data D2 from the image data storage unit 11, and performs secondary correction. Processing is executed (step S107).
By this secondary correction process, the secondary correction unit 14 displays the correction factor outline data S3 generated by the correction preparation unit 13 on the display unit 7, and the outline data (correction) selected by the user via the operation input unit 8. The corrected image data is generated from the rate contour data S3).
Then, the secondary correction unit 14 outputs the corrected image data to an external device (such as a PC or a server) via a storage unit such as an HDD or the I / F unit 9 (step S108). Then, the image reading apparatus 1 ends the process.

<Correction preparation process>
Next, the correction preparation processing operation (step S106 in FIG. 13) of the correction preparation unit 13 will be described with reference to FIG. 14 (refer to FIGS. 1 to 13 as appropriate).
The outer shape extraction unit 21 included in the correction preparation unit 13 acquires the read image data D1 and the distortion corrected image data D2 from the image data storage unit 11 (step S201). Then, the outer shape extraction unit 21 first performs outer shape extraction processing on the read image data D1 (step S202). Thereby, the outer shape extraction unit 21 acquires the read outer shape data S1. Then, the acquired reading outline data S1 is output to both the component point arrangement unit 22 and the correction factor setting unit 24.

Here, a case where the outer shape extraction unit 21 performs the outer shape extraction processing operation (step S202 in FIG. 15) on the read image data D1 will be described with reference to FIG. 15 (refer to FIGS. 1 to 14 as appropriate). ).
<External shape extraction processing>
The outer shape extraction unit 21 performs a smoothing process on the read image data D1 (step S301), and further applies a differential filter to the smoothed read image data D1 to detect an edge. Then, the pixels on the edge are extracted (step S302).
Here, the smoothing process executed by the outer shape extraction unit 21 is executed by a known smoothing algorithm. In addition, the process of applying the differential filter to detect the edge is executed by a filter process using a known Laplacian filter.

Then, the outer shape extraction unit 21 performs binarization processing on the pixels on the edge (step S303). At this time, a pixel having a pixel value larger than the linear pixel value is extracted as compared with a predetermined linear pixel value (threshold value), and a pixel after an isolated point is removed from all the extracted pixels is defined as an edge pixel. It should be noted that isolated points are removed using a known algorithm.
Then, all the straight lines L are detected using the coordinates of the edge pixels (step S304). Here, the detection of the straight line L is executed by a known Hough transform algorithm. Thereby, the external shape extraction part 21 detects a straight line (L11, ..., L17, ...) from the read image data D1 of Fig.4 (a).
Then, the outer shape extraction unit 21 sets the straight line (L11) closest to the center of the image as the page boundary line in all the detected straight lines L (sets the page boundary line) (step S305).

Then, the outer shape extraction unit 21 divides the image at the page boundary line L11 (D1L and D1R) (step S306).
The outer shape extraction unit 21 compares the number of straight pixels out of all the straight lines L extracted in step S305 from one divided image (D1L) with a number equal to or greater than the number of straight pixels. A straight line L including the edge pixel (passing the coordinates of the edge pixel) is extracted (step S307). As a result, six straight lines L from straight lines L12 to L17 are extracted.
Then, the outer shape extraction unit 21 rasterizes the straight lines and curves that form the contour line of one page, and extracts the pixels (outer shape pixel Mz ₁ ) that form the contour line (step S308).

Hereinafter, following the procedure of extracting the outer shape pixels Mz ₁ in step S308. FIG. 6 is a diagram illustrating a curve that approximates an edge pixel.
First, the outer shape extraction unit 21 performs a process of extracting pixels on a distorted curve. The outer shape extraction unit 21 sets the edge pixel closest to the page boundary line L11 as the point Pc1 among the edge pixels on the straight line L14. In addition, in the edge pixel on the page boundary line L11, the edge pixel closest to the straight line L14 is set as a point Pc2. The intersection of the page boundary line L11 and the straight line L14 is set as a point Pc0.
Next, a triangle having points Pc0, Pc1, and Pc2 as vertices is formed, edge pixels in the triangular region are extracted, and a curve Lc1 that approximates these edge pixels is represented by a known minimum two Obtain using multiplication.
Then, the intersection point D1 _d between the acquired curve Lc1 and the page boundary line L11 and the intersection point D1 _c between the straight line L14 are acquired. The curve Lc1 between the intersection points D1 _{c to} D1 _d is rasterized to extract pixels forming the curve Lc1 (extraction pixel A: a part of the external shape pixel Mz ₁ ).

Similarly, the outer shape extraction unit 21 sets the edge pixel closest to the page boundary line L11 as the point Pc4 (boundary point) among the edge pixels on the straight line L12. In addition, in the edge pixel on the page boundary line L11, an edge pixel closest to the straight line L12 is set as a point Pc5 (boundary point). The intersection of the page boundary line L11 and the straight line L12 is defined as a point Pc3.
Next, a triangle having vertices of point Pc0, point Pc1, and point Pc2 is formed, edge pixels within the triangular region are extracted, and a curve Lc2 that approximates these edge pixels is represented by a known minimum two Obtain using multiplication.
Then, an intersection point D1 _e between the acquired curve Lc2 and the page boundary line L11 and an intersection point D1 _f between the straight line L12 are acquired. By rasterizing curves Lc2 between the intersections D1 _e ~ D1 _f, extracts pixels that form a curve Lc2 (extracted pixel B: part of the outer shape pixels Mz _1).

The outer shape extraction unit 21 acquires an intersection D1 _a between the straight lines L12 and L13 and an intersection D1 _b between the straight lines L13 and L14. Next, the intersections D1 _{d to} D1 _e on the page boundary line L11 are rasterized, and pixels forming a line segment between the intersections D1 _{d to} D1 _e are extracted (extraction pixel C: part of the outer shape pixel Mz ₁ ). ).
Further, the line segments are rasterized between the intersection points D1 _{a to} D1 _f on the straight line L12, between the intersection points D1 _{a to} D1 _b on the straight line L13, and between the intersection points D1 _{b to} D1 _c on the straight line L14. Are extracted (extracted pixel D: part of the outer shape pixel Mz ₁ ).
As described above, the outer shape extraction unit 21 extracts pixels on the line segment and the curve (outer shape pixel Mz ₁ ) that form the outline of one page by the extraction pixels A to D.

  Next, the outer shape extraction unit 21 similarly performs the processing of step S307 to step S308 for the other divided image (D1R), and the outer shape extraction processing (FIG. 14) for the read image data D1. Step S202), the process of step S203 (FIG. 14) is executed.

  For example, even if a page (photo G in FIG. 4A) on which a photo, a frame, a straight line, or the like is loaded is read, the outer shape extraction unit 21 can display the photo G as shown in FIG. The frame is extracted as a straight line (L181 to L184). Then, the outer shape extraction unit 21 determines, by the outer shape extraction process, that none of the straight lines is the straight line closest to the center of the image, and further determines that the straight line does not form the outer shape in the divided image. It becomes.

Returning to the description of FIG.
Next, the outer shape extraction unit 21 similarly executes outer shape extraction processing on the distortion corrected image data D2 (step S203, FIG. 14). Thereby, the external shape extraction part 21 acquires distortion correction external shape data S2. Then, after the outer shape extraction unit 21 finishes the outer shape extraction processing on the distortion corrected image data D2, the process of step S204 is executed next.

Here, the process which performed the external shape extraction process to distortion correction image data D2 is demonstrated.
The outer shape extraction unit 21 detects straight lines (L21,..., L27,...) From the distortion corrected image data D2 of FIG. Then, L21 is set as a page boundary line. Also, profile-shape extractor 21 sets the linear L22~L27 to the outer shape straight Lz _2, further sets the edge pixels on the contour straight line Lz ₂ to the outer shape pixel Mz _2. Then, a page boundary line L21, the outer shape straight Lz _2, the outer shape pixel Mz _2, the distortion correction profile data S2. Then, the distortion correction outer shape data S <b> 2 is output to both the component point arrangement unit 22 and the correction factor setting unit 24.

  Next, the component point arrangement unit 22 acquires the read outer shape data S1 and the distortion-corrected outer shape data S2 from the outer shape extraction unit 21, and first executes the component point arrangement process on the read outer shape data S1 (step S204). .

Here, with reference to FIG. 16, the case where the component point arrangement unit 22 performs the component point arrangement processing operation (step S204 in FIG. 16) on the read outline data S1 will be described (refer to FIGS. 1 to 14 as appropriate). ).
<Composition point arrangement processing>
The component point arrangement unit 22 extracts the outer shape pixel Mz ₁ from the read outer shape data S1, converts the coordinates of each of the outer shape pixels Mz ₁ to the xy plane, and acquires the xy plane coordinate values of the outer shape pixels Mz _1. (Step S401). Next, the configuration point placement section 22, the extracting outer shape pixels Mz ₁ as a corner (corner constituent points) (step S402), further, the outer shape pixel Mz ₁ (angular configuration point) serving as the extracted corner y-coordinate value is extracted smallest outer shape pixel Mz _1, the starting point of this square configuration point (step S403).
Here, in step S402, when extracting outer shape pixels Mz ₁ as a corner, it may be the intersection of profile-shape extractor 21 obtains the (D1 _a ~D1 _d) as a candidate corner control points.

Next, the component point arrangement unit 22 searches for the outer shape pixel Mz ₁ adjacent to the corner point of the start point based on the xy plane coordinates, and extracts one detected outer shape pixel Mz ₁ (step S404). .
The search direction at this time is a predetermined direction. For example, the search direction when the image of the read outline data S1 to be processed is the left image when divided in step S306 is counterclockwise. In the case of the right side image, the search direction is preferably clockwise.
In addition, the corner composing point extracted as the start point may be the one having the largest y coordinate value, and it is only necessary to extract one corner composing point. Any point is acceptable.

Then, the constituent point arrangement unit 22 determines whether or not the xy plane coordinates of the extracted outer shape pixel Mz ₁ coincide with the xy plane coordinates of the corner constituent point of the start point (step S405).
If they do not match (Step S405, No), composing point arrangement unit 22, the extracted contour pixels Mz ₁ configuration point as (Configuration point P1) (step S406), to impart an identification number. Next, the identification number of the component point and the xy plane coordinate value of the component point are associated with each other and stored in the outer shape storage unit 26 (step S407). Then, again, it performs the process of step S404, searches the external shape pixels Mz ₁ and the adjacent control points (control points P1) and became outer shape pixel Mz ₁ (extracted contour pixels Mz _1). Thereby, an identification number is provided in order from the component point close to the corner component point of the start point.
On the other hand, if they match (step S405, Yes), the component point arrangement unit 22 ends the component point arrangement process (step S204 in FIG. 14) for the read outline data S1. Then, the process of step S205 is executed.

By the component point arrangement process, the component point arrangement unit 22 converts the contour line formed by the straight lines (page boundary line L11 and the outer shape straight line Lz ₁ (straight lines L12 to L14)) included in the read outline data S1 into the component points. _{P1 (P 11, P 12,} ..., _{and, P 1a, P 1b, ...} ) is expressed by (Figure 7).

Returning to the description of FIG.
Next, the constituent point placement unit 22 similarly executes the constituent point placement processing on the distortion correction outline data S2 (step S205, FIG. 14). As a result, contour lines formed by straight lines (page boundary line L21 and external shape straight lines Lz ₂ (straight lines L22 to L24)) included in the distortion-corrected outer shape data S2 are converted into constituent points P2 (P ₂₁ , P ₂₂ ,. Or P _2a , P _2b ,...) (FIG. 7).

  Next, the correspondence relationship construction unit 23 acquires the configuration point P1 and the configuration point P2 from the outer shape storage unit 26, and executes a correspondence relationship construction process (step S206). By this correspondence construction processing, the correspondence construction unit 23 associates all the constituent points P1 constituting the outline of the read outline data S1 with all the constituent points P2 constituting the outline of the distortion correction outline data S2. .

Here, with reference to FIG. 17, the correspondence construction processing operation (step S206 in FIG. 14) of the correspondence construction unit 23 will be described (refer to FIGS. 1 to 14 as appropriate).
<Correspondence building process>
First, the correspondence construction unit 23 extracts a boundary point between a straight line and a curve from the constituent point P1 on the contour line of the read outer shape data S1 (step S501). Through this process, D1 _c , D1 _d , D1 _e , and D1 _f are extracted as boundary points.
Similarly, the correspondence construction unit 23 extracts a boundary point between a straight line and a curve from the constituent point P2 on the contour line of the distortion correction outer shape data S2 (step S502). Through this processing, D2 _c , D2 _d , D2 _e , and D2 _f are extracted as boundary points.

Next, the correspondence construction part 23 selects the boundary points D1 _e (D1 _d ) on the page boundary line L11 out of the boundary points extracted from the read outline data S1 and the boundary points extracted from the distortion corrected outline data S2. Among these, the boundary point D2 _e (D2 _d ) on the page boundary line L21 is associated (step S503).
Next, the correspondence construction unit 23 associates the corner component point D1 _a (D1 _b ) of the read outer shape data S1 with the corner component point D2 _a (D2 _b ) of the distortion corrected outer shape data S2 (step S504).
The process of step S503～S504, correspondence construction unit 23, as indicated by a broken line in FIG. 7, correlates the angular configuration point D1 _a and D2 _a, and lies on the page boundary line L11 or L21 The boundary point D1 _e is associated with the boundary point D2 _e .

Then, the correspondence relationship construction unit 23 projects the boundary point D1 _f (D1 _c ) that is not on the page boundary line L11 among the boundary points extracted from the read outer shape data S1 to the distortion corrected outer shape data S2 (step S505). .

Then, the correspondence construction unit 23 associates the boundary point D1 _f not on the page boundary line L11 of the read outer shape data S1 with the nearest constituent point P2 (D2 _f ) at the position projected on the distortion corrected outer shape data S2. (Step S506).
By the processing in steps S505 to S506, the correspondence relationship building unit 23 associates the boundary point D1 _f with the projection point D2 _{f as} indicated by a broken line in FIG. Similarly, the correspondence construction unit 23 creates a boundary point D1 _c that is not on the page boundary line L11, and the nearest constituent point P2 (D2 _c ) at the position where the boundary point D1 _c is projected onto the distortion correction outline data S2. Associate.

Then, the correspondence establishing section 23 includes a number of control points P1 located on the contour line of the section from the border point D1 _e in the read contour data S1 to the boundary point D1 _f, projected from the boundary point D2 _e in distortion correction profile data S2 Adjustment is performed by increasing the number of constituent points of the smaller outer shape data so that the number of constituent points P2 on the contour line of the section up to the point D2 _f is the same (step S507). Then, the correspondence relationship construction unit 23 associates the constituent points in the section in the read outer shape data S1 with the constituent points in the section in the distortion correction outer shape data S2 (step S508). The processing of steps S507 to S508 will be described later.
Then, the correspondence construction process (step S206 in FIG. 14) is terminated, and then the process of step S207 is executed.

Here, when the correspondence construction unit 23 adjusts the number of constituent points in the section, first, the number of constituent points P1 in the section (D1 _{e to} D1 _f ) in the read outer shape data S1 and the distortion correction outer shape data. comparing the number of sections (D2 _e ~D2 _f) constituent points in P2 in S2. Then, a process of newly adding the difference number (the number of constituent points is insufficient) in the outline data section having the smaller number of constituent points to adjust the number of constituent points is performed in step S507. ˜S508 is performed.

(Description of processing in steps S507 to S508)
Here, the processing of steps S507 to S508 performed by the correspondence construction unit 23 will be described with reference to FIGS.
When the correspondence construction unit 23 adjusts the number of constituent points in the section, first, the number of constituent points P1 in the section (D1 _{e to} D1 _f ) in the read outer shape data S1 and the section in the distortion corrected outer shape data S2 ( D2 _{e to} D2 _f ) and the number of constituent points P2 are obtained, and the number of constituent points is compared. Then, the number of constituent points in the section of the outline data having a small number of constituent points is added by the number that is insufficient.
For example, in the case of FIG. 7, the number of component points P1 = 3 and the number of component points P2 = 7, and the read outline data S1 has only four component points.

  Next, the correspondence construction unit 23 performs processing for newly adding the difference number (the number of configuration points is insufficient) in the section of the external data having the smaller number of configuration points. At this time, a new component point is added to the same position as the existing component point sequentially from the existing component point located closest to the section center point. Even if a new component point is added at the same position as the existing component point that is located farthest away, if there is a surplus in the number of new component points to be added, it is again positioned closest to the section center point Add composition points in order from existing composition points.

The addition process will be described with reference to FIG.
First, the correspondence relationship building unit 23 refers to the outer shape storage unit 26 and extracts a position that bisects the section (D1 _{e to} D1 _f ) from the read outer shape data S1 having a smaller number of constituent points. The position is defined as a section center point M. Then, the nearest constituent point P ₁₂ is extracted from the section center point M, and the first new constituent point P ₁₄ is added at the same position as the constituent point P ₁₂ . Then, the next closest configuration point P ₁₁ is extracted from the section center point M, and a second new configuration point P ₁₅ is added at the same position as the configuration point P ₁₁ . Then, the third closest configuration point P ₁₃ from the section center point M is extracted, and a third new configuration point P ₁₆ is added at the same position as the configuration point P ₁₃ .
Here, the interval (D1 _e ~D1 _f) within, adding a new configuration points composing point P ₁₃ which is positioned farthest from the section center point M Note that the configuration point than distortion correction profile data S2 Since the number is one, the fourth new component point P ₁₇ is added at the same position as the component point P ₁₂ .

As described above, the process of step S507 is performed, and the number of constituent points P1 of the read outer shape data S1 and the number of constituent points P2 of the distortion corrected outer shape data S2 are adjusted. As shown in FIG. There are seven constituent points P _{11 to} P ₁₇ in the section (D1 _{e to} D1 _f ) of the outer shape data S1. Then, P ₁₅ is added at the same position as P ₁₁ , P ₁₄ and P ₁₇ are added at the same position as P _12, and P ₁₆ is added at the same position as P ₁₃ .
In this way, the number of constituent points P1 of the read outer shape data S1 and the number of constituent points P2 of the distortion-corrected outer shape data S2 are adjusted, and the process of step S507 ends.

Then, the correspondence relationship construction unit 23 performs the process of step S508. The processing in step S508 will be described with reference to FIG.
Here, the configuration point P1 in the section (D1 _{e to} D1 _f ) in the read outline data S1 is P ₁₁ , P ₁₅ (P ₁₁ ), P ₁₂ , P ₁₄ (from the configuration point nearest to D1 _{f. P 12), P 17 (P} 12), are arranged in the order of _{P 13, P 16 (P 13} ). On the other hand, the constituent point P2 in the section (D2 _{e to} D2 _f ) in the distortion corrected outer shape data S2 is P ₂₁ , P ₂₂ , P ₂₃ , P ₂₄ , P ₂₅ , P ₂₅ from the constituent point closest to D2 _{f. 26,} are arranged in the order of P _27.
First, the correspondence construction unit 23 associates the constituent points that are closest to D1 _f and D2 _f on the basis of the coordinate values of the constituent points stored in the outer shape storage unit 26, and then corresponds to the arrangement order. To do. That is, the correspondence relationship construction unit 23 first associates the configuration point P ₁₁ that is closest to D1 _f with the configuration point P ₂₁ that is closest to D2 _f , and then configures the configuration point P ₁₅ (P ₁₁ ) and associates the configuration point P _22. Finally, the configuration point P ₁₆ (P ₁₃ ) is associated with the configuration point P ₂₇ .
By the above processing, as shown in FIG. 9, the constituent point P1 (P _{11 to} P ₁₇ ) in the section in the read outer shape data S1 and the constituent point P2 (P _{21 to} P ₂₇ in the section in the distortion corrected outer shape data S2). ) Are associated with each other, and the process of step S508 ends.

  The correspondence relationship construction unit 23 performs the correspondence relationship construction processing shown in steps S501 to S508 on all the constituent points P1 constituting the contour line of the read outer shape data S1 and all the constituent points constituting the contour line of the distortion corrected outer shape data S2. Perform for point P2. Thereby, all the configuration points P1 and all the configuration points P2 are associated with each other.

Returning to the description of FIG.
The correction factor setting unit 24 sets the state of the distortion correction outer shape data S2 to the correction factor R = 1.0, and sets a plurality of correction factors R whose values are changed at a predetermined interval (for example, 0.2). Setting is made (step S207).

The process of step S207 will be described with reference to FIG.
The correction rate setting unit 24 sets the length of the line segment connecting the boundary point D1 _d and the boundary point D1 _e on the page boundary line L11 in the read outline data S1 as the correction rate R = 0.0, and corrects the distortion. on the profile data S2, the line segment connecting the boundary point D2 _d and the boundary point D2 _e on the page boundary L21 state length as the correction factor R = 1.0. Then, by changing the correction factor R, performs correction to be multiplied by the length of a line connecting the boundary point D2 _d and the boundary point D2 _e.
Here, if the correction rate R is “R <1”, the length of the line segment connecting the boundary point D2 _d and the boundary point D2 _e becomes short, and when R = 0, 0, the boundary point D1 _d and the boundary point This coincides with the length of the line segment connecting the point D1 _e . On the other hand, if “R> 1”, the length of the line segment connecting the boundary point D2 _d and the boundary point D2 _e is further increased.

Returning to the description of FIG.
The shape correction unit 25 generates a plurality of correction factor outer shape data S3 obtained by correcting the distortion correction outer shape data S2 with each value of the correction factor R (step S208), and corrects the read outer shape data S1 and the distortion correction outer shape data S2. The rate outline data S3 is stored in the outline storage unit 26 (step S209). Then, the correction preparation unit 13 ends the correction preparation process (step S106 in FIG. 13).

In the process of step S208, the shape correcting unit 25 performs correction as follows, and generates correction factor outline data S3.
First, the shape correction unit 25, from the outer storage section 26, distortion correction acquires profile data S2, the line distortion correcting boundary point in the contour of the outer shape data S2 D2 _d ~ page boundary line L11 between boundary points D2 _e The coordinate value of the component point P2 on (FIG. 5B) is corrected with the correction rate R, and the coordinate value of the component point P2 after correction is stored in the outer shape storage unit 26.
Furthermore, construction of the line segment of the curve between the boundary points D2 _c ~ boundary point D2 _d (FIG. 5 (b)) and the line segment of the curve between the boundary points D2 _e ~ boundary point D2 _f (FIG. 5 (b)) The coordinate value of the point P2 is corrected based on the correction rate R, and the coordinate value of the corrected component point P2 is stored in the outer shape storage unit 26. Correction based on the correction factor R is as as the correction amount is increased close to a page boundary L11, so that the more the correction amount near the boundary point D2 _c and boundary point D2 _f Conversely decreases. At this time, the coordinate values of the boundary points D2 _c and boundary point D2 _f is corrected based on the correction factor R is not performed.
This correction can be realized, for example, by performing field morphing, which is known two-dimensional morphing.

<Secondary correction processing>
Next, the secondary correction processing operation (step S107 in FIG. 13) of the secondary correction unit 14 will be described with reference to FIG. 18 (refer to FIGS. 1 to 13 as appropriate).
The feature region extraction unit 27 included in the secondary correction unit 14 acquires the read outer shape data S1, the distortion correction outer shape data S2, and the correction factor outer shape data S3 from the outer shape storage unit 26, and further reads the read image from the image data storage unit 11. Data D1 and distortion corrected image data D2 are acquired (step S601). Then, a feature region extraction process is executed.

  In the process of step S306, the feature region extraction unit 27 combines two images (D1L and D1R) obtained by dividing the read image data D1 by the outer shape extraction unit 21 with images having the same correction rate R, One image data is set (step S602). Similarly, the outer shape extraction unit 21 combines two images (D2L and D2R) obtained by dividing the distortion-corrected image data D2.

  Then, the feature region extraction unit 27 uses the correspondence relationship construction unit 23 to set all of the constituent points associated with the read image data D1 and the distortion corrected image data D2 (the constituent points P1 and P2). The distance between the component points in the set) is calculated (step S603). Then, the feature region extraction unit 27 detects a set of component points that maximizes the distance between the component points (the rate of change is high) (step S604). At this time, the coordinate values of the constituent points of the set are acquired.

The feature region extraction unit 27 extends a line segment connecting the constituent points, and extracts a rectangular feature region having the line segment as a diagonal line (step S605). At this time, the coordinates of the end points of the line segment are acquired.
By the feature region extraction processing described above, the feature region extraction unit 27 extracts a region (feature region) having a predetermined size with a high rate of change between the contour line of the read outer shape data S1 and the contour line of the distortion corrected outer shape data S2. To do.

Next, the feature region extraction unit 27 extracts an image (partial image) corresponding to the feature region extracted in step S605 from all the correction factor outline data S3 (step S606).
As a result, the feature region extraction unit 27 performs the processing of steps S603 to S606 after combining the images with the same correction rate R in step S602, so that the feature region extraction unit 27 differs for each correction rate R. A plurality of partial images are extracted.

Then, the display control unit 28 acquires the partial images extracted by the feature region extraction unit 27, and causes the display unit 7 to display a list of different partial images for each correction factor R (see FIG. 10) (step S607).
The user operates the operation input unit 8 to select one partial image (step S608).
The input processing unit 29 acquires selection information input from the operation input unit 8 (step S609).
Then, the corrected image generation unit 30 executes a corrected image generation process (step S610).

<Corrected image generation processing>
Here, the corrected image generation processing operation (step S610 in FIG. 18) of the corrected image generation unit 30 will be described with reference to FIG. 19 (refer to FIGS. 1 to 18, particularly FIG. 12 as appropriate).
First, based on the selection information, the corrected image generation unit 30 acquires the correction rate outline data S3 of the correction rate R corresponding to the selected partial image (feature area) from the outline storage unit 26 (step S701). Then, the correction factor outline data S3 is rasterized, and the contour line of the correction factor outline data S3 is acquired (step S702).

The corrected image generation unit 30 acquires the coordinates (x ₃ , y ₃ ) of the virtual point U3 (u _3n ) that is not on the contour line of the correction factor outer shape data S3 but is located within the region surrounded by the contour line (Step 3 In step S703, the constituent point P3 of the nearest correction factor outline data S3 is detected from the virtual point U3 (step S704).

The corrected image generation unit 30 acquires the configuration point P2 (P _2n ) of the contour line of the distortion correction outline data S2 associated with the configuration point P3 (P _3n ) (step S705), and configures the configuration point from the configuration point P3. A movement vector V (x _t , y _t ) up to P2 is calculated (step S706).

Furthermore, the corrected image generation unit 30 calculates the weighting factor ω from the distance from the point Q (intersection Q _LL or intersection Q _LW ) that satisfies a predetermined condition to the constituent point P3 and the distance from the point Q to the virtual point U3. Is calculated (step S707).

The weighting factor ω is calculated as follows.
Let Q _LL be the intersection of a straight line connecting the constituent point P2 and the constituent point P3 and a straight line LL that cuts through the center of the page. Further, an intersection of a straight line connecting the configuration point P2 and the configuration point P3 and a straight line LW crossing the center of the page is defined as Q _LW .
Then, the distance from the intersection point Q _LL to the component point P3 or the distance from the intersection point Q _LW to the component point P3, whichever is shorter, is L _max, and the point selected as the distance L _max (the intersection point Q _LL or the intersection point Q 3 _Let L _u be the distance from _LW ) to the virtual point U3. Then, using Equation (1), the ratio of L _u to L _max is set as a weighting factor ω.

Returning again to the description of FIG.
The corrected image generation unit 30 estimates that the virtual point U3 (x ₃ , y ₃ ) was in the distortion corrected image data D2 before correction from the movement vector V and the weighting coefficient ω using Equation (2). Estimated point U2 (x ₂ , y ₂ ) (estimated point u _2n ) is calculated (step S708).

The corrected image generation unit 30 uses the equation (3) to calculate the pixel value I ₃ of the pixel corresponding to the position of the virtual point U3 and the distortion corrected image data corresponding to the position of the estimated point U2 (x ₂ , y ₂ ). corrected pixel value I ₂ of D2 pixel (step S709).

The corrected image generation unit 30 determines whether or not the coordinates of all the points (virtual points U3) located in the region surrounded by the outline of the correction factor outline data S3 have been acquired (step S710).
If there are coordinates of points that have not been acquired (step S710, No), the process of step S703 is performed again.
On the other hand, when the coordinates of all the points have been acquired (step S710, Yes), the corrected image generation process (step S610 in FIG. 18) ends and the secondary correction process (step S107 in FIG. 13) ends. . Then, the process of step S108 (FIG. 13) is executed.

As described above, since the image reading apparatus 1 according to the first embodiment performs the secondary correction on the image with the insufficient correction amount according to the correction rate R selected by the user, the distorted image with the appropriate correction amount. Can be corrected.
Further, in the image reading apparatus 1, the corrected image generation unit 30 generates image data for each corrected correction rate R for the corrected contour line (correction rate outer shape data S <b> 3). By displaying the list, it is possible to give the user the right to select the correction factor R. Then, the user can easily determine a desired contour line after correction (correction factor outer shape data S3).

  Further, the image reading apparatus 1 uses the contour line (distortion correction outer shape data S2) of the medium extracted from the medium image data (distortion correction image data D2) as the contour (correction rate outer shape data S3) of the correction factor R selected by the user. ), The distorted shape of the medium can be corrected based on the correction rate R selected by the user by correcting the medium image data (distortion correction image data D2).

<< Second Embodiment >>
Next, the configuration of the image reading apparatus according to the second embodiment will be described.
As shown in FIG. 20, the secondary correction unit 14A of the image reading apparatus 1A according to the second embodiment is further added to the secondary correction unit 14 (FIG. 3) of the image reading apparatus 1 according to the first embodiment. The correction shape evaluation unit 51 and a display control unit 28A to which a function of acquiring data from the correction shape evaluation unit 51 and performing processing are added.

(Correction shape evaluation unit 51)
The correction shape evaluation unit 51 is a component that acquires all the correction factor outline data S3 from the outline storage unit 26 and executes the correction shape evaluation process.
By this correction shape evaluation process, the correction shape evaluation unit 51 calculates the evaluation value of the outline of each correction factor outline data S3. Then, the correction rate R applied to the highest evaluation value correction rate profile data S3 is output to the display control unit 28A.

(Display control unit 28A)
The display control unit 28A in the second embodiment generates image data in which the partial image corresponding to the correction rate R acquired from the correction shape evaluation unit 51 is emphasized, and causes the display unit 7 to display a plurality of partial images. FIG. 21 shows an example in which the display control unit 28A emphasizes and displays the partial image with the correction ratio R = 1.4 of J3 on the display panel 86 (display unit 7).

<Secondary correction processing>
Next, with reference to FIG. 22, the secondary correction processing operation (step S107 in FIG. 13) of the secondary correction unit 14A will be described.
Here, the processes in steps S601 to S606 and steps S608 to S610 that overlap with the secondary correction processing operation (FIG. 18) of the image reading apparatus 1 (secondary correction unit 14) in the first embodiment are described. The description is omitted.

After the feature region extraction unit 27 extracts a plurality of different partial images for each correction rate R (step S606 in FIG. 18), the correction shape evaluation unit 51 stores all the correction rate outer shape data S3 stored in the outer shape storage unit 26. Then, a corrected shape evaluation process is executed (step S801). By the correction shape evaluation process, the correction shape evaluation unit 51 outputs the correction rate R (optimum correction rate) performed on the correction rate outer shape data S3 of the highest evaluation value to the display control unit 28A.
The display control unit 28A acquires the partial images extracted by the feature region extraction unit 27, and causes the display unit 7 to display a list of different partial images for each correction factor R. At this time, the display control unit 28A generates image data in which the partial image corresponding to the correction rate R (optimum correction rate) input from the correction shape evaluation unit 51 is emphasized, and causes the display unit 7 to display the partial image. (See FIG. 21) (step S802).

<Correction shape evaluation process>
Next, the corrected shape evaluation processing operation (step S801 in FIG. 22) of the corrected shape evaluating unit 51 will be described with reference to FIGS. 23 and 24 (refer to FIGS. 20 to 22 as appropriate).
The correction shape evaluation unit 51 first acquires the correction factor outer shape data S3 from the outer shape storage unit 26 (step S901), and the coordinate values of the two corner composing points (D3 _a and D3 _b ) from the correction factor outer shape data S3. The coordinate value of the boundary point (D3 _e and D3 _d ) on the page boundary line L31 is extracted, and the angle between the straight line L33 connecting the corner component point D3 _a and the corner component point D3 _b and the page boundary line L31 θ is calculated (step S902).
Then, the correction shape evaluation unit 51 calculates the evaluation value E _grad (parallelism) of the inclination of the page boundary line L31 using Expression (4) (Step S903). However, it is assumed that the angle θ is an acute angle.

Next, the correction shape evaluation unit 51, a line connecting the length L _edge of the segment (straight line L33) connecting the corner control points D3 _a and angular configuration point D3 _b, and a boundary point D3 _e and the boundary point D3 _d The length L _{border of the} minute (page boundary line L31) is calculated (step S904).
Then, the corrected shape evaluation unit 51 calculates the evaluation value E _length (degree of coincidence) of the _{length of} the page boundary line L31 using Equation (5) (Step S905).

And the correction | amendment shape evaluation part 51 calculates _total evaluation value _Etotal from evaluation value _Egrad and evaluation value _Elength using Numerical formula (6) (step S906). Here, α is a predetermined number representing the weight of the evaluation value for the length.

The correction shape evaluation unit 51 determines whether or not all the correction factor outer shape data S3 has been acquired from the outer shape storage unit 26 (step S907).
When there is correction rate outline data S3 that has not yet been acquired (No at Step S907), the correction shape evaluation unit 51 executes the process at Step S901 again.

On the other hand, it has acquired all the correction factor profile data S3 (step S907, Yes), the correction shape evaluation unit 51, a comprehensive evaluation value E _total calculated from the correction factor profile data S3, the highest total evaluation value E _total Extract (step S908). Then, the corrected shape evaluation unit 51 outputs the correction rate R performed on the correction rate outer shape data S3 of the highest evaluation value (the highest overall evaluation value _Etotal ) to the display control unit 28A (step S909). Then, after the corrected shape evaluation unit 51 ends the corrected shape evaluation process (step S801 in FIG. 22), the process of step S802 is executed next.

As described above, in the image reading apparatus 1 </ b> A according to the second embodiment, the correction shape evaluation unit 51 calculates the correction factor outline data S <b> 3 with the least distortion with the highest evaluation value (the highest total evaluation value E _total ). Thereby, the correction rate R (optimum correction rate) of the correction rate outer shape data S3 is output to the display control unit 28A.
Then, the display control unit 28A generates image data in which the partial image corresponding to the correction rate R (optimum correction rate) acquired from the correction shape evaluation unit 51 is emphasized. Therefore, the image reading apparatus 1 </ b> A can present the optimal correction amount to the user on the display unit 7.

Although the embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments, and can be implemented with appropriate modifications.
For example, an external device such as a scanner including the image reading unit 2 and the image data generation unit 3 transmits the generated image data via a network such as a telephone line or a LAN, a recording medium such as a USB memory or a CD-RAM, or the like. Thus, it may be configured to input to an image processing apparatus (such as a PC).
This image processing apparatus includes a data input unit to which image data generated by an external device is input, an operation input unit that receives an input operation from a user, a display unit that displays an image, a CPU, a ROM, a RAM, an HDD (Hard Disc Drive) ) And the like. The computer can be realized by a program (image correction program) that functions as the pre-correction processing unit 61, the display processing unit 62, and the image correction unit 63 shown in FIG.
Here, the data input unit may be a communication unit that is connected to an external device via a network and receives image data transmitted by the external device.

  Further, in the present embodiment, as shown in FIG. 1, the case where the booklet-shaped document 91 is placed on the medium placing table 81 in a state where the booklet-shaped document 91 is opened is shown, but the cover and back cover of the booklet-shaped document 91 are separated. In a state of being bent so as to be in contact, the page on one side may be placed on the medium placing table 81 as a reading surface. In this case, the outer shape extraction unit 21 of the image reading apparatus 1 (1A) does not divide the image into two in step 306 (FIG. 15).

DESCRIPTION OF SYMBOLS 1 Image reading apparatus 2 Image reading part 3 Image data generation part 7 Display part 8 Operation input part 10 Image processing part (4 CPU)
DESCRIPTION OF SYMBOLS 11 Image data memory | storage part 12 Primary correction | amendment part 13 Correction | amendment preparation part 14 Secondary correction | amendment part 21 Outline shape extraction part 22 Composition point arrangement | positioning part 23 Correspondence relationship construction part 24 Correction rate setting part 25 Shape correction part 26 Outline storage part 27 Feature area extraction Unit 28 Display control unit 29 Input processing unit 30 Correction image generation unit 61 Pre-correction processing unit 62 Display processing unit 63 Image correction unit D1 Read image data D2 Distortion correction image data

Claims (5)

An input unit that accepts a selection input from the user, and a display unit;
An image reading device that reads a surface image of a medium having a predetermined shape to generate medium image data, and corrects the distortion of the medium image data generated in a shape in which the surface of the medium is distorted when read.
An outline of the medium surface is extracted from distortion corrected image data obtained by first correcting the medium image data generated by reading the medium surface in a distorted shape, and the correction values are different from each other at a plurality of correction rates. A pre-correction processing unit for correcting the contour line;
A display processing unit that generates an image of the corrected correction factor outline contour for each correction factor, enables the selection of the image for each correction factor, and displays the list on the display unit;
An image correction unit that performs second correction on the distortion-corrected image data so that a contour line extracted from the distortion-corrected image data matches the correction factor outline contour line that is selected and input to the input unit. An image reading apparatus. The correction preprocessing unit
A primary correction unit configured to generate the distortion correction image data by performing the first correction on the medium image data generated in a distorted shape of the medium surface;
An outer shape extraction unit that extracts a reading outer contour line of the medium surface from the medium image data, and extracts a distortion correction outer contour line of the medium surface from the distortion correction image data;
A correction rate setting unit for setting a plurality of correction rates based on the distortion-corrected outer contour line;
A shape correction unit that corrects the distortion-corrected outer contour line with the correction factor and generates a correction factor outer contour line for each correction factor;
The image reading apparatus according to claim 1, further comprising: The correction preprocessing unit
A configuration placement unit that places a configuration point on the read outline contour extracted by the profile extraction unit, and a configuration point that arranges the configuration point on a distortion-corrected profile contour extracted by the profile extraction unit;
A correspondence relationship construction unit that associates component points on the read contour outline and component points on the distortion-corrected contour outline that are close to each other on the same plane;
The display processing unit
Detecting a set of component points having the maximum distance between the associated component points, extending a line segment connecting the detected component points, generating a rectangular frame having the line segment as a diagonal line, and generating the shape A feature region extraction unit that extracts a feature region that is within the rectangular frame from a plurality of correction factor outer contour lines generated by the correction unit;
The display control part which produces | generates the image of the characteristic area extracted for every said correction factor, makes the image for every said correction factor selectable, and makes it display on a list on the said display part. The described image reading apparatus. The medium has a shape having at least a pair of line segments,
The display processing unit extracts a pair of line segments from all the correction factor outline contour lines for each correction factor, acquires parallelism from an angle formed by the pair of line segments, and lengths of the respective line segments. A correction shape evaluation unit that acquires the degree of coincidence from the ratio of, and obtains the correction factor outline contour of the optimum correction factor from the parallelism and the degree of coincidence,
The image reading apparatus according to claim 3, wherein the display control unit further causes the display unit to highlight an image of the feature region extracted from the correction rate outline of the optimal correction rate. . A computer including an input unit that receives a selection input from a user, a display unit, and a data input unit that acquires medium image data generated in a shape in which the medium surface is distorted.
A first correction is performed on the medium image data acquired by the data input unit to generate distortion-corrected image data, an outline of the medium surface is extracted from the distortion-corrected image data, and a plurality of corrections having different values are provided. Correction preprocessing means for correcting the contour line at a rate,
Display processing means for generating an image of a correction rate outline contour line obtained by correcting the contour line for each correction rate, enabling selection of an image for each correction rate, and displaying the list on the display unit;
Image correction means for performing second correction on the distortion-corrected image data so that the contour line extracted from the distortion-corrected image data matches the correction factor outer contour line selected and input to the input unit;
An image correction program characterized by functioning as

Images