JPH114325A - Image reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow the image reader to provide an output of an image without distortion by allowing an image pickup means to match focusing at reading of an original image in response to a height of an original surface detected by a height detection means to correct distortion of the read image caused by distortion of the shape of the original. SOLUTION: At first a CCD line sensor acting as a preliminary scanner moves from a home position and reads an original surface and an upper side face of the original reflected in a range finding mirror 5, then discriminates a shape of the original and the contrast of the original surface to obtain height distribution of the original surface from an image reflected in the range finding mirror 5. Then based on height data of the original obtained by the movement of the CCD line sensor acting as a main scanner, all faces of the original surface are focused and the image is corrected in matching with a curved state of the original in response to the obtained height distribution. Thus, a fogged image on the original is avoided and the image without distortion is obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image reading apparatus, and more particularly, to an image reading apparatus for reading a document image from above with a document placed thereon.

[0002]

2. Description of the Related Art As one of image reading apparatuses, there is an image reading apparatus in which an original to be read is placed upward and an image on the surface of the original is read by an image sensor. As such an image reading apparatus, for example, Japanese Patent Application Laid-Open No. 7-79324 discloses that when reading an image of a book original such as a book or a file, first, the upper side (upper side) of the book original whose thickness is known, such as a book. Is projected on a mirror, the boundary between the image on the upper surface of the book document reflected on the mirror and its background is detected from the contrast difference, a preliminary scan is performed to detect the height distribution of the document surface from this boundary, and At the time of the main scan, which reads the original image, focuses on the image on the original surface based on the obtained height distribution, and corrects the image distortion based on the original height to obtain an image that is out of focus or distorted. I'm trying to get.

[0003]

In the book document reading apparatus described in the above-mentioned publication, the height distribution of the book document is such that the vertical center line (center line) of the double-page spread, and in the case of a normal book, the height distribution thereof, It is performed based on the assumption that it changes uniformly in the direction along the binding portion. Then, based on the height distribution detected under such an assumption, the image obtained by the main scan is one-dimensionally distorted by the vertical bending of the original surface and the character shrinkage in the horizontal direction. Has been corrected.

However, in the case of a book original such as a book, the right and left sides of the original are parallel to the center line of the binding part of the original due to the binding tendency of the binding portion on the bookbinding and the opening tendency when the book original is opened. In some cases, the open surface of the original is distorted like twisting and placed on the original platen.

In such a case, since the surface of the original has a complicated three-dimensional shape, the height distribution of the book original changes uniformly with respect to the vertical center line of the double-page spread as in the related art. There is a problem that the method of detecting the height distribution of the document based on the assumption and correcting the image based on the result cannot be dealt with.

SUMMARY OF THE INVENTION The present invention has been made in view of the above points, and detects a distortion including a shape distortion due to a twist unique to a book original, and corrects a distortion of a read image caused by the shape distortion. An object of the present invention is to provide an image reading device capable of outputting an image without distortion.

[0007]

According to a first aspect of the present invention, there is provided an image reading apparatus for reading an image on the surface of a document placed upward on a document table from above the document. Imaging means for reading a document placed on the document table, and a lateral side edge position detecting means for detecting the position of the lateral side edge of the document from the image of the document itself read by the imaging means, Document center line detection means for detecting the position of the center line of the document, the position of the lateral side edge detected by the lateral side edge position detection means, and the position of the center line detected by the document center line detection means An image reading device, comprising: a correcting unit that corrects a length of the document in a horizontal direction from an upper end to a lower end of the image of the document.

According to the present invention, the position of the center line and the lateral side edge of the document is detected from the image of the document itself, and further from the detected center line and the position of the lateral side edge to the upper end and lower end of the document. By correcting the length in the horizontal direction by the correction means, the original image is made to have no distortion.

Further, according to the present invention, in the configuration according to the first aspect, the image pickup means can read an upper surface of the document, and the image reading device further includes an image forming device for reading the upper surface of the document from the image of the upper surface. Height detecting means for detecting the height,
An image reading apparatus, comprising: focusing means for performing focusing when reading the document image by the imaging means in accordance with the height of the document surface detected by the height detecting means.

According to the present invention, the configuration according to claim 1 is applied to an image reading apparatus having a document surface height detecting means, whereby the height obtained by the document surface height detecting means is obtained. 2. An original image which is read without being defocused by focusing according to data, and wherein
The configuration described above can also correct the shape distortion of the document.

[0011]

BRIEF DESCRIPTION OF THE DRAWINGS FIG.
An embodiment of the present invention will be described.

FIG. 1 is a diagram showing the overall configuration of an image reading apparatus according to the present embodiment. FIG. 2 shows a book (book original) as an original placed on the original platen shown in FIG. FIG.

The image reading apparatus shown in FIG. 1 is called a book original reading apparatus. The book document reading device (hereinafter simply referred to as a device) 10 places the book document on a document table 1 that can be individually moved up and down on the left and right sides in a state where the book document is opened. Is illuminated, and an image is read by the CCD line sensor 8 via the lens 7 in the scanner unit 2. When reading the image, the shape of the upper surface of the book document reflected on the distance measuring mirror 5 is read, and the height of the document surface is detected. Then, focusing on the scanner unit 2 and distortion correction of the read image are performed based on the detected height of the document surface.

As shown in FIG. 2, the document reading operation can be started on the document table 1 of the apparatus 10 even when both sides of the book document are pressed with both hands in order to keep the book document B open. Is provided with a start key 6. An operation panel 4 for performing various settings is provided at the back of the document table 1, and a power switch 11 is provided on a side surface of the apparatus.

The CCD line sensor 8 provided inside the scanner unit 2 is arranged such that photo sensors (pixels) are arranged from the front to the back of the apparatus with respect to the image (FIG. 1).
It is moved in the left-right direction (the direction of arrow B in FIG. 1 (the sub-scanning direction indicated by B 'in FIG. 1 for an image)) by a motor.

As shown in FIG. 2B, the book document B is placed so that the binding portion of the book document is aligned with the center mark M shown on the apparatus 10, and
The D-line sensor 8 moves in the sub-scanning direction, receives light reflected from the surface of the document, and reads an image. At this time, the surface of the document is illuminated by a lamp (light source) in the lamp unit 3, and an illuminating device that does not fluctuate in light amount such as flicker is used as the light source.

A shading detection plate 9 for removing shading due to the influence of external light is provided above the distance measuring mirror 5 along the sub-scanning direction of the CCD line sensor 8.
Is provided. The shading detection plate 9 is a white plate having a constant density. Shading detection plate 9
The density fluctuation in the sub-scanning direction due to external light such as flicker is recognized from the signal of the CCD line sensor 8 when the image is read, and the effect of external light such as flicker is removed by automatically correcting the image output according to the density fluctuation. doing.

The basic image reading operation of this apparatus is as follows. First, as a preliminary scan, the CCD line sensor 8 moves from the home position and reads the surface of the document and the upper side of the document reflected on the distance measuring mirror 5, and will be described later. As described above, while determining the shape of the original and the density of the original surface,
The height distribution of the document surface is obtained from the image projected on the distance measuring mirror 5. Then, as a main scan, the CCD line sensor 8 moves, focuses on all surfaces of the document surface based on the obtained height data of the document, and further adjusts the curvature of the document according to the obtained height distribution. By correcting the image in accordance with the state, an image free of distorted original images and free of distortion is obtained.

In the present embodiment, the upper surface of a book or the like is shown as the side of the original to be projected on the distance measuring mirror 5, and the side of a portion usually called "heaven" in a book. It goes without saying that the present invention can be carried out in exactly the same manner by reading not only the lower surface of a book or the like, that is, the side usually called “ground”.

The basic structure and operation of the book document reading apparatus are already known, and are disclosed, for example, in Japanese Patent Application Laid-Open No. 7-7 / 1995.
Since this is the same as the image reading apparatus described in Japanese Patent No. 9324, a detailed description thereof will be omitted.

Here, the names of each part of the book document used in the following description will be defined with reference to FIG. 3, which will be described later. It should be noted that each name is to be compared with the name of each part of the manuscript used in the claims of the present invention, and in this specification, it follows this definition, but it is used in ordinary books, files, and other manuscripts. Even if different names are used, the same parts as those defined herein indicate the same parts.

First, the lateral side end d in the present invention is the end of the page on the side of opening the book called "edge" in the book as shown in FIG. These are the left and right edges of the document image. Next, the manuscript center line
In the case of a book, this is a line formed at a portion corresponding to the binding line (however, in the case of the present apparatus, it may be the position of the center mark M). Next, the upper side b is “heaven” in the book.
This is the side on which the pages on the side referred to overlap. Next, the upper end b 'is the page end on the "top" side. Next, the lower side surface e is a side surface on which pages on the side called “ground” in the book are seen overlapping. Next, the lower end e ′
This is the end of the page on the “ground” side.

Hereinafter, correction of a document image in the present embodiment will be described.

FIG. 3 shows a center mark M on the document table 1.
(See FIG. 2 (B)), when the book document B is placed with its bindings aligned, the height of the document surface a is parallel and uniform to the center line c appearing at the binding of the book document B. FIG. 4A is a schematic diagram of a case where the document B is placed in a changing state. FIG. 4A shows an image of the upper surface b of a book document B reflected on a distance measuring mirror 5, and FIG. Each of the images represents the original document itself. When the book document B is placed without distortion as described above, the height of the document surface a changes uniformly in the vertical direction along the center line c as the height reflected on the distance measuring mirror 5 as it is. On the assumption that the height distribution detected from the image of the upper surface b uniformly changes in the direction along the center line c on the document surface a, as in the related art, the height obtained by the image of the upper surface b By correcting the image of the document surface a from the distribution data, a clear image can be obtained.

On the other hand, FIGS. 4 and 5 show an example in which the book document B is distorted and placed.

FIG. 4 is a schematic diagram showing a case where the right page of the book document B is placed on the platen 1 with a wide distortion.
2A shows an image of the upper surface b of the book document B reflected on the distance measuring mirror 5, and FIG. 2B shows an image of the document B itself. 5A and 5B are schematic diagrams when the right page of the document is placed on the document table 1 in a widely distorted manner. FIG. 5A shows a book document B reflected on the distance measuring mirror 5. FIG. 4B shows an image of the original B itself, and FIG.

As described above, depending on the book document B placed on the document table 1, the entire document is distorted, and the center line c corresponding to the binding of the book document B is shifted to the right, left, or both sides. The side end d is not parallel,
The entire page may be distorted as if twisted.

In the original placed distorted in this manner, the height distribution of the entire original surface a does not match the height distribution obtained from the upper side surface b of the book original B. Even if the image is corrected from the height distribution data based on the assumption that the height distribution of the document B changes uniformly along the center line c, the image distortion may not be completely corrected. For example, when the image is corrected using the height distribution data obtained from the image of the upper surface b as in the related art, the image is narrowed in a portion where the height change is abrupt. Therefore, in the image shown in FIG. 4, the image near the lower end e ′ is unnecessarily spread over the image near the upper surface b, and in the image shown in FIG. An image close to the lower end e 'is narrowed because the width of the image is not enough to expand the image, and the obtained image has poor appearance.

FIG. 6 is a diagram for explaining in more detail how the original surface a is placed in a distorted state. The original surface a of the book original B having the lower width shown in FIG.
3 is an enlarged view of the right page.

Since a book original such as a book or a file usually has a stiffness on paper, it does not wrinkle as long as the opened pages themselves do not have paper wrinkles. Three-dimensional shape. Therefore, if the height distribution of the edge (upper end b ') of the upper surface b and the height distribution of the lower edge (lower end e') of the lower surface are known, each side surface is spaced at a fixed ratio (for example, a page). (The actual length in the horizontal direction is equally divided), and the points are connected to form a line segment, whereby the three-dimensional shape of the document surface a can be obtained.

However, although the height of the upper end b 'can be detected by the image projected on the distance measuring mirror 5, a distance measuring mirror is provided on the lower side e due to the configuration of the book document reading apparatus. It is difficult to directly obtain the height distribution of the lower end portion e '. This is because, for example, when a distance measuring mirror for projecting the lower surface e is provided, when the book document B is pressed by a finger, the arm or hand of the operator obstructs the optical path of the distance measuring mirror for the lower surface, and the lower end e ′ Cannot be obtained accurately, or if there is a protrusion such as a distance measuring mirror on the side of the lower side surface e, it becomes difficult to place the book original B, and the book original B is distorted and mounted. This is because the frequency of the operation increases. It is also conceivable that the distance measurement mirror is dirty because the distance measurement mirror is close to the operator, resulting in poor distance measurement due to a decrease in contrast of an image reflected on the mirror.

In this embodiment, three-dimensional distortion on the surface of a book document is corrected without obtaining height distribution data on the lower surface.

FIG. 7 is a diagram for explaining data necessary for correcting an image on the surface of a book document having three-dimensional distortion without using data on the lower surface.

This data is obtained from the image of the book document B itself. The position and length L1 of the lateral side edge d of the image of the document itself, and the length of the upper end b 'of one page in the image of the document itself L2 and the length L3 of the lower end portion e 'of one page in the image of the original document itself. These L1 to L3 are calculated based on the contrast difference between the original background density and the density of the surface of the original table from the image of the original document obtained during the preliminary scan.
The edges around the document are detected, and the number of pixels in the main scanning direction and the sub-scanning direction of the edge portions corresponding to L1 to L3 are counted from the detected edges around the document, and are set as lengths L1 to L3.

Specifically, in the case of L1, the position is L
The edge of the lateral side end d that is 1 is detected, and the position of each pixel at that edge (the address of each pixel in the shape detection memory in the main scanning direction and the address in the sub-scanning direction) becomes the position L1. . The length of L1 is determined by detecting the pixel positions of the upper end d1 and the lower end d2 of the edge, and calculating the trigonometric function or Pythagoras from the number of pixels in the main scanning direction and the sub-scanning direction (address value of each pixel in a shape detection memory described later). L1 is calculated using the theorem of For L2, the number of pixels (address value) in the sub-scanning direction between center line c and d1
Similarly, L3 is determined from the number of pixels (address value) in the sub-scanning direction from the center line c to d2.

In the case of a book original such as a book, the center line c is usually darkened because illumination hardly enters the binding portion, so that the center portion is darkened. Since the density of the binding portion is extremely low, this portion is detected as the center line c. However, in the case of a book manuscript in which relatively open pages such as a weekly magazine can be flattened and the binding portion is sufficiently illuminated, the density of the binding portion is small so that the center line may be adjusted by the binding. May not be detected. In such a case, the position of the center mark M specified in advance is used as the position of the center line c, and the number of pixels of the document in the main scanning direction at that portion is determined as the center line c. Length.

L1 obtained here is the true length of the document in the vertical direction. L2 is the length of the upper side on the image, and L3 is the length of the lower side on the image.

In the case of a normal book, the center line c is equal to the length L1 of the lateral end d of the page. Therefore, as for the length of L1, the length of the center line c may be directly used as the length of L1.

In the actual processing, the data representing the three-dimensional shape of one such page is obtained for each of the left and right pages centering on the center sign c for both of the opened pages.

The original image obtained at the time of the main scan is corrected from the data representing the three-dimensional shape thus obtained. FIG. 8 is a schematic diagram showing the state of this correction. First, FIG. 8A is a drawing showing a read image in which three distortions have occurred due to a change in height specific to a book document placed on the document table 1 as described above.

One of the three distortions is an image enlargement due to a decrease in the distance between the original and the CCD line sensor at a high portion of the original surface, that is, a distortion caused by the occurrence of elongation in the vertical direction. This is why the upper and lower portions of the center of the image are bulged round. This is called "line bending". Next, distortion occurs in a portion where the height change is abrupt, especially in a rising portion of the binding of the book document, in which the document is not perpendicular to the CCD line sensor and thus contracts in the horizontal direction. This is called "character shrinkage". Finally,
This is a vertically asymmetric distortion caused by the document being twisted and placed. This is called "torsion distortion".

Of these three distortions, "torsion distortion" is corrected first. FIG. 8B shows a document image after this torsional distortion has been corrected. This correction will be described with reference to FIG. 7. First, among the data L1 to L3 obtained for the correction as described above, the length L2 of the upper end b 'and the length L3 of the lower end e' Are compared with each other, and the length L3 of the lower end portion e 'is adjusted to the length L2 of the upper end portion b'. In the example of FIG. 7, L3 is reduced to match L2. Therefore, L
L3 is reduced at a magnification of the ratio L2 / L3 of 2 and L3.

Next, the intermediate portion from L2 to L3 is represented by L2
Adjust to For this purpose, the enlargement or reduction magnification is obtained for each correction position of the intermediate portion and the enlargement or reduction is performed. Here, assuming that the vertical correction position along the center line c is P, in the example of FIG. 7 (when L3> L2), the magnification X at the P position is: X = 1 / [1 + ｛(L3-L2 ) × (L1−P) / L1｝] (1) Then, the magnification X is set to the length of the original in the sub-scanning direction at the correction position P, that is, the length Lp from the center line c at the correction position P to the lateral side end d when L1 is obtained. Reduce by applying.

When L3 <L2, the magnification is obtained, and the magnification X is obtained by the following equation: X = 1 + {(L2-L3) × (L1-P) / L1} (2) Then, in the same manner as described above, the magnification X is multiplied by the length Lp of the document in the sub-scanning direction at the correction position P to enlarge the image. By adjusting the length L3 of the lower end portion e 'to the length L2 of the upper end portion b' in this manner, the torsional distortion is corrected as shown in FIG. With the height distribution data obtained from the image of the side surface b, it is possible to correct line bending and character shrinkage on the assumption that the height of the document surface a changes uniformly over the entire surface.

Next, correction of "line bending" is performed. As described above, the line bending is caused by a shift in the image forming position caused by the difference in the height of the document depending on the position of the document. This shift occurs in the main scanning direction of the CCD line sensor.

When the deviation of the image forming position between the flat state and the raised state of the document is defined as the deviation amount in the object plane direction, the deviation amount X is expressed as follows: X = dh / (b + h) (3) (Where h is the height of the document surface, b is the distance from the height to the lens, and d is the distance from the center position of the lens to the read pixel in the main scanning direction on the document surface).

This shift amount can be expressed by the number of pixels of the CCD line sensor. Here, when reading a point on a document, the number of pixels from the pixel at the center position of the CCD line sensor to the pixel to be read at this point is represented by n.
Where Δc is the shift amount between the flat and curved original surfaces on the pixels of the CCD line sensor, and Δn is the number of pixels corresponding to the shift amount, Δn = nh / (b + h) .. (4) holds. Therefore, if the height h of the document and the distance b from the height to the lens are known, the shift amount can be corrected.

In the present embodiment, the height h is obtained from the image data of the upper surface b as described above, and the distance b from the height to the lens is determined by the imaging magnification and the focal length of the lens. Since the height of the document surface is corrected for torsional distortion as described above, distortion occurs even if the height on the document surface is assumed to be the same in the main scanning direction. Therefore, based on the height data obtained from the image of the upper surface b, Δn is calculated using the above equation (4) over the entire surface of the document, and image data obtained during the main scan by Δn pixels is calculated. Line shifting is corrected by shifting. The basic principle of the correction of the line bending is disclosed in Japanese Patent Application Laid-Open No. 5-161002, so that the detailed description is omitted.

Next, "character shrinkage" is corrected. As described above, the character shrinkage is such that the character is shrunk and read when the height of the document changes sharply. Therefore, the character contraction is corrected by extending the read original in the sub-scanning direction.

Specifically, as shown in FIG. 9, for example, from the image of the upper surface b reflected on the distance measuring mirror, the actual length ΔL in the document lateral direction in a certain section Δx is determined in advance in the sub-scanning direction. The correction rate (enlargement rate) ΔL / Δx is obtained from the divided section Δx and the height difference Δh in the section.
Then, the correction rate ΔL / Δx is multiplied by the number of lines in the sub-scanning direction actually read during the section Δx,
The number of lines when the distance of the actual length ΔL of the document is assumed to be read is obtained. Then, the number of lines actually read in this section Δx is evenly allocated to the number of lines assuming that the distance of the actual length ΔL of the document has been read, and repeatedly output, thereby outputting an image in the sub-scanning direction. Is stretched, and character shrinkage is corrected. The basic operation of this specific example is the same as that disclosed in Japanese Patent Application Laid-Open No. 7-254980, and a detailed description thereof will be omitted.

Since the correction of the character shrinkage is performed on the image data after the correction of the torsional distortion as described above (and the data after the correction of the line bending), Δh used here is measured. The height distribution data obtained from the image of the upper surface b reflected on the distance mirror may be used.

As described above, by performing the correction of the “twist distortion”, the “line bending”, and the “character shrinkage”, FIG.
A rectangular image without distortion is obtained as shown in FIG.

Next, the configuration and operation of a control unit that performs the above-described correction processes will be described.

FIG. 10 is a block diagram showing the configuration of the control unit of the present apparatus. In this control unit, first, an image read by the CCD line sensor 8 at the time of the preliminary scan is stored in the shape detection memory 21. The images read at the time of the preliminary scan are an image of the original surface a and an image of the upper surface b of the original reflected on the distance measuring mirror 5. In this embodiment, an image is read every eight dots in both the main scanning direction and the sub-scanning direction. This is because, at the time of the preliminary scan, the image data required therefor is for calculating various correction data, so that the entire image data is not necessarily required. This is because the point data is sufficient, and the capacity of the memory for storing the preliminary scan data can be reduced.

The CPU 22 has a ROM 23 connected thereto.
By controlling the image reading operation at the time of the preliminary scan by executing the program in the above, the respective lengths L ＄＄ to 4 shown in FIG. Calculation is performed, and based on the calculated correction parameters, correction parameters for performing the correction of “twist distortion”, “line bending”, and “character shrinkage” are obtained. Also, CP
U22 denotes a scan motor 30 for moving the CCD line sensor 8 in the sub-scanning direction and A for focusing.
The F motor 31 is controlled. The AF motor 31 does not move as a fixed focus at the time of the preliminary scan before the document height is detected, but at the time of the main scan after the height distribution is detected from the image data obtained by the preliminary scan, the AF motor 31 does not move. Based on the distribution data, successive focusing is performed in accordance with movement in the sub-scanning direction. Note that C
The RAM 24 connected to the PU 22 is a memory used by the CPU 22 for various processes.

On the other hand, during the main scan, the image read by the CCD line sensor 8 is temporarily stored in the image memory 26 in the torsion distortion correction circuit 25.
The data stored at this time is an image of the document surface. An address is assigned to each pixel of the image read in the main scanning direction and the sub-scanning direction, and the stored data is stored. This address is an address corresponding to each pixel of the CCD line sensor 8 in the main scanning direction, and is an address for each line in the sub-scanning direction accompanying movement in the sub-scanning direction.

Then, using the data in the image memory 26, the CPU 22 causes the shape detection memory 2 to perform a preliminary scan.
The above-described torsion distortion correction is performed using a correction parameter (corresponding to the above-described enlargement magnification X) for torsion distortion correction calculated based on the data stored in 1. At this time, an address is generated by the read address generation circuit 27, and the image is read out in the vertical direction along the center line c of the image, and the torsional distortion is corrected.

Thereafter, the CPU 22 instructs the data after the twist distortion correction to be sent to the character shrinkage correction circuit 28 to correct the character shrinkage, and then sends the data after the character shrinkage correction to the line skew correction circuit 29. Is performed, and the line is corrected, and then output to the outside.

FIG. 11 is a flowchart for explaining the operation procedure in the control unit.

First, a preliminary scan is started (S1), and an image on the distance measuring mirror 5 and an image on the document table 1 are read.
The data is loaded into the shape detection memory 21 (S2). Then, after all necessary data has been fetched, the preliminary scan ends (S3).

Next, the image of the distance measuring mirror 5 in the image data fetched into the shape detection memory 21 by the preliminary scan is analyzed, and the image of the upper surface b of the document B reflected on the distance measuring mirror 5 is used as the document surface. Is obtained (S4).

Next, the position of the center line c is detected from the image of the book document B on the document table 1 and its length is obtained (S5).

Next, the position and length of the lateral side edge d are determined for each of the right and left pages with respect to the center line c (S6).

Next, for each of the right and left pages with respect to the center line c, the true length L # in the horizontal direction obtained by extending the upper edge portion is obtained (S7).

Next, based on the position data of the center line c obtained in step S5 and the position data of the left and right lateral side ends b obtained in step S6, the length L2 and the lower end of the upper end b 'of the read image are obtained. The length L3 of e 'is obtained (S
8).

Next, based on the obtained L ＄＄ to L3, a correction amount for torsional distortion correction, that is, L2 / L
3 and the magnification X shown by the equation (1) are obtained (S9). Subsequently, the correction amount of the line bending and the correction amount of the character contraction are obtained (S
10).

Next, the main scan is started (S11), and the image of the book document B is read while focusing the lens according to the height distribution data, and all the images are stored in the image memory 26 (S12). After reading all the images, the main scan ends (S13).

At the same time as the end of the main scan, the image of the book document B stored in the image memory 26 is read, and
The correction of torsional distortion, the correction of line bending, and the correction of character shrinkage are performed (S14), and the corrected image data is output to the outside (S15).

As described above, in the present embodiment, even if the book original placed on the original platen has a twisted distortion, this is determined by obtaining data representing the three-dimensional shape of the original. By correcting the read image based on this to a planar image, a clear image without distortion can be output.

The present invention is not limited to the above-described embodiment, but can be variously modified without departing from the scope of the present invention.

[0071]

As described above, according to the first aspect of the present invention, the center line and the side edge of the original are detected from the image of the original itself, and the detected center line and the side edge are further detected. From the top to the bottom of the document, the length of the document in the horizontal direction is corrected by the correction means, so that one or both pages of the book document such as a book are distorted and placed. Even in such a case, it is possible to output a clear image by correcting the image distortion due to the twist distortion of the original document itself.

According to the second aspect of the present invention, in addition to the configuration of the first aspect, the height of the surface of the original is detected, and the focus of the imaging means is adjusted according to the height. Therefore, in addition to the effect of the configuration of the first aspect, it is possible to obtain a clear image with no defocus.

[Brief description of the drawings]

FIG. 1 is a perspective view showing an overall configuration of an image reading apparatus to which the present invention is applied.

FIGS. 2A and 2B are diagrams showing a state in which a book document is set in the above-described apparatus, wherein FIG. 2A is a view of the present apparatus as viewed from the front, and FIG. FIG.

FIG. 3 is a schematic diagram illustrating a case where a document is placed on a document table of the apparatus without distortion.

FIG. 4 is a schematic view of a case where a right page of a document is placed on the document table of the above-described apparatus so as to be distorted downward.

FIG. 5 is a schematic diagram showing a case where a right page of a document is placed on the document table of the above-described apparatus while being distorted widely;

6 is an enlarged view of the right page of the document as shown in FIG.

FIG. 7 is a view for explaining a method of correcting a distortion of a document by the above-described apparatus.

FIG. 8 is a diagram showing a state of correction by the above device.

FIG. 9 is a view for explaining character shrinkage correction by the above device.

FIG. 10 is a block diagram illustrating a configuration of a control unit that performs correction in the device.

FIG. 11 is a flowchart showing a procedure of an image reading operation by the apparatus.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Original platen, 2 ... Scanner unit, 7 ... Lens, 8 ... CCD line sensor, 21 ... Shape detection memory, 22 ... CPU, 25 ... Torsion distortion correction circuit, 26 ... Image memory, 27 ... Read address generation circuit, 28 ... Character contraction correction circuit, 29 ... Line bending correction circuit.

Claims (2)

[Claims] 1. An image reading apparatus for reading an image on the surface of a document placed upward on a platen from above the document, an image pickup unit for reading the document placed on the platen, and the image pickup unit A lateral side end position detecting means for detecting a position of a lateral side end of the document from an image of the document itself read by the document; a document center line detecting means for detecting a position of a center line of the document; The position of the lateral side edge detected by the end position detecting means and the vertical direction from the position of the center line detected by the document center line detecting means to the upper end to the lower end of the image of the document. An image reading device comprising: a correction unit configured to correct a length in a direction. 2. The image reading means can read an upper surface of the document; the image reading means further includes a height detecting means for detecting a height of the surface of the document from an image on the upper surface; 2. The image reading apparatus according to claim 1, further comprising: focusing means for performing focusing when the image pickup means reads a document image according to the height of the document surface detected by the document detection means.