JP2001127984A - Image processing apparatus and image-reading apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make the step of joining part in the partial images of a composite image inconspicuous by correcting slight deviations in the reading position of the partial images which is caused by plays, etc., in a driving part in an image processing apparatus for composting the image from the parting images. SOLUTION: The first divided image is photographed (S1), the luminance data of a feature image is calculated from the image pickup data (S2), the second divided image is photographed (S3) in advance and the luminance data of the feature image is calculated from the data (S4). A position deviating amount between the feature images is calculated from the luminance data of the feature images (S5), an image division input means and an original stand are relatively moved by the position deviation amount (S7), the second division image is photographed (S8), and the both division images are composited (S9). Thus, the slight position deviation between the feature image is corrected.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus and an image reading apparatus, and more particularly to a technique for dividing an original into a plurality of partial images and taking them into one whole image.

[0002]

2. Description of the Related Art In the field of a conventional image processing apparatus, there is an image processing apparatus in which a document is divided into a plurality of partial images, taken in, and combined into one whole image. In this type of apparatus, a plurality of feature points are provided on a document table, and feature points in each partial image read in the absence of a document and feature points in each partial image read in the presence of a document are superimposed. In some cases, these divided images are combined into one whole image (see, for example, JP-A-8-263626). In addition, there is an image processing apparatus that shortens the time required for synthesizing images by restricting a range to be determined as a displacement of an image to a common overlap portion when synthesizing images (for example,
See JP-A-11-195104). Furthermore,
By providing reference point display means between the document surface and the solid-state imaging device, and correcting the shift amount of the partial image of the document based on the imaging information of the reference point displayed on the reference point display means, automatically In Japanese Patent Application Laid-Open No. 10-13661, a correction is made at the time of joining partial images. Also, by comparing the contents of the overlapping regions of the respective partial images, it is possible to determine whether or not there is a feature portion of the image that can be used for superimposing and combining in the overlapping regions. (See, for example, JP-A-7-273976). Further, by setting the values of the x-coordinate or y-coordinate of the first feature point and the second feature point in the reference partial image to be the same, the detection speed of the shift amount when the respective partial images are pasted together is reduced. There has been an improvement (see, for example, JP-A-7-284013).

[0003]

However, the image processing apparatus disclosed in Japanese Patent Application Laid-Open No. 8-263626 has a problem in that it is necessary to read twice with and without a document. Further, in any of the apparatuses disclosed in the above publications, the subtle pixel-by-pixel (unit of one pixel or less) displacement of the reading position of each partial image due to the play and vibration of the driving unit, the variation of the driving actuator, and the like. Cannot be corrected, and
There is a problem that a step occurs at a joint portion of each partial image in the composite image.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problem, and when an original is divided into a plurality of partial images and taken in and combined into one image, the play of a driving unit is reduced. An image processing apparatus and an image reading apparatus capable of correcting a slight shift of the reading position of each partial image caused by the correction so as to make a step of a joint portion of each partial image in the composite image inconspicuous. It is intended to provide a device.

[0005]

According to a first aspect of the present invention, a document set on a document table having one or more characteristic images outside a document area is divided into a plurality of partial image areas. An image processing apparatus comprising: an image division input means for dividing the image into a plurality of partial images captured by the image division input means; and an image synthesis means for synthesizing a plurality of partial images into an entire image. A plurality of adjacent images are captured and captured so as to include one or more outer characteristic images and have portions overlapping each other, and the first partial image area captured by the image division input means First data calculating means for calculating data of the characteristic image of the second partial image area, and a second partial image based on data acquired by preliminary imaging before imaging the second partial image area adjacent to the first partial image area A second data calculating means for calculating data of the characteristic image in the area, a data of the characteristic image in the first partial image area calculated by the first data calculating means, and a second data calculating means for calculating the data of the characteristic image in the first partial image area. 2 based on the data of the characteristic images in the partial image area, calculating the amount of positional deviation of these characteristic images, based on the calculation results by the positional deviation amount calculating means, Moving means for moving the relative position of
The image division input means captures an image of the second partial image area after the relative movement by the moving means, and repeats capturing until the capture of the images of all the partial image areas is completed.

In the above arrangement, the data of the characteristic image in the first partial image area taken by the image division input means is calculated by the first data calculating means, and the second image adjacent to the first partial image area is calculated. The data of the characteristic image in the partial image area is calculated by the second data calculation unit based on the data acquired by the preliminary imaging. Then, from the data of these characteristic images, the amount of positional deviation between these characteristic images is calculated by the positional deviation amount calculating means, and the relative position between the image division input means and the document table is moved by the calculated positional deviation amount. After correcting the positional deviation between these characteristic images by moving the image, the second partial image area is imaged by the image division input means. By repeating this series of processing until the capture of the images of all the partial image areas is completed, it is possible to correct a subtle shift in the reading position of each partial image area based on the shooting position of the adjacent partial image area. it can.

Further, the first and second data calculation means include:
The image density or brightness of the edge portion of each feature image may be calculated. This makes it possible to accurately calculate the positions of the respective feature images in the first and second partial image areas. Therefore, based on these calculated values, the amount of delicate displacement between the two feature images can be calculated. Can be calculated.

[0008] The position shift amount calculating means calculates a position between the characteristic images based on a ratio of an image density or luminance of an edge portion of each characteristic image in the first partial image region and the second partial image region. The shift amount may be calculated. This makes it possible to accurately calculate the amount of displacement between these characteristic images.

According to a fourth aspect of the present invention, the entire image region is divided into a plurality of partial image regions so that the peripheral portions of the division boundary overlap each other, and the images are taken, and the partial images are synthesized so that the peripheral portions overlap each other. In the image reading device that outputs the entire image, the image pickup device and a pixel in a subpixel unit of the image pickup device between one partial image region and the other partial image region in a combined portion around the division boundary A pixel position shift amount detecting means for detecting a position shift amount, and at least one partial image of overlapping partial image regions according to a pixel position shift amount in sub-pixel units detected by the pixel position shift amount detecting means. Image forming means for forming an image displaced in sub-pixel units with respect to the image of the area.

In this configuration, in the combined portion around the division boundary, a pixel position shift amount in a sub-pixel unit between one partial image region and the other partial image region is determined by a pixel position shift amount detecting means. The image forming means forms an image which is displaced in units of sub-pixels with respect to the image of at least one of the overlapping partial image areas in accordance with the shift amount. As a result, it is possible to remove the influence of a shift in the pixel position in sub-pixel units between one of the overlapping partial image areas and the other partial image area.

[0011] The pixel position shift amount detecting means detects a pixel between these partial image areas based on an image density or luminance ratio of an edge portion of each characteristic image of the one partial image area and the other partial image area. The position shift amount may be detected. This makes it possible to accurately detect the shift amount of the pixel position between one partial image region and the other partial image region in subpixel units.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an image processing apparatus according to an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of a control system of the image processing apparatus according to the present embodiment. The image processing apparatus according to the present embodiment divides the entire image region into a plurality of partial image regions so that the peripheral portions around the division boundary overlap each other, takes an image, and synthesizes the peripheral portions of each partial image so that they overlap each other. An image reading device that outputs an image. The image reading apparatus 1 includes: an image division input unit 10 for dividing and photographing a document placed on a document table into a plurality of images; an image memory 20 for storing divided images acquired by the image division input unit 10; It comprises an arithmetic and control unit 100 for performing various arithmetic and control processes for creating a composite image from the divided images, and a moving unit 45 for moving the image division input unit 10. The image division input means 10 has an image sensor for capturing an image and an optical system for forming an image on a document table on the image sensor. Further, the arithmetic control unit 100 calculates the characteristic image data calculating means 30 (first data calculating means in claims) for calculating the data of the characteristic image 62a (see FIG. 4 (4)) on the platen included in the divided image. A second data calculation unit), a shift amount calculation unit 40 that calculates a shift amount of an adjacent divided image with respect to a reference divided image based on the data calculated by the characteristic image data calculation unit 30 (position shift in claims) (Amount calculating unit, pixel position shift amount calculating unit), and an image synthesizing unit 50 for synthesizing a plurality of divided images captured by the image division input unit 10 with the entire image.

At the time of reading an image, the shift amount of the adjacent divided image with respect to the reference divided image calculated by the shift amount calculating unit 40 is fed back to the image division input unit 10 via the moving unit 45. That is,
The arithmetic and control unit 100 moves the positional relationship between the image division input unit 10 and the original platen by the moving unit 45 in accordance with the amount of deviation calculated by the amount of deviation calculating unit 40 by the amount of the deviation. The divided image is taken in again by the divided input means 10. Then, the data of the re-captured divided image is sent to the image synthesizing means 50, and a synthesizing process such as masking of the area outside the document and erasing of the overlapping area is performed.

FIG. 2 shows a flowchart of the image processing method of the image reading apparatus 1 according to the present embodiment. Here, a case where the number of divided images is 2 is shown. Arithmetic control unit 100
When a start key (not shown) is pressed, the first divided image is photographed by the image division input means 10 (S1), and the calculation of the data of the characteristic image included in the first divided image is performed. This is performed by the image data calculation means 30 (S2).
Subsequently, the image division input unit 10 is relatively moved by a predetermined fixed amount with respect to the document table by the movement unit 45, and the second division image is preliminary photographed by the image division input unit 10 (S3). ). Here, since only data for examining the positional relationship between the first divided image and the second divided image is required, the data in the document area is not necessarily required. Next, the arithmetic control unit 100 uses the characteristic image data calculation means 30 to calculate the data of the characteristic image included in the second divided image that has been preliminarily photographed (S4). And
Based on the characteristic image data of the first and second divided images calculated in S2 and S4, the deviation amount δ between the first and second divided images is calculated by the deviation amount calculating means 40 (S5). The arithmetic control unit 100 compares the shift amount δ with a preset allowable amount (S6). If the shift amount δ is equal to or larger than the allowable amount (NO in S6), the arithmetic control unit 100 After the platen is relatively moved by δ by the moving unit 45 (S7), the second divided image is photographed by the image division input unit 10 (S8). If the shift amount δ calculated by the shift amount calculating means 40 is smaller than the allowable amount (YE in S6)
S), the moving operation is omitted. When the shift amount δ is smaller than the allowable amount, the data of the preliminary photographing (S3) includes the data in the document area, and if the data remains in the image memory, the divided image data at the time of the preliminary photographing is used. This eliminates the need for re-shooting the second divided image (S8). The arithmetic control unit 100 controls the first
The divided image and the second divided image are combined by the image combining means 50 (S9), and the image processing ends.

FIG. 3 shows a flowchart of the processing when the number of divided images is three or more. The processing of S11 to S19 in the figure respectively corresponds to the processing of S1 to S9 in FIG. Here, the number of divided images is n, and the I-th (I = 2 to n)
Is assumed to be the (I-1) -th divided image. This assumption corresponds to synthesizing these divided images after dividing and photographing an image that is long in one direction. In the case where the image is divided and photographed two-dimensionally (in two directions), an image as a bonding reference of the I-th divided image is determined according to the photographing order. In this case, there is a possibility that a plurality of standards exist depending on the number of bonded portions. Since the flow of the image processing in FIG. 3 is the same as that in FIG. 2, the details are omitted.

FIGS. 4A to 4D show how a document placed on a document table is divided and photographed. Fig. 4 (1)
Indicates a state where nothing is placed on the document table 60. On a document table 60, a plurality of characteristic image portions 62 serving as a reference for pasting the divided images are provided outside a document region 61 indicated by hatching.
Is provided. The number of divisions of the document image area is determined by setting the document size and the resolution at the time of reading. At this time, each divided image area always includes one or more feature image portions 62.
Are arranged so as to be commonly included in adjacent divided image areas. FIG. 4B shows a state where the document 63 is set on the document table 60. In the figure, the document 63 is set so as to be aligned with the end of the document area 61.
May be set anywhere within the document area 61, and the document 63 may be inclined. This is because such a case can be dealt with by performing an unnecessary portion masking process, an image rotation process, or the like as necessary after the image combining process. FIG. 4C shows a state where the document 63 is photographed by being divided into two images. In the figure, the first divided image area 64 and the second divided image area 65 are indicated by broken lines, respectively. These divided image areas have overlapping areas 66.
The characteristic image portion 62 exists in a portion of the overlapping area 66 outside the document area. FIG. 4 (4)
The state when the first divided image 64a and the second divided image 65a obtained by the preliminary photographing are stored in the image memory 20 is shown. After photographing the first divided image 64 a by the image division input means 10, the arithmetic control unit 100 moves the image division input means 10
(Not shown), the image division input means 1
According to 0, the second divided image 65a is photographed. The relative movement amount X of the image division input means 10 is a value theoretically calculated based on the number of image divisions determined according to the document size and the reading resolution. However, in practice, the relative movement amount X of the image division input unit 10 often does not become the theoretical value due to the play and vibration of the driving unit and the variation of the driving actuator. The main object of the present invention is to provide a subtle deviation of the relative movement amount X of the image division input means 10 due to these causes, and the fact that the cell size of the image sensor of the image division input means 10 and the relative movement amount X are always integral multiples. An object of the present invention is to obtain a good image by correcting a delicate shift of a photographing position caused by the fact that the image does not become blurred.

FIG. 5A shows the first state shown in FIG.
It is an enlarged view around each characteristic image 62a of the divided image 64a and the second divided image 65a. The cells in the figure correspond to the size of the image pickup device of the image division input means 10. FIG. 5A shows an example in which the feature image portion 62 is black and the background is white, but the combination of the feature image portion 62 and the background color or density is not particularly limited. FIG.
(2) shows the output value of each pixel on the line indicated by the arrow H in FIG. 5 (1), and is an output example of the characteristic image data calculating means 30 described above. Here, a luminance value is used as an output value. The brightness represents the brightness of a pixel. The larger the luminance value, the brighter (white), and the smaller the luminance value, the darker (black). In the image reading apparatus 1, the user can set an output value (luminance value) at an arbitrary stage. The above-mentioned displacement amount calculating means 4
In the case of 0, a smaller difference can be detected as the number of steps of the output value increases. Here, the output value is represented by 256 levels from 0 to 255.

Next, an example of calculating the shift amount by the shift amount calculating means 40 will be described. Here, it is assumed that no deviation occurs in a direction perpendicular to the direction indicated by the arrow H in FIG. By comparing the respective output data shown in FIG. 5 (2), the line C of the first divided image 64a and the line c or d of the second divided image 65a obtained by preliminary photographing are compared.
Can be estimated to be a line to be superimposed. However, if the output value of the line C of the first divided image 64a is expressed as V (C) (the same applies hereinafter), the output value V (C) becomes the second value. Neither the output value V (c) of the line c nor the output value V (d) of the line d of the divided image 65a match. Here, focusing on the line c of the second divided image 65a, the first divided image 6
The amount of deviation from line C of 4a is calculated. Assuming that the size of the pixel of the image sensor of the image division input means 10 is α, the shift amount δ can be calculated by the following equation.

Δ = {(V (C) −V (c)) / 255} × α (1)

When the shift amount δ shown in the above equation (1) is negative, the output value V (C) of the line C of the first divided image 64a is changed to the second divided image 65a by the preliminary photographing.
Is smaller than the output value V (c) of the line c (line C is darker than the line c). A good image having no step is obtained. On the other hand, when the shift amount δ is positive, the output value V (C) of the line C of the first divided image 64a is equal to the second value.
Is larger than the output value V (c) of the line c of the divided image 65a (the line C is brighter than the line c), and if the second divided image 65a is shifted to the right by δ, the step of the bonding portion No good image is obtained. Here, in order to facilitate understanding, based on the output value V (C) of the line C of the first divided image 64a and the output value V (c) of the line c of the second divided image 65a, Although the shift amount and the shift direction of the relative position between the first divided image 64a and the second divided image 65a have been calculated, actually, the characteristic image 62a
It is necessary to calculate the shift amount and the shift direction based on the output data of a plurality of lines constituting the edge portion of the edge portion. Therefore, the shift amount is calculated based on the output values of about three consecutive pixels.

FIG. 6A shows a composite image created by superimposing the line C of the first divided image 64a and the line c of the second divided image 65a obtained by preliminary photographing in FIG. 5A. 6 (2) is a line C of the first divided image 64a
And a line c of the second divided image 65a captured by moving the image division input unit 10 by the shift amount δ, and a composite image created by superimposing the line c. As shown in FIG. 6A, when the first divided image 64a and the second divided image 65a obtained by the preliminary photographing are synthesized as they are, a step occurs at the image joining portion indicated by the arrow J. As shown in FIG. 6 (2), the image division input means 10 is moved from the photographing position at the time of preliminary photographing by the shift amount δ, and the second divided image 65a is moved.
Is re-photographed, and the re-photographed second divided image 65a and the first divided image 64a are combined, so that there is no step in the combined image.

As described above, according to the image reading apparatus 1 of the present embodiment, the brightness value of the edge portion of the characteristic image 62a in the first divided image 64a captured by the image division input means 10 and the first From the luminance value of the edge portion of the feature image 62a in the second divided image 65a adjacent to the divided image 64a, a subtle positional deviation amount δ between these characteristic images is calculated, and only the calculated deviation amount δ is calculated. , Image division input means 10
The second divided image 65a is taken after the relative position between the first divided image 64a and the original platen 60 is moved, so that the driving unit may cause the play based on the photographing position of the adjacent first divided image 64a. It is possible to correct a slight shift in the shooting position of the second divided image 65a. For this reason, it is possible to make the step at the joint of the divided images 64a and 65a in the composite image inconspicuous.

The present invention is not limited to the above embodiment, and various modifications are possible. For example, in the above-described embodiment, the image division input means 10 is moved to correct the relative position deviation between the image division input means 10 and the document table 60. The platform 60 may be moved. Further, in the above-described embodiment, the image division input unit 10 is moved from the photographing position at the time of the preliminary photographing by the shift amount δ to photograph the second divided image 65a again, and the second divided image 65a thus photographed is used. Although the composite image is created, the composite image may be created using the divided image 65a by displacing the second divided image 65a obtained by the preliminary photographing by the shift amount δ. Furthermore, in the above embodiment, the brightness value of the edge portion of the feature image 62a is used as the data for detecting the position of the feature image 62a, but the image density value of the edge portion of the feature image may be used.

[0023]

As described above, according to the first aspect of the present invention,
When combining divided images and combining them into a single image, the amount of misalignment between the characteristic images is calculated from the data of the characteristic images in the divided image region, and the read position of the divided image region is delicate. Since the misalignment is corrected, the step of the bonded portion can be made inconspicuous, and a good composite image can be obtained.

Further, by calculating the image density or luminance of the edge portion of the characteristic image, the position of each characteristic image in the partial image area can be calculated accurately.

Further, the amount of positional deviation between the characteristic images is calculated based on the image density or the luminance ratio of the edge portion of the characteristic image, so that the amount of positional deviation can be accurately calculated.

According to the fourth aspect of the present invention, the shift amount of the pixel position in the sub-pixel unit between the partial image areas is detected, and one of the overlapping partial image areas and the other partial image area are detected. Since the effect of the displacement of the pixel position in the sub-pixel unit with the region is removed, the step of the joint portion of each partial image in the composite image is made inconspicuous, and a good composite image is obtained. Can be.

Further, by detecting the shift amount of the pixel position between the partial image areas based on the image density or the luminance ratio of the edge portion of the characteristic image, the shift amount can be accurately detected.

[Brief description of the drawings]

FIG. 1 is a block diagram of a control system of an image processing apparatus according to an embodiment of the present invention.

FIG. 2 is a flowchart illustrating an image processing method of the image processing apparatus when the number of divided images is two.

FIG. 3 is a flowchart illustrating an image processing method when the number of divided images of the image processing apparatus is three or more.

4A is a diagram showing a state where nothing is placed on a platen, FIG. 4B is a diagram showing a state where a document is placed on a platen, and FIG. FIG. 4D is a diagram illustrating a state in which the first divided image and the second divided image obtained by the preliminary photographing are stored in the image memory.

5A is an enlarged view around a characteristic image of each of a first divided image and a second divided image, and FIG. 5B is an output value of each pixel on a line indicated by an arrow H in FIG. FIG.

6A is a line C of a first divided image in FIG. 5A and a line c of a second divided image obtained by preliminary photographing.
FIG. 2B is a diagram illustrating a composite image created by superimposing the image and the image. FIG. 2B illustrates a line c of the second divided image and a line C of the first divided image photographed by moving the image division input unit by the shift amount δ. It is a figure showing the synthetic picture created by superimposing.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Image reading apparatus (image processing apparatus) 10 Image division input means 30 Characteristic image data calculation means (first data calculation means, second data calculation means) 40 Shift amount calculation means (position shift amount calculation means, pixel position shift) Amount calculating means) 45 moving means 50 image synthesizing means 62a characteristic image 64a first divided image (first partial image) 65a second divided image (second partial image)

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5B047 AA01 BA02 BB06 BC15 CA14 CB09 CB30 DC06 DC09 5B057 AA11 CA16 CB16 CC02 CE08 CE10 DA07 5C076 AA14 AA19 AA36 BA01 BA03 BA04 BA05 BA06

Claims (5)

[Claims] An image division input unit configured to divide a document placed on a document table having one or more characteristic images outside a document region into a plurality of partial image regions and to capture an image; An image processing apparatus comprising: an image synthesizing unit that synthesizes a plurality of captured partial images into an entire image, wherein the image division input unit includes one or more characteristic images outside a document area, and A first data calculation unit that captures and captures a plurality of adjacent images so as to have an overlapping portion, and calculates data of a characteristic image in a first partial image region captured by the image division input unit; Calculating the data of the characteristic image in the second partial image area based on the data acquired by preliminary imaging before imaging the second partial image area adjacent to the first partial image area; Data calculation means, data of the characteristic image in the first partial image area calculated by the first data calculation means, and data of the characteristic image in the second partial image area calculated by the second data calculation means A position shift amount calculating unit for calculating the position shift amounts of these characteristic images from the data; and a moving unit for moving a relative position between the image division input unit and the platen based on a calculation result by the position shift amount calculating unit. The image division input means captures a second partial image area after the relative movement by the moving means, and repeats the imaging until the capture of the images of all the partial image areas is completed. Image processing device. 2. The first and second data calculation means,
2. The image processing apparatus according to claim 1, wherein an image density or a brightness of an edge portion of each characteristic image is calculated. 3. The image processing apparatus according to claim 1, wherein the position shift amount calculation unit calculates a characteristic image of each of the first partial image area and the second partial image area based on a ratio of image density or luminance of an edge portion of the characteristic image. The image processing apparatus according to claim 1, wherein an amount of misregistration is calculated. 4. A whole image area is divided into a plurality of partial image areas so that the peripheral parts of the divisional boundary overlap each other, imaged, and synthesized so that the peripheral parts of each partial image overlap each other to output an entire image. In the image reading device, a shift amount of a pixel position of the image sensor in a sub-pixel unit between one partial image area and the other partial image area is detected in an image sensor and a combined portion around a division boundary. A pixel position shift amount detecting unit that performs, according to a shift amount of a pixel position in a sub-pixel unit detected by the pixel position shift amount detecting unit, an image of at least one of the overlapping partial image regions. And an image forming means for forming an image displaced in units of sub-pixels. 5. The image forming apparatus according to claim 1, wherein the pixel position shift amount detecting unit determines a difference between the partial image areas based on an image density or a luminance ratio of an edge portion of each characteristic image of the one partial image area and the other partial image area. The image reading device according to claim 4, wherein a shift amount of a pixel position is detected.