JP2727983B2 - Document reading device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document reading apparatus,
More particularly, the present invention relates to an original reading apparatus suitable for an OCR apparatus or a facsimile having a CCD linear image sensor.

[0002]

2. Description of the Related Art An image reading apparatus such as an OCR apparatus, an image sensor, and a facsimile includes a reading light source (LED, cold-cathode tube, fluorescent light, or the like) for irradiating a reading original with light;
A lens array for forming an image of reflected light from a read original, and a reading element array (CCD on which a photoelectric conversion element is mounted) for converting the reflected light imaged by the lens array into an electric signal
Linear image sensor).

That is, the density of a read document is output from a reading sensor as an electric signal.

In a conventional image reading apparatus, the image quality is improved by shading correction or gamma correction of density information from a reading sensor.

Further, as disclosed in Japanese Patent Application Laid-Open No. 4-358468, a deviation from a correct pitch is defined as a moiré by using an equal-magnification detection chart formed so that a black-and-white pitch matches a pitch of a read pixel. There has been proposed a method for detecting and correcting the amount of deviation to correct the reading level of each pixel.

[0006]

However, in the above conventional example, since the lens distortion characteristic is not linear as shown in FIG. 8 due to optical design or manufacturing distortion of the lens array, FIG. As shown in the figure, there is a problem that the image is distorted and the recorded image quality such as moire is deteriorated.

This inconvenience can be solved by designing the curvature of the lens with high precision and manufacturing it faithfully, or by using a method of shifting the reading pixels of the reading sensor by the amount of lens distortion based on the lens distortion characteristics. The problem that cost becomes very high arises.

[0008]

SUMMARY OF THE INVENTION It is an object of the present invention to improve the disadvantages of the prior art, and in particular, to correct a pixel shift in the main scanning direction due to lens distortion by image processing automatically and without causing image deterioration. A reading device is provided.

[0009]

Therefore, according to the present invention, there is provided a reading sensor for converting density information of a read document into an electric signal,
An image processing unit is provided for generating correction data for correcting a positional deviation due to lens distortion of the reading sensor and for correcting image data from the reading sensor based on the correction data. This aims to achieve the above-mentioned object.

[0010]

In the correction mode, when the reading sensor reads in advance an oblique original with a clear inclination and notifies the image processing unit, the image processing unit compares the theoretical image data with the image data read by the reading sensor and reads the document. Correction data for correcting a positional shift amount in the main scanning direction caused by lens distortion of the sensor is created.

In the reading mode, the image processing section shifts the image data from the reading sensor in the main scanning direction based on the correction data, or corrects the line spacing, so that the image data faithful to the read original without image distortion or moiré can be obtained. To fix.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to FIGS.

In the embodiment shown in FIG. 1, a reading sensor 3 for converting density information of a read document into an electric signal, an A / D converter 4 for digitizing a reading signal from the reading sensor 3, and various data are stored. The image processing unit 1 includes a RAM 2, an image processing unit 1 that corrects image data from the A / D converter 4 based on correction data stored in the RAM 2, and a ROM 5 that stores processing procedures in the image processing unit 1. .

Also, a printer control unit for printing the corrected image data, a modem for transmitting the corrected image data to another device using a public telephone line, and an NCU (Network Communication)
Unit: a network control device), an operation display control unit, an LCD display, and operation buttons for guiding the operation procedure of the user and for the user to input various information such as a transmission destination.

As shown in FIG. 2, the reading sensor 3 includes a reading light source 3a for irradiating the reading document with light, and a lens array 3b for imaging reflected light from the reading document.
And a read element array 3c for converting the reflected light imaged by the lens array 3b into an electric signal.

That is, the light emitted from the reading light source 3a and reflected by the read original becomes a light amount corresponding to the density of the image at the reflection position on the read original, and is read by the lens array 3b on the read element array 3c having the photoelectric conversion element. And is converted into an electric signal proportional to the amount of light received by the reading element.

As the reading light source 3a, an LED, a cold cathode tube, a fluorescent lamp, or the like is used.

As the reading element row 3c, a CCD linear image sensor or the like on which a photoelectric conversion element is mounted is used.

The RAM 2 has an image memory section 2a for storing image data from the A / D converter 4, a correction data storage section 2b for storing pitch shift amount correction data for each reading element, and a pitch shift amount correction data. And a corrected image memory unit 2c for storing corrected image data obtained by subjecting the image data to pixel shift correction based on the corrected image data.

It is also possible to add an image processing IC dedicated to correction processing to the image processing section 1 in order to improve the processing speed.

Next, a description will be given, with reference to the flowchart of FIG. 3, of a process of creating pitch shift amount correction data in the image processing section 1 in the present embodiment. Here, the reading element row 3
The direction parallel to c is the main scanning direction, and the direction perpendicular to the reading element row 3c is the sub-scanning direction.

(1). First, an original for correction on which one oblique line is drawn is read. Here, the width of the correction document is larger than the reading width of the reading sensor 3.

The inclination K of the oblique line is preferably such that the original feed amount in the sub-scanning direction and the pixel pitch in the main scanning direction are 1: 1.

(2). The read signal from the read sensor 3 is A
It is stored as image data in the image memory unit 2a via the / D converter 4 (step S1 in FIG. 3).

(3). The image processing unit 1 calculates an oblique line inclination k from the image data using the following equation (1) and compares it with a preset value K (step S2 in FIG. 3,
S3).

[0026]

(Equation 1)

This is for correcting a displacement that occurs when the correction document is inserted into the reading device.

In equation (1), Y (t) is the original feed amount in the sub-scanning direction at time t, P (Y (t)) is the black detection start pixel position in the main scanning direction at time t,
Y (t + x) is a document feed amount in the sub-scanning direction at time t + x, and P (Y (t + x)) is a black detection start pixel position in the main scanning direction at time t + x.

(4). If the slope k of the oblique line calculated based on the image data is significantly different from the preset value K, the image processing section 1 returns to the above-described process (1) again (step S4 in FIG. 3). ).

On the other hand, if the inclination k of the oblique line calculated based on the image data is not significantly different from the preset value K, the inclination of the oblique line calculated based on the image data is defined as K. The following processing is performed (step S5 in FIG. 3).

(5). The theoretical value of the hatched position at the time of the document feed amount Y (t) is obtained by the following equation (2) (step S in FIG. 3).
6).

[0032]

(Equation 2)

Here, mp is a position where white changes to black, and mo is a position where black changes to white. Also, A
Is the width of the oblique line, and C is a constant.

(6). Based on the image data stored in the image memory unit 2a, the document feed amount Y
The position P at which the color changes from white to black and the position O at which the color changes from black to white at the time (t) are obtained (step S7 in FIG. 3).

[0035]

(Equation 3)

The reading element 3c is set to output a signal of level 1 when the reading position of the document is pure white, and to output a signal of level 0 when the reading position of the document is pure black. When a white portion and a black portion are mixed, a signal having a level proportional to the ratio is output.
That is, when the ratio of the white portion to the black portion is 3: 1, the signal of level 0.75 is used. When the ratio is 1: 1, the signal of level 0.5 is used. When the ratio is 1: 3, the signal is used. Outputs a signal of level 0.25.

Therefore, when oblique lines are read, FIG.
As shown in the figure, the image data near the boundary between the black portion and the white portion of the document has a level between level 0 and level 1.

In the equation (3), the position of the image data in the main scanning direction at which the level becomes 0 first is defined as the point B, and the image data in the main scanning direction from the scanning start position to the point B is integrated. From black to black can be determined.

Further, the image data from the point B to the scanning end position in the main scanning direction is integrated, and the total number of pixels N in the main scanning direction is calculated.
From the point of change O from black to white in the main scanning direction.
Can be requested.

(7). The difference Δmp between the theoretical value and the measured value at the position where the color changes from white to black and the difference Δmo between the theoretical value and the measured value at the position where the color changes from black to white are calculated from the following equation (4) (FIG. 3). Step S8).

[0041]

(Equation 4)

△ mp and △ mo indicate the amount of positional relationship deviation.

(8). Δmp and ΔMo offset from mo △
Find m.

Here, the positional deviation amount Δm is represented by Δmp and Δm.
whether to average o or use only one of the data,
If neither data is valid, 0 or the data of the closest element is quoted and set as Δm.

The detection of the edges on both sides of the oblique line is for improving the accuracy.

(9). The displacement amount Δm is stored in the correction data storage unit 2b (Step S9 in FIG. 3).

To create correction data as described above,
The effect of the lens distortion on the image can be accurately quantified.

Next, the pixel shift correction processing in the image processing section 1 in the present embodiment will be described with reference to the flowchart in FIG.

Here, it is assumed that the displacement amount Δm for all the pixels in the main scanning direction has already been stored in the correction data storage unit 2b. L (m) is image data by the m-th reading element in the main scanning direction. Hereafter, m is referred to as a pixel number.

(1). When the user sets a document to be read at a predetermined position and instructs to start reading, the image processing unit 1 instructs a document feeding unit (not shown) to take in the document and instructs the reading sensor 3 to start reading.

(2). The read signal from the read sensor 3 is A
The image data is stored as image data in the image memory unit 2a via the / D converter 4.

(3). The image processing section 1 refers to the correction data #m stored in the correction data storage section 2b and corrects the image data stored in the image memory section 2a (see FIG. 5).
(Step S11), and stored in the correction image memory 2c (Step S12 in FIG. 5).

This makes it possible to match the pixel pitch of the document.

That is, the correction data Δm is separated into an integer Q and a decimal R, and a correction operation is performed using the following equation (5). For example, when R = 0 for the pixel number m, the image data L (m) in the image memory unit 2a is stored at the storage position of the pixel number (m + Q) in the corrected image memory unit 2c.

[0055]

(Equation 5)

In the case of R ≠ 0, if 正 m is positive, the image data L (m) and L (m) stored in the pixel number m and the pixel number (m−1) of the image memory unit 2a. -1) is corrected between lines according to the value of R, and is stored in the pixel number (m + Q) of the corrected image memory unit 2c.

On the other hand, if Δm is negative, the image memory unit 2a
Of the image data L (m) and L (m + 1) stored in the pixel number m and the pixel number (m + 1) are corrected according to the value of R to obtain the pixel number (m + Q) of the corrected image memory unit 2c. Store.

When correction is sequentially performed for each pixel number, there is a case where the change point of Q does not correspond to the corrected image memory unit 2c. In this case, the image memory unit 2a corresponding to the pixel number has Line spacing is corrected from image data.

Next, the above-described line-to-line correction will be described with reference to schematic diagrams. As shown in FIGS. 6 and 7, when the original density of the read document at a certain document feed amount is a thick line,
Assuming that Δm = −0.5, the image data is quantized as shown by the solid line, and becomes 0.
The image is read shifted by 5 pixels to the left. Therefore, when the pixel is shifted to the right by 0.5 pixel by the correction, the position shown by the dotted line is obtained. Get the data.

(4). The image processing unit 1 includes an image memory unit 2a
The maximum value L (BK) of the continuous black signal and the minimum value L (WH) of the continuous white signal of the pixel density are compared with the density of neighboring pixels, and the pixel number M (BK) or M ( WH).

Then, the pixel number M (BK) or M (W
With respect to H), a pixel number is obtained by shifting the correction data Δm by the number of rounded pixels by rounding to an integer value.

The correction image memory unit 2 corresponding to this pixel number
In the storage position of c, the pixel number M (BK) or M (WH)
Is stored as priority data.

This is called a peak holding function (step S13 in FIG. 5).

As shown in FIG. 6, assuming that the pixel number (m + 1) is the peak position of the black signal in the image data and that the correction data Δm is 1.1, the correction data Since the result is 1 when rounded off, the image data L (m) at the pixel number (m + 1)
+1) is stored at the position of the pixel number (m + 2) in the correction image memory unit 2c.

This is because the peak value is lost due to the line-to-line correction, and as shown in FIG. 7, the peak value of the image data indicated by the solid line is present to restore the peak value. The pixel numbers are stored preferentially as indicated by hatching near the positions of the corrected pixel numbers.

As described above, according to the present invention, in order to correct the pixel shift in the main scanning direction due to the distortion of the lens by image processing, the accuracy required for the design and manufacture of the lens array 3b can be relaxed. Thus, an inexpensive document reading device can be supplied.

Further, since the correction value of the pixel shift amount can be automatically created, the handling is simplified, and
It is possible to prevent the image from deteriorating due to the correction of the pixel shift.

[0068]

Since the present invention is constructed and functions as described above, according to this, it is possible to automatically correct the pixel shift in the main scanning direction due to lens distortion by image processing.
This makes it possible to relax the required accuracy in the design and manufacture of the lens array, supply it at low cost, and prevent the degradation of the grayscale signal of the image due to the correction of the pixel shift. Can be provided.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing one embodiment of the present invention.

FIG. 2 is an explanatory diagram for explaining a reading sensor in the embodiment of FIG. 1;

FIG. 3 is a flowchart illustrating a correction data creation process for correcting a displacement due to lens distortion in the embodiment of FIG. 1;

FIG. 4 is an explanatory diagram for explaining correction data creation processing in the embodiment of FIG. 1;

FIG. 5 is a flowchart illustrating a process of correcting image data in the embodiment of FIG. 1;

FIG. 6 is an explanatory diagram showing an example of a process of correcting image data in the embodiment of FIG. 1;

FIG. 7 is an explanatory diagram showing an example of processing for correcting a line gap in the embodiment of FIG. 1;

FIG. 8 is an explanatory diagram for explaining a positional shift due to lens distortion.

FIG. 9 is an explanatory diagram for explaining image distortion due to lens distortion.

[Explanation of symbols]

 Reference Signs List 1 image processing unit 2 RAM 2a image memory unit 2b correction data storage unit 2c correction image memory unit 3 reading sensor 3a reading light source 3b lens array 3c reading element row 4 A / D converter 5 ROM

Claims (3)

(57) [Claims] In a document reading apparatus provided with a reading sensor for converting density information of a reading document into an electric signal, correction data for correcting a position shift due to lens distortion of the reading sensor is created and based on the correction data. A document reading apparatus, comprising: an image processing unit that corrects image data from the reading sensor. 2. The image processing unit according to claim 1, further comprising: a unit configured to generate correction data for correcting a position shift due to a lens distortion of the reading sensor based on a position shift amount in read data of the oblique original from the reading sensor. The document reading device according to claim 1, wherein: 3. The document reading apparatus according to claim 1, wherein said image processing unit includes means for holding image data at a density peak position when correcting the image data.