JP2904258B2 - Shading correction method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a shading correction method for a reading method of an image reading apparatus for reading an original with an image sensor, and more particularly to a shading correction method capable of obtaining a good image quality.

[0002]

2. Description of the Related Art Usually, when an image is read by an image reading apparatus such as a facsimile, shading correction is performed using shading correction data based on white reference data in order to obtain read pixel data with little image quality deterioration. The variation in the brightness of each pixel due to the characteristics of the image sensor, the illumination light source, and the like is corrected.

In the conventional shading correction method, white reference data is first stored in a memory as white reference data, and a fixed value of the white reference data stored in the memory is read out and used when reading a document. In addition, a white reference sheet disposed opposite to the read image sensor, or a white reference roller similarly disposed opposite to the read image sensor, and There is a method of using data obtained by scanning with an image sensor before reading.

In the latter method of using the white reference data obtained by scanning a sheet or roller serving as a white reference with an image sensor, there is a problem of image quality deterioration due to dirt and aging. .

A technique relating to a shading correction method for preventing the image quality from being deteriorated due to dirt or aging is disclosed in, for example, Japanese Patent Application Laid-Open No. Hei 5-14712.
In this method, shading correction is performed by using either the averaging information or the white reference information in accordance with the reduction rate of the white reference information with respect to the white reference averaging information.
The problem has been solved.

Another technique is disclosed in Japanese Unexamined Patent Application Publication No. Hei 5-19164.
No. 5 discloses this. In this method, a photoelectric conversion element for shading correction and a white member for covering the photoelectric conversion element for shading are provided at least outside one of the photoelectric conversion elements for document reading, and the output of the photoelectric conversion element for correction is stored. The problem is solved by comparing a reference value with a temperature change or an aging change of the output, and correcting image information obtained from the original reading photoelectric conversion element based on the change.

[0007]

Among the above-mentioned conventional shading correction methods, the former, which is a shading correction method using a fixed value stored in a memory in advance, is deteriorated due to long-term use of an image sensor, an illumination light source and the like. In addition, there is a problem that shading correction data becomes inappropriate and image quality deteriorates.

In the latter method of performing shading correction before reading a document, even if the influence of deterioration due to long-term use of an image sensor, an illumination light source, and the like can be avoided, dust or a white reference sheet or a white reference roller may be removed. Since dust or the like is attached, the reflection density changes entirely or partially, so that proper shading correction cannot be performed, and a problem that image quality deteriorates occurs.

Further, in any of the above cases, in the shading correction method for preventing the image quality from being deteriorated due to the white reference stain or the white reference aging, if the degree of the deterioration is not uniform over all the pixels to be read, an appropriate There is a problem that shading correction cannot be performed.

An object of the present invention is to make shading correction improper due to deterioration of an image sensor, an illumination light source, or the like, or dirt or dust attached to a white reference sheet or a white reference roller. It is an object of the present invention to provide a shading correction method which can prevent the occurrence of a shading and obtain a good image quality.

[0011]

The shading correction method according to the present invention detects the peak value of each pixel while reading the original for each pixel, and updates the peak value for each pixel, thereby reducing the total number of pixels. create a shading correction data consisting of peak value of the pixel, the shading correction method of performing shading correction on the basis of this data,
A correction coefficient is calculated such that the maximum value data in the shading correction data is equal to the updated maximum value data in the peak value, and the correction coefficient is calculated by applying the correction coefficient to each pixel of the shading correction data. Calculating means for obtaining data; comparing the correction value data obtained by the calculating means with the peak value data for each pixel to obtain difference data; comparing the difference data with a predetermined threshold; When the difference value is larger than the threshold value, the correction value data determined by the arithmetic unit is determined as appropriate data. When the difference data is equal to or smaller than the threshold value, the peak value data for each pixel is determined. And determining means for determining suitability data, wherein the suitability data determined by the determining means is used for shading correction.

The shading correction method for a reading method in which a document is scanned in a plane and an image sensor reads the document is used as an illumination light source for illuminating the document, and an optical signal read from the document is converted into an electric signal. And an analog / digital converter that converts an analog signal from the image sensor into a digital signal corresponding to each pixel.
A D converter, a peak value memory storing peak value data of a read image signal for each pixel, and a read image signal data of each pixel and a peak value data of a corresponding pixel stored in the peak value memory. A comparison circuit that compares and outputs a larger (whiter) value to update and store the peak value data in the peak value memory; and a shading memory that stores correction data for shading correction. ,
The read image signal data output from the A / D converter is
A shading correction circuit for performing shading correction with the correction data of the shading memory for all pixels and outputting proper read pixel data, and a maximum value (whitest data) from the peak value data stored in the peak value memory And the maximum value (whitest data) is also detected from the correction data stored in the shading memory, a correction coefficient is calculated from the detected maximum value data, and the correction coefficient is stored in the shading memory. An operation circuit for multiplying the correction data of each pixel by the correction coefficient to obtain correction value data, and the correction corresponding to the correction data of each pixel output from the operation circuit and stored in the shading memory. The value data is compared with the peak value data of the corresponding pixel stored in the peak value memory. If the difference is larger than a predetermined threshold, the correction value data input from the arithmetic circuit is determined as appropriate data and selected, and if smaller than the predetermined threshold, the correction value data is input from the peak value memory. A determination circuit for determining and selecting the peak value data as the appropriateness data and outputting the selected appropriateness data as update data for updating the shading memory;

[0013]

Next, the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of a shading correction method according to the present invention.

In the shading correction method shown in FIG. 1, the illumination light source 1, the original 2, the image sensor 3, the A / D converter 4, the comparison circuit 5, the peak value memory 6, the shading correction circuit 7, the shading memory 8, Arithmetic circuit 9,
Further, a determination circuit 10 is provided, and the original 2 is read by the image sensor 3 by planar scanning.

First, the light emitted from the illumination light source 1 illuminates the document 2. The light information of the document surface reflected by the document 2 is incident on the image sensor 3, is photoelectrically converted inside the image sensor 3, and is extracted as read information of the document 2 by an analog electric signal. The read information of the document 2 based on the analog electric signal output from the image sensor 3 is input to the A / D converter 4 and is output from the A / D converter 4 as a digital read image signal.

The digital read image signal output from the A / D converter 4 is input to a shading correction circuit 7 and a comparison circuit 5. The read image signal input to the shading correction circuit 7 is output as read pixel data subjected to shading correction with reference to the shading memory 8.

On the other hand, the read image signal input to the comparison circuit 5 is compared in the comparison circuit 5 with reference to the peak value data read from the peak value memory 6. The comparison circuit 5 updates the data in the peak value memory 6 by outputting the larger data (whiter data) of the read image signal and the peak value data and writing the data in the peak value memory 6.

Here, the peak value data read from the peak value memory 6 and referred to is the same pixel as the read image signal. Further, the data stored in the peak value memory 6 has the maximum value of the peak value for each pixel detected during the previous reading operation. The updating process of the shading memory 8 by the peak value data over all the pixels stored in the peak value memory 6 is performed, for example, for 1000 pages because the deterioration speed of the image sensor 3, the illumination light source 1 and the like is much slower than the document reading speed. Is performed in a cycle for each original reading.

The arithmetic circuit 9 reads out the data stored in each of the peak value memory 6 and the shading memory 8, detects the respective maximum values, and determines the maximum value of the correction data stored in the shading memory 8 as the peak value memory. Then, a correction coefficient for equalizing the maximum value of the peak value data stored in No. 6 is calculated. That is, the maximum value of the peak value data stored in the peak value memory 6 is equal to the maximum value of the correction data stored in the shading memory 8 multiplied by this correction coefficient. Further, the arithmetic circuit 9 reads the correction data for each pixel from the shading memory 8, multiplies it by the calculated correction coefficient, and outputs it to the determination circuit 10.

The determination circuit 10 receives the data output from the peak value memory 6 and the data output from the arithmetic circuit 9 for each pixel, compares the data by reference, calculates the difference between the data, and determines a predetermined threshold value. When the calculated difference is larger than the threshold value, the data input from the arithmetic circuit 9 is corrected. When the calculated difference is smaller than a predetermined threshold, the data input from the peak value memory 6 is corrected. The update data is selected as appropriate update data, sent to the shading memory 8 and written, and updated to new data for shading correction.

Next, the operation of updating the shading correction data, which is a main feature of the present invention, will be described with reference to FIG.

The peak value memory 6 stores the peak value (brightest data) of the previously read document for each pixel. For example, when one pixel is represented by 8-bit data and has 2368 pixels, the peak value memory 6 stores
A memory capacity of “2368 × 8” bits is required.

When the original 2 is read, the data of the read image signal of the n-th pixel of one line scanned and inputted to the comparison circuit 5 corresponds to the n-th pixel of the peak value data stored in the peak value memory 6. Is compared with the peak value data. If the data of the read image signal exceeds the peak value data (when the read image signal is data whiter than the peak value data), the comparison circuit 5 outputs the data of the read image signal, and The image signal data is less than the peak value data (when the peak value data is white data that is higher than the read image signal)
If so, the peak value data is output.

The data output from the comparison circuit 5 is stored in the peak value memory 6 as new peak value data.
The peak value data in the peak value memory 6 is updated. In this way, the peak value memory 6 is stored for all the pixels.
Is updated. This operation is performed for all the lines in the document 2 for each line to be scanned, and the peak value data is updated each time the reading operation is performed.

As described above, the updating of the shading correction data is performed in the peak value memory 6 for each line for reading and scanning the peak value, while the shading memory 8 reads, for example, a total of 1000 pages. Done after. Such a long cycle is because the deterioration speed of the image sensor 3, the illumination light source 1, and the like is very slow as compared with the reading speed of the original, and is also to avoid erroneous detection of the peak value due to the character information of the original. is there.

Next, the operation of the arithmetic circuit 9 for updating the shading correction data will be described with reference to FIG.

First, the maximum value data D _P for all the pixels of the peak value data stored in the peak value memory 6.
Is detected. Similarly, to detect the data D _S of the maximum value for all the pixels of the correction data stored in the shading memory 8. Next, the relationship between the data D _P and the data D _S is calculated as α = D _P / D _S (Equation 1) as the correction coefficient α.

[0029] That is, the correction coefficient α is multiplied by the data D _S of the maximum value stored in the shading memory 8,
A value that is equal to the maximum value data D _P stored in the peak value memory 6 is calculated. Further, the arithmetic circuit 9
The determination circuit 1 determines the update data to be sent to the shading memory 8
In order to make the determination with 0, the correction data for each pixel of the shading memory 8 is multiplied by a correction coefficient α, and the determination circuit 1
Output to 0. The data A _i output from the arithmetic circuit 9 to the determination circuit 10 is obtained by comparing the data S _i of the _i -th pixel of the correction data stored in the shading memory 8 with A _i =
α × S _i (Equation 2).

Next, the operation of the determination circuit 10 for updating the shading correction data will be described. Judgment circuit 1
To 0, data is input from the peak value memory 6 and the arithmetic circuit 9 for each pixel. Difference between the data A _i input from the arithmetic circuit 9 to the determination circuit 10 corresponding to the i-th pixel and the data P _i input from the peak value memory 6 to the determination circuit 10 corresponding to the i-th pixel Data B _i
Is determined by B _i = | A _i −P _i | (Equation 3).
It is calculated at zero.

The determination circuit 10 calculates the difference data B _i
This is compared with a predetermined threshold Th. "B _i >Th", that is, the peak value data Pi of the _i- th pixel stored in the peak value memory 6 and the shading data A obtained by correcting the data of the i-th pixel stored in the shading memory 8 with the correction coefficient α. _If the difference from _i is greater than a predetermined threshold,
The data P _i is determined as inappropriate data and discarded, and the value of the data A _i is determined as appropriate data, stored in the shading memory 8 as update data, and updated. On the other hand, when “B _i ≦ Th”, the value of the data P _i is determined as appropriate data, stored in the shading memory 8 as update data, and updated.

By performing the above operation for all the pixels, the updating operation of the shading memory 8 is completed.

The arithmetic circuit 9 and the judgment circuit 10 as described above
With the function described above, a specific pixel is darkened due to dust or the like on the reading position of the image sensor 3, and as a result, even if the data of the peak value memory becomes inappropriate, the shading memory is not used by the inappropriate data. It is not updated, and it is possible to prevent the overall image quality from being deteriorated due to dirt such as dust and dust on the image sensor 3.

Also, the threshold value compared in the determination circuit 10 is
For example, the image quality is prevented from deteriorating by setting the white level to about several percent.

After the above-mentioned updating operation of the shading memory 8 is completed for all the pixels, the peak value memory 6
Clear the memory (set the peak value data of all pixels to the minimum value,
That is, the peak value is updated again each time the document is read.

That is, in this configuration, in the initial state,
First, the determination circuit 10 sets the threshold value to the maximum value, the peak value memory 6 clears the memory, and sets all the peak value data to the minimum value (black data), so that the first document is read as a white reference, The peak image memory 6 stores read image data of a document based on white.

Next, an updating operation of the shading memory 8 is performed. Since the threshold value of the determination circuit 10 is the maximum value, all the data stored in the peak value memory 6 is stored in the shading memory 8, and the initial shading correction data is obtained.

In the present invention, in order to use data in the original 2 as shading correction data when reading the original,
Conventionally, there is no need for a white reference sheet or a white reference roller used for shading correction.

The shading correction is performed as described above, and the shading correction data becomes unsuitable due to deterioration of the image sensor, the illumination light source, etc., or dirt or dust attached to the reading position of the white reference roller or the image sensor. Can be prevented, and as a result, deterioration of image quality can be suppressed.

Next, the main operation flow of the present invention will be described with reference to FIG. 1, FIG. 2 and FIG.

In FIGS. 2 and 3, reference numeral i is i
The reference numeral j indicates the j-th page of the document 2. Data P _i indicates data corresponding to the i-th pixel of the peak value memory 6, and data Q _i indicates read data of the _i- th pixel. Here, a period X in which data is written from the peak value memory 6 to the shading memory 8. (The updating operation of the shading memory 8 is performed each time X sheets are read.)

In FIG. 3, data S _i corresponding to the i-th pixel of the shading memory 8, data D _{P of the} maximum value stored in the peak value memory 6, and data of the maximum value stored in the shading memory 8 shall each data D _S is defined. The correction coefficient calculated by the arithmetic circuit 9 by an operation represented by Equation 1 alpha, shading data A _i obtained by multiplying the correction factor alpha and the data S _i represented by the above equation 2, the data represented by the above formula 3 It is assumed that the absolute value of the difference between P _i and data A _i , the difference data B _i , and the threshold value Th for the determination circuit 10 to determine the update data from the value of the data B _i are defined in the same manner as described above. I do.

2 and 3, data P _i , Q _i ,
In S _i respectively, shows that as the color of the pixel is white is a large value. That is, for example, "P _i <Q _i" indicates that the direction of data Q _i is the white color value from the data P _i.

The flow of repeating steps S2 to S8 shown in FIG. 2 and "No" in step S9 is performed by the illumination light source 1, document 2, image sensor 3, A / D shown in FIG. Converter 4, comparison circuit 5,
The processing is performed by the peak value memory 6, the shading correction circuit 7, and the shading memory 8.

First, as shown in FIG.
At step 1, the value 0 is written to the reference code j, and the next step S2
Then, the code j becomes the value (j + 1). Next, step S
At 3, the reference code i becomes the value 1 (becomes the first pixel).

Subsequently, in step S 4, the data Q _i of the read image signal corresponding to the i-th pixel is compared with the data P _i of the peak value memory 6. If better data Q _i is larger than the data P _i (Yes), step, after the contents of the peak value memory 6 corresponding to the i-th pixel shifts to step S5 is updated from the data S _i to the data P _i The process proceeds to the next step S6. If “No” in the step S4, the step skips the step S5 and shifts to the step S6. In step S6, the code i becomes the value (i + 1) (becomes the next pixel).

Next, in step S7, it is determined whether or not reading of one line has been completed. Step S7 is "Ye
s ", the step moves to step S8, where" N "
If "o", the process returns to step S4. In step S8, it is determined whether the reading of the j-th sheet of the document 2 has ended. If step S8 is "Yes", the step moves to step S9. , “No”, the process returns to step S3.

In step S9, it is determined whether or not the j-th page of the document 2 is longer than the period X. Step S9 is "Y
es ", the step moves to step S10,
If “No”, the step returns to Step S2, and the flow of Steps S2 to S9 is repeated.

After the content updating operation of the shading memory 8 shown in FIG. 3 is performed in step S10, the process returns to step S1 and repeats the flow of steps S1 to S10.

The above-described steps S2 to S8 are operations during reading of the original 2, and the steps S2 to S8 after the reading of the X originals 2 are completed.
In the flow of 10, the content updating operation of the shading memory 8 is performed, and the peak value memory 6, the shading memory 8, the arithmetic circuit 9, and the determination circuit 10 of FIG. 1 operate.

As shown in FIG. 3, first, in step S
At 101, the data D of the maximum value stored in the peak value memory 6
_P is detected, in the subsequent step S102, similarly, the data D _S stored maximum peak value memory 8 are detected.
Next, in step S103, steps S101 and S10
The correction coefficient α is calculated from the data D _P and D _S detected in step 2 as described above.

In step S104, the code i is set to the value 1 (1
Pixel). Further, in next step S105, the data S _i of the i-th pixel of the shading memory 8, the data A is calculated by multiplying the correction coefficient alpha, the subsequent step S10
In step 6, the absolute value of the difference between the data P _i and A _i is obtained as the value of the difference data B _i corresponding to the i-th pixel.

Subsequently, in step S107, the difference data Bi and the threshold value Th are compared for the _i- th pixel. If the person of the differential data B _i is greater than the threshold Th (Ye
s), the process proceeds to step S108, and the data A _i is stored and updated in the _i- th pixel of the shading memory 8. If step S107 is "No", the step moves to step S109, where i
The data _Pi is stored and updated at the pixel. Step S10
8. In steps S109 and S109, the process proceeds to step S110, and it is determined whether the update operation has been completed for all the pixels. If step S110 is "Yes", the flow ends. On the other hand, if "No", the flow shifts to step S111 to set the code i to the value (i + 1) (next pixel). Step S1
The step following Step 11 returns to Step S105 and repeats the flow of Steps S105 to S111.

In the above description, the block configuration and the operation flow are illustrated. However, the order of the illustrated and described functions such as merging, separation and the like, the permutation of steps, and the parallel processing are as follows. The present invention is free as long as the above functions are satisfied, and the above description does not limit the present invention.

[0055]

As described above, according to the present invention,
Update the peak value of each pixel data when reading the original,
Since the data of the peak value is used as data for shading correction, it is possible to prevent deterioration of image quality due to deterioration of an image sensor, an illumination light source, and the like.

Further, according to the present invention, when updating the shading correction data, the peak value data and the shading correction data are compared to determine which data is to be used. Even if the peak value data has an abnormal value due to dust or dirt attached to the reading position, the abnormal data is not updated as shading correction data, so that dust or dirt attached to the reading position of the image sensor is not updated. This can prevent the image quality from deteriorating.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of a shading correction method according to the present invention.

FIG. 2 is a flowchart showing one embodiment of a main operation flow according to FIG. 1;

FIG. 3 is a flowchart showing an example of a part of the operation flow of FIG. 2 in detail;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Illumination light source 2 Original 3 Image sensor 4 A / D converter 5 Comparison circuit 6 Peak value memory 7 Shading correction circuit 8 Shading memory 9 Operation circuit 10 Judgment circuit

Continuation of the front page (58) Field surveyed (Int.Cl. ⁶ , DB name) H04N 1/40-1/409 H04N 1/46 H04N 1/60

Claims (3)

(57) [Claims] 1. While reading a document for each pixel, a peak value of each pixel is detected, and the peak value is updated for each pixel, so that shading composed of the peak values of all pixels is performed.
Create the correction data, the shading correction method of performing shading correction on the basis of this data, obtains the correction coefficient equal to the maximum value data in said peak value the maximum value data is updated in the shading correction data Calculating means for calculating the correction value data by applying the correction coefficient to each pixel of the shading correction data, and comparing the correction value data obtained by the calculation means with the peak value data for each pixel. When the difference data is larger than the threshold value by comparing with a predetermined threshold value, the correction value data obtained by the calculation means is determined as the appropriate data, while the difference data is If the threshold value is equal to or smaller than the threshold value, the peak value data for each pixel is determined as appropriate data. And a judging means, a shading correction method characterized by using the proper data which the determination means determines the shading correction. 2. A shading correction method for a reading method in which a document is scanned in a plane and read by an image sensor, an illumination light source for illuminating the document, an image sensor for converting a light signal read from the document to an electric signal, and the image sensor. An A / D converter for converting an analog signal from the sensor into a digital signal corresponding to each pixel, a peak value memory for storing peak value data of a read image signal for each pixel, and read image signal data for each pixel; Comparing the peak value data of the corresponding pixel stored in the peak value memory, outputting a larger (whiter) value, and updating and storing the peak value data in the peak value memory. A circuit, a shading memory storing correction data for shading correction, and a read image signal data output from the A / D converter. A shading correction circuit that performs shading correction on all pixels with the correction data of the shading memory and outputs appropriate read pixel data; and a maximum value (the most white color) from the peak value data stored in the peak value memory. ), A maximum value (whitest data) is also detected from the correction data stored in the shading memory, and a correction coefficient is calculated from the detected maximum value data to obtain the shading. An arithmetic circuit for performing an operation of multiplying the correction data of each pixel stored in the memory by the correction coefficient to obtain correction value data; and an output circuit for outputting the correction data of each pixel stored in the shading memory. The corresponding correction value data is converted to the peak value data of the corresponding pixel stored in the peak value memory. In comparison, when the difference is larger than a predetermined threshold, the correction value data input from the arithmetic circuit is determined as appropriate data and selected, and when smaller than the predetermined threshold, the correction value data is read from the peak value memory. A determination circuit for determining the input peak value data as suitability data and selecting the suitability data, and outputting the selected suitability data as update data for updating the shading memory. 3. The arithmetic circuit according to claim 2 , wherein the peak value data stored in the peak value memory is calculated by the calculation circuit using correction data stored in the shading memory, and a result of the calculation. The shading correction method is characterized in that the judgment by the judgment circuit using the correction value data and the updating of the shading memory by the appropriateness data based on the judgment are executed in a cycle of a predetermined period.