JPH11275321A - Picture reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce a level difference occurred between picture elements different in channels in a picture reader where the driving-type CCD linear sensor of a plurality of channels is used and to improve the picture quality. SOLUTION: In a picture reader using the driving-type CCD 8 of a plurality of channels, an inter-channel level difference correction means 20 correcting a level difference existing between the channels of CCD 8 with picture data of a picture read result by CCD 8 as an object is installed. For correcting the level difference, the picture elements of a plurality of continuous channels are averaged and smoothed or the level difference between the channels is detected in the reading of a white reference sheet and the offset of the picture elements of the respective channels are corrected by using the detected level difference.

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 using a CCD linear sensor, such as an image scanner, and more particularly to a multi-channel driving type for switching and selecting a plurality of channels in pixel units and synthesizing and outputting. CC
The present invention relates to an image reading device using a D linear sensor. In order to respond to the needs for increasing the number of pixels and increasing the speed of image processing in recent years, the use of a CCD linear sensor driven by a plurality of channels in an image reading apparatus tends to increase. The present invention solves the problem of the level difference between pixels inherently present in a multi-channel drive type CCD linear sensor,
The purpose is to improve image quality.

[0002]

2. Description of the Related Art Many image scanners use a CCD linear sensor which performs one-dimensional reading by line scanning. FIG. 8 shows an example of a mechanism of a conventional image scanner device using a CCD linear sensor.

In FIG. 8, 1 is a housing, 2 is a platen glass, 3 is a document, 4 is a reading unit, 5 is a light source, 6 is a mirror, 7 is a lens, 8 is a CCD linear sensor, and 9 is a white reference sheet. is there.

The original 3 is placed on the platen glass 2 with the image to be read facing down. The light source 5 illuminates the lower surface of the document 3 in the reading unit 4. The reflected light from the document surface is bent 90 degrees by the mirror 6,
The light is incident on the CD linear sensor 8. The reading unit 4 moves in the direction indicated by the arrow during the document reading operation, and performs sub-scanning on the lower surface of the document 3. At this time, the white reference sheet 9 is read prior to reading the document 3, and shading correction for setting a white reference value pattern at each scanning position in the main scanning direction, that is, a direction perpendicular to the paper surface is performed.

FIG. 9 shows a signal system of the image scanner of FIG. In FIG. 9, 8 is a CCD linear sensor,
Reference numeral 10 denotes a video amplifier, 11 denotes an A / D converter, and 12 denotes a digital image processing circuit.

A CCD linear sensor 8 holds an analog image signal of one line of pixels by a shift pulse CCDSH synchronized with line scanning, and outputs a transfer clock CCD.
The analog image signals are serially output and transferred according to CK. An analog image signal output from the CCD linear sensor 8 is amplified by a video amplifier 10 and is subjected to A / D conversion.
The data is converted into a digital image signal by the converter 11. A /
The D converter 11 has an A / D converter clock ADC.
A / D conversion operation is performed at the cycle of CK. The digital image signal is then input to the digital image processing circuit 12, where white level tracking processing and binarization processing for adapting to changes in the ground color of the document are performed.

The CCD linear sensor 8 detects light corresponding to each pixel on a line, and stores an array of photosensitive cells for storing electric charges and the electric charges stored in each of the photosensitive cells in parallel, and outputs the serially output electric charges. A CCD analog shift register that performs one-channel driving, using one CCD analog shift register for one photosensitive cell array, and a plurality of CCD analog shifts for one photosensitive cell array. There is a CCD linear sensor of a multi-channel drive type in which registers are connected in parallel. In a conventional low-speed, low-gradation image scanner,
A one-channel drive type CCD linear sensor 8 has been used. FIG. 10 shows a circuit configuration of a one-channel drive type CCD linear sensor.

In FIG. 10, 13 is a photosensitive cell array,
14, a CCD analog shift register; and 15, a shift gate. Each cell of the photosensitive cell array 13 stores an electric charge corresponding to the input light. When a shift pulse SH synchronized with the horizontal scanning is input, the shift gate 15 is closed and each cell of the photosensitive cell array 13 is closed. The charge of
The data is transferred to the corresponding stage of the CCD analog shift register 14 and held there. Next, when the CCD analog shift register 14 is driven by the two-phase transfer pulses φ1 and φ2, the charges of each stage are sequentially shifted and transferred in one direction in the CCD analog shift register 14 and output in series. As a result, an analog image signal in which the level of the pixel position changes according to the amount of charge stored in each photosensitive cell of the photosensitive cell array 13 is obtained.

In such a one-channel drive type CCD linear sensor, when the number of pixels is increased, the number of photosensitive cells in the photosensitive cell array and the number of stages of the CCD analog shift register are correspondingly increased. , The number of shifts required to transfer the data is also increased. However, there is a limit in increasing the speed of the CCD analog shift register, and in many cases, a large increase in the number of pixels causes a significant decrease in operation speed. Multi-channel drive type CCD
A linear sensor is an effective means for solving this problem. The CCD linear sensor of the multi-channel drive type has a plurality of CCD analog shift registers in parallel, distributes the photosensitive cells sequentially to each CCD analog shift register, and reduces the apparent number of shifts to a multiple of a high speed. It is intended to make it. FIG. 11 shows a circuit configuration of an example of a two-channel drive type CCD linear sensor.

In FIG. 11, 16 is a photosensitive cell array,
17a and 17b are CCD analog shift registers, 18
a and 18b are shift gates. The number of cells in the photosensitive cell array 16 is twice as large as the number of cells in the photosensitive cell array 13 in FIG.
The number of stages of 7b is the same as the number of stages of the CCD analog shift register 14 in FIG.

The odd-numbered cells in the photosensitive cell array 16;
, Are connected to each stage of the CCD analog shift register 17a via a shift gate 18a, and the even-numbered cells of the photosensitive cell array 16 are
It is connected to each stage of the CCD analog shift register 17b via the shift gate 18b.

When the shift pulse SH is applied to the shift gates 18a and 18b simultaneously, the photosensitive cell array 1
The charge of the odd-numbered cell 6 is transferred to each stage of the CCD analog shift register 17a, and the charge of the even-numbered cell of the photosensitive cell array 16 is transferred to the CCD analog shift register 1.
7b.

[0013] CCD analog shift register 17a, 1
7b, the transfer pulses φ1 and φ2 are simultaneously applied, so that the charges of each stage of both shift registers are transferred in parallel and output to buffer stages (not shown), respectively, and the odd-numbered and even-numbered charges are respectively transferred. Are merged into the original order of alternating odd and even numbers, and output in series.
As a result, in the two-channel driven CCD linear sensor of FIG. 11, twice as many photosensitive cells are read out in the same operation time as the one-channel driven CCD linear sensor of FIG. 10 without increasing the speed of each shift register. Becomes possible. In the same manner, C such as a three-channel drive type or a four-channel drive type
A CD linear sensor is realized.

[0014]

In a CCD linear sensor of a plurality (n) channel drive type, charges of a plurality of (n) cells of one photosensitive cell array are sequentially distributed to CCD analog shift registers of different channels. After that, the pixels are merged again, so that a level difference occurs between pixels having different channels in an image signal output from the CCD linear sensor due to variations in characteristics of individual channel systems.
FIG. 12 shows an example. FIG. 12 shows a level change that occurs when an area where there is essentially no change in gradation is read by a CCD linear sensor of a four-channel drive type.

The level difference between pixels of different channels is usually small, but has a periodicity. For example, image data of a low gradation such as 16 gradations or 32 gradations is not so conspicuous. Such as high-gradation images, especially in areas where there is little gradation change,
There was a problem in image quality.

SUMMARY OF THE INVENTION It is an object of the present invention to reduce the level difference generated between pixels of different channels and improve image quality in an image reading apparatus using a CCD linear sensor driven by a plurality of channels.

[0017]

The image reading apparatus of the present invention is configured as follows to solve the above-mentioned problems. (1) In an image reading apparatus using a CCD linear sensor of a multi-channel drive type, an inter-channel correction for correcting a level difference existing between respective channels of the CCD linear sensor for image data of an image read result by the CCD linear sensor. A level difference correction device is provided. (2) In the above (1), the CCD linear sensor is n
Is an integer value of 2 or more, and is an n-channel drive type CCD linear sensor. The inter-channel level difference correction device inputs image data obtained by performing image reading using the n-channel drive type CCD linear sensor. A pixel selection unit that selects, for each successive pixel of interest, values of consecutive n pixels including the pixel of interest, and n based on the value of the selected n pixels.
A correction operation unit that performs an operation for correcting a level difference existing between pixels of the channel and outputs the corrected value of the target pixel. (3) In the above (2), the correction operation unit of the inter-channel level correction device performs an averaging operation on the values of the n pixels selected by the pixel selection unit, and sets the operation result as the corrected target pixel value. It is output. (4) In the above (2), the correction operation unit of the inter-channel level correction device performs a smoothing operation on the value of n pixels selected by the pixel selection unit, and uses the operation result as a corrected target pixel value. It is output. (5) In the above (1), the inter-channel level difference correction device holds an inter-channel level difference detecting section for detecting a level difference existing between each channel of the CCD linear sensor, and holds data of the detected inter-channel level difference. And a correction operation unit that corrects each pixel value of the image data for each channel using the stored data of the inter-channel level difference.

The basic configuration of the image reading apparatus of the present invention will be described with reference to FIGS. The image reading apparatus shown in FIG. 1A sequentially selects a plurality of continuous pixels of the same number as the number of channels of the CCD linear sensor from image data obtained by reading using a CCD linear sensor of a multi-channel drive type. Then, the level difference between channels is reduced by performing the averaging process or the smoothing process.

In FIG. 1A, reference numeral 8 denotes a CCD linear sensor of an n (≧ 2) channel drive type; 10 a video amplifier for amplifying a video signal (analog image signal) output from the CCD linear sensor 8; Is a video amplifier 10
An A / D converter for converting the video signal amplified by the above into a multi-valued digital image signal, a digital image processing circuit 12 for performing a white level tracking process or a binarization process on the multi-valued digital image signal, and 20 for the present invention. Is a pixel selection unit that receives the digital image signal output from the A / D converter 11 and selects successive n-pixel values, and 22 is a pixel selection unit 2
Reference numeral 1 denotes a correction operation unit that performs an averaging operation or a smoothing operation on the sequentially selected n pixel values and outputs the result.

A video signal output while the CCD linear sensor 8 is reading an image is transmitted to a video amplifier 10.
, And is converted into a multi-valued digital image signal by the A / D converter 11, and the inter-channel level difference correction device 2
0 is input in pixel series. In the inter-channel level difference correction device 20, the pixel selection unit 21 selects the input digital image signal while moving it by n pixels sequentially from the top pixel, and supplies the selection to the correction operation unit 22. The correction operation unit 22 performs a correction operation for each of the supplied n pixels, and outputs a result to the digital image processing circuit 12.

FIG. 1B shows a four-channel drive type CC.
4 shows a data flow in the case of performing correction by averaging using a D linear sensor. Here, the address of the pixel of interest to be corrected in the pixel column of the input digital image signal is i,
The pixel value before the correction is s (i), and the pixel value after the correction is S
When expressed by (i), since n = 4, values s (i−1), s (i), and s of four consecutive pixels including the target pixel (i)
(I + 1) and s (i + 2) are averaged, and the correction operation expression is given by the following expression.

S (i) = [s (i-1) + s (i) + s
(I + 1) + s (i + 2)] / 4 FIG. 1C shows a data flow in the case of performing correction by smoothing using a CCD linear sensor of a 4-channel drive type. As in the case of the averaging, the values s (i−1), s (i), and s (i +) of four consecutive pixels including the target pixel (i)
1), smoothing is performed by s (i + 2), and when K is a coefficient indicating the degree of smoothness, a correction operation expression is given by the following expression.

S (i) = s (i) -K ｛s (i)-[s
(I-1) + s (i) + s (i + 1) + s (i + 2)]
As the value of / 40K, for example, a fixed value of 0.5 is used.

The image processing apparatus shown in FIG. 2A detects a level difference between channels in advance for a CCD linear sensor of a multi-channel drive type, and uses data of the detected level difference to generate a digital image signal. In this method, each pixel value is corrected for each channel to eliminate a level difference.

In FIG. 2A, 9, 10, 11,
Reference numeral 12 is the same as that shown in FIG. 23 is an inter-channel level difference correction device added according to the present invention;
Reference numeral 24 denotes an inter-channel level difference detecting unit for detecting a level difference between pixels of different channels based on a digital image signal output from the A / D converter 11, and reference numeral 25 denotes a level difference detected by the inter-channel level difference detecting unit 24 for each channel. An inter-channel level difference data holding unit that holds the corresponding offset is a correction operation that corrects the level difference of each pixel in the digital image signal using the offset held in the inter-channel level difference data holding unit for each channel. Department.

In the image reading apparatus, the CCD linear sensor 8
Before the image reading of the document is started, the white reference sheet is read and shading correction is performed. At this time, the inter-channel level difference detection unit 24 of the inter-channel level difference correction device 23 designates a predetermined address on the line, for example, an address at the center position, and takes in the values of n consecutive pixels, Detect the level difference.
The detection result becomes an offset value to be corrected for the pixels of the respective channels 1 to n, and is held in the inter-channel level difference data holding unit 25 as offset 1 to offset n.

When the CCD linear sensor 8 starts reading an image of a document, the correction operation unit 26 of the channel-to-channel level difference correction device 23 operates on the sequential pixel values of the digital image signal output from the A / D converter 11. The offset correction is performed by sequentially applying the corresponding ones of the n offset values (offset 1 to offset n) held in the inter-channel level difference data holding unit 25 and output to the digital image processing circuit 12.

FIG. 2B shows a 4-channel drive type CCD.
5 shows a data flow of a level difference correction calculation by offset correction when a linear sensor is used. Since n = 4, the same offset value is applied every four pixels. The pixel addresses of four consecutive pixels corresponding to channels 1 to n are 4j, 4j + 1, 4j + 2, 4j + 3 (j = 0, 1,
2, 3,...), The correction operation expression is given by the following expression.

S (4j) = s (4j) + offset 1 S (4j + 1) = s (4j + 1) + offset 2 S (4j + 2) = s (4j + 2) + offset 3 S (4j + 3) = s (4j + 3) + offset 4 (j = 0, 1, 2, 3,...)

[0030]

FIG. 3 is a circuit diagram of an embodiment of the inter-channel level difference correcting device 20 shown in FIG.
4 shows a configuration for performing a correction operation by averaging.

In FIG. 3, reference numeral 27 denotes a pixel selection unit.
4 of the shift register configuration indicated by 28, 29, 30, 31
It is composed of stage registers. Reference numeral 32 denotes a four-input correction operation unit, which performs addition of four inputs and １ ／ division.

The digital image signal output from the A / D converter 11 shown in FIG. 1A is input to a register 28 of an image selector 27, and an A / D converter clock ADC
The next register 29,
It is shifted to 30, 31. Therefore, register 28,
Four consecutive pixel values s (i +
2), s (i + 1), s (i), s (i-1) are set. The correction operation unit 32 inputs these four pixel values in parallel, adds them, shifts the addition result right by two bits, and subtracts 1 /
Fourth division is performed and [s (i-1) + s (i) + s (i +
1) + s (i + 2)] / 4 and outputs the corrected pixel value S (i). This operation is performed when i = 0, 1,
Are executed continuously for each address of 2,.

FIG. 4 is a circuit diagram of another embodiment of the inter-channel level difference correction device 20 shown in FIG. 1A, and shows a configuration for performing a correction operation by smoothing. In FIG.
Is a correction operation unit for performing smoothing, and is composed of three operation units indicated by 34, 35, and 36. The operation unit 34 is 4-input addition and ８ division, the operation unit 35 is １ ／ division, and the operation unit 36
Performs two-input addition.

The arithmetic unit 34 has four consecutive pixel values s (i + 2) and s (i +
1), s (i), and s (i-1) are added, and the addition result is 1
/ 8 division. The operation unit 35 receives s (i) and inputs 1 /
The arithmetic unit 36 divides by two, and adds the respective arithmetic results of the arithmetic unit 34 and the arithmetic unit 35 to obtain s (i) / 2 + [s (i−1) + s
(I) + s (i + 1) + s (i + 2)] / 8, and outputs the corrected pixel value S (i).

FIG. 5 is a circuit diagram of an embodiment of the inter-channel level difference correction device 23 shown in FIG. 2A, and shows a configuration for performing an offset correction operation. In FIG. 5, reference numeral 37 denotes a pixel selection unit, and the pixel selection unit 2 shown in FIGS.
7 has the same function as that of FIG.
j), s (4j + 1), s (4j + 2), s (4j +
Select 3). Reference numeral 38 denotes an inter-channel level difference detection unit which detects a level difference between pixels of different channels. 3
Reference numeral 9 denotes an inter-channel level difference data holding unit, which holds the detected level difference as an offset value for each channel. Reference numeral 40 denotes a correction operation unit which selects an offset value by a selector 41 and adds / subtracts the offset value to / from an input pixel value by an adder / subtracter 42 to perform offset correction. 43 is a counter for timing control,
A / D is reset by the horizontal synchronizing pulse CCDSH.
The converter clock ADCCK is counted, the level difference detection timing, the offset value setting timing,
Generates a pulse such as an offset value switching selection timing. An MPU 44 controls to preset the counter 43.

The MPU 44 counts the timing of detecting the level difference between channels by reading the white reference sheet, using a counter 4.
3 is set by presetting a predetermined pixel address. The counter 43 counts ADCCK,
When the preset value is reached, an inter-channel level difference detector 38
Performs the level difference detection, and causes the inter-channel level difference data holding unit 39 to set an offset value.

The inter-channel level difference detecting section 38 uses the pixel values s (4j) of the channel 1 as a reference for the channels 2, 3,
4, each pixel value s (4j + 1), s (4j + 2), s (4
j + 3) is detected and set in the inter-channel level difference data holding unit 39 as offset 1, offset 2, and offset 3, respectively.

When the CCD linear sensor reaches the original reading line, the counter 43 synchronizes with the pixel position on the line and instructs the selector 41 to select the offset value selection pulse C.
H1, 2, 3, and 4 are applied. This selection pulse does not select an offset value when the input pixel value s (4j) to the adder / subtracter 42 belongs to the channel 1, and s (4j)
If offset belongs to channel 1, select offset value 1, if s (4j) belongs to channel 2, select offset value 2, and if s (4j) belongs to channel 3, select offset value 3. Is generated as follows. Thus, the adder / subtractor 42 can always perform the correction using the offset value corresponding to the level difference between the channels. The adder / subtractor 42 performs addition or subtraction according to the positive or negative polarity of the offset value.

FIG. 6 shows a circuit of another embodiment. 6 basically replaces the function of the inter-channel level difference detection unit 38 in the embodiment circuit of FIG. 5 with firmware 45 provided in the MPU 44. Reference numeral 46 denotes four consecutive pixel values s (4) for detecting the level difference between channels when the firmware 45 reads the white reference sheet.
j), s (4j + 1), s (4j + 2), s (4j +
This is a register for capturing 3).

The firmware 45 of the MPU 44 obtains four pixel values s (4j), s (4j + 1), s (4j + 1), s (4j) from the register 46 for detecting the level difference between channels.
j + 2), the maximum value among them is detected, the level difference from the other pixel values is determined based on the pixel value indicating the maximum value, and the offset value is stored in the inter-channel level difference data holding unit 39 as an offset value. Set. If the offset value is determined from the level difference based on the maximum value in this manner, the offset value to the channel corresponding to the pixel value indicating the maximum value is 0, and the offset values to the other channels are all negative. Therefore, the offset correction by the adder / subtracter 42 only needs to be added, and the circuit can be simplified by using the adder / subtracter 42 as a simple adder. Of course, the offset value can be set in the same manner as in the embodiment of FIG.

FIG. 7 shows an example of image data obtained by performing a level difference correction between channels using the circuit of the embodiment shown in FIG. FIG.
7A shows a level difference and an offset value detected between channels, FIG. 7B shows image data before level difference correction, and FIG. 7C shows image data after level difference correction. .

[0042]

According to the present invention, in an image reading apparatus using a CCD linear sensor driven by a plurality of channels, the level difference between channels can be corrected by a relatively simple circuit configuration, and the number of pixels can be increased and the speed can be increased. While the tendency of image formation is increasing, it is possible to easily improve image quality.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of a first basic configuration of an image reading apparatus according to the present invention.

FIG. 2 is an explanatory diagram of a second basic configuration of the image reading apparatus according to the present invention.

FIG. 3 is a configuration diagram of an embodiment of an inter-channel level difference correction device that performs correction by averaging.

FIG. 4 is a configuration diagram of an embodiment of an inter-channel level difference correction device that performs correction by smoothing.

FIG. 5 is a configuration diagram of an embodiment of an inter-channel level difference correction device that performs offset correction.

FIG. 6 is a configuration diagram of another embodiment of an inter-channel level difference correction device that performs offset correction.

FIG. 7 is an explanatory diagram of an example of correcting a level difference between channels of image data.

FIG. 8 is a cross-sectional view of an example of a mechanism section of the image scanner device.

FIG. 9 is a configuration diagram of an example of a signal system circuit of the image scanner device.

FIG. 10 is a diagram showing a circuit configuration of a one-channel drive type CCD linear sensor.

FIG. 11 is a diagram illustrating a circuit configuration of a two-channel drive type CCD linear sensor.

FIG. 12 is an explanatory diagram showing an example of a level difference between channels in a plurality of channel driven CCD linear sensors.

[Explanation of symbols]

 8: a plurality of channel driven CCD linear sensors 10: video amplifier 11: A / D converter 12: digital image processing circuit 20: channel level difference correction device 21: pixel selection unit 22: correction calculation unit

Claims (5)

[Claims] 1. An image reading apparatus using a CCD linear sensor of a multi-channel drive type, wherein image data of an image read result by the CCD linear sensor is used as a target.
An image reading apparatus comprising an inter-channel level difference correction device for correcting a level difference existing between respective channels of a CD linear sensor. 2. The CCD linear sensor according to claim 1, wherein n is an integer of 2 or more, and the n-channel driving type C
The device is a CD linear sensor, and the inter-channel level difference correction device inputs image data obtained as a result of image reading using the n-channel drive type CCD linear sensor and includes the target pixel for each successive target pixel. A pixel selection unit for selecting the value of the continuous n pixels, and an operation for correcting the level difference existing between the pixels of the n channels based on the selected values of the n pixels, and outputting as a corrected value of the target pixel An image reading apparatus comprising: 3. The value of a target pixel in which the correction operation unit of the inter-channel level correction device performs an averaging operation on the values of n pixels selected by the pixel selection unit, and corrects the operation result. An image reading apparatus, wherein the image is output as an image. 4. The correction operation unit of the inter-channel level correction device according to claim 2, wherein the correction operation unit performs a smoothing operation on the value of the n pixel selected by the pixel selection unit, and corrects the operation result to the value of the target pixel. An image reading apparatus, wherein the image is output as an image. 5. An inter-channel level difference correction device according to claim 1, wherein the inter-channel level difference correction device detects an inter-channel level difference existing between channels of the CCD linear sensor.
An inter-channel level difference data holding unit that holds the detected inter-channel level difference data, and a correction operation unit that corrects each pixel value of the image data for each channel using the held inter-channel level difference data. An image reading apparatus, comprising: