JPH0779342A - Picture processor - Google Patents

Abstract

PURPOSE:To eliminate a level difference of an odd number picture element output and an even number picture element output even when a level of an original read output is changed in the picture processing unit correcting a black level of picture data read by an image sensor. CONSTITUTION:Black level is corrected by using a subtractor 15 to subtract a correction value from a selector 13 from an output from an A/D converter 3. An odd number picture element output and an even number picture element output being a black level correction value formed by reading a black reference level are formed by using an output of an image sensor element arranged to a dummy bit section of an image sensor 1. However, since a level difference between an odd number picture element output and an even number picture element output is subject to an output level at original reading, only the black level correction value is not enough to make proper correction. Then a 2nd correction value in response to an output level at original reading is prepared in a fixed correction value selection circuit 7 and the correction value is selected by a detection signal of the level detection section 6. The 2nd correction values are added by adders 11, 12 to set the correction value and then a level difference between the odd number picture element output and the even number picture element output is properly corrected even when the read output level is changed.

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 having a portion for black-correcting image data read by an image sensor.

[0002]

2. Description of the Related Art FIG. 9 is a diagram showing a general configuration of an image processing apparatus. In FIG. 9, 1 is an image sensor, 2
Is an amplifier, 3 is an AD converter, 4 is a black correction unit, and 5 is a white correction unit. The image data (analog value) read by the image sensor 1 is amplified by the amplifier 2 and the AD converter 3
Is converted to a digital value. However, due to various reasons, the value corresponding to black (for example, 0 in 8-bit value) may not match exactly even when reading a black original, and even when reading a white original, white may not be obtained. There is a case where the value corresponding to (e.g., 255 with an 8-bit value) does not exactly match. A black correction unit 4 and a white correction unit 5 are provided to correct and match them. After performing these corrections, digital processing such as a digital filter is performed if necessary.

The image sensor 1 is equipped with a large number of image sensor elements, but when main scanning of one line is performed, image data read by these elements is once sent to a shift register and serially sent from there. Taken out. Some image sensors 1 are equipped with two shift registers. FIG. 10 shows a connection relationship between an image sensor element and a shift register in such an image sensor. C is an image sensor element, S
R ₁ and SR ₂ are shift registers. When two shift registers are mounted, one shift register SR ₁ is connected to an odd-numbered image sensor element (hereinafter, referred to as an “odd pixel image sensor element”), and the other shift register SR ₂ is connected to an even-numbered image sensor element. Image sensor element (hereinafter, referred to as “even pixel image sensor element”) is connected.

When such an image sensor is used,
Due to a slight difference in characteristics between the two shift registers, a slight difference may occur between the output level of the odd pixel and the output level of the even pixel. FIG. 11 is a diagram for explaining the odd-even pixel output level difference. The horizontal axis represents the position of the image sensor element disposed in the image sensor, and the vertical axis represents the image sensor output (analog value; upper direction is white, zero is black).

L indicates one line of main scanning. G
[Fig. 4] is an enlarged view of a portion E which is a part of the image sensor output curve. Reference numeral 21 is an odd-numbered pixel output (hereinafter, referred to as “odd pixel output”), and 21 is an even-numbered pixel output (hereinafter, referred to as “even pixel output”). here,
The case where the odd-numbered pixel output appears to be larger is shown. H is the level difference between the odd pixel output and the even pixel output. This level difference does not adversely affect the image quality to be output when the image data is binary-processed or when the number of gradations is low. However, high gradations (eg 1
If processing is performed with more than 100 gradations, a vertical stripe pattern is formed and the image quality is deteriorated.

Therefore, various proposals have been made to prevent the vertical stripe pattern from being formed due to the level difference of the odd-even pixel output. One of such proposals is Japanese Patent Laid-Open No. 61-196676. This is done by reading a black reference with a constant density (or a black original with a constant density) placed on the platen over one main scan, and calculating the difference between the average value of the odd pixel read data and the average value of the even pixel read data. Is disclosed, and the difference is used as a correction value when an actual document is read to eliminate the level difference.

[0007]

[Problems to be Solved by the Invention]

(Problem) However, in the above-mentioned conventional technique, since the level difference between the odd pixel output and the even pixel output changes depending on the read output level of the image sensor, the correction value obtained by reading the black reference is always used for correction. There was a problem that it was not always possible.

(Explanation of Problems) FIG. 5 is a diagram for explaining the problems of the present invention, that is, the difference in output level difference between odd and even pixels depending on the read output level. 21A indicates an average of odd pixel outputs, 21A indicates an average of even pixel outputs, and H _{1 to} H ₃ are odd-even pixel output level differences. (I)
Indicates the read output when the black reference is read, and the odd-even pixel output level difference at this time is H ₁ . The above-mentioned correction value is obtained based on the output at this time. It is a value that either raises the even pixel output by H ₁ or lowers the odd pixel output by H ₁ . (B) and (C) show the read output when a portion of density other than the black reference (for example, a gray portion close to black or a gray portion close to white) is read. The level difference in these cases is H ₂ and H ₃ .

Then, the odd-even pixel output level difference H ₂ , H ₃
Are often different in size from H ₁ . That is, the odd-even pixel output level difference may differ according to the black density of the portion read by the image sensor (in other words, according to the read output level). In such a case, the correction value generated by the odd-even pixel output level difference obtained by reading the density portion called black reference cannot be completely corrected. An object of the present invention is to solve the above problems.

[0010]

In order to solve the above-mentioned problems, according to the present invention, an image sensor, an amplifier for amplifying the output of the image sensor, an AD converter for converting the output of the amplifier into a digital value, In an image processing device comprising a black correction unit for black-correcting an output of the AD converter, the black correction unit uses an output from an image sensor element arranged in a dummy bit part of the image sensor. First correction value generation means for generating first correction values for odd-numbered pixel output and even-numbered pixel output, and second correction value generation means for generating second correction value according to the output level at the time of reading the original. , And a correction execution means for performing correction with the first correction value and the second correction value.

An image sensor, an amplifier for amplifying the output of the image sensor, an AD converter for converting the output of the amplifier into a digital value, and a black correction unit for black-correcting the output of the AD converter. In the image processing device that comprises,
An odd-even division unit for allocating the output of the image sensor element in the dummy bit unit of the image sensor to an odd pixel output and an even pixel output for each line, and an odd pixel output from the odd-even division unit. An odd pixel averaging circuit for averaging outputs, an even pixel averaging circuit for averaging even pixel outputs from the odd-even division unit, a level detection unit for detecting an output level at the time of reading an original, and a level detection unit from the level detection unit. A fixed correction value selection circuit that selects and outputs a fixed correction value according to a detection signal, a first adder that adds the average of the odd pixel outputs and the selected fixed correction value, and an even pixel output A second adder that adds the average and the selected fixed correction value, a selector that alternately selects the outputs from the first and second adders, and an output from the AD converter from the selector. Output of A subtracter, a bit signal for each pixel is sent to the odd-even division unit and the selector, and addition in the odd pixel averaging circuit and the even pixel averaging circuit is performed from an image sensor element in the dummy bit unit of the image sensor. It is also possible to adopt a configuration having a timing control circuit that sends a signal that specifies that only output is performed.

[0012]

[Operation] Black correction values for odd pixel output and even pixel output, which are created by reading the black reference, are created by using the output from the image sensor element arranged in the dummy bit part of the image sensor. However, since the odd-even pixel output level difference varies depending on the output level at the time of reading the original, the black correction value alone cannot be appropriately corrected. Therefore, in the present invention, a second correction value corresponding to the change in the odd-even pixel output level difference is prepared according to the output level at the time of reading the original, and is added to the correction value for black correction. Then, even if the read output level changes, the odd-even pixel output level difference can be appropriately corrected.

[0013]

【Example】

(First Embodiment) Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 is a diagram showing an image processing apparatus according to the first embodiment of the present invention. In FIG. 1, 1
Is an image sensor, 2 is an amplifier, 3 is an AD converter, 4 is a black correction unit, 6 is a level detection unit, 7 is a fixed correction value selection circuit, 8 is an odd-even division unit, 9 is an odd pixel averaging circuit, and 10 is an even number. Pixel averaging circuit, 11 and 12 adders, 13 selectors, 14 timing control circuits, 15 subtractors, 16 sampling enable signal lines, 17 bit signal lines,
Reference numeral 18 is an averaging enable signal line, and 19 is a bit signal line.

It is assumed that the image sensor 1 has two shift registers as described with reference to FIG. 10, and that there is an output level difference between the odd pixel output and the even pixel output. The image data read by the image sensor 1 is amplified by the amplifier 2 and converted into a digital value by the AD converter 3. The image data converted into a digital value is black-corrected by the black correction unit 4, and then sent to a white correction unit (not shown) and further sent to a digital signal processing unit for edge enhancement and the like. The level detector 6 detects the level range of the output of the AD converter 3.

The black correction is for correcting the value of the image data to a predetermined value (eg, zero) that corresponds to black when reading black. Usually, the output as read is often a little larger than the predetermined value. Therefore, the read output from the AD converter 3 and the correction value from the selector 13 are input to the subtracter 15 of the black correction unit 4, and the latter is subtracted from the former to perform the correction. Among the constituent elements of the black correction unit 4, the constituent elements in the preceding stage of the selector 13 are provided for producing a correction value.

The timing control circuit 14 issues various timing signals required for main scanning of a line. The odd-even division unit 8 distributes the output from the AD converter 3 into odd-numbered ones and even-numbered ones. The distribution is performed by the bit signal line 1 from the timing control circuit 14.
7 signal. Details will be described with reference to FIG. The odd pixel averaging circuit 9 averages the odd pixel outputs, and the even pixel averaging circuit 10 averages the even pixel outputs. The signal on the averaging enable signal line 18 determines the period during which the pixel output for averaging is input.

FIG. 2 shows the configurations of the odd-even division unit, the odd pixel average circuit, and the even pixel average circuit. Reference numerals correspond to those in FIG. 1, 81 and 82 are selectors, 91 and 92 are adders, and 9
3, 94 are latches, 95, 96 are dividers, A, B
Is an input terminal, and S is a select terminal. Zero is always input to the input terminal B of the selector 81 and the input terminal A of the selector 82. Then, each time the bit signal from the bit signal line 17 is input, the selectors 81 and 82 alternately select and output the input in the input terminal A and the input in the input terminal B. As a result, the selector 81 outputs only odd-numbered pixels, and the selector 82 outputs only even-numbered pixels.

The adders 91 and 92 perform the adding operation only during the period when the averaging enable signal is supplied through the averaging enable signal line 18. The number of outputs (the number of pixels) to be added is determined by the length of this period. The adder 91 adds the output of the selector 81 and the value latched in the latch 93, and the adder 92 adds the output of the selector 82 and the value latched in the latch 94.
The added results are held as new values in the latches 93 and 94, respectively. Therefore, in the adders 91 and 92, the output of the pixel of interest and the cumulative value of the output of the pixels before that are added.

The divider 95 divides the added value of the odd pixel outputs to calculate the average value. The divider 96 divides the added value of the even pixel outputs to calculate the average value. The division can be easily performed by bit shifting. For example, if the number of added outputs is eight, the bits of the added value are shifted by 3 bits in the upper direction. The average value thus obtained is used as one component of the correction value (first correction value).

Returning to FIG. 1, in the present invention, the black correction is appropriately performed even if the odd-even pixel output level difference changes due to the output level change when the document is actually read. In addition to the above-mentioned first correction value, a second correction value according to the output level is used. As for the second correction value, when the actual document is read, the level detection unit 6 first detects how large the output level is, and then the fixed correction value selection circuit according to the output level. Choose from 7 to get. Next, the level detector 6 and the fixed correction value selection circuit 7 will be described.

FIG. 3 shows an example of the structure of the level detector 6.
Reference numerals correspond to those in FIG. 1, 61 to 63 are registers, and 6
Reference numerals 4 to 66 are comparators, and 67 is an encoder. In this example, the output from the AD converter 3 is assumed to be sent in 8 bits, and the output 0 to 255 is equally divided into four ranges. Registers 61, 62 and 63 have the comparison reference values “63”, “127”,
"191" is a CPU (central processing unit) not shown
Is written in advance by. Comparators 64 and 6
Reference numerals 5 and 66 compare the output from the AD converter 3 with these comparison reference values. The encoder 67 creates a signal indicating the output level range based on the comparison result, and sends it to the fixed correction value selection circuit 7.

FIG. 8 shows an example of a correspondence table of the output level range, the level detection output and the fixed correction value. For example, if the output of the AD converter 3 is 70,
Since this belongs to the output level range of 64 to 127, the outputs of the comparators 64, 65 and 66 are H, L and L, respectively, and the output of the encoder 67 (this is the level detection signal) is LH. When this detection signal LH is input to the fixed correction value selection circuit 7, "2" is selected as the odd pixel fixed correction value and "3" is selected as the even pixel fixed correction value.

FIG. 4 shows the configuration of the fixed correction value selection circuit 7. Reference numerals correspond to those in FIG. 1, 71 is a sample circuit,
Reference numerals 72 and 73 are selectors, and 74 to 79 are registers.
The registers 74 to 76 are stored in the registers shown in FIG.
The odd pixel fixed correction value of the table of (1) is written in advance, and the even pixel fixed correction value is written in the registers 77 to 79 in advance. The sample circuit 71 captures the level detection signal emitted from the level detection unit 6 when the sampling enable signal comes. The sampling enable signal is issued for each pixel within the period in which correction is required (that is, within the document reading period).

Then, the selectors 72 and 73 are operated according to the level detection signal, and one of the registers is selected to determine the fixed correction value. For example, when the level detection signal is LH as described above, the selector 72 selects the register having the fixed correction value of “2”, and the selector 73 selects the register having the fixed correction value of “3”. . The sample circuit 71 may be provided in the input path of the comparator of FIG. In this case, the sampling enable signal 16 from the timing control circuit 14 must be sent to the level detection unit 6 instead of the fixed correction value selection circuit 7.

The adders 11 and 12 shown in FIG. 1 add the above-mentioned two correction values (that is, the first correction value obtained by reading the black reference and the second correction value selected according to the output level). It is provided to do so.

(Generation of First Correction Value ... Generation of Correction Value Obtained by Reading Black Reference) By the way, in order to read the black reference and generate a correction value, in the prior art, the black reference is placed on the original table. Although an object to be provided (for example, a black reference plate or a black reference document) was prepared and read, the present invention uses the dummy bit portion originally included in the image sensor.

The dummy bit portion is a portion which originally exists at the end portion of the image sensor, and is a portion which is provided with an image sensor element but which does not receive light from the outside. The image sensor element in this portion is in an environment equivalent to that in which the black reference is always read. In the present invention, paying attention to this point, the output of the image sensor element in the dummy bit portion is used to obtain the correction value.
That way, you don't have to bother preparing something that provides a black standard.

FIG. 7 is a diagram showing the relationship between the image sensor element position and the image sensor output. Reference numeral 1-1 is a dummy bit portion of the image sensor 1, 1-2 is a document reading portion of the image sensor 1, and C is an image sensor element. The output from the dummy bit section 1-1 is D. The document reading section 1-2 is an area other than the dummy bit section 1-1,
An output corresponding to the read light amount is output from the image sensor element arranged here. Black correction is made by subtracting D from the output.

FIG. 6 is a diagram showing a timing waveform for averaging the outputs of the dummy bit portion and the image sensor. The reference numerals correspond to those in FIG. 7. 6A shows the image sensor 1, and FIG. 6B shows the averaging enable signal flowing through the averaging enable signal line 18 shown in FIGS. Only when the signal is high (H), the pixel outputs are added by the adders 91 and 92 in FIG.

FIG. 6C shows the sampling enable signal flowing through the sampling enable signal line 16 of FIG. As described above, the signal is output for each pixel within the period that requires correction (that is, within the document reading period). Then, for example, the value of the output level detection signal is captured (sampling) at the rising edge of this signal. FIG. 6D shows the dummy bit section 1-1.
An image sensor element C provided in a portion other than the above indicates a line synchronization signal when a document is read.

In the present invention, the correction value (first correction value) created by reading the black reference is created line by line in the dummy bit section 1-1. Then, the portion which is still insufficient is supplemented by the fixed correction value which is the second correction value.

[0032]

As described above, according to the image processing apparatus of the present invention, in addition to the black correction values for the odd pixel output and the even pixel output, which are created by reading the black reference, the black correction value changes depending on the output level. In order to cope with the odd-even pixel output level difference
The correction value of is also used. Therefore, even if the output level of the document reading output changes, the level difference of the odd-even pixel output can be appropriately corrected.

[Brief description of drawings]

FIG. 1 is a diagram showing an image processing apparatus according to a first embodiment of the present invention.

FIG. 2 is a diagram showing a configuration of an odd-even division unit, an odd pixel averaging circuit, and an even pixel averaging circuit.

FIG. 3 is a diagram showing a configuration of a level detection unit.

FIG. 4 is a diagram showing a configuration of a fixed correction value selection circuit.

FIG. 5 is a diagram showing that the odd-even pixel output level difference differs depending on the read output level.

FIG. 6 is a diagram showing timing waveforms for averaging outputs of a dummy bit part and an image sensor.

FIG. 7 is a diagram showing a relationship between an image sensor element position and an image sensor output.

FIG. 8 is a diagram showing an example of a correspondence table of output level ranges, level detection outputs, and fixed correction values.

FIG. 9 is a diagram showing a general configuration of an image processing apparatus.

FIG. 10 is a diagram showing a connection relationship between an image sensor element and a shift register.

FIG. 11 is a diagram illustrating an odd-even pixel output level difference.

[Explanation of symbols]

1 ... Image sensor, 1-1 ... Dummy bit part, 1-2
... original reading unit, 2 amplifier, 3 AD converter, 4 black correction unit, 5 white correction unit, 6 level detection unit, 7 fixed correction value selection circuit, 8 odd-even division unit, 9 odd number Pixel averaging circuit,
10 ... Even pixel averaging circuit, 11, 12 ... Adder, 13 ...
Selector, 14 ... Timing control circuit, 15 ... Subtractor,
16 ... Sampling enable signal line, 17 ... Bit signal line, 18 ... Averaging enable signal line, 19 ... Bit signal line

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Atsushi Takahashi 3-7-1, Fuchu, Iwatsuki-shi, Saitama Fuji Zerox Co., Ltd. (72) Koji Yormoto 3-7-1, Fuchu, Iwatsuki, Saitama Prefecture Fuji Zelocks Co., Ltd. (72) Inventor Ken Kajikawa, 3-7-1, Fuchu, Iwatsuki City, Saitama Prefecture Fuji-Zerox Co., Ltd. (72) Inventor, Yasuo Komatsu 3-7-1, Fuchu, Iwatsuki City, Saitama Prefecture Within Rocks Co., Ltd. (72) Inventor Hiroshi Hayashi 3-7-1 Fuchu, Iwatsuki City, Saitama Prefecture

Claims (2)

[Claims] 1. An image sensor, an amplifier for amplifying the output of the image sensor, an AD converter for converting the output of the amplifier into a digital value, and a black correction unit for black-correcting the output of the AD converter. In the image processing device, a first correction value for odd pixel output and even pixel output is output to the black correction unit by using an output from an image sensor element arranged in a dummy bit portion of the image sensor. A first correction value generating means for generating a second correction value, and a second correction value generating means for generating a second correction value according to the output level at the time of reading the original.
An image processing apparatus comprising: a correction execution unit that performs a correction based on the first correction value and the second correction value. 2. An image sensor, an amplifier for amplifying the output of the image sensor, an AD converter for converting the output of the amplifier into a digital value, and a black correction unit for black-correcting the output of the AD converter. In the image processing device, the configuration of the black correction unit includes an odd-even division unit that divides an output from an image sensor element in a dummy bit unit of the image sensor into an odd pixel output and an even pixel output for each line,
An odd pixel averaging circuit that averages the odd pixel outputs from the odd and even division unit, an even pixel averaging circuit that averages the even pixel outputs from the odd and even division unit, and a level detection unit that detects the output level during document reading And a fixed correction value selection circuit for selecting and outputting a fixed correction value according to a detection signal from the level detection unit, and a first addition for adding the average of the odd pixel output and the selected fixed correction value. Unit, a second adder for adding the average of the even pixel outputs and the selected fixed correction value, a selector for alternately selecting the outputs from the first and second adders, and the AD conversion A subtracter for subtracting the output from the selector from the output from the selector, a bit signal for each pixel is sent to the odd-even division unit and the selector, and addition in the odd pixel averaging circuit and the even pixel averaging circuit is performed in the image sensor. The image processing apparatus being characterized in that a configuration and a timing control circuit for sending a signal defining that performs only the output from the image sensor element in the in Mibitto unit.