JPH09224156A - Image processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processor eliminating the influence of a noise part at the time of detecting the density of the ground and reading fast by providing a subtracter storing shading data generated for a ground removing function in a first-in and first-out (FIFO) circuit and then subtracting a value equivalent to the noise part from a shading data value. SOLUTION: This processor consists of a reading means 7 reading an original, a peak detection means 25 detecting a peak value, a comparing and outputting means 27 receiving the output of the reading means 7 and the output of the peak detection means 25 to determine an output value, and a shading correction device 41 shading-operating by shading data from the comparing and outputting means 27. After storing shading data at the shading correction circuit 41, the value equivalent to the noise part is subtracted from the value of the stored shading data.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electrophotographic copying machine,
The present invention relates to an image processing apparatus that digitally processes original image data in a scanner, a fax machine, and the like, and particularly to an image processing apparatus that eliminates the influence of noise when detecting the background density of the background removal function and enables high-speed reading.

[0002]

2. Description of the Related Art In a conventional image processing apparatus which has a background removal function and digitally processes original image data, when the background removal function is activated, the reading value of a reference white board is compared with a predetermined value. Then, the shading data is generated, and the gain of the background density detecting unit is adjusted so as to match the background reading data of the document. At that time, even if noise is superimposed on the background data at the time of detecting the background density, the effect of the noise is not exerted, so the gain of the background density detection unit was adjusted to be large so as to absorb the noise. Adjustment is troublesome, and gain changes due to changes over time after adjustment, etc., and the effects of noise have come out. Also, when the shading data is generated using the same circuit as the background detection, variable means such as a volume used for gain adjustment can be eliminated.
With this method, the gain of the background density detection unit cannot be made to absorb the noise component, so the noise component is output as data in the image data of the background detection unit, and it is impossible to skip the background with high accuracy. Was occurring.

[0003]

In the above prior art,
In order to eliminate the influence of noise when the background density is detected and to carry out the background fly with high accuracy, the variable means such as the volume used for gain adjustment is adjusted. However, in this method, it is troublesome to perform accurate adjustment, and gain changes due to changes over time due to adjustment and the like, and the influence of noise has come out. Further, if the variable means can be omitted, it is not possible to absorb the noise component by the gain of the background density detecting unit, so that the noise component is output as data in the image data of the background detecting unit, which is highly accurate. There was a problem that the skin could not be removed. The present invention has been made to solve the above-mentioned conventional problems, and its purpose is to:
After storing the shading data generated for the background removal function in a first-in first-out (FIFO) circuit, by providing a subtractor that subtracts a value corresponding to noise from the shading data value, the effect of noise at the time of detecting the background density is affected. It is an object of the present invention to provide an image processing apparatus that eliminates the problem and enables high-speed reading.

[0004]

In order to achieve the above object, the invention according to claim 1 is, in an image processing apparatus for digitally processing document image data, a reading unit for reading a document and the reading unit. A peak detecting means for detecting a peak value, a comparing output means for receiving an output of the reading means and an output of the peak detecting means to determine an output value, and storing the decided output data from the comparing output means. It is characterized by having a storing means and a data changing means for changing the data from the storing means. According to a second aspect of the present invention, in the image processing apparatus according to the first aspect, the output data determined by the comparison output means is:
The shading data is effective shading data for the background removal function, and the storage means is a first-in first-out (FIFO) circuit for storing the shading data in a shading correction circuit that performs a shading operation based on the shading data from the comparison output means. Is composed of subtracting means for subtracting a value corresponding to noise from the value of the shading data from the first-in first-out circuit.

According to the image processing apparatus described in claims 1 and 2, it is possible to perform high-accuracy background skipping that does not affect noise at the time of background density detection, and after the FIFO storing the shaded data. Since the subtraction circuit for shading data correction is added, the shading data generation process can be omitted even when the reading operation is repeatedly performed while changing the mode of the reading condition, and thus high-speed reading is possible. Further, since there is a subtraction circuit for shading data correction after the FIFO for storing the shading data, it is possible to change the shading data even in the case where the process of generating the shading data is performed once and the document is read a plurality of times. , It is possible to change the copy density (dynamic range).

[0006]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of a reading unit of an image processing apparatus embodying the present invention. Here, a scanner unit of a typical electrophotographic copying machine (hereinafter referred to as a copying machine) will be described as an example of the image processing apparatus, but the same idea can be applied to other image processing apparatuses such as a scanner and a fax machine. In FIG. 1, a document 3 placed on the contact glass 1
Before being read, the variations due to the light source itself of the light source 5 and the variations due to temperature and aging are corrected, and the main scanning direction of the light source 5 and the reading element 7a such as CCD (vertical direction from the front to the back in the figure). Correct the variation of one line. Therefore, in order to read on the reference white board 9,
An optical system including the light source 5, the first mirror 11, the second mirror 13, and the third mirror 15 is provided at the rightmost end in the main scanning direction scale 17.
Move to the home position (indicated by HP in the figure) located in the vicinity, and move to the left (sub scanning direction) at a predetermined speed from there, and when the optical system comes on the reference white plate 9, the reference One line in the main scanning direction on the white plate 9 is irradiated by the light source 5, the reflected light is guided to the lens 19 by the first mirror 11, the second mirror 13, and the third mirror 15, and read by the reading element 7a. At that time, first, in order to correct the variation due to the light source itself of the light source 5 and the variation due to temperature and aging, the GCA amplifier in the reading means 7 is set so that the peak value of the reading output becomes a predetermined target value. The gain of 7c is changed (see FIG. 2).

FIG. 3 is a view of the original 3 placed on the contact glass 1 as seen from above. The original 3 is placed on the scale 17 in the main scanning direction and the scale 23 in the sub scanning direction so that the two sides of the original 3 are along. Has been. The shaded area 3a on the original 3 is a portion where the background is detected. A reference white plate 9 is set in a portion sandwiched between the main scanning direction scale 17 and the contact glass 1. In FIG. 4, the reading output of one line in the main scanning direction on the reference white board 9 is shown by a curve f1, and the peak value Vp1 of the reading output (the output value of the P1 point) becomes the target set value Vp0 (P1 point is P0. It shows how the gain is changed (moving to a point). When the above correction is completed, shading data is generated by reading one line in the main scanning direction on the reference white board 9. That is, as shown in FIG. 5, the output curve f2 after the peak value correction is sampled for each read pixel, and the output values (b1, b2, b3 to bn) are stored and used as shading data.

Next, when the optical system of FIG. 1 moves in the sub-scanning direction to reach the leading edge of the original 3, the reading of the original 3 is started, and the shading correction of the read data of the original image is read first. It is carried out by the shading data stored and sent to the next image processing unit. At that time, when the background removal function is valid, the background portion of the document is also obtained by detecting the peak value of the read pixel data of the document image. The peak value Vp0 of the shading data f2 after the above-mentioned peak value correction is defined as a, and a fixed comparison reference value _{Vref0 for it is set.}
The b value of is given as in FIG. Further, the peak value Vp2 (= a ') is set at the point P2 of the output curve f3 (output of the reading unit 1) of the read data of the original image, and the peak detecting unit 25
_Assuming that the value V _refAE taken at the point P2 in the output curve f3 that has passed through (see FIG. 2) is b ′, the gain of the peak amplifier 25b is conventionally adjusted to be b ′ in the following equation. a ′ ÷ b ′ = a ÷ b Therefore, in order to perform accurate adjustment, the peak amplifier 25b
It is necessary to be able to make fine adjustments to the gain, and after the setting, the gain varying means of the peak amplifier 25b loses reliability, and if the gain changes, the background can be accurately skipped. Disappear.

This embodiment will be described by taking an example of an image processing apparatus which can eliminate the adjustment of the gain of the peak detecting means 25 and can perform the adjustment-free operation to remove the background with high accuracy. The main part of the present invention will be described in detail with reference to the block diagram of the main part of the image processing apparatus of FIG. 2 and FIGS. 7 and 8. The original image and the reference white plate 9 are read by the reading means 7 and transmitted to one end Vin of the input of the comparison output means (A / D) 27. On the other hand, in the comparison output means 27, a reference value is input to the input terminal V _ref of the comparison reference value, and the reference value is input Vi.
It is determined what level n has reached and is digitally output. That is, the comparison output means 27 is an A / D
It has a function as a converter. The reading means 7 is
A reading element 7a such as a CCD, a sample hold circuit 7b that samples the output from the reading element 7a and holds the output until the next sampling, a variable gain reading amplifier 7c, and a zero clamp circuit 7d that creates a black (zero) level reference. It consists of

Further, in order to achieve the background removal function, the read output information is transmitted from the output of the reading means 7 to the peak detecting means 25, and the peak value is detected. When the output of the peak detecting means 25 is selected by the selecting circuit 29, the selecting circuit 2
9 is transmitted to the output terminal Q, and is transmitted to the comparison reference input terminal V _ref of the comparison output means 27. The peak detecting means 25 comprises a peak hold (P / H) circuit 25a for holding and storing the peak value and a peak amplifier 25b having a predetermined gain. Conventionally, as described above with reference to FIG.
The gain variable device of the peak amplifier 25b was adjusted so that a '÷ b' = a ÷ b. However, this method makes it difficult to make a fine adjustment, and if the gain of the peak amplifier 25b changes after the adjustment, the accuracy of I wasn't able to remove the good background.
Therefore, a method that does not require the gain variable device of the peak amplifier 25b described below is used.

Considering the timing of operation of the background removal function of FIG. 7, when the optical system starts moving from the home position and comes over the reference white plate 9, the WTGT gate which makes the reading of the reference white plate 9 effective. Is the gate generation circuit 31
AGCGT, which is opened by a gain control (AGC) circuit 33 that controls the gain of the read amplifier 7c at the same time.
The gate also opens. When the WTGT gate opens, the selection circuit 29
The input terminal A is selected, and the CPU 35
Is applied with V _ref0 . In order that the maximum output value obtained by reading the reference white plate 9 becomes Vp0, gain control (A
GC) D for controlling the gain of the read amplifier 7c from the circuit 33
-Give AGC1 and close AGCGT gate. By this operation, variations due to the light source of the light source 5 itself, variations due to temperature and aging, and variations in sensitivity of the reading element 7a are corrected.

Next, AEMO for enabling the background removal function
The DO gate and the SHGT gate corresponding to the shading data creation period are opened to generate shading data. At that time, as in the case of detecting the background of the document, the reference white plate 9
Output of the reading means 7 of the peak hold (P / H) circuit 2
5a and a peak amplifier 25b, which leads to the peak detecting means 25, detects the peak value, and inputs it to the comparison reference input terminal V _ref of the comparison output means 27. Of course, the output terminal Q of the selection circuit 29 at that time is controlled so that the input terminal C is selected. SEL-Q in FIG. 7 indicates the output state of the output terminal Q of the selection circuit 29. Further, although not explained later, the PWIND gate representing the background detection period is open for a period corresponding to the shaded area 3a for detecting the background of FIG. When the shading data has been generated from the reference white board 9 in this way, the WTGT gate and the SHG are generated.
Close the T gate.

Next, when the optical system comes over the original 3, the FGATE gate showing the image effective period in the sub-scanning direction, that is, the reading period of the original 3, is opened, and the peak detecting means 25 is operated to detect the background of the original. It will be. Since the peak value is detected using the same path and is input to the comparison reference input terminal V _ref of the comparison output means 27 when the shading data is generated as described above and also in the background detection during the reading period of the document 3, the peak value is detected. Peak amplifier 25b of the peak detecting means 25
It becomes possible to ignore the effect of gain changes due to changes over time. Also, in the initial stage, the peak amplifier 25b
The work efficiency can be improved without adjusting the gain.

Next, theoretical consideration will be added. As a precondition, it is assumed that the comparison output means 27 is converted into a digital value from 0 to 255 and output according to the value input to the input terminal Vin. Thus, whether the input value Vin is equal to the reference value V _ref, is greater than V _ref, the output value of the comparison output unit 27 becomes 255. When D _G is the original reading data and D _SH is the shading data, the image reading expresses the density as a relative value to the shading data. Thus, the image data GD is, GD = shown in _{(D G ÷ D SH) ×} 255 ... (1), it is a shading correction. When the original is read when the background removal function is not operating, V _G is the value of the input terminal Vin of the comparison output means 27 when reading the original, and V _G is CC.
The output value of the reading element 7a such as D, V _REF1 is input to the comparison reference input terminal V _ref of the comparison output means 27, A _v0 is G
Assuming the gain of the CA amplifier 7c, D _G = INT [(V _G ÷ V _REF1 ) × 255] = INT [(V _G × A _V0 ÷ V _REF1 ) × 255] ... (2) ). Here, INT [] indicates an adjustment unit of a calculation formula up to 255.

Similarly, when the reference white board 9 is read, D _SH = INT [(V _SH ÷ V _REF0 ) × 255] = INT [(V _SH × A _V0 ÷ V _REF0 ) × 255] (3) INT [] only means quantization when A / D conversion is performed by the comparison output means 27, and therefore it is omitted, and equations (2) and (3) are substituted into equation (1) to obtain GD. = (D _G ÷ D _SH ) × 255 = [(V _G ÷ V _REF1 ) ÷ (V _SH ÷ V _REF0 )] × 255] (4) (V _SH ÷ V _REF0 ) determines the dynamic range of document reading, and if the subscript _N is added to the read data that has the lowest density (highest reflectance), (V _GN ÷ V _REF1 ) = (V _SH ÷ V _REF0 ) (5) so that the comparison reference input value V of the comparison output means 27 is satisfied.
_{Since REF0} is determined, the equation (4) is GD = (V _G ÷
V _GN ) × 255 (6) holds, and it is understood that the image data from the comparison output unit 27 is generated as the relative value of the document density.

When the conventional background removal function is used, the input terminal C of the selection circuit 29, that is, the peak value of the peak detection means 2 is selected only when the document is read.
If the comparison reference input value is V _REF-AE , then D _AE-G = INT [(V _G ÷ V _REF-AE ) × 255] (7) Here, as can be seen from FIG. 2, in V _REF-AE , the output of the reading means 7 has passed through the peak detecting means 25. Therefore, the maximum image data value of the background portion of the document is set to V _GJ, and the gain of the peak amplifier 25b is set. the When _{a V1, D AE-G =} INT [(V G × 255) ÷ (V GJ × a V1)] = INT [(V G × a VG × 255) ÷ (V GJ × a V0 × a V1 _{)] = INT [(V G} × 255) ÷ (V GJ × A V1)] ... (8) , and the (1) to (3) _{(8), GD AE ÷ 255 = D AE} -G ÷ D SH _{= [V G ÷ (V GJ} × a V0 × a V1)] becomes _{÷ (V SH ÷ V REF0)} ... (9). Here, even when the reference white board 9 is read, the value at the comparison reference input end of the comparison output means 27 is obtained as the peak value of the peak detection means 25. That is, V _{REF0 in} equation (9)
Indicates that the read output of the reference white board 9 has passed the peak detection means 25 so as to correspond to V _REF-AE which has passed the peak detection means 25 at the time of reading the original when the input terminal C of the selection circuit 29 is selected. Therefore, V _REF0 = V _SH × A _V0 × A _V1 (10) and substitute it into the equation (9) to obtain GD _AE = (V _G ÷ V _GJ ) × 255 (11) Therefore, the background can be blown by a method that is not affected by the gain of the amplifier. In general, the peak hold circuit 25a of the peak detecting means 25 for detecting the background is 1. It is not affected by noise or the brightness of small dots on the original. 2. I do not want to affect the peak even if there is a relatively large black solid image. For a reason such as the above, a large time constant (τ ₀ = R1 × in FIG. 1)
It has C1). However, since the width of the reference white plate 9 in the sub-scanning direction is shortened as much as possible due to cost problems and improvement in the reading processing capability of the image processing apparatus, it is advantageous to make the time constant shorter when the peak of the reference white plate 9 is detected. Is.

Here, looking at the method of generating the shading data, the data D _G0 obtained by digitizing the read data of the reference white board 9 by the comparison output means 27 by the above method is obtained. Since this data D _G0 includes a zero level variation due to the dark current of the CCD, the data D _OPS of the black offset level value (V _OPS comparison output means 2 in FIG. 5).
The output value of 7) is detected by the black level detection means 37, and the data D _OPS is subtracted from the output value D _G0 of the comparison output means 27 by the first subtractor 39. When the background removal function is not operating, D _OPS is subtracted from the data D _G0 by the first subtractor 39 to generate the shading data D ′ _G0. However, when the background removal function is operating, the background can be skipped. This is not enough because it is the purpose. Previous (1
Since the density of the background part is the denominator from the equation (1), the data GD _AEJ when reading the background part is GD _AEJ = INT [(V _GJ ÷ V _GJ ) × 255] = 255. However, since the actual image data includes the noise component V _N , GD _AEJ = INT [(V _GJ −V _N ) ÷ V _GJ × 255] = 255−GD _N (12) Therefore, in the present invention, the shading data is reduced by the output correction value D _OFF that suppresses the noise component V _N only during the operation of the background removal function so that the noise component V _N does not appear as data. That is, GD _AEJ = INT [D _GJ ÷ (D _AF-SH = D _OFF ) × 255] (13) Here, if D _GJ -D _N > D _AE-SH = D _{OFF DN} <D _OFF is satisfied, GD _AEJ always outputs 255 even if the noise amount V _N is included in the data of the background portion. Therefore, the noise component V _N is not output to the image data.

Therefore, as shown in FIG. 2, in the present invention, the above-mentioned D _OFF is supplied from the CPU 35 to the shading correction circuit 41.
Output to the output correction value D _OFF as described below.
Is configured to reduce the shading data. That is, as shown in FIG. 8, the shading correction circuit 41 sets the data subtracted by the comparison output correction means 39 as D ′ _G , during the period when the SHGT gate is open,
The average value circuit 41a calculates an average value for each pixel in the main scanning,
All the data is stored in the first-in first-out (FIFO) circuit 41b. Then, the data D _sh1 corrected by subtracting the output correction value D _OFF from the shading data D _sh from the FIFO circuit 41b by the second subtractor 41c.
Then, the shading data is reduced by the output correction value D _OFF so that the noise component V _N does not appear in the data. After that, a pair of the read data D ′ _G and the corresponding data from the second subtractor 41c is stored in the ROM 41d.
The shading operation is performed by outputting the calculation result which is input to the address of and written in the ROM 41d in advance. Here, as a general value, D _OFF is set to about 5 only when the background is removed, and 0 is set in other cases.

The applicant of the present invention has proposed a method of subtracting the output correction value D _OFF before the shading correction circuit 41, that is, in the first subtractor 39. Then, turn on the shading data correction function when generating shading data.
Since the shading data is changed because it is turned off, the method of omitting the shading data generation process is used as a means to perform high-speed reading with a model that does not require light intensity correction for each lamp using a halogen lamp or xenon lamp. When adopted, in order to read a document multiple times while changing the mode, it is necessary to generate shading data each time, and high-speed reading may not be possible. On the other hand, according to the above-described configuration of the present invention, the F that stores the shading data is stored.
Since the subtraction circuit 41c for shading data correction is added after the IFO 41b, the shading data generation process can be omitted even when the reading operation is repeatedly performed while changing the mode of the reading condition, and thus high-speed reading becomes possible. .

Further, in the above-mentioned technique proposed by the applicant of the present invention, a method of adding a noise component to image data is proposed as a method of performing without changing shading data. Since the addition function of the noise removal output is turned ON / OFF, an offset will be given to the data when the background removal function is turned ON, and there will be problems such as the black side level rising and the line thinning and solid filling not occurring. There was a nature. On the other hand, according to the above-described configuration of the present invention, the subtraction circuit 41c for shading data correction is provided after the FIFO 41b that stores shading data, so that the process of generating shading data is one time and the document is read multiple times. Even in such a case, the shading data can be changed, and the copy density (dynamic range) can be changed.

Further, a calculation result of GD = INT [(D ' _G ÷ D _sH ) × 255] is input to the ROM 41d. Then, when the shading data is generated, the WTGT gate and the SHGT gate are closed. However, in this state, as can be seen from the configuration shown in FIG. 2, the capacitor C of the peak hold circuit 25a is still present.
Since 1 is still charged, if the background of the original 3 is darker than the reference white plate 9, the background density of the original 3 cannot be accurately detected. Therefore, as shown in FIG.
A before reading (before opening the FGATE gate)
The EMODO gate is closed and SW1 is closed to discharge the electric charge of the capacitor C1. When the discharge ends, A
The EMODO gate is opened again, and after waiting for the optical system to reach a position where the original 3 can be read, the FGATE gate is opened and the reading of the original 3 is started. When the reading range of the original 3 is completed, the FGATE gate is closed and the PWIND gate indicating the background detection period is also closed, and the reading amplifier 7c is closed.
The gain D-AGC of is returned to a predetermined value. After that, AEMO
The DO gate is closed to discharge the electric charge of the capacitor C1 of the peak hold circuit 25a, and the output of the selection circuit 29 is returned from the input end C which is the output of the peak detection means 25 to the normal input end B, and the reading operation is completed. To do.

[0022]

As described above, according to the present invention, both when the reference white plate is read to generate the shading data and when the background density is detected, the reading means reads and uses it as the input of the comparison output means, and the other reference input. Is supplied with the data peak value read by the peak detection means, the output value is obtained by comparison from the comparison output means, and the black value detected by the black level detection means from the output of the comparison output means by the first subtraction means. Since the offset level value is subtracted, and the correction value for suppressing the noise component of the background data is added / subtracted by the second subtraction means provided after the FIFO in the shading correction circuit, a high value that does not affect the noise when the background density is detected. Shading data can be stored after the FIFO that stores shading data while enabling accurate background removal. Since you are adding a subtraction circuit Tadashiyo, even when performing repetitive reading operations while changing the mode of reading conditions, it is possible to read fast since it is possible to omit the shading data generation process. Further, since there is a subtraction circuit for shading data correction after the FIFO for storing the shading data, it is possible to change the shading data even in the case where the process of generating the shading data is performed once and the document is read a plurality of times. , The copy density (dynamic range) can be changed.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a reading unit of an image processing apparatus embodying the present invention.

FIG. 2 is a block diagram of a main part of an image processing apparatus in which the present invention is implemented.

FIG. 3 is an explanatory diagram illustrating a background reading unit for a document on a document table of the reading unit illustrated in FIG.

4 is an explanatory diagram illustrating a correction method when reading a reference white plate in the image processing apparatus illustrated in FIG.

5 is an explanatory diagram illustrating a method of reading a reference white board in the image processing apparatus illustrated in FIG.

6 is an explanatory diagram illustrating a method of reading a background portion of a document in the image processing apparatus illustrated in FIG.

FIG. 7 is a timing chart of a main part in reading in the image processing apparatus shown in FIG.

8 is a block diagram of a shading correction circuit in the image processing apparatus shown in FIG.

[Explanation of symbols]

1 ... contact glass, 3 ... manuscript,
5 ... Light source, 7 ... Reading means, 7a ... Reading element, 7b ...
Sample and hold circuit, 7c ... reading amplifier,
7d ... Zero clamp circuit, 9 ... Standard white board,
11 ... First mirror, 13 ...
Second mirror, 15 ... Third mirror, 17 ... Main scanning direction scale, 19 ...
Lens, 23 ... Sub-scanning direction scale, 2
5 ... Peak detecting means, 25a ... Peak hold circuit,
25b ... Peak amplifier, 27 ... Comparison output means (A / D), 29 ... Selection circuit, 31 ... Gate signal generation circuit, 33 ... Gain control circuit, 3
5 ... CPU, 37 ... Black level detecting means, 39 ... First subtractor,
41 ... Shading correction circuit, 41a ... Average value circuit, 41b ... First-in first-out (FIFO) circuit, 41c
... second subtractor, 41d ... ROM,

Claims (2)

[Claims] 1. An image processing apparatus for digitally processing original image data, comprising reading means for reading an original, peak detecting means for detecting a peak value of the reading means, output of the reading means and the peak. It has a comparison output means for receiving the output of the detection means and determining an output value, a storage means for storing the output data determined by the comparison output means, and a data changing means for changing the data from the storage means. An image processing device characterized by the above. 2. The shading correction circuit, wherein the output data determined by the comparison output means is effective shading data of the background removal function, and the storage means performs a shading operation based on the shading data from the comparison output means. 2. A first-in first-out circuit for storing shading data, wherein said data changing means comprises subtracting means for subtracting a value corresponding to noise from the value of the shading data from said first-in first-out circuit. Image processing device.