JPH02177669A - Picture reader - Google Patents

Abstract

PURPOSE:To execute normalization with high accuracy by comparing picture signals for plural lines in the same picture element of a black reference plate with each other, storing the minimum value as a black reference signal, and reading out the stored black reference signal. CONSTITUTION:In a digital comparator 20, smaller data out of inputted digital data are selected and stored in the memory space of the same address in a line memory 19. This operation is executed for all the picture elements of one line, and the smaller data out of the respective same picture elements of two lines are stored in the line memory 19. Writing in and reading out operations for the line memory 19 are accurately executed by an address signal and a write enable signal WE from a memory timing control circuit 21. Thus, the black reference signal can be generated by removing the influence of a scratch, dirt, and dust on the black reference plate, and the normalization with high accuracy can be executed.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to an image reading device commonly used in digital copying machines, facsimile machines, electronic filing systems, etc.
In particular, the present invention relates to an image reading device featuring a black correction circuit.

[Conventional technology]

When reading an image on a document and outputting it to a printer, etc.
Photoelectric conversion elements such as CCDs are arranged in a line (one dimension),
A line image sensor that extracts optical information as an electrical signal (image signal) is used as the imaging means.

The image signal obtained by such an image sensor is
Even when a document surface with uniform reflectance is read, the signal is not at a constant level, and variations and undulations occur. This phenomenon is generally called shading.

Furthermore, this phenomenon also changes due to temperature fluctuations and the like. The factors of shading can be broadly modeled by variations and fluctuations in the offset and gain of each photoelectric conversion element in the imaging system.

When there is shading in an imaging system using a line image sensor, the image signal has a -
Various errors are added. If an image is directly reproduced from this image signal, a distribution of brightness and darkness will occur in the main scanning direction (the direction in which the photoelectric conversion elements are arranged).

Furthermore, if there is a high frequency component in the shading, linear noise parallel to the sub-scanning direction will occur in the reproduced output image, visually degrading the image quality more than the size of the noise. In particular, in the case of a color image reading device, the image quality of the output image deteriorates significantly because the image signal is required to have a higher S/N ratio than that of a monochrome image signal.

In order to eliminate the effects of such shading and obtain an image signal that is proportional to the reflectance of the document surface regardless of the position of the photoelectric conversion element (pixel), normalization (also called shading correction) is usually applied to the image signal. Ru. To explain the principle of this standardization, first, before reading an image on a document, a black reference plate with a reflectance of 0 and a white reference plate with a reflectance of I are read to obtain a black reference signal and a white reference signal. Here, the black reference signal represents the offset for each pixel in the imaging system, and the subtraction of the black reference signal from the white reference signal represents the gain for each photoelectric conversion element in the imaging system. Therefore, these black and white reference signals are memorized, and the black reference signal (offset correction signal) is subtracted from the image signal obtained during actual image reading.
Furthermore, by dividing by a gain correction signal obtained by subtracting the black reference signal from the white reference signal, the image signal is normalized so that when the pixel on the document is black, O2 is 1, and when the pixel is white, it is 1, and shading is removed.

[Problem to be solved by the invention]

FIG. 3 shows a block diagram of an example of the standardization circuit.

The image data (analog signal) read by the CCD 31 is subjected to sample and hold, amplification, predetermined correction, etc. in an analog processing section 32, and then converted to a digital signal (8 bits) by an A/D converter 33. When reading the white reference plate, the switch 34 is ONL to the contact A side, and the switch 35 is ONL to the contact C side.
The white reference signal is stored in RAM+1136. When reading the black board and V board, the switch 34 is set to contact A side O.
With the switch 35 remaining in the N position, the switch 35 is set to the ONL contact side, and the black reference signal is stored in the RAM (2137).While reading the white/black reference board, the upper limit of the input level of the A/D converter 33 is determined by the +R
Determine EF and the lower limit - REF is input with a constant value that has been adjusted in advance so that all input signals fall within that range. Therefore, the switch 38 is turned on to the E side. When reading a document, the switch 34 is turned on to the contact B side, and the switch 38 is turned on to the contact F side (switch 35 is optional). At this time, white reference data and black reference data corresponding to the original image data input to the A/D converter 33 are stored in the RAM.
fl) 36 and RAM+2) 37, and the D/A
After being converted into analog signals by converters 39 and 40, the white reference data is input to +REF of the A/D converter 33, and the black reference data is input to -REF. As a result, A/D converter 3
The output of No. 3 becomes a normalized digital signal (8 bits), which is subjected to T-conversion in the γ conversion section 41 and then outputted from the I/F section 42 to the outside. As mentioned above, RAM (1136, RAM+2
) 37 are accurately performed by the address signal and write enable signal WE from the memory timing control circuit. Note that 43 is a memory timing control circuit.

By the way, if there are scratches, dust, or other deposits on the black reference plate, the reflectance of the scratched or dusty part is generally higher than that of the black part, so the output read by the image sensor will be
The output of the latter is higher (and the digital output of the A/D converter also has a larger value for areas with scratches or dust. If this normal value is used as the black reference signal for that pixel, normal black correction will be performed. For example, as shown in the circuit of Fig. 3, when the black reference signal is input to the -REF terminal of the A/D converter 33, the lower limit of the conversion range is raised, so the From the brighter original density to the level of the black reference signal, the value gradually shifts to the darker side compared to normal black-corrected pixels, and quickly saturates to the output O (Figure 2). As a result, black streaks appear on the image.

This occurs when a scratch, dust, or dirt spot is used as the black reference signal in a method in which one line of the signal read from the black reference board is used as the black reference signal.

In addition, in order to eliminate the influence of noise components included in the black reference signal, multiple lines of the read signal of the black reference board are cumulatively added for each pixel, then divided by the number of read lines, and the averaged value is used as the black reference signal. Techniques to do this have also been proposed.

FIG. 4 is a block diagram of a device for reading the image. Light source 1. While the optical system consisting of the lens array 4 and the like is scanning the black reference plate 8, the output of the image sensor 5 is amplified by the amplifier 6, converted to a digital signal by the A/D converter 7, and then converted to the first signal. The signal is cumulatively added to the line memory 14 through the adder 13. When the reading of the black reference plate 8 is completed, the cumulative addition result is stored in the second line memory 15. Thereafter, the data in the second line memory 15 is divided by the cumulatively added number of lines by the CPU 16 and the averaged value is stored in the second line memory 15 again. Next, the white reference plate 9 is read, and the read white reference signal is
The signal is input from the D converter 7 to the subtracter 11 in the standardization circuit 10, and the offset correction signal obtained from the second line memory 15 is subtracted therefrom. As in the case of the black reference signal, the calculation result is cumulatively added for each pixel by the adder 13 and the first line memory 14, and is stored in the line memory 14 as a gain correction signal. When reading the original 3, the standardization circuit 10 first uses a subtracter 11.
, the offset correction signal from the second line memory 15 is subtracted, and then input to the multiplier 18 , where the reciprocal of the gain correction signal from the first line memory 14 is obtained by using the reciprocal table 17 . Multiplyed. Through this multiplication, the original image signal is divided by the gain of the imaging system,
Shading is corrected.

However, even with this black reference signal generation method, it is not possible to completely eliminate the effects of scratches and dust on the black reference board.
Furthermore, since the degree of influence varies depending on the number of lines to be sampled, the line pitch, and the size of scratches and dust on the black reference plate, there is a possibility that abnormal images may occur. Furthermore, the averaging circuit requires a memory capacity equal to (number of sampling lines)×(number of pixels in l line), which also has the disadvantage of increasing costs.

The present invention has been made based on this background, and eliminates the drawbacks of the above-mentioned prior art, eliminates the effects of scratches and dust on the black reference plate, creates a black reference signal, and provides a highly accurate black reference signal. An object of the present invention is to provide an image reading device that can perform standardization.

[Means to solve the problem]

In order to achieve this object, the present invention provides an image reading device equipped with a standardization circuit that performs standardization using a black reference signal obtained from an image signal read from a black reference board. The present invention is characterized by comprising a comparison means for comparing image signals for lines, a storage means for storing the minimum value of the comparison result as a black reference signal, and a reading means for reading the stored black reference signal.

(Function) In order to minimize the adverse effects of scratches and dust on the black reference plate on black correction, the present invention samples multiple lines of the black reference plate and selects the minimum value among the image signals corresponding to the same pixel. This is adopted as the P base signal of that pixel.

If you increase the number of lines to be sampled, set the sampling pitch appropriately, and sample from the black reference sheet metal area, the black reference signal of the black reference plate will be much better than the conventional black reference signal obtained by sampling only one line or averaging multiple lines. The probability of being affected by scratches or dust is extremely low. Therefore, it is possible to perform accurate black correction.

〔Example〕

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows an embodiment of a black reference signal generating circuit according to the present invention. The circuit operation will be explained below.

After the image sensor starts reading the black reference board, the A/D converter output of a certain line is stored in the black reference signal line memory 19 at a predetermined timing. Next, at the timing of sampling the second line, the A/D
The output is input to the input side (A) of the digital comparator 20. At the same time, data in the line memory 19 corresponding to the A/D output pixel is read out and input to the input side (B) of the digital comparator 20. The digital comparator 20 selects the smaller one of the digital data manually entered in (A) and (B), and the line memory 1
9 is stored in the memory space at the same address. This operation is performed for all pixels in one line, and the smaller data of the same pixels in the two lines is stored in the line memory 19. The same operation is performed for each of the plurality of lines sampled on the black reference plate, and when reading of the black reference plate is completed, the minimum value of the plurality of lines for each pixel is stored in the line memory 19. The data in the line memory 19 becomes a black reference signal and is used as a white reference signal and a correction signal for the original image signal.

Writing and reading operations of line memory 19 are performed accurately by an address signal and write enable signal WE from memory timing control circuit 21.

Incidentally, the comparison means described in the claims is the digital comparator 20, and the storage means is the line memory 19.
Accordingly, the reading means is further constituted by the memory timing control circuit 2I.

〔Effect of the invention〕

As described above, according to the present invention, it is possible to generate a black reference signal while excluding the effects of scratches, dust, and lumps on the black reference plate, and to perform highly accurate standardization.

[Brief explanation of the drawing]

FIG. 1 is a control block diagram of the main parts of an image reading apparatus according to an embodiment of the present invention, FIG. 2 is a characteristic diagram showing the relationship between document density and A/D output value, and FIGS. 3 and 4 are FIG. 2 is a block diagram of a standardization circuit according to a conventional example. 19... Line memory, 20... Digital comparator, 21... Memory timing control circuit. Figure 1 Figure 4 Figure 2

Claims (1)

[Claims] In an image reading device equipped with a standardization circuit that performs standardization using a black reference signal obtained from an image signal read from a black reference board, a comparison means for comparing image signals for a plurality of lines at the same pixel of the black reference board; An image reading device comprising: storage means for storing the minimum value of a comparison result as a black reference signal; and means for reading out the stored black reference signal.