JP4121981B2 - Image reading device - Google Patents

Description

  The present invention relates to an image reading apparatus mounted on a copying machine.

  Digital copying machines have become popular in recent years, and CCD line sensors are used as image reading means in popular machines. Generally, a charge signal photoelectrically converted by a photodiode is transferred to an analog shift register for each odd / even pixel, and an image signal is output for each odd / even pixel (2ch output). Furthermore, due to the demand for higher productivity of image output, higher reading speed was required. However, the limit to speeding up with a CCD with 2ch output was reached, and one line sensor was divided in half (here, left and right). Note) CCDs with a total of 4 channels of output, left and right odd / even outputs, have come to be used.

  In such a 4ch CCD, even if images of the same density are read due to a difference in left and right CCD characteristics due to a difference between left and right devices (occurred during device fabrication), mainly due to a linearity difference, a difference in density occurs between the left and right. Conventionally, as a countermeasure, first, a chart (hereinafter referred to as a gray scale chart) having a plurality of stages of different densities in the sub-scanning direction (scanning direction) is read and stored in the memory. Based on either the left or right read data, image processing (γ correction) is applied to the other read data to adjust the density.

  As an example of the prior art, the variation in the left and right CCD characteristics is eliminated, and a stable image density without a difference in left and right density between the aircraft is reproduced. A pattern with gradation in the sub-scan direction is output from the test pattern generation circuit, and when it is read by the reading unit, the left data is used as a reference and correction is made from the variation in the right data. In addition, γ correction data is obtained from the left side data, and variations between machines are eliminated. These can be corrected with right and left variations and density variations with a single adjustment means, and can be adjusted efficiently (see, for example, Patent Document 1). There is an “image information processing apparatus, image information processing method, and image forming apparatus” that calculates correction data from a signal deviation and corrects the data after shading by a data conversion table for signal values (see, for example, Patent Document 2). ).

  In addition, as another conventional example, while adjusting the fluctuation of the offset component included in the CCD output signal to obtain a good image, the left and right division (FL type) CCD is taken as an example, and a slight offset difference of each channel is obtained. As a solution to the problem of differences in the read signal level on the left and right, causing a step, an adder circuit for adding an arbitrary section (pixel area) for each signal, subtracting the black offset There is a “subtracting circuit for performing subtraction, a register for setting a subtraction value, and an“ image input device and an image input system using the same ”that adjust the offset (see, for example, Patent Document 3).

Furthermore, it has a CCD check mechanism in the process and adjusts the transfer gate voltage of the CCD to adjust the linearity "CCD image sensor and image reading apparatus" (for example, see Patent Document 4), CCD check in the process There is an “image reading device and CCD image sensor” that has a mechanism and adjusts the linearity of the output from the transfer register by adjusting the voltage level of the switch gate to each transfer register of the CCD (see, for example, Patent Document 5). .
JP 2000-188686 A JP-A-11-215298 JP 2001-61104 A JP-A-10-276368 Japanese Patent Laid-Open No. 10-276305

  In the case of the above prior art, each reading apparatus has γ correction data for adjusting the left and right densities, and further, the read image data is corrected. However, when printing out the read image data (such as paper as an output image), γ correction may be strengthened in order to provide gradation on the high density (dark original image) side. This tendency is particularly strong for color images.

  For example, a gray scale chart is read within a normal reading range, and image data is digitized to generate correction data. If the left and right image density correction is performed on the normal read data from this correction data, a considerable density difference is corrected. However, if the γ correction corresponding to the print image is performed on the corrected image data, and the γ correction on the high density side is particularly strong, the slight difference in accuracy of the digital value after the correction is output as the density difference. Will appear.

  The present invention has been made in view of the above circumstances, and is characterized in that the correction accuracy on the high density side is improved with respect to the right and left density difference correction, and an image reading apparatus having no density variation in the left and right images is provided. Objective.

In order to achieve this object, the invention according to claim 1 is characterized in that a storage means for receiving optical image information and photoelectrically converting it is stored in one line of elements, and a potential signal stored in the storage means is stored in one line. Means for transferring every even number from the end in the first direction of the line, among means for one line, means for transferring every odd number from the end in the first direction, and one element for one line A photoelectric conversion element having means for transferring every even number from the end in the second direction opposite to the first direction and means for transferring every odd number from the end in the second direction of one line is used. An image reading apparatus configured to perform predetermined odd number and even number for each direction with respect to the even and odd analog image signals in the first direction of the photoelectric conversion elements and the even and odd analog image signals in the second direction. Adjust the gain and level to adjust Synthesizing the even signals and odd signals after, means for performing digital processing of the two systems with respect to the analog image signal synthesized for each direction, have a, using a photoelectric conversion element, a first direction of the image The high-density side digital image data digitized by the second system is used for the left-right density difference of the output image due to the linearity difference between the signal and the image signal in the second direction. High-accuracy image correction data is generated. For an image brighter than a high density, image correction data is generated from normal first-system digital image data. Correction data satisfying the above is generated, and the left-right density difference is corrected .

  The invention described in claim 2 is characterized in that, in the invention described in claim 1, means for performing two systems of digitization processing can set the range of digitization for each image signal.

  According to a third aspect of the present invention, in the first or second aspect of the present invention, the first system of digitization processing is set to a normal setting for digitizing an image signal in a range satisfying a reading density range of the image reading device, The second type of digitization processing is characterized in that it is set to digitize on the high-density document side.

  According to the present invention, it is possible to digitize the high density side (dark document) with high accuracy from the left and right image signals. The A / D converter can also be easily realized because it can use a device that has been conventionally used in the high-speed region with 2ch (odd / even).

  The best mode for carrying out the present invention will be described below in detail with reference to the accompanying drawings.

First, the structure and driving of the 4ch CCD will be described.
FIG. 2 shows a configuration of a general 4-channel CCD line sensor. This is mainly used for monochrome reading. In order to increase the reading speed, the sensor of one line is divided into two to form an odd / even pixel output of 4ch. As a result, the reading speed of the conventional 2-channel CCD can be read up to twice as fast as the operation clock.

  The drive terminals of the CCD are provided with an RS terminal, a CP terminal, and φ2B, φ1, and φ2 terminals (not shown) on the left and right sides inside the CCD. In general, signals are generated by a dedicated timing generation circuit for these driving operations, and one signal is output for each signal, and the signals are supplied to the left and right terminals by two drivers (buffers).

  In the analog shift register of each channel, charges are sequentially shifted by φ1 and φ2 clocks, and are output as electrical signals from the CCD by φ2B signals. The charge shift state is shown in FIG. The state of T11 is a situation where the shift state is stable as a result of the previous shift operation. From T11 to T12, the charge is transferred (transferred) to the adjacent register, and at T13, the transfer is completed. Further, in order to convert the charge signal into an electric signal (voltage) and output it to the outside of the CCD, this can be achieved by supplying a φ2B signal to the output gate.

  It has been clarified by a conventional experiment that a linearity difference between left and right appears depending on the waveform shape of the φ2B signal supplied here.

  As shown in FIG. 4, the conventional general signal processing configuration is such that OS1 (odd) and OS2 (even) 2ch signals output from the CCD 10 are input to Amp11, where each channel is designated. The gain is adjusted. In Amp 11, although not shown, the signals of OS1 (odd) and OS2 (even) are combined into a 1ch signal. This signal is input to the A / D 13a and digitally converted. The dynamic range of the digital conversion is set such that the Ref voltage on the Top side is Vr-T and the Ref voltage on the Bottom side is Vr-B, corresponding to the reading density range defined by the reading device.

  Similarly, 2ch signals of OS3 (odd) and OS4 (even) output from the CCD 10 are input to the Amp 12, where the designated gain is adjusted for each channel. Although not shown in the Amp 12, the signals of OS3 (odd) and OS4 (even) are combined into a 1ch signal. This signal is input to the A / D 14a and digitally converted. The dynamic range of digital conversion is set to Vr-T and Vr-B as the same Ref voltage as that of the A / D 13a.

A first embodiment of the present invention is shown in FIG.
The output up to Amp11 is the same as before, and the combined signal of OS1 (odd) and OS2 (even) is input to A / D 13a and A / D 13b. The Ref voltage of the A / D 13a is set to a voltage corresponding to the reading density range defined by the reading device. The Ref voltage of the A / D 13b is set according to the required resolution. To increase the resolution by n times,
Top side Ref voltage = Vr−B + (Vr−T−Vr−B) / n
Set. Thereby, the resolution of the digital value on the high density side can be increased. For example, when n = 8, it is possible to increase the resolution of 3 bits. Further, the resolution of A / D conversion will be described later as 8 bits.

  Next, the bit conversion means 15 that matches the number of bits of the output of the A / D 13a is passed through the resolution corresponding to n times the A / D 13b. When n = 8, the image data 1a is converted from 11 bits to 11 bits data shifted left by 3 bits. Here, the lower 3 bits are set to 0. The image data 2 output from the A / D 13b is data having an 11-bit resolution obtained by multiplying the 8-bit resolution by 8 because n = 8.

Next, the image data 1b and the image data 2 are arranged into one line of data by the data selection means 17. This is because when the value of the image data 2 is close to an 8-bit full scale, for example, 250 or less, the image data 2 is selected, otherwise the image data 1b is selected and aligned as one system of data. It is done. Further, as a function of the data selection means 17, there is also a function capable of arbitrarily selecting the image data 1b or the image data 2. Thereafter, shading is performed with 11 bits, and thereafter, the left and right images are corrected with 11 bits. When this series of processing is completed, the number of bits is changed again and data flows to the image processing block.
The processing after Amp 12 is the same as described above.

  In the second embodiment, the A / D conversion dynamic range shown in the first embodiment is realized by varying the A / D Ref voltage. A D / A converter is used to vary the Ref voltage. The output voltage of D / A21 is connected in common to A / D13a and 14a, and the voltage value of 1 / n times is connected to A / D13b and A / D14b from D / A22 and D / A23. In this way, two digital ranges are realized.

  In addition, by using individual D / A 22 and D / A 23 for A / D 13b and A / D 14b, it is possible to individually adjust the accuracy for n times. Since Vt-B is a standard for A / D conversion, they are all common. Here, the output of the D / A 24 is used, but a fixed voltage value may be used.

In the third embodiment, the voltage corresponding to the reading density range defined by the reading device is generated from the D / A 21 and connected to the Vr-T of the A / D 13a and the A / D 14a in the first and second embodiments. To do. Since the analog image signal on the high density side (dark document) is a signal close to Vr−B, Vr−B + (Vr−T−Vr−B) / A / D13b and A / D14b have Vr−T. n
By setting, analog signals on the high density side (dense density) can be digitized with high resolution.

  In the fourth embodiment, left / right density correction data is created from the image data 5 and the image data 6 which are arranged into one system of data by the data selection means 17 described in the first embodiment. The method of creation is to obtain correction data for each of the left and right of the reference image data from the image data 5 and the image data 6 read from the gray scale chart, and create correction curve data. This data is provided as a conversion means in the data conversion unit 19 and the data conversion unit 20, and corrects image data obtained by normal image reading. As a reference for correction, the above-described image data serving as a reference, or left and right image data may be used.

  As described above, according to the first embodiment of the present invention, the high density side (dark document) can be digitized with high accuracy from the left and right image signals. Also in the A / D converter, since a device that has been conventionally used in a high-speed region with 2ch (odd / even) can be used, it can be easily realized.

  According to the second embodiment of the present invention, the dynamic range (the digitization range) can be arbitrarily set, and can be set separately for the left and right, so that the left and right dynamic ranges can be matched, and correction data for the left and right density difference is obtained. The generation accuracy can be increased.

  According to the third embodiment of the present invention, it is possible to easily realize digitization of the entire reading area and high-precision digitization of the high density side (dark document) area.

  According to the fourth embodiment of the present invention, when generating the correction data for the left and right density differences, the density of the left and right images is corrected by using data obtained by increasing the resolution of the data on the high density side. The accuracy of correction can be increased. As a result, even if printer γ conversion is performed to increase γ on the high density side at the stage of obtaining an output image, it is possible to suppress the occurrence of density unevenness on the left and right, contributing to image stability. Particularly for color images, there is a tendency to strengthen (stand up) the printer γ on the high density side, so the effect of reducing the left-right density difference is great.

  As mentioned above, although the Example of this invention was described, it is not limited to the said Example, A various deformation | transformation is possible in the range which does not deviate from the summary.

  The present invention can be applied to all techniques for reading a document having a sheet through and a flat bed scanner.

1 is a block diagram illustrating a signal processing configuration according to an image reading apparatus that is an embodiment of the present invention. It is a block diagram which shows the structure of a general 4ch CCD line sensor. It is a figure which shows the general shift state of an electric charge. It is a block diagram which shows the signal processing structure concerning the conventional general image reading apparatus.

Explanation of symbols

10 CCD
11, 12 Amp
13a, 13b, 14a, 14b A / D
15, 16 bit length modulation xn
17, 18 Selector 19, 20 Image data processing block 21, 22, 23, 24 D / A

Claims (3)

Storage means for receiving optical image information and photoelectrically converting a potential signal in one line of elements;
Means for transferring the potential signal stored in the storage means from the end of the line in the first direction among the elements of the one line every even number;
Means for transferring every other odd number from the end in the first direction among the elements of the one line;
Means for transferring every even number from the end in the second direction opposite to the first direction among the elements of the one line;
An image reading device using a photoelectric conversion element having a means for transferring every odd number from an end in the second direction of the one line,
For the even and odd analog image signals in the first direction of the photoelectric conversion element and the even and odd analog image signals in the second direction, predetermined gain adjustment and level adjustment are performed for each odd number and even number for each direction. combines the even signal and odd signal after the adjustment, possess means for performing digital processing of the two systems, the relative analog image signal synthesized for each direction,
Using the photoelectric conversion element, a high density digitized by the second system with respect to the left-right density difference of the output image due to the linearity difference between the image signal in the first direction and the image signal in the second direction High-precision image correction data is generated for the high-density side using the digital image data on the side, and for images brighter than the high-density, the image correction data is obtained from the normal first-system digital image data. An image reading apparatus that generates and generates correction data that satisfies the reading density range of the image reading apparatus from the two correction data, and corrects the left-right density difference .   2. The image reading apparatus according to claim 1, wherein the means for performing the two systems of digitization processing can set a digitization range for each image signal.   The first system digitization processing is a normal setting for digitizing the image signal within a range that satisfies the reading density range of the image reading device, and the second system digitization processing is digitized to the high-density document side. The image reading apparatus according to claim 1, wherein the image reading apparatus is set.

Images