JP2005341097A - Image reading apparatus, image forming apparatus, and image reading method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image reading apparatus of a 4-channel division type wherein pixels are divided into two of, a first half and a second half in addition to 2 division of even number and odd number pixels, and the appearance of a linear density difference of information in a subscanning direction in the middle part of the main scanning direction is prevented to the utmost, or the density difference is made inconspicuous. <P>SOLUTION: For further dividing an even number pixel group and an odd number group that have been divided into two into a former half and a latter half, respectively, a dividing position of each of even number pixel group analog shift registers 204, 206 and odd number pixel group analog shift registers 203, 205, that is the number of output pixels to be divided, is variably and freely set by a pixel number setting circuit 213, thus preventing a linear density variation from being generated in a sub-scanning direction at the center of a main scanning direction, or thus making the density difference hardly conspicuous. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an image reading apparatus, an image forming apparatus, and an image reading method using a 4-channel division type photoelectric conversion element that further divides each of the image reading element into even-numbered / odd-numbered two divisions into a first half and a second half. About.

  Conventionally, in CCD linear image sensors used in image reading apparatuses used in digital copying machines, facsimiles, etc. in addition to a single image scanner, output is alternately performed at odd / even pixel numbers such as ODD / EVEN output. Sorting is done. In other words, an ODD / EVEN two-channel output image sensor and an analog signal processing circuit are provided, and analog image signals output from the image sensors are odd-numbered / even-numbered independently and processed in parallel independently, thereby increasing the reading speed. It has been.

  However, at present, there is an increasing demand for an image reading apparatus having a higher reading speed than before, and it is necessary to realize a reading speed that cannot be achieved by such an ODD / EVEN 2-channel output image sensor.

  For this reason, recently, as a sensor capable of realizing a reading speed twice as high as that of an ODD / EVEN 2-channel output image sensor, in addition to the ODD / EVEN separate reading, the light receiving element array is shifted left and right at the center in the main scanning direction. Dividing into two parts, the first half (First) and the latter half (Last), and dividing into 4 parts as a whole, so that the pixel frequency becomes 1/4, so-called 4-channel output image sensor (FL type image sensor) Has been proposed (see, for example, Patent Documents 1 to 4).

JP 2002-158837 A JP-A-11-215298 JP 2000-188686 A JP-A-11-261760

  However, since the outputs in the four areas of pixel sending of such an FL type image sensor pass through separate CCD analog shift registers and buffer amplifiers, the transfer efficiency of the four types of CCD analog shift registers and the output of the buffer amplifier It is conceivable that the offset or linearity shift due to the delay difference affects the image density. For example, when a halftone image such as a gray chart is read from the viewpoint of gradation, the offset or linearity due to the output delay difference of the buffer amplifier from the CCD analog shift register is restored to the original image density in the four-divided pixel area. A shift is added, resulting in four gradations in reading the same image. Further, this tendency is not noticeable because the difference becomes one pixel difference in ODD / EVEN division, but it becomes clear in the first half / second half division as it is ½. It becomes density difference.

  In the case of such a 4-channel output CCD, the difference occurs at the center pixel position, and as an image, the difference occurs in a straight line in the sub-scanning direction at the central portion in the main scanning direction. It will be. Even in the case of the conventional ODD / EVEN 2-channel output CCD, even if the same thing occurs, it is a difference in units of one pixel, so the difference is hardly understood, but in the case of the first half / second half division type, the difference in density is It may be clearly recognized.

  An object of the present invention is to provide a density difference that is linear in the sub-scanning direction in the central portion of the main scanning direction in the 4-channel division type in which each is further divided into the first half and the latter half in addition to the even / odd two divisions. Is to prevent the difference in density as much as possible or make the difference in density inconspicuous.

  An object of the present invention is to realize the above object with simple control.

  An object of the present invention is to make it difficult for human eyes to recognize the divided joints, thereby preventing the occurrence of a linear density difference in the sub-scanning direction at the central portion in the main scanning direction as much as possible, or reducing the density difference. It is to make it inconspicuous.

  An object of the present invention is to realize the above object with a simple configuration.

  An object of the present invention is to prevent an abnormality in output decrease in realizing the above object.

  An object of the present invention is to prevent the occurrence of a linear density difference in the sub-scanning direction at the central portion in the main scanning direction by using read image data as much as possible, or to make the density difference less noticeable.

  An object of the present invention is to prevent random density differences from appearing in the central portion of the main scanning direction as much as possible by using random numbers in addition to image data to be read, or to make the density differences less noticeable. It is.

  An object of the present invention is to improve halftone density reproducibility during color image reading.

  An object of the present invention is to provide a configuration capable of reducing a signal difference between a front half side and a rear half side of a photoelectric conversion element and reducing a density difference of image data.

  An image reading apparatus according to a first aspect of the present invention is a photoelectric conversion element in which a plurality of pixel sensors that receive optical image information that is reflected light from a document are arranged in the main scanning direction, and this photoelectric conversion element is an even pixel sensor. 4 channels that divide into even-numbered pixel groups and odd-numbered pixel groups composed of odd-numbered pixel sensors, and divide these even-numbered pixel groups and odd-numbered pixel groups into two parts, the first half and the second half in the main scanning direction, respectively. And a transfer circuit for transferring the optical image information read by the photoelectric conversion element to a 4-channel output unit, and an even-numbered pixel group and an odd-numbered pixel group with respect to the transfer circuit, respectively. And output pixel number setting means for variably setting the number of output pixels to be divided into two parts in the latter half.

  According to a second aspect of the present invention, in the image reading apparatus according to the first aspect, each pixel sensor of the photoelectric conversion element includes a photodiode that receives optical image information and performs photoelectric conversion, and a potential signal converted by the photodiode. The transfer circuit stores a charge stored in each storage unit of the photoelectric conversion element and a 4-channel analog shift register that performs data transfer to each output unit. A shift gate circuit for transferring to the analog shift register, wherein the shift gate circuit is capable of selectively switching the transfer destination for the analog shift register according to the setting of the number of output pixels of the output pixel number setting means. A plurality of pixel sensor units for transferring the charges accumulated in the pixel shift unit to the selected analog shift register. It comprises a gate electrode.

  According to a third aspect of the present invention, in the image reading apparatus according to the second aspect, the analog shift register is the pixel sensor located in the center region in the main scanning direction of the photoelectric conversion element for each of the even-numbered pixel group and the odd-numbered pixel group. Have overlapping areas that overlap.

  According to a fourth aspect of the present invention, in the image reading apparatus according to the second or third aspect, the output pixel number setting unit is a connection selection unit that selectively switches connection of the shift electrode to the analog shift register.

  According to a fifth aspect of the present invention, in the image reading apparatus according to any one of the first to fourth aspects, the output pixel number setting means variably sets the number of output pixels for each line read by the photoelectric conversion element. To do.

  According to a sixth aspect of the present invention, in the image reading apparatus according to any one of the first to fifth aspects, the output pixel number setting means variably sets the number of output pixels by serial data output to the transfer circuit. To do.

  According to a seventh aspect of the present invention, in the image reading apparatus according to any one of the first to sixth aspects, the output pixel number setting means is configured so that the number of output pixels to be divided in two does not overlap between the first half and the second half. Set.

  According to an eighth aspect of the present invention, in the image reading apparatus according to any one of the first to seventh aspects, the pixel sensor position for recognizing the feature of the image data of the read original and dividing it into two based on the feature is determined. Image data recognition means for determining the number of output pixels set by the output pixel number setting means is provided.

  According to a ninth aspect of the present invention, in the image reading apparatus according to the eighth aspect, the image data recognizing means has an edge detection function, and a plurality of pixel sensor positions whose data changes are large in the recognized image data. Each is determined as the number of output pixels.

  A tenth aspect of the present invention is the image reading apparatus according to the eighth or ninth aspect, further comprising prescan means for prescanning a document using the photoelectric conversion element, wherein the image data recognition means is acquired by prescan. Recognize the characteristics of the image data.

  According to an eleventh aspect of the present invention, in the image reading apparatus according to the eighth or ninth aspect, the image data recognizing means recognizes the feature of the read image data of the previous line.

  A twelfth aspect of the present invention is the image reading apparatus according to any one of the ninth to eleventh aspects, further comprising random number generating means, wherein the image data recognizing means has a constant level of data change in the recognized image data. In the following cases, the random number generated by the random number generator is determined as the number of output pixels.

  A thirteenth aspect of the present invention is the image reading apparatus according to any one of the ninth to eleventh aspects, further comprising random number generating means, wherein the image data recognizing means has a constant level of data change in the recognized image data. If it is larger, a plurality of pixel sensor positions where the data change is large is determined as the number of output pixels, and if it is below a certain level, the random number generated by the random number generating means is determined as the number of output pixels.

  According to a fourteenth aspect of the present invention, in the image reading device according to any one of the first to thirteenth aspects, the photoelectric conversion element has a three-line color CCD configuration having sensors for each of RGB colors.

  According to a fifteenth aspect of the present invention, in the image reading device according to the second aspect, the arrangement and wiring length of the analog shift register and the shift electrode are for the first half and for the second half with respect to the photodiode row. Are set to be the same.

  An image forming apparatus according to a sixteenth aspect of the present invention forms an image on a recording medium on the basis of the image reading apparatus according to any one of the first to fifteenth aspects of reading an image of a document and the read image of the document. A printer engine.

  According to the seventeenth aspect of the present invention, there is provided a photoelectric conversion element in which a plurality of pixel sensors that receive optical image information that is reflected light from a document are arranged in the main scanning direction; A four-channel transfer channel that divides a group into an odd-numbered pixel group composed of an odd-numbered pixel sensor and divides these even-numbered pixel group and odd-numbered pixel group into a first half and a second half in the main scanning direction, respectively. An image reading method using an image reading apparatus including a transfer circuit that forms and transfers optical image information read by the photoelectric conversion element to a 4-channel output unit, and reads a document image by the photoelectric conversion element And an output pixel number setting step for variably setting the number of output pixels for dividing the even-numbered pixel group and the odd-numbered pixel group into the first half and the second half of the transfer circuit, respectively, And a step of transferring processing optical image information to the output section of the four-channel read by forming a transfer channel of 4-channel the photoelectric conversion element to the transfer circuit in accordance with the number of output pixels.

  According to an eighteenth aspect of the present invention, in the image reading method according to the seventeenth aspect, a step of recognizing the feature of the image data of the read document and a pixel sensor position to be divided into two based on the feature of the recognized image data are determined. And determining the number of output pixels.

  According to a nineteenth aspect of the invention, in the image reading method according to the eighteenth aspect, the image data recognizing means for executing the recognizing step has an edge detection function, and the determining step is performed by the image data recognizing means. A plurality of pixel sensor positions having a large data change in the recognized image data are determined as the number of output pixels.

  The invention according to claim 20 is the image reading method according to claim 18 or 19, further comprising a pre-scanning step of pre-scanning a document using the photoelectric conversion element, wherein the step of recognizing is acquired by pre-scanning. Recognize the features of image data.

  According to a twenty-first aspect of the present invention, in the image reading method according to the eighteenth or nineteenth aspect, the step of recognizing recognizes the characteristics of the read image data of the previous line.

  According to a twenty-second aspect of the present invention, in the image reading method according to any one of the nineteenth to twenty-first aspects, the determining step includes generating a random number when the amount of data change in the recognized image data is below a certain level. The random number generated by the means is determined as the number of output pixels.

  According to a twenty-third aspect of the present invention, in the image reading method according to any one of the nineteenth to twenty-first aspects, the determining step includes the step of determining the data when the amount of data change in the recognized image data is greater than a certain level. A plurality of pixel sensor positions having a large change are determined as the number of output pixels, respectively, and when the level is below a certain level, a random number generated by the random number generator is determined as the number of output pixels.

  According to the inventions of claims 1 and 17, when the even-numbered pixel group and the odd-numbered pixel group divided into two are further divided into two parts, ie, the first half and the second half, respectively, these division positions, that is, should be divided. Since the number of output pixels can be variably set, it is possible to prevent the occurrence of a linear density variation in the sub-scanning direction at the center in the main scanning direction, or to make the density difference less noticeable.

  According to the second aspect of the present invention, since the photoelectric conversion element is composed of a photodiode array for one line, for example, since the reading light is different at the time of color reading, a one-line photodiode array configuration is provided for each RGB. The effect of not being affected by variations in sensitivity characteristics for each line of the multi-line photodiode array can be obtained. Further, as the transfer means, it is possible to realize high speed by having an analog shift register for 4 channels, and as a transfer means to the analog shift register for 4 channels, it is performed by opening and closing the shift electrode of the shift gate circuit. Therefore, a transport error can be prevented with a simple configuration.

  According to the third aspect of the invention, by limiting the overlap area required for the analog shift register to the center area in the main scanning direction, the function of changing the division position near the center is exhibited while taking advantage of high speed. Can be made.

  According to the fourth aspect of the present invention, it is possible to easily realize the control related to the change setting of the division position by changing the connection destination of the shift electrode.

  According to the fifth aspect of the invention, since the number of output pixels is variably set for each line reading, the joint portion at the center is changed, so that the joint portion does not continue in units of one line. It can be variably set, which makes it difficult for the joint part to be recognized by the human eye, and prevents the occurrence of linear density fluctuations in the sub-scanning direction at the center part in the main scanning direction, or the density difference can be reduced. It can be inconspicuous.

  According to the invention described in claim 6, since the number of pixels can be set by one or two serial data, the input control lines for setting to the photoelectric conversion element can be greatly reduced, and simple Can be realized with a simple configuration.

  According to the seventh aspect of the present invention, if the connection of the shift electrode overlaps in the overlap region of the analog shift register, the charge of the one-line photodiode array is dispersed, but overlaps in the first half and the second half. Since the number of output pixels is set so as not to occur, such an abnormality in output reduction can be prevented.

  According to the invention described in claims 8 and 18, the read image data is used, the feature is recognized by the image data recognition means, the appropriate pixel sensor position is determined as the division position based on the recognition result, and the corresponding. Since the division position is variably set by determining the number of output pixels, it is possible to prevent the occurrence of linear density fluctuations in the sub-scanning direction in the central part of the main scanning direction according to the actual image data, or the density difference Can be made inconspicuous.

  According to the ninth and nineteenth aspects of the present invention, the edge detection function of the image data recognition means is used to determine the pixel sensor position having a large data change in the image data as the division position. Therefore, it is possible to prevent the density variation from appearing in the central portion in the main scanning direction, or to prevent the density difference from being noticeable.

  According to the tenth and twentieth aspects of the present invention, the recognition / determination process of the pixel sensor position having a large data change in the image data is performed based on the prescan of the document image. Can be realized.

  According to the invention described in claims 11 and 21, paying attention to the fact that a normal document image has a correlation also in the sub-scanning direction, pixels having a large data change in the image data based on the read image data of the previous line. Since the sensor position recognition / determination process is performed, the inventions of claims 9 and 19 can be reliably realized without requiring time-consuming pre-scanning.

  According to the inventions of claims 12 and 22, in the case of a document having a uniform density, there is a possibility that regularity will occur and become noticeable even if the position where the data change is large is recognized / determined. When the amount of data change is below a certain level, the number of output pixels for the division position is randomly set using random numbers with no regularity, so the density fluctuation is straight in the sub-scanning direction at the center of the main scanning direction. Can be prevented, or the density difference can be made inconspicuous.

  According to the inventions of claims 13 and 23, it is possible to always determine the optimum number of output pixels for the division position.

  According to the fourteenth aspect of the present invention, when the photoelectric conversion element has a three-line color CCD configuration, the above-described configuration including a four-channel analog shift register is adopted for one-line photodiode array for each color separation filter. Accordingly, there is an effect of preventing density fluctuations that are straight in the sub-scanning direction at the center in the main scanning direction even during color reading, and halftone density reproducibility during color image reading can be improved.

  According to the fifteenth aspect of the present invention, the arrangement of the analog shift register and the shift electrode and the wiring length are set to be the same on the first half side and the second half side, so that the influence due to the difference in the wiring length and the like can be reduced. It is possible to cancel each other out, and it is possible to reduce the signal difference between the first half side and the second half side and reduce the density difference of the image data.

  According to the sixteenth aspect of the present invention, since the image reading device according to the first to fifteenth aspects of the present invention is provided, an image forming apparatus having the same effects as the first aspect of the present invention can be provided. it can.

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

[Prerequisite configuration example]
First, a schematic configuration of a full-color digital copying machine 1 to which an image reading apparatus and an image forming apparatus of the present invention are applied will be described with reference to FIG. The digital copying machine includes a scanner unit 2 that is an image reading device that reads an image from a document, and a printer unit 3 that forms an image on printing paper.

  In the printer unit 3, a toner cleaner 5, a charging charger 6, a laser scanner 7, four developing devices 8, a transfer belt 9 and the like are arranged around a drum-shaped photoconductor 4 arranged in an upper part inside. The transfer belt 9 and the fixing device 10 are arranged in the paper transport path 11 to form an electrophotographic mechanism 12 as a printer engine.

  A plurality of paper feed cassettes 13 and manual feed trays 14 for supplying printing papers having different sizes and directions are provided at positions where the electrophotographic mechanism 12 communicates with the paper conveyance path 11. A drive control mechanism (not shown) supplies the printing paper set in the difference tray 14 or the paper feed cassette 13 to the electrophotographic mechanism 12. Since the printer unit 3 of the present embodiment forms a full color image on the printing paper by the electrophotographic mechanism 12, each of the four developing devices 8 has YMCBk (Yellow, Magenta, Cyanide, Black). Color toners (not shown) are individually stored.

  In the scanner unit 2, a contact glass 16 is provided on the upper surface of the main body housing 15, and a reading document (not shown) is placed on the upper surface of the contact glass 16. The first scanning unit 17 is movably supported at a position facing the contact glass 16, and the second scanning unit 18 is movably supported at a position facing the first scanning unit 17. ing. Here, the first scanning unit 17 is formed of a halogen lamp 19 and a reflection mirror 20 whose reflection surface is inclined at 45 degrees, and the second scanning unit 18 is inclined at 45 degrees and has an inner angle of 90. It is formed by a pair of reflecting mirrors 21 and 22 that face each other at a degree.

  A three-line CCD 24 as a photoelectric conversion element is fixedly arranged via the imaging optical system 23 at a position facing the reflection mirror 22 of the second scanning unit 18. Consists of a CCD array, and B line, G line, and R line (none of which are shown) for reading B light, G light, and R light, respectively, are connected at intervals of several lines.

  Here, since the scanning speed of the first and second scanning units 17 and 18 is set to 2: 1, from the contact glass 16 to the 3-line CCD 24 via the first and second scanning units 17 and 18. The optical path length of the imaging optical path that communicates is constant even if the first and second scanning units 17 and 18 move. Then, the three-line CCD 24 photoelectrically converts the reflected light of the read original placed on the contact glass 16 and illuminated by the halogen lamp 19 into image data by such a fixed length imaging optical path.

  Next, a hardware configuration related to a scanner IPU (Image Processing Unit) for processing image data obtained by photoelectric conversion by the three-line CCD 24 will be described with reference to FIG. FIG. 2 shows a configuration example of a conventional block diagram applied to the scanner unit 2 of such a full-color digital copying machine 1. The CPU 92, which is a control unit on the control unit of the scanner IPU, executes the program stored in the ROM 93 and writes data or the like in the RAM 94, thereby controlling the entire scanner IPU. The CPU 92 is connected to the entire system of the digital copying machine 1 by serial communication with the system control unit 47 side, and executes an operation instructed by transmission / reception of commands and data. Further, the system control unit 47 is connected to the operation panel 25 by serial communication, and sets an operation mode and the like according to a key input instruction from the user.

  Further, an I / O (document detection sensor, home position sensor, document pressure plate open / close sensor, cooling fan, etc.) 95 is connected to the CPU 92, and the I / O 95 is detected and controlled on / off. The motor driver 96 is driven by the PWM output from the CPU 92 to generate an excitation pulse sequence, and drives a pulse motor 97 that scans the first and second scanning units 17 and 18. A lamp regulator 98 for lighting the halogen lamp 19 is also connected to the CPU 92.

  Various processing circuits for sequentially processing image data output from the 3-line CCD 24 are provided on the scanner IPU. First, the 3-line CCD 24 is supplied with each driving clock as a timing signal by a timing circuit (timing signal generating means) 99 on the control unit of the scanner IPU, and at a predetermined timing, each RGB first (first half) side and Last ( Analog signals of odd (odd pixels) and even (even pixels) on the second half side are output to the emitter follower circuits 100RF, 100RL, 100GF, 100GL, 100BF, and 100BL. The analog signals input from these emitter follower circuits 100RF,..., 100BL to the analog processing circuits 101RF,..., 101BL constituting the analog processing means are subjected to sampling processing by a subtraction method CDS (correlated double sampling) method as analog processing. The line clamp is performed at the optical black part of the 3-line CCD 24, and the output difference between the ODD and EVEN on the first (first half) side and the last (second half) side is corrected, so that the amplifier for each system Adjust the gain. After gain adjustment, the ODD and EVEN systems on the first (first half) side and the last (second half) side are combined in a time series by a multiplexer to form one analog signal, and finally the DC level Are input to the A / D converters 102RF,..., 102BL. An analog signal processing system is constituted by the elements from the 3-line CCD 24 to the A / D converters 102RF,.

  The analog signals input to the A / D converters 102RF,..., 102BL are converted into digital signals and then input to the shading correction circuit and time series conversion circuit 103. In the shading correction circuit, the FL output of the first (first half) side and the last (second half) side is made into one line, and the shading correction that corrects the uneven light quantity of the illumination system and the pixel output (sensitivity) of the 3-line CCD 24 is corrected. Perform processing functions. Further, the time series conversion circuit portion has a function of making the output of each First (first half) side and Last (second half) side into one line using a memory such as FIFO or LIFO. Of the image data that has been subjected to shading correction and made into one line, the image data for G and R is input to the inter-line correction memories 104R and 104G and delayed by the number of lines between the RGB lines on the three-line CCD 24. Thus, a process for performing alignment on the line is performed and input to the dot correction circuit 105. In the dot correction circuit 105, dots within one line are associated with the R and G image data output from the interline correction memories 104R and 104G and the B image data output from the shading correction circuit and the time series conversion circuit 103. A shift correction process is performed. Next, the scanner γ correction circuit 106 corrects the reflectance linear data by a look-up table method.

  The digital data corrected by the scanner γ correction circuit 106 is converted into an RGB filter, color conversion process, scaling process, create circuit 110, printer γ via an automatic document color determination circuit 107, an automatic image separation circuit 108, and a delay memory 109. It is input to the correction / write processing circuit 111.

  The automatic document color determination circuit 107 performs ACS (chromatic / achromatic determination) processing. In this ACS processing, black / gray determination is performed. In the automatic image separation circuit 108, as image area separation processing, edge determination (determined by the continuity of white pixels and black pixels), halftone dot determination (determination by the repetitive pattern of peak / valley peak pixels in the image), photo determination ( (When there is image data outside the character / halftone dot), the area of the character and the printing part (halftone dot part) and the photograph part is judged and transmitted to the CPU 92, and the subsequent RGB filter / color conversion, printer γ correction , YMCBk filter, used to switch parameters and coefficients in gradation processing.

  In the RGB filter, filter coefficients such as RGB MTF correction, smoothing, edge enhancement, and through are switched and set according to the previous region determination result. In the color conversion process, YMCBk conversion, UCR, and UCA processes are performed from RGB digital data. The scaling processing circuit performs enlargement / reduction processing in the main scanning direction of the image data. An image display unit 112 is connected to the RGB filter, color conversion process, scaling process, and create circuit 110 so that digital data after the enlargement / reduction process can be displayed. The create circuit performs create editing and color processing. In create editing, italics, mirroring, shadowing, and hollow processing are performed, and in color processing, color conversion, specified color deletion, undercolor processing, and the like are performed. The printer γ correction and writing processing circuit 111 performs printer γ conversion and setting of filter coefficients based on the previous area determination result. In gradation processing, dither processing is performed, and in video control, writing timing settings, image areas, white areas can be set, test patterns such as gray scales and color patches can be generated, and final image data can be written. Processing is performed so as to output to the laser diode in the laser scanner 7.

  Each of these function processes is connected to the CPU 92, and the setting and operation of each process is executed according to an instruction from the system control unit 47 by a program stored in the ROM 93.

[Conventional configuration example of color 3-line CCD]
The color 3-line CCD 24 is configured to perform 4-channel output for each photosensitive pixel line for each of RGB, and can be handled at high speed by being parallelized. At the same time, the CCD drive frequency can be lowered as compared with the conventional even / odd two-channel CCD, so that it is possible to secure a timing margin in the CCD drive. The RGB output has a three-line configuration with an RGB filter applied to the surface of the photosensitive pixel for each line. Since the configuration of each color is the same, the configuration of the 4-channel CCD unit of the output unit for R will be described as an example with reference to FIG.

  FIG. 3 shows a configuration of a four-channel output CCD 24R of the output unit for R. A photoelectric conversion element 200 in which light receiving elements (photosensitive pixels, photodiodes, etc.) S1 to S7400 are arranged in a row in the center. The shift gate 1 (201), the shift gate 2 (202) and the analog shift registers 1 to 4 (203 to 206) constituting the transfer circuit, and the output buffers 1 to 4 (207 to 210) serving as output units of four channels. ).

  As shown in FIG. 3, in the case of the 4-channel output CCD 24R, the signal output is divided into an even-numbered pixel group consisting of even-numbered pixel sensors and an odd-numbered pixel group consisting of odd-numbered pixel sensors. As a whole, the output configuration is four systems (four channels), so there are four independent CCD analog shift registers 1 to 4 indicated by 203 to 206, respectively. Therefore, the CCD analog shift register 1 (203) sequentially transfers and outputs the signals from the left end light receiving element belonging to the first half of the odd pixel group, and the analog shift register 2 (204) sequentially outputs the left end of the even pixel group belonging to the first half. The signals from the light receiving elements are sequentially transferred and output by the analog shift register 3 (205), and the signals by the right end light receiving elements belonging to the second half of the odd pixel group are sequentially transferred and output by the analog shift register 4 (206). The signals are transferred and output in order from the signal from the rightmost light receiving element belonging to the second half of the group.

  The last signals output from the first half and the second half (left and right) of each of the odd-numbered pixel group and even-numbered pixel group are the light-receiving elements S3699, S3700, S3701, and S3702 that are arranged side by side at the center of the light-receiving elements S1 to S7400. Signal. Control signals (φ1A, φ2A two-phase transfer clock, φ2B final transfer clock, SH shift gate signal, RS reset signal, CP clamp signal) necessary to drive such a 4-channel output CCD 24R are as follows: It is generated by the timing circuit 99 via a driver (not shown).

  Here, the joint area (divided area) of the first half and the second half of the elliptical broken line part in FIG. 3 will be described in more detail with reference to FIG. FIG. 4 is an enlarged circuit configuration diagram showing an example of the detailed configuration of the joint region. The light receiving pixel (pixel sensor) in the center is a photodiode and a MOS capacitor (accumulation) as shown as a principled equivalent circuit. Part) and a parallel circuit. The input light reflected from the original is photoelectrically converted by a photodiode to become a photocurrent, accumulated in a MOS capacitor to become a signal charge, and this charge becomes each photosensitive pixel in the shift gate 1 (201) or the shift gate 2 (202). It is transferred to any one of the CCD analog shift registers 1 to 4 (203 to 206) through the shift electrode 211 connected every time. The shift gate signal SH is a pulse for turning on and off the shift electrode 211. When turned on, the charge of the photosensitive pixel is transferred. This charge is sequentially transferred to the CCD analog shift registers 1 to 4 (203 to 206) by the two-phase transfer clocks φ1A and φ2A, and via the output buffer output buffers 1 to 4 (207 to 210) by the final transfer clock φ2B. The voltage signals OS1 to OS4 are output from the OS terminal.

  Such a configuration of the 4-channel CCD 24R itself is a conventional technique in a monochrome digital copying machine or the like.

  Here, in a black-and-white machine, it is desired for general users to perform image processing (AE, background removal) so that black and white can be clearly understood, and the density in halftones has not been a problem. . However, in color CCDs using three-line CCDs, there is a high demand for halftone color reproduction due to the pursuit of color reproducibility, so the following problems peculiar to FL CCDs may occur. That is, in the case of a four-channel output CCD, the difference occurs at the pixel at the center division position, and as an image, the difference occurs in a straight line in the sub-scanning direction at the central portion in the main scanning direction. . Even if the same occurs in the conventional CCD with even / odd 2-channel output, the difference is in the unit of one pixel, so the difference is hardly understood, but in the case of the first half / second half division type, the difference in density May be clearly recognized.

[Embodiment]
The present invention bisects an even pixel group composed of such even pixel sensors and an odd pixel group composed of odd pixel sensors, and divides these even pixel group and odd pixel group into the first half in the main scanning direction. With respect to the 4-channel output CCD that divides the output into the second half and the second half, the division position (joint position) of each first half / second half is changed in units of one reading line, that is, between the even pixel group and the odd pixel group. For each, the number of effective output pixels divided into two in the first half and the second half is variably set, so that density fluctuations that are linear in the sub-scanning direction are prevented or inconspicuous in the central portion in the main scanning direction.

  FIG. 5 shows a configuration example of the 4-channel output CCD 24R according to the present embodiment. In addition, as a code | symbol in drawing, it shows using the same code | symbol as the code | symbol shown in FIG.

  The ODD-side CCD analog shift registers 1 and 3 (203, 205) of OS1 and OS3 outputs and the EVEN-side CCD analog shift registers 2 and 4 (204 and 206) of OS2 and OS4 outputs are respectively center pixels in the main scanning direction. The arrangement configuration has an overlap region 212 that overlaps the region. Although it is possible to adopt a configuration in which the total number of pixels is taken as the overlap area 212 at the maximum, there is no merit in speeding up, so that the position of the joint near the center pixel in the main scanning direction can be moved as in this embodiment. It is desirable to set the number of pixels of the joints within a range where it is difficult for human eyes to move. In the present embodiment, an example is shown in which the overlap region 212 is set to 400 pixels as ± 200 pixels from the center.

  The position of the joint (output pixel number) is set by the pixel number setting circuit (output pixel number setting means) 213, and this setting is sent from the CPU 92 to the pixel number setting circuit 213 of the 3-line CCD 24 via the timing circuit 99. The number of effective pixels is controlled by the serial data to be connected.

  FIG. 5 shows an example in which the EVEN-side division setting is set to 3698 pixels or less as the first half side output and 3700 pixels or more as the second half side output for easy explanation. By giving such a set number of divided pixels (number of output pixels) by serial data 1, the EVEN side is set. Similarly, with respect to the ODD side division setting, an example is shown in which 3701 pixels or less are set as the first half side output and 3703 pixels or more are set as the second half side output. By giving such a set number of divided pixels (number of output pixels) by the serial data 2, the ODD side is set. In the present embodiment, for the sake of simplicity, an example is shown in which no difference is given to the divided pixel positions on the EVEN side and the ODD side, but the divided pixels on the EVEN side and the ODD side are overlap regions 212. Can be freely set by the serial data 1 and 2.

  The characteristics of the physical configuration of the 4-channel output CCD 24R according to the present embodiment will be described with reference to FIG. In this embodiment, the CCD analog shift registers 1, 2, 3, 4 (203 to 206) are arranged with respect to the center photodiode 200, and the wiring length is made the same on the front half side and the rear half side. Therefore, the difference in signal delay time due to the difference is reduced to reduce the difference in image data density. That is, the arrangement relationship of the CCD analog shift registers 1 and 2 (203, 204) on the first half side and the arrangement relationship of the CCD analog shift registers 3 and 4 (204, 205) on the second half side with respect to the central photodiode 200 are shown. This can be realized by using a symmetrical arrangement. Specifically, when the CCD analog shift register 2 (204) for the EVEN output (OS2) on the first half side is arranged near the photodiode 200, the CCD analog shift register for the ODD output (OS3) on the second half side 3 (S205) is similarly arranged close to each other so that the influence of the difference in wiring length and the like cancel each other out on the front half side and the rear half side. Conversely, when the CCD analog shift register 1 (203) for the ODD output (OS1) on the first half side is disposed near the photodiode 200, the CCD analog shift register 4 for the EVEN output (OS4) on the second half side. Similarly, (S206) may be arranged close to each other.

A joint area (divided area) in the first half and the latter half of the elliptical broken line part with respect to the overlap area 212 in the central part in FIG. 5 will be described in more detail with reference to FIG. FIG. 6 is an enlarged circuit configuration diagram showing a detailed configuration example of the joint region.
The overlap region 121 of the CCD analog shift registers 1, 3 (203, 205) on the ODD side can be selected by the shift electrode 2 (214) of the shift gate 1 (201), and the H, L of the pixel number setting circuit 213 can be selected. The CCD analog shift register 1, 3 (203, 205) to which the charge is transferred is determined according to the level. That is, the pixel number setting circuit 213 according to the present embodiment is configured as a connection selection unit that selectively switches the connection of the shift electrode 2 (214). In this example, the ODD side division setting is set such that 3701 pixels or less are set as the first half side output and 3703 pixels or more are set as the second half side output.

  When the shift gate signal SH is input in this state, the charge of the photosensitive pixel is transferred to the CCD analog shift register 1 or 3 (203 or 205) via the shift electrode 1 (215) and transferred by the transfer clock. Serial data 1 is input so as to change the next pixel number setting position during the OFF time of the shift gate signal SH, and the ON / OFF position of the shift electrode 2 (214) is controlled. By repeating this, it is possible to variably change and set the position (division position) of the joints for each line, and the density change of the joint parts can be changed by not arranging them continuously on a straight line. It is possible to make the level unrecognizable. The ON side of the shift electrode 2 (214) on the ODD side becomes an effective pixel, and this performs the effective pixel number setting (output pixel number setting). The OFF side is treated as an invalid pixel, and is transferred as idle feed by the CCD analog shift register 1 or 3 (203 or 205). This is deleted and synthesized based on the serial data 1 set in the pixel number setting circuit 2 when time series conversion on the first half side (First side) and the second half side (Last side) is performed by the time series conversion circuit 103 in the subsequent stage. Will be. Since the setting of the joint position for the EVEN-side CCD analog shift registers 2 and 4 (204 and 206) can be performed in the same manner, the description is omitted.

  The serial data 1 and 2 of the pixel number setting circuits 1 and 2 must be such that continuity cannot be recognized in the sub-scanning direction by the joint portion. Therefore, a method for determining the number of output pixels to be a joint portion (division position) will be described. In general, the serial number of the pixel number setting circuit 213 is not regular so that the continuity in the sub-scanning direction by the joint portion cannot be recognized, and the feature of the image data read from the document is recognized. However, by using a method of variably setting a pixel position having a large density change as a joint portion, in the present embodiment, it is difficult to be recognized by human eyes.

  FIG. 7 shows a block diagram configuration example applied to the scanner unit 2 of the present embodiment shown corresponding to FIG. In FIG. 7, the emitter follower circuit 100 or the scanner γ correction circuit 106 portion is the pre-stage processing unit 114, and the RGB filter / color conversion processing / magnification processing / create circuit 110 and the printer γ correction / write processing circuit 111 are the post-stage processing unit. 115 are collectively shown. In the scanner section 2 of the present embodiment, as shown in FIG. 7, an image data recognition circuit (image data recognition means) 301 and a random number generation circuit (random number generation means) 302 are added after the scanner γ correction circuit 106. ing.

  The image data recognition circuit 301 has a function of performing edge detection in the overlap region 212 as a feature of the image data of the read document. The edge detection function is a function that detects a change from white to black or from black to white, detects the maximum data change point position for the image data in the overlap area 212, and sets the number of pixels on the data bus. To the CPU 92. In response to this notification, the CPU 92 converts the number of pixels at the maximum change point into the serial data 1 and 2 as the number of output pixels and transmits it to the CCD 24 via the timing circuit 99.

  An example of the configuration of such an image data recognition circuit 301 is shown in FIG. Each RGB image data after the scanner γ correction is input to the image data recognition circuit 301. In the illustrated example, the image data recognition circuit 301 performs a difference calculation with respect to the previous pixel of the main scanning line by the previous pixel difference calculation circuit 303 that can be realized with the simplest configuration as a circuit that detects the edge of the image. In the previous pixel difference calculation circuit 303, the number of pixels in the overlap region 212 (in this example, 400 pixels) is set by the control logic unit 304 by an address bus and a data bus connected from the CPU 92. An image clock and a line synchronization signal from the timing circuit 99 are input to the control logic unit 304. By synchronizing with this signal, the pixel position of the edge portion of the image data in the overlap region 212 is counted, and the above difference is obtained. Arithmetic can be performed. The difference calculation result is input to the maximum value detection circuit 305 in the next stage. In the maximum value detection circuit 305, the maximum value of the difference result is input to the maximum value holding register 1 (306) for each main scanning line, and the next smallest value is input to the maximum value holding register 2 (306). The pixel position having the maximum value is input to the maximum pixel counter value holding register 1 (307), and the pixel position having the next smallest value is input to the maximum pixel counter value holding register 2 (307). These register values are read by the CPU 92 through the control logic circuit 304 by interrupt processing or the like. Here, the function of the determining means is executed by the CPU 92. Thus, the number of output pixels for the joint portion can be determined for each ODD and EVE from the pixel counter value of the maximum value and the next smallest value.

  By the way, in the case of a document having a uniform density in the overlap area 212, regularity occurs on the contrary by setting the maximum density change point recognized by the image data recognition circuit 301 as a joint. It is possible to make it stand out. In such a case, as an example of a method for preventing the serial data of the pixel number setting circuit 213 from having regularity in which the continuity in the sub-scanning direction by the joint portion cannot be recognized, a random number of output pixels is used. It is desirable to variably set. Therefore, in this embodiment, by providing the random number generation circuit 302 connected to the above-described image data recognition circuit 301, in such a case, a random random number can be set as the number of output pixels of the joint portion. Yes. When the change amount of the density change point is not more than a certain level in the image data recognition circuit 301, the random number generation circuit 302 is instructed to generate a random number by setting the random number generation request signal to the L level.

  That is, in the image data recognition circuit 301, the comparison circuit 308 compares the change point of the edge portion with a level set by the CPU 92 or more, and the maximum value holding register values 1 and 2 (306) are compared. By comparing each line, a random number generation request signal for each of RGB and ODD / EVEN is input to the random number generation circuit 302 via the control logic unit 304. If the change point is equal to or higher than the level set by the CPU 92, an H level signal is generated. If the change point is below the level set by the CPU 92, an L level signal is generated.

  The random number generation circuit 302 has a function of outputting random values of pixel positions in the overlap region 212 by the random bit generation circuits 1 and 2 (310) for each color. A random pixel position within the set range of the pixel counter in the random bit generation circuits 1 and 2 (310) is output. The setting range of the pixel counter is set via the control logic unit 311 by an address bus and a data bus connected to the CPU 92. An image clock and a line synchronization signal from the timing circuit 99 are input to the control logic unit 311, and the pixel position of the overlap region 212 can be counted by synchronizing from this signal. When the above random number generation request signals are H level signals, the random bit generation circuits 1 and 2 (310) do not operate. On the other hand, when each random number generation request signal is an L level signal, the random bit generation circuits 1 and 2 (310) can execute an operation of outputting a random value to transmit random joint information to the CPU 92.

  Therefore, the CPU 92 converts the read pixel position value of the maximum pixel counter value 1, 2 (306) or the pixel position value of the joint from the random bit generation circuit 1, 2 (310) into serial data 1, 2. Input to the CCD 24.

  FIG. 9 shows the image data of the overlap area 212 input to the image data recognition circuit 301 and the difference calculation result of the previous pixel difference calculation circuit 303 when a certain original image is read. When there are three edge portions in the overlap area 212 such as the 128th pixel of the edge portion 1, the 220th pixel of the edge portion 2, and the 315th pixel of the edge portion 3 as in this document, the maximum value holding register 1 (306). Is set to -60, which is the maximum value (absolute value) of the previous pixel difference, and the 128th pixel at that time is set to the maximum pixel counter value holding register 1 (307). The maximum value holding register 2 (306) is set to the next smallest value (absolute value) 40 after the maximum value of the previous pixel difference, and the maximum pixel counter value holding register 2 (307) is set to the 220th pixel at that time. Is done. As an example of the predetermined value in the comparison circuit 308, for example, ± 5 is preset from the CPU 92 and the comparison is performed. Since the change point of the edge portion is equal to or greater than a predetermined value, the joint portion is set to the 128th pixel and the 220th pixel in the overlap region 212. Each of the serial data 1 and 2 of the pixel number setting circuit 213 of the CCD 24 is set.

  Next, an example of processing for determining the number of output pixels for the joint position accompanied by image data recognition processing and the like executed by the CPU 92 will be described with reference to a schematic flowchart shown in FIG. The image data recognition mode is set by the user or service person according to the operation display unit 25 or each key input in the SP mode. First, if the pre-scan mode key is pressed (Y in step S1), the scanner unit 2 performs pre-scan of the document image (pre-scan means, pre-scan process). If the pre-scan mode key is not pressed (N in S1), serial data constant transmission processing setting is performed to perform recognition processing while reading the original image line by line (S2). Perform a read operation. When image data recognition processing as described above is performed by the image data recognition circuit 301 while performing normal reading in this way, a delay of one line or more occurs at least in the detection operation. Actually, a delay of several lines occurs because it passes through various image processing circuits such as shading correction, line-to-line correction, dot correction, and scanner γ correction. Here, since a normal document image has a correlation also in the sub-scanning direction, a delay of several lines is not a problem. Therefore, in such a case, it is possible to cause the image data recognition circuit 301 to execute recognition processing on the image data of the previous line read in the actual document image reading operation.

  However, such a method may cause a problem when reading a special document having no correlation in the sub-scanning direction, so the pre-scan mode is executed. In this pre-scan mode (Y in S1), serial data accumulation processing is set for image data read by pre-scan (S3). In this case, since image data is acquired in advance by pre-scanning prior to actual document reading, the image data recognition circuit 301 can set the number of output pixels for the joint without line delay.

  Even in such a pre-scan process, the series of operations described for the image data recognition circuit 301 and the random number generation circuit 302 are performed. The CPU 92 stores the read pixel position value of the maximum pixel counter value 1, 2 (306) or the joint pixel position value from the random bit generation circuits 1, 2 (310) in the RAM 94 for each line. During the main scan, the stored pixel position values are converted into serial data 1 and 2 and input to the pixel number setting circuit 213 of the CCD 24 via the timing circuit 99.

  After setting the presence / absence of such a pre-scan mode, selection processing of the output pixel number setting mode for three types of joints is executed. First, in the case of the random mode (Y in S4), the random number is generated by setting the connection signal line to the L level so that the random number is always generated as the number of output pixels by setting the random mode in the image data recognition circuit 301. The circuit 302 sets a random number of output pixels in the overlap region 212 (S5). In the maximum change detection mode (Y in S6), maximum change detection is always performed, and the number of output pixels at the maximum change point in the overlap region 212 is set (S7). In (maximum change detection mode + random mode) (Y in S8), the maximum change detection mode is executed when the change point is greater than a certain level depending on the level of the change point, and when the change point is less than or equal to a certain level. A mode for switching to execute the random mode is performed (S9).

1 is a schematic front view showing an example of the internal configuration of a full-color digital copying machine. It is a block diagram which shows the example of the conventional general hardware constitutions of a scanner part. It is a schematic block diagram which shows the structural example of the conventional general 4-channel output CCD. It is a circuit block diagram which expands and shows the detailed structural example of the joint area | region. It is a schematic block diagram which shows the structural example of 4 channel output CCD of this Embodiment. It is a circuit block diagram which expands and shows the detailed structural example of the joint area | region. It is a block diagram which shows the hardware structural example of a scanner part. It is a block diagram which shows the one example of a structure. It is explanatory drawing which shows the difference calculation result of the image data of an overlap area | region, and a previous pixel difference calculation circuit. It is a schematic flowchart which shows the example of an output pixel number determination process for the joint positions accompanied by an image data recognition process.

Explanation of symbols

2 Image reading device 12 Printer engine 200 Photoelectric conversion elements 201 and 202 Shift gate circuits 203 to 206 Analog shift registers 207 to 210 Output section 212 Overlapping area 213 Output pixel number setting means 214 and 215 Shift electrode 301 Image data recognition means 302 Random number Generation means

Claims (23)

A photoelectric conversion element in which a plurality of pixel sensors that receive optical image information that is reflected light from a document are arranged in the main scanning direction;
This photoelectric conversion element is divided into an even pixel group composed of even pixel sensors and an odd pixel group composed of odd pixel sensors, and these even pixel group and odd pixel group are divided into the first half and the second half in the main scanning direction, respectively. A transfer circuit that forms a transfer channel of 4 channels that divides into two parts, and transfers optical image information read by the photoelectric conversion element to an output part of 4 channels;
Output pixel number setting means for variably setting the number of output pixels that divides the transfer circuit into even-numbered pixel groups and odd-numbered pixel groups into a first half and a second half, respectively;
An image reading apparatus comprising: Each pixel sensor of the photoelectric conversion element comprises a photodiode that receives light image information and performs photoelectric conversion, and an accumulation unit that accumulates a potential signal converted by the photodiode as a charge,
The transfer circuit includes a 4-channel analog shift register that transfers data to the output units, and a shift gate circuit that transfers charges accumulated in the storage units of the photoelectric conversion elements to the analog shift register. Consists of
In the shift gate circuit, the transfer destination for the analog shift register can be selected and switched in accordance with the setting of the number of output pixels of the output pixel number setting means, and the charge stored in the storage unit is transferred to the selected analog shift register. The image reading apparatus according to claim 1, further comprising a plurality of shift electrodes for each pixel sensor to be transferred.   3. The image according to claim 2, wherein the analog shift register has an overlap region that overlaps the pixel sensor located in a central region in the main scanning direction of the photoelectric conversion element in each of the even-numbered pixel group and the odd-numbered pixel group. Reader.   The image reading apparatus according to claim 2, wherein the output pixel number setting unit is a connection selection unit that selectively switches connection of the shift electrode to the analog shift register.   5. The image reading apparatus according to claim 1, wherein the output pixel number setting unit sets the output pixel number to be variable every time one line is read by the photoelectric conversion element.   The image reading apparatus according to claim 1, wherein the output pixel number setting means sets the output pixel number to be variable by serial data output to the transfer circuit.   The image reading apparatus according to claim 1, wherein the output pixel number setting unit sets the number of output pixels to be divided in two so that they do not overlap in the first half and the second half.   2. An image data recognizing unit that recognizes a feature of image data of a read original, determines a pixel sensor position to be divided into two based on the feature, and determines the number of output pixels set by the output pixel number setting unit. 8. The image reading device according to any one of items 7.   9. The image reading apparatus according to claim 8, wherein the image data recognizing unit has an edge detection function and determines each of a plurality of pixel sensor positions having a large data change in the recognized image data as the number of output pixels. Pre-scanning means for pre-scanning a document using the photoelectric conversion element;
The image reading apparatus according to claim 8, wherein the image data recognition unit recognizes characteristics of image data acquired by prescanning.   The image reading apparatus according to claim 8, wherein the image data recognition unit recognizes the feature of the read image data of the previous line. Equipped with random number generation means,
12. The image data recognizing means determines the random number generated by the random number generating means as the number of output pixels when the data change amount in the recognized image data is below a certain level. An image reading apparatus according to claim 1. Equipped with random number generation means,
When the data change amount in the recognized image data is larger than a certain level, the image data recognizing means determines a plurality of pixel sensor positions where the data change is large as the number of output pixels, and The image reading apparatus according to claim 9, wherein the random number generated by the random number generation unit is determined as the number of output pixels.   The image reading apparatus according to claim 1, wherein the photoelectric conversion element has a three-line color CCD configuration having sensors for each of RGB colors.   The image reading according to claim 2, wherein the arrangement and wiring length of the analog shift register and the shift electrode are set to be the same for the first half and for the second half with respect to the photodiode row. apparatus. An image reading apparatus according to any one of claims 1 to 15, which reads an image of a document;
A printer engine that forms an image on a recording medium based on the read image of the document;
An image forming apparatus comprising: A photoelectric conversion element in which a plurality of pixel sensors that receive optical image information that is reflected light from a document are arranged in the main scanning direction, and this photoelectric conversion element is an odd-numbered pixel group that includes even-numbered pixel sensors and odd-numbered pixel sensors. A four-channel transfer channel that divides the pixel group into two and divides the even-numbered pixel group and the odd-numbered pixel group into the first half and the second half in the main scanning direction is formed. An image reading method using an image reading apparatus including a transfer circuit that transfers read optical image information to a 4-channel output unit,
Reading a document image with the photoelectric conversion element;
An output pixel number setting step for variably setting the number of output pixels for dividing the even-numbered pixel group and the odd-numbered pixel group into the first half and the second half with respect to the transfer circuit;
Forming a transfer channel of 4 channels in the transfer circuit according to the set number of output pixels and transferring the optical image information read by the photoelectric conversion element to the output unit of 4 channels;
An image reading method comprising: Recognizing the characteristics of the image data of the read document;
Determining a pixel sensor position to be divided into two based on the recognized feature and determining the number of output pixels;
An image reading method according to claim 17. Image data recognition means for performing the recognition step has an edge detection function,
The image reading apparatus according to claim 18, wherein the determining step determines a plurality of pixel sensor positions having a large data change in the recognized image data as the number of output pixels. A pre-scanning step of pre-scanning a document using the photoelectric conversion element;
20. The image reading method according to claim 18 or 19, wherein the recognizing step recognizes the feature of image data acquired by prescanning.   20. The image reading method according to claim 18 or 19, wherein the recognizing step recognizes the feature of the read image data of the previous line.   22. The determining step according to any one of claims 19 to 21, wherein the random number generated by the random number generating means is determined as the number of output pixels when the data change amount in the recognized image data is below a certain level. Image reading method. In the determining step, when the amount of data change in the recognized image data is larger than a certain level, a plurality of pixel sensor positions where the data variation is large are determined as the number of output pixels, respectively. The image reading method according to any one of claims 19 to 21, wherein the random number generated by the random number generation means is determined as the number of output pixels.

Images