JPS596666A - Reading method - Google Patents

Abstract

PURPOSE:To read out pictures of high quality at a high speed, by arranging picture elements in respective line of a sensor like a zigzag pattern by shifting them each other in the main scanning direction to interpolate the reading-out of an original and multiplying the interval between the lines of in the subscanning direction by an integer times the size of a systematic dither pattern in the subscanning direction. CONSTITUTION:Plural odd photosensitive element lines (O lines) consisting of plural odd elements S1-S4751 which are prepared for reading out an original and an even photosensitive element array (E array) S2-S4752 are arranged with a prescribed interval in the subscanning direction (Y) respectively like a zigzag pattern so as not to be overlapped to the Y direction along the main scanning direction (X). A linear array sensor is constituted by O-line analog CCD sensors 1, 3, E-line analog CCD sensors 4, 6 and photosensitive element arrays 2, 5. The interval between the lines in the Y direction is systematically fixed on an integer times the size of a dither pattern in the subscanning direction to read out the pictures of high quality at a high speed.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a reading method suitable for digital copying machines, facsimile machines, etc., and particularly to a zigzag pattern CCD sensor as a reader module for digital copying. '(FF)

Conventionally, digital document reading devices generally use CCD
(Charge-coupled device) A linear array sensor is used, but the focus and number of reading pixels are limited to, for example, 20'l per 1 sensor due to the sensor manufacturing process.In this case, the A3 To read a plate-sized original with a resolution of 16 mm/mm, it would require 11,752 focal points, so three of the above-mentioned CCDs had to be used in parallel. In this method of reading a plurality of CCD sensors side by side, a problem arises in the connection of pixels between the sensors. Solving this problem with a chemical automatic connection method would require a complicated configuration, which would increase costs.
'L Tx was good. Furthermore, another important problem with such conventional devices is that the original document is lifted by 1. If there are wrinkles or creases in the document (゛ is a curved surface such as the binding part of a book), the viewing angle of the reading lens changes, and at the junction between the sensor and the sensor, both sensors read the same area, resulting in overlapping images. will be output.

As a result, a copy may be obtained that differs from the original, and in some cases serious copying errors may occur. It is very difficult to correct this problem, and it is a problem that must be avoided.

On the other hand, the above-mentioned conventional problems can be essentially solved by creating a line sensor that can cover the entire document size per document. However, in order to read an A3 size image at 76 dots/sec, a linear sensor with a minimum focus of 1I752 is required. The current linear CCD sensor is
In a configuration with a 2ovr focus, the pixel pinch is 73 μm~/gμm. For example, if the pinch is narrowed to % in the design process, an element with a tO zoro focus will be obtained, and as an extension of this process, the pixel pinch will be 73 μm~/gμm. It is also conceivable to develop a new sensor. However, in the latter 1752 focus sensor, n
A channel diameter of 1 μm is always required for this purpose.
It becomes necessary to apply the width of the position as the Nayanel stonba. Therefore, in such a pixel pinch method, the effective pixel element sensitivity range is (/μm x 3'i) - daμm =
It became 78m, and the reading impression dropped significantly,
It could not be implemented. Furthermore, even if the CCD optical sensor for analog shift is placed on both sides, it is necessary to configure 2 ott stages on one side, so the transfer efficiency/MTF (
A sensor that is very disadvantageous in terms of spatial circumference*h's+cow] and f
I was unable to put it into practice because I was running out of time.

The present invention uses a plurality of COD sensors arranged in a zigzag pattern to solve the problem of connection between each sensor, IQ, and duplicate reading due to floating paper.
The purpose of this invention is to provide a reading method that is faster and has a higher resolution than the previous digital four-image reading system.

Hereinafter, the present invention will be described in detail with reference to the drawings.

FIG. 1 shows an example of the arrangement configuration of the original sensitivity density of a staggered pattern CCD sensor array to which the d-type method of the present invention is applied, where 81+83+...+84749 and S4
751 is an odd-numbered photosensitive If element row (hereinafter referred to as 0 row) I
S2184,...-* b4750 and 847
Reference numeral 52 denotes an even-numbered photosensitive element row (hereinafter referred to as E row).The 0 row and E row are spaced apart from each other by a predetermined distance in the Y direction (sub-scanning direction), and are arranged in the X direction (main scanning direction) as shown in the figure. Arrange the elements in parallel along the G line, n, and the elements in the 0 and E columns in the X direction (sub-scanning direction) in a staggered pattern so that the reading areas in the X direction are staggered. The gaps between the photosensitivity elements in the 0 and E rows (diagonal aMI in the figure) function as a Teyanel tone bar, and are masked so as not to have any photosensitivity. Therefore, the total photosensitive I-A array 81.82+ 83+
84+ “”” + 84761 +547sz is X
It is a device that can read the total number of pixels in the direction and the resolution of 76 dots/square, that is, the A3 size. As device dimensions that are generally applicable at present, for example, if the pinch between each pixel is /μm, and the Nayanel tombstone and masking are 7μm, the photosensitive element size of the fifth embodiment is 7μm, which means that the conventional element is % X when arranged in a row with pixel pinch
A significant increase in photosensitivity can be obtained compared to those in the 3-1I μm photosensitivity region.

FIG. 2 shows an example of the configuration of a staggered pattern arrangement of analog CCDs to which the reading method of the present invention is applied, where l and 3 are analog CCDs with 0 rows, respectively, and jN and 5 are photosensitivity lines extending in the X direction. between elements, lI and 6 are each 8
This is the analog COD in column 1.

CCD / and C on both sides of one photosensitive element row 2
CD 3 are arranged parallel to each other along the main scanning X direction.

Further, on both sides of the remaining photosensitive element row S, CCDq and CCD6 of row E are arranged in parallel along the main scanning X direction. With these five arrangement configurations, each CCD /, j.

The Nayanel transfer stages of V and 6 are 10241 stages,
Transfer efficiency despite multiple focus/M'l'
In terms of F, good results similar to those of the conventional method can be obtained.

Therefore, as a CCD sensor device, 201
Using an 11-pin CCD IJ near array sensor,
Regarding the direction (main running progeny), the sensor area is narrowed by Teyanel stomper and masking, and the Y direction (
lil II scanning direction), the elements are arranged at a predetermined interval (guyang), as in the embodiment of FIG. This creates a multi-pin linear sensor array with a staggered pattern.

This gap in the Y direction is usually an integer multiple of the reading resolution in the Y direction, so the systematic dither matrix for manning with the halftone processing method is
It is an integer multiple of the prime number m of the direction of music. Therefore, for example <z
In the dither matrix of xe, m=+, and the gap in the Y direction is +n (where n is an integer).In general,
If it is mn, it will be convenient in the processing described later. Since digital copies are usually reproduced with halftones, the size of the dither matrix is, for example, + The gap size between the line sensors is preferably an integral multiple of the number 2g.

FIG. 3 shows an example of the configuration of a reading thread using the zigzag pattern sensor according to the present invention shown in the 11th case or FIG. 2.

Here, 7 is the zigzag pattern CCD sensor v in Figure 1 or Figure 2.
, r is its driver, which drives the SENTU 7s in the 0 and E columns in parallel to increase speed.

As a result, the driver circuit R is simplified, and the clock noise of the read video signal occurs simultaneously in the sensors 7 of the 0 column and the E column, so that analog/digital (hereinafter referred to as A/D) The narrowness of the sampling pulse also increases in the conversion section.

The video signals read from the sensors 7 in the 0th row and the E row are amplified by /θ in separate G video amplifiers 9 and 1.
The A/D converter // is digitized by /2 and compared with the reference signal from the digital reference signal by /j. It is binarized to 0''. If the original is binary data equivalent to normal line copying, the Digital Reference/Hiro generates a predetermined threshold value according to the operator's instructions, and if the original is -7 tones, the operator If halftone Morton specifies, the systematic dither matrix threshold value 7 is used.

For example, in the case of a dither matrix of ψ×≠.

As shown in the figure, the threshold Pll for each pixel.

PI2 + """, P44 and - I1
Processed by α bronc. However, for pixels in the X direction, the threshold value given as a reference (matching value) from the Magnina Udo Converter /3 is pH* Plg + PI□+”13・ at the moment of timing 1=/.
. . . are repeated, whereas the threshold given by the same reference in one Madanitude Converter/S is the sensor gap in the Y direction, and 6. Since it is a common multiple of , P at the moment of elapsed time (timing) 1 = /
I3 + PI3 + PI2 + PI3...
will be repeated. Therefore, digital reference/F
For example, the moment when +XII threshold elapsed time 1=/ is read out in rows and divided into odd and even numbers corresponding to each sensor 7. Similarly, when the elapsed time is 1.lambda., 3, and l, the elapsed time is 1.lambda., 3, and l are processed. =l
The same threshold value is read out, and the same processing continues in a circular manner.

The signals subjected to the binarization process by the digital converter /3 or /S are stored in the digital memories /2 to 2/ as gap panzer memories. These memories 12 to 2/ are read/write controlled by the memory control/B.
It will be done. At that time, for example, the gap between each sensor in the 0th row and the E row of the staggered CCD sensor sensor is
If we multiply the resolution by 2g (this value corresponds to the least common multiple of IJ links), the time chart of the read signal is shown in Figure 5. Be the number one to show me. That is, compared to the 0th row sensor, the 1st row sensor has a 2ttM childization time lead (7) in terms of the quantization line in the Y direction.
・・・・・・+ Binarization processing done by ME26 @
This is because it is necessary to keep it in memory. Next, at elapsed time t = , 25, the signal from the 0th column sensor enters the y1 line electronic processing.

0 row banfa memo IJ M. Ki 1: @! be done. During the following elapsed time t-26, while the memory IJMo2 is being inserted (memory write), the odd-numbered pixel content Y of the 3'+ line is written to the memo IJMol at the elapsed time t=21 Note 1: @
Read out Uchitani's Memo'Jy+Mot. Similarly, while the contents of the 3'za line '2W are being stored in the first-column sensor memory ME26, a memo Uy1Mt1 is read out in which the contents of even-numbered pixels of the y1 line are written at an elapsed time of 1=/. At the elapsed time t = , 26, the content of the read TcoylMol and the content of yIMEI are combined in this way, and the signal yI of the /line in the Y direction is synthesized, and this n'i;e serial No. 16 is obtained. supply If this serial signal is given as a video signal to a laser beam printer, for example, the original image can be reproduced and recorded.

In this way, in the O row upper sensor, 2
Memories MO1 and M for lines. 2 lights alternately/
Lead control is used, and the other E-row sensor is used as a camp pants memory, with a number of lines corresponding to the number of lines between camps, in this example (2'1 plus 2).
Ill! 26 line bread 7 amemory MEI, MB2, ・-・corresponding to the H1 value of II
・.

MB26 will be used repeatedly.

It should be noted that a normal static memory (tAM) is suitable as this line panzer memory. However, if a relatively large number of line memories are required as buffers for an 8-row sensor, Dynamink RAM is better in terms of price, and especially in this case, the refractive indexing and read/write cycles are long. Therefore, it becomes possible to use Dynamink RAM g.

In addition, each party sensitivity element 81 (1=/~'f752
) is changed from the iE rectangular force shown in the figure to a rectangle that is long in the X direction, for example, if the dimension is j μm You can improve your performance.

J: As explained above, according to the present invention, linear array sensors consisting of CCDs, etc. are monolithically fabricated in two rows, and the elements in each row are arranged in a staggered pattern so as to interpolate each other in the array direction. Since the row spacing of the sensors formed on the monolithic array is an integral multiple of the dimension of the systematic dither pattern in the direction of the sub-travel force, the digital copying machine glue
As a module, it is possible to take high-speed, high-resolution, and high-quality images.

Furthermore, according to the present invention, a high-resolution multi-focus line sensor can be realized even with the current meridian manufacturing process. According to the present invention, by driving the 0 row and E row of the zigzag pattern sensor in parallel, high speed and sufficient light sensitivity are ensured. CCD
It is also possible to perform masking and configure l with an o0 position phase shifter, but it is preferable to make the gap as small as possible, as this will reduce the line pan spam memory.

The present invention has the advantage of being able to cope with six systematic dithered IJ links. In this case, if it is not the least common multiple, it can be done by changing the recombination of the dithered IJ links at that period. In this case, it is necessary to manufacture and design the camp pants in such a way that the amount of memory is reduced as much as possible.

[Brief explanation of the drawing]

FIG. 1 is an enlarged schematic diagram showing an example of the configuration of a linear array sensor according to the present invention, and FIG. An example of the structure of the reading thread of the present invention using the sensor shown in FIG. 11 or the sensor shown in FIG. 'f explanatory diagram, Fig. 5 is a timing chart of the blade signal generated in the reading system of Fig. 3. St, 8s+ 85187~821-1~8475
1-odd element sensitivity cumulative column (0 column), ``2+ 84+ sa, ssゝ821 '54752
``Even numbered photosensitive element row (E row), X...main running shoe direction, Y...secondary running shoe direction, /
, J...0 row analog CCD sensor,! , j... Original search element sequence. ≠, 6... E-row analog CCD sensor, 7... - staggered pattern CCD sensor, t... driver, ri, lθ... video amplifier. //, /2...A/D converter, /J, 1! ...Digital Magnina Udo Converter, lg...Digital Reference, /6...Memory Control, /7~Node...Tessital Memory (Gambu Banfa).

Claims (1)

[Claims] A linear sensor has a plurality of rows of pixel elements arranged in the main scanning direction for reading the original, and a plurality of rows arranged in the sub-scanning direction. The dither pattern is arranged in a staggered pattern so as to perform reading and interpolation of the original, and the inter-column gap in the sub-scanning direction of each column is an integral multiple of the dimension of the systematic dither pattern in the sub-scanning direction. Reading method.