JPS63115462A - Method for correcting sensitivity variance of optical sensor - Google Patents

Abstract

PURPOSE:To avoid unevenness of reproduced pictures and to improve picture quality by having multiplication between the uncorrected picture signal and a correction coefficient or addition between them after converting them into logarithms for correction of the picture signal and rearranging at random the order of plural adjacent picture elements of the corrected picture signal to obtain the reproduced picture signals. CONSTITUTION:An A/D converter 22 performs A/D conversion for each picture element clock PCLK and delivers an 8-bit digital picture signal to a terminal at one side of a digital multiplier 24. While the white level correction data is supplied to the other terminal of the multiplier 24 from a correction memory 26 for each clock PCLK. The address signal ADR beta data is outputted from an output terminal of an adder 36 and supplied to an address input terminal A2 of a line memory 28. Then the reproduction picture data PLDATA undergone the random number correction is outputted from an output terminal D3. Thus it is possible to rearrange at random plural adjacent picture data of those picture data undergone the white level correction. As a result, the unevenness can be avoided with reproduced pictures.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for correcting variations in sensitivity of optical sensors for reading characters, images, etc. displayed on manuscripts in facsimile machines, image reading devices for plate making, etc.
- Layer In detail, it is an optical sensor (hereinafter referred to as a line sensor) composed of a large number of photoelectric conversion units arranged in a line.
In particular, before reading the original image, first create and store correction data based on the output signal of each photoelectric conversion unit when the white level reference incident light is input to the line sensor;
Then, when scanning the original image, the correction data is read out, the image signal is multiplied or logarithmized, and the reproduction image signal is subjected to addition processing, and rearrangement correction processing is performed on the image signal for high quality. The present invention relates to a method for correcting sensitivity variations in an optical sensor for obtaining reproduced image signals.

For example, in the fields of printing and plate-making, image scanning recording and reproducing systems are widely used to electrically process image information carried on manuscripts and create film masters for the purpose of streamlining work processes and improving image quality. ing.

This image scanning, recording and reproducing system basically consists of an image reading device and an image reproducing device. That is, in the image reading device, image information of a document that is sub-scanned and conveyed in one direction in an image reading section is main-scanned by an optical sensor in a direction substantially perpendicular to the sub-scanning direction and converted into an electrical signal.

Next, the image information photoelectrically converted by the image reading device is subjected to arithmetic processing such as gradation correction and edge enhancement according to plate-making conditions in an image reproducing device, and then converted to an optical signal such as a laser beam and film. The images are recorded and reproduced on an image recording carrier made of a photosensitive material such as. Note that this image recording carrier is developed by a predetermined developing device and used for printing or the like as a film original plate.

By the way, when the image reading device main-scans a document and reads its image information, a CCD in which a large number of photoelectric conversion units are arranged in a line along the main-scanning direction is used as an optical sensor.
(charge coupled device) is suitably employed. Here, the photoelectric conversion characteristics of each photoelectric conversion section constituting the CCD vary due to manufacturing errors and the like. Therefore, for example, even when an original with uniform density is scanned, there is a drawback that the obtained image signal is uneven.

Therefore, in order to eliminate such defects, various correction methods have been proposed. An example of this is the correction method disclosed in Japanese Patent Application Laid-Open No. 60-189372.

In this conventional technique, a correction coefficient for variations in sensitivity of each photoelectric conversion unit is determined by reading an original with uniform density in advance using a CCD, and the data is stored in a memory. In other words, when performing main scanning of a document, it is possible to obtain a corrected and even image signal by multiplying the image signals obtained from each photoelectric conversion section of the CCD by the correction coefficient.

However, even with this method, there are problems such as quantization error (difference between the value rounded by quantization and the sample value) at the stage of quantizing the image data read into each photoelectric conversion unit, so it is not completely accurate. However, the disadvantage of unevenness in reproduced images has not yet been resolved.

The present invention has been made in order to overcome the above-mentioned disadvantages, and the present invention reads a white level reference original with a line sensor before reading a desired original image, and based on the output signal of each photoelectric conversion unit of the line sensor. Correction data is created and stored in memory, and when a desired original image is read and scanned, the correction data is read out and multiplied or logarithmized and added to the image signal to create an image for reproduction. A method for correcting sensitivity variations in an optical sensor is provided in which a signal is obtained, and then a rearrangement correction process is performed on the image signal for reproduction to obtain reproduced image data for reproducing an even and high-quality image. The purpose is to provide.

In order to achieve the above object, the present invention corrects variations in sensitivity of the photoelectric conversion units by reading a reference original in advance when reading an original image using an optical sensor in which a large number of photoelectric conversion units are arranged in a line. Obtain the correction factor for
A corrected image signal is obtained by multiplying the uncorrected image signal obtained during the main scanning of the original image by the correction coefficient or adding the logarithmized image signal, and further calculates the value of the adjacent plurality of image signals of the corrected image signal. The method is characterized in that the order of pixels is rearranged at random to obtain a reproduced image signal for recording an original image on an image recording carrier.

Next, the method for correcting sensitivity variations in optical sensors according to the present invention will be described in detail below with reference to the accompanying drawings, citing preferred embodiments in relation to an apparatus for implementing the method.

In FIG. 1, reference numeral 10 indicates a reading section of an image reading device to which the method for correcting sensitivity variations in photosensors according to the present invention is applied. In this reading section 10, image information carried on the document S is transferred to a lens 1 by a CCD line sensor 12.
4 is read out photoelectrically. In this case, the original S is sub-scanned and conveyed in the direction of the arrow X, and main-scanned in the direction of the arrow Y by the CGD 12, so that image information on the entire surface thereof is read. Here, the CCD 12 has a large number of photoelectric conversion units arranged in a line along the main scanning direction (arrow Y direction), and is usually composed of several thousand photoelectric conversion units. The photoelectric conversion characteristics of each photoelectric conversion section are
As mentioned above, it is not constant due to manufacturing errors and the like.

For this reason, the sensitivity of each photoelectric conversion section varies.
For example, in order to obtain N-bit size correction data for the CD line sensor, first scan the document S using a white level document, and then calculate the image signal level P6 (0) of each photoelectric conversion unit obtained at this time. <P6≦PoMax), find 2' +p0xpHMax, and convert the value into an integer. Here, PoMa
x is the estimated maximum value of the image signal output when reading the white level original. The correction data obtained in this case will be hereinafter referred to as white level correction data.

Therefore, the method for correcting sensitivity variations in optical sensors according to the present invention will be described in more detail below.

FIG. 2 is a block diagram of an apparatus for carrying out the method of the present invention, in which arrows input to or output from each component, that is, a block, represent the flow of signals, and there are A display in which one short diagonal line intersects and a number (M=2.5.7.8.13) is attached near the short diagonal line indicates an M-bit signal.

Reference numeral 12 is a schematic CCD line sensor that outputs an analog image signal every time a pixel clock PCLK is input to a control terminal.

The analog image signal is introduced into an analog input terminal of an A/D converter 22 via an amplifier 20.

The control terminal of the A/D converter 22 is connected to the pixel clock PCLK.
is introduced, and the A/D converter 22 performs A/D conversion for each pixel clock PCLK and outputs an 8-bit digital image signal to one terminal of the digital multiplier 24. The other terminal of the multiplier 24 is connected to the CCD line sensor 12.
White level correction data from the correction memory 26 in which white level correction data for each of the photoelectric conversion units is stored is introduced every pixel clock PCLK. In this case, the multiplier 24 has the pixel clock PCLK at its control terminal.
Each time the 8-bit A/D converter 22 is introduced, the 8-bit image signal data of the A/D converter 22 and the 8-bit white level correction data of the correction memory 26 are multiplied for each photoelectric conversion section, and the resulting 8-bit The primary correction data PDATA is introduced into the data input terminal D2 of the line memory 28.

As described above, the pixel clock PCLK is introduced into each control terminal of the CCD line sensor 12) A/D converter 22) correction memory 26 and multiplier 24, and the pixel clock PCLK is introduced into each block. It operates every time.

On the other hand, reference numeral 30 is a reordering circuit for implementing the method of the present invention, and a quaternary counter 32 constituting an input section of the reordering circuit 30 is connected to the pixel clock PCLK.
will be introduced. The counter output terminal of the quaternary counter 32 is connected to the lower two bits of the 7-bit address input terminal A of the sequential memory 38. Furthermore, 5-bit output data from a random number generation circuit 40 that generates a random number every time the quaternary counter 32 carries out a carry is introduced into the upper 5 bits of the address input terminal AI of the sequential memory 38. The 2-bit output terminal of the sequential memory 38 is connected to the adder 3.
The adder 36 is connected to one input terminal of the adder 36, and the other input terminal of the adder 36 receives a 13-bit output signal of a block counter 34 which counts and amplifies each carry output of the quaternary counter by 4. A 13-bit address signal ADRβ data is output from the output terminal of the adder 36.
Address input terminal A! of the line memory 28! will be introduced in The output terminal D3 of the line memory 28 outputs the reproduction image data PLDATA that has been subjected to random number correction.

The apparatus for carrying out the method for correcting sensitivity variations in photosensors according to this embodiment is basically constructed as described above, and its operation and effects will be explained below.

First, the method of the present invention was carried out under the following conditions.

That is, (1) The number of pixels of the CCD line sensor 12 is 5ooo.

■ Consider one main scanning line.

(2) Data rearrangement is limited to four adjacent pixels of the CCD line sensor 12.

In this case, as shown in FIG. 3, the order memory 38 stores address data (5-bit upper address data) from the random number generation circuit (factorial of 4 = 24) and the output of the quaternary output terminal 32. 2-bit order data corresponding to address data (2-bit lower address data) is stored.

Incidentally, the corrected image data for reproduction by the conventional correction circuit is expressed as shown in FIG. 4a.

That is, in FIG. 4a, each grid is numbered 1 to 5 corresponding to the number of pixels of the CCD line sensor 12 in order from the left.
Addresses up to 000th are shown, and the numbers in the grid represent the pixel numbers of the corrected image data for reproduction multiplied by the white level correction coefficient. That is, the number l is CCD line sensor 1
The number N is the corrected image data for reproduction based on the Nth pixel of the CCD line sensor 12. As can be easily understood from the arrangement in FIG. 4a, the image data for reproduction obtained by the conventional correction circuit is stored in the line memory 28 along the pixel numbers of the CCD line sensor 12.

However, the image data for reproduction obtained by the rearrangement circuit 30 shown in this embodiment is expressed as shown in FIG. As is clear from the number arrangement in the grid in Figure 4, the image data in the line memory 28 according to the present invention has a structure in which four adjacent pieces of image data are randomly exchanged. .

Next, a method of arranging in such a random order will be described in detail below with reference to the drawings.

In FIG. 2, the quaternary counter 32 is a pixel clock PC.
When LK is introduced, its output data is 0.12.3
．． Outputs count data corresponding to 2-bit binary numbers such as 0.1.2.3, . . . 0.1.2.3. This count data forms the lower two bit address of address input A of the sequence memory 38 (see lower address in FIG. 3).

Next, the block counter 34 receives the carry signal from the quaternary counter 32 and inputs the count output to the adder 36 in units of blocks. That is, the block counter 34 is 4.
0.4 each time the carry output of the decimal counter 32 is introduced.
．． 8.12)... Outputs data that increases by 4. Next, the random number generation circuit 40 generates random numbers for 24 numbers from 0 to 23. That is, each time the carry output of the quaternary counter 32 is introduced, O1■, 2.3
, . . 22, 23, one random number is output as address data of the top five pins of the order memory 38. An example of this 5-bit upper address data is shown in FIG. 3 above.

Therefore, first, the rearrangement of the first to fourth pixels of the CCD line sensor 12 will be explained.

In this case, since the previous random number data remains as the output data of the random number generation circuit 40, that data is introduced into the address input terminal A1 of the sequential memory 38. Now let's assume that the random number is 3. That is, it is assumed that the upper address data (00011) shown in FIG. 3 is input as the upper address of the sequential memory 38. At this time, as is clear from the right column of the upper address data (00011) in FIG.
The lower two bits of address data (OO), (01), (10), and (11) of the bit address signal ADRα are introduced into the address input terminal A1 of the sequential memory 38 in synchronization with the pixel clock PCLK. Address data ADRα consisting of these 5-bit upper address data and 2-bit lower address data is stored in the sequential memory 38.
When introduced into the data output terminal.

2-bit output data (11), (00), (10), (01) are output according to the order data column of FIG. On the other hand, in this case (for the first to fourth pixels of the line sensor CCD 12), the output signal of the block counter 34 introduced to the other input terminal of the adder 36 is 0, so the output of the adder 36, That is, the 13-bit address data ADRα has a value equal to the output data of the output terminal D1 of the sequential memory 38, and when expressed in decimal notation, the value is 3.
It becomes 0.2.1. When this address data ADRβ3.0.2)■ is introduced into the address input terminal A2 of the line memory 28, the output image data PDATA of the multiplier 24 is transferred to the data input terminal of the line memory 28 in synchronization with this address data. Since it is introduced in D2, the memory map in the line memory 28 is configured as shown in Table 1.

Table 1 Memory map in line memory 28 Next, image data for reproduction from address 4 onwards in line memory 28 PLDAT
Consider the arrangement order of A. 5 in quaternary counter 32
Since the th pixel clock PCLK is introduced, the quaternary counter 32 outputs a carry. This carry signal is input to the block counter 34 and the random number generation circuit 4o. At this time, the output of the block counter 34 becomes 4 and a new random number is generated from the random number generation circuit 40. For example, the random number generated at this time has an upper address of (
10110), by repeating the same procedure as above, the data from the 5th pixel to the 8th pixel is 8.7.5
．． It is understandable that they are arranged in the order of 6 (see Figure 3). 1250 for 5000 pixels
By repeating random number generation twice, all pixels can be rearranged, and as a result, image data for reproduction in which four adjacent images are randomly rearranged as shown in FIG. 4 can be obtained.

As described above, according to the present invention, it is possible to randomly rearrange four pieces of image data adjacent to the white level corrected image data, so even if white level correction is conventionally performed due to quantization error etc. This provides the effect of eliminating unevenness in the reproduced image that could not be eliminated.

Although the present invention has been described above with reference to preferred embodiments, the present invention is not limited to these embodiments.
It goes without saying that various improvements and design changes can be made without departing from the scope of the present invention, such as increasing the number of adjacent image data items to be rearranged to 3 or 5, for example.

[Brief explanation of the drawing]

FIG. 1 is a perspective view of an image reading device to which the method of the present invention is applied; FIG. 2 is a schematic configuration diagram of an apparatus for carrying out the method of the present invention, and is a block diagram including a pixel sorting circuit; The figure shows memory data including random number addresses (upper addresses) stored in advance in the sequential memory, and FIG. ) is an explanatory diagram showing. 12...CCD line sensor 28...Line memory 32...Quadary counter

Claims (2)

[Claims] (1) When reading a document image using an optical sensor in which a large number of photoelectric conversion units are arranged in a line, a correction coefficient for correcting sensitivity variations of the photoelectric conversion units is obtained by reading a reference document in advance, and then, A corrected image signal is obtained by multiplying the uncorrected image signal obtained during the main scanning of the original image by the correction coefficient or adding the logarithmized image signal, and further calculates the value of the adjacent plurality of image signals of the corrected image signal. A method for correcting sensitivity variations in an optical sensor, the method comprising randomly rearranging the order of pixels to obtain a reproduced image signal for recording an original image on an image recording carrier. (2) In the amendment method described in claim 1,
A method for correcting sensitivity variations in an optical sensor, in which the number of adjacent pixels to be rearranged and corrected is set to 3 to 5.