JP3105936B2 - Image reading device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image reading apparatus for reading an image using an image sensor, such as a facsimile, a digital copying machine, a video camera and the like.

[0002]

2. Description of the Related Art In general, in an apparatus for reading reflected light from a document by photoelectrically converting the reflected light from an image sensor such as a CCD, the illuminance distribution characteristic of a light source (that is, a distribution in which the illuminance is higher at the center than at both ends of the light source). The light-collecting characteristics of the lens (there is a so-called cosine fourth law in which light converges at the center of the lens) Sensitivity correction called shading correction is performed due to unevenness in sensitivity of each pixel of the image sensor.

[0003] There are various types of correction methods, and in addition to a mechanical correction method using a shading correction plate or the like, there is an electrical correction method using shading correction data. This electrical correction method uses a white reference plate, creates shading correction data for each pixel based on the read image data, and stores the shading correction data for each pixel in a ROM. When reading an image, the image data of each pixel is corrected and output based on the shading correction data.

[0004] When the shading correction is performed electrically as described above, the flattening rate becomes a problem. The “flattening rate”
It means what percentage of shading correction is applied to the peak value of the read image data on the main scan.
The higher the flattening rate is, the better the S / N ratio is, but the correspondingly, it becomes impossible to correct the pixels at both ends in the main scanning direction with a small amount of received light. That is, it is desirable that the flattening ratio is a ratio of the minimum value to the maximum value in the image effective range when the white reference plate is read. Therefore, even if the shading correction data for a certain flattening rate specified as, for example, 40% is stored in the ROM, it is not known whether or not the flattening rate itself is appropriate, so that appropriate shading correction cannot be performed. Will be.

In view of the above, a plurality of ROMs for storing shading correction data are provided, and 30%, 40%
Japanese Patent Application Laid-Open Publication No. Hei 9 (1995) -19764 discloses a method in which correction data groups corresponding to a plurality of flattening rates such as%, 50%, and 60% are stored, and a flattening rate is selected and a corresponding correction data group is selected and used. It has been proposed by the present applicant as Japanese Patent Application No. 1-295269. In this case, the flattening rate is selected, for example, by detecting the maximum value and the minimum value in the main scanning direction from the image data obtained by reading the white reference plate and obtaining the flattening rate by the ratio. The set flattening rate closest to the flattening rate is selected. According to such a plural flattening rate method, unlike the single flattening method, the flattening rate is too high to correct the portion where the amount of light is small at both ends, or the flattening rate is too low and the S / N ratio is too low. Inconveniences such as worsening are avoided.

[0006]

However, the following problems are also caused by such a proposed example system. That is,
Since the ratio between the maximum value and the minimum value when the white reference plate is read is not known in advance, in order to perform more appropriate correction, a large number of correction data groups are assumed assuming a large number of averaging rates. Must be prepared, a large-capacity ROM is required, and the cost is high. In particular, when a fluorescent lamp is used as the light source, it is known that the ratio of the amount of light between the center part and the end part greatly changes depending on the elapsed lighting time. Is required. Also,
Even if a large number of averaging rates are prepared in this way, the number of the averaging rates is limited, and considering that the ratio between the maximum value and the minimum value is essentially arbitrary, the optimal correction for the ratio at that time is considered. A data group cannot be obtained, and the values are inevitably skipped, so that optimal correction cannot always be performed.

[0007]

A white reference plate for shading correction, an image sensor having a large number of elements for reading an image and the white reference plate, and a read image signal for A
An A / D converter for performing A / D conversion, a RAM for storing a digital value obtained by converting the image signal read from the white reference plate by the A / D converter, and an image signal read from the white reference plate. A maximum / minimum value detector for detecting a maximum value and a minimum value, a ratio calculator for calculating a ratio between the detected maximum value and the minimum value, and for each image signal of each element read from the RAM. A correction data generator for generating shading correction data for the element using the ratio calculated by the ratio calculator, and correction using shading correction data generated for an image signal for each element read from the image And a computing unit for computing.

[0008]

The maximum / minimum value detector detects the maximum value and the minimum value of the read image signal for the white reference plate, calculates the ratio with a ratio calculator, and uses the calculated ratio to generate a correction data generator. Creates shading correction data for each element of the image sensor, and uses the correction data to correct the image signal for the original image with the computing unit. Correction can be performed. In this case, since a ROM is not required, the cost can be reduced.

[0009]

An embodiment of the present invention will be described with reference to the drawings. In the figure, a solid line path indicates a path at the time of image reading, and a broken line path indicates a path at the time of reading shading data using a white reference plate (not shown). First, the D / A that is the arithmetic unit
A converter 1, an amplifier 2, a D / A converter 3, and a RAM 4 are provided in this order. Further, a maximum / minimum value detector 5 for detecting a maximum value and a minimum value of the image signal read from the white reference plate and a ratio calculator 6 for calculating the ratio are provided. A correction data generator 7 for generating shading correction data based on the output ratio data and the image signal for each element stored in the RAM 4 and outputting the shading correction data to the D / A converter 1 is provided. Each of these handles 8-bit data.

In such a configuration, first, when reading shading data, the white reference plate provided outside the image area and on the reading start position side is read by the image sensor and input to the analog input terminal of the D / A converter 1. Let it.
Here, D / A converter is a digital-to-analog multiplier example, the analog input voltage v _i, 8-bit digital input value D, and between the analog output voltage v _o is, v _o = v _i - D / 255 …………………… (1) That is, v _o is a value obtained by multiplying v _i by D / 255.

An image signal _vo output from the D / A converter 1
Is amplified by the amplifier 2 and then input to the analog input terminal V _IN of the A / D converter 3 to be converted into 8-bit digital data for each pixel. At the time of actual image reading, this is output to the outside as image data. However, at the time of shading data reading, RA is output for each pixel in accordance with address information.
It is stored in M4.

[0012] When such shading data read, the image signal v _i is also input to the maximum-minimum detector 5, for example, a peak hold circuit a peak value (maximum value) in the image signal in the main scanning line And the edge output (minimum value) of the image reading effective range is detected by the edge hold circuit.

[0013] After the scanner scans the reference white plate and the reading is completed, the actual image reading is started. At this time, shading correction data generated for each pixel by the correction data generator 7 is input to the digital input terminal of the D / A converter 1. That is, the output from the maximum / minimum value detector 5 is input to the ratio calculator 6, and the ratio between the maximum value and the minimum value is calculated as the flattening ratio. The calculated ratio = flattening ratio is input to the correction data generator 7 together with each pixel data obtained by reading the white reference plate from the RAM 4, and shading correction data is generated for each pixel (each element). Now, assuming that the flattening rate is a and the pixel data stored in the RAM 4 is D _IN , the correction data D _OUT generated by the correction data generator 7 is D _OUT = 225 (0 ≦ D _IN <255a). .. (2) D _OUT = 225 ² · a / D _IN (255a ≦ D _IN ≦ 255)... (3) The D / A converter 1 performs a correction operation process using such correction data D _OUT for each pixel, obtains a corrected output, and outputs the output through the amplifier 2 and the A / D converter 3.

When D = D _{OUT in} equation (1) given by the correction data generator 7 when reading the white reference plate,
A value of D _OUT = 225 · a (0 ≦ D _IN ≦ 255) using the previous flattening rate a is used.

In this embodiment, the correction data of each pixel is sequentially generated in parallel with the actual image reading. However, if the processing cannot be performed in time, the correction data generator 7 stores the correction data in the RAM. A function may be provided so that shading correction values for all pixels are created and stored after completion of reading the white reference plate and before actual image reading starts.

[0016]

According to the present invention, as described above, the maximum / minimum value detector detects the maximum value and the minimum value of the read image signal for the white reference plate, and the ratio calculator calculates the ratio. Since the shading correction data is generated for each element of the image sensor by the correction data generator using the calculated ratio, and the image signal for the original image is corrected by the arithmetic unit using the correction data, Shading correction using optimal correction data can be performed for any flattening rate detected at that time, and good image quality with good S / N can be obtained. It does not require a ROM and can be realized at low cost.

[Brief description of the drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention.

[Explanation of symbols]

 Reference Signs List 1 arithmetic unit 3 A / D converter 4 RAM 5 maximum / minimum value detector 6 ratio arithmetic unit 7 correction data generator

Claims (1)

(57) [Claims] 1. A white reference plate for shading correction,
An image sensor having a number of elements for reading an image and the white reference plate, and an A / D converter for A / D converting the read image signal
A / D converter, a RAM for storing a digital value obtained by converting an image signal read from the white reference plate by the A / D converter, and a maximum value and a minimum value among image signals read from the white reference plate A maximum / minimum value detector for detecting the maximum and minimum values, a ratio calculator for calculating the ratio between the detected maximum value and the minimum value, and the ratio calculator for each image signal of each element read from the RAM. A correction data creator that creates shading correction data for the element using the calculated ratio, and a calculator that performs a correction operation using the shading correction data created for the image signal of each element read from the image. An image reading apparatus, comprising: