JPH10164364A - Image reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce a memory, to use the multiplier of a small input signal number and to simplify a shading correction circuit by making the arithmetic operations of the integer part and decimal part of white reference data be independent. SOLUTION: In a white reference data memory 9, the data of reading a white rference original are stored beforehand. The white reference data are read synchronized with the read of an original and inputted to an address part of an arithmetic LUT 14. In the arithmetic LUT 14, LUT data for reciprocating the white reference data and converting them to bit constitution for expressing the integer part by one bit are stored beforehand. A most significant bit for indicating the integer part of the white reference data is inputted to a bit shift circuit 15 as a shift signal and the circuit shifts input image data to the left for one bit by the state of the shift signal. In the meantime, remaining signals excluding the most significant bit are inputted to the multiplier 16 and multiplied with input image signals, an arithmetic result is added with the output of the bit shift circuit 15 in an adder 17 and a shading correction arithmetic operation is completed.

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 having a shading correction function used for a digital copying machine, an image scanner, a facsimile, and the like.

[0002]

2. Description of the Related Art Conventionally, in an image reading apparatus using a one-dimensional image sensor, a light distribution characteristic of a light source lamp for irradiating a document, a change in light quantity with time, a sensitivity variation among a plurality of light receiving elements of the image sensor, and Shading distortion exists due to the influence of the characteristics of the optical system and the like, and has been a main factor causing deterioration of the read image.

In order to read a document image with high image quality, so-called shading correction for correcting deterioration due to this distortion is essential image processing.

FIG. 4 is a block diagram of a general digital correction process from image reading to shading correction by an image reading apparatus using a one-dimensional image sensor.

In FIG. 4, a light source lamp 1 irradiates an original 2 and its reflected light enters a one-dimensional image sensor 3 and is photoelectrically converted into an analog electric signal.

[0006] The photoelectrically converted image is amplified by a signal amplifier 4 and then input to an A / D converter 5 to be converted into a digital signal.

Thereafter, the dark offset removal circuit 6 subtracts the black reference data previously stored in the black reference data memory 7 from the read data of the document 2 to remove the dark offset, and the shading correction circuit 8 executes the white reference data memory 9.
After performing shading correction using the white reference data stored in the processing block, the data is sent to the next processing block.

In practice, it is necessary to acquire black reference data and white reference data as preparation for shading correction.

That is, in order to remove a dark current generated when no optical signal is incident on the image sensor and an offset noise generated in a circuit system of the image signal, the image sensor in a state where no optical signal is incident on the image sensor in advance. Is stored in a black reference data memory as black reference data, and data obtained by reading a white reference document is stored in a white reference data memory as white reference data.

It goes without saying that the black reference data is reduced by the dark offset removing circuit when storing the black reference data in the white reference data memory.

In order to reduce errors due to various kinds of noise when acquiring black reference data and white reference data, the reference data generation circuit 10 performs a noise reduction process.

To summarize the above, the arithmetic expression for shading correction is:

[0013]

(Equation 1)

As shown in FIG. In (Equation 1), each data corresponds to the coordinate position of the input image data in the main scanning direction.

FIG. 5 is a block diagram showing in detail the shading correction circuit 8 of FIG. 4 and its periphery.

In FIG. 5, the value of the white reference data stored in the white reference data memory 9 is converted by the data converter 11 so that the shading operation can be executed only by multiplication.
The white correction data is input to one of the multipliers 12.

The other input is original read data.
When the result of the multiplication exceeds the target white density level, the clipping circuit 13 clips to the target white density level.

As is well known, the shading operation is to divide the image data by the white reference data to normalize the data. However, in the case of a hardware configuration, it is generally realized by multiplying the reciprocal value of the white reference data. is there.

[0019]

However, in the above-described conventional configuration, it is necessary to use a multiplier for the shading operation and transform the data obtained by reading the white reference original into image data.

That is,

[0021]

(Equation 2)

As shown in (Equation 2), a value obtained by multiplying the reciprocal of the white reference data by the data of the target white density level is input to the multiplier as white correction data. The converted white correction data is divided into an integer part and a decimal part, and a dedicated memory suitable for each data bit number is required.

Also, a multi-input type expensive multiplier must be used. For example, when the image data is represented by a multi-value of 8 bits per pixel, WT =
The white correction data WC converted when 255 is set is 1.
In addition to being able to take a value in the range of 0 to 255.0, if the precision of the decimal part uses a maximum of 8 bits, it becomes data of a maximum of 16 bits including the integer part, and the multiplier used for shading correction is 16 × An 8-bit input type is required.

In particular, the bit configuration of a general-purpose memory is set so as to be 4 bits, 8 bits, 16 bits, etc., and when it exceeds 8 bits, 16 bits or a plurality of memories are used in parallel to create the reference data. Had to be stored.

These problems have led to an increase in the scale and complexity of the shading correction circuit and an increase in cost.

An object of the present invention is to solve the above-mentioned conventional problems, and an object of the present invention is to provide an image reading apparatus in which a memory can be reduced, a multiplier having a small number of input signals can be used, and a shading correction circuit can be simplified. And

[0027]

In order to solve this problem, an image reading apparatus according to the present invention comprises a storage means for storing data obtained by reading a white reference original before reading image data; A conversion unit for converting the read data into an irregular bit configuration, and performing shading correction using a multiplier on the original image data read based on the read data of the white reference original converted by the conversion unit. Correction means,
By making the operation of the integer part and the decimal part of the white reference data independent, the bit configuration of the storage memory is reduced to reduce the number of bits, thereby reducing the capacity and greatly simplifying the operation circuit.

[0028]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The invention according to claim 1 of the present invention is a storage for storing data obtained by reading a predetermined white reference original before reading an image by a one-dimensional image sensor having a light receiving element of a plurality of pixels. Means for calculating and processing the read data of the white reference original, and setting the integer part of the calculated value to 1
A conversion unit for converting into a bit configuration expressed in bits, and a correction unit for performing shading correction using a multiplier on a document image read based on the correction data converted by the conversion unit. This has the effect of reducing the memory capacity and simplifying the multiplier.

The invention according to claim 2 is provided with means for separating the shading correction calculation into an integer part and a decimal part of the correction data and performing the calculation in different systems, and has an effect of simplifying the multiplier. Having.

According to a third aspect of the present invention, read data of a predetermined white reference original is processed and the integer part of the calculated value is set to 1
A look-up table is used as a conversion means for converting into a bit configuration expressed by bits, and has an effect of improving the operation speed.

According to a fourth aspect of the present invention, as a conversion means for processing and processing read data of a white reference original and converting the integer part of the calculated value into a bit configuration expressing one bit, the CPU of the system performs the calculation. This has the effect of simplifying the arithmetic circuit.

According to a fifth aspect of the present invention, the operation of the integer portion of the correction data and the original image is performed by bit shifting the original image data, and has an effect of simplifying the multiplier.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of the present invention, and the same components as those of the conventional example are denoted by the same reference numerals.

In the figure, a white reference data memory 9 stores data obtained by reading a white reference document before reading an image of the document by a one-dimensional image sensor having a plurality of light receiving elements.

When performing the shading correction, the white reference data stored in the white reference data memory 9 is read out in synchronization with the reading of the original, and an arithmetic look-up table (hereinafter referred to as an LUT) composed of a general RAM is used. Is called)
14 is input to the address part.

The operation LUT 14 previously stores LUT data for reversing the white reference data and converting the white reference data into a bit configuration in which the integer part is represented by one bit.

The data output of the operation LUT 14, that is, the most significant bit indicating the integer part of the white correction data is input to the bit shift circuit 15 as a shift signal.

This circuit has the function of shifting the input image data left by one bit, that is, doubling it by the state of the shift signal.

On the other hand, the remaining signals of the data output of the operation LUT 14 except for the most significant bit are input to the multiplier 16 and multiplied by the input image signal.

Then, the operation result of the multiplier 16 is added to the output of the bit shift circuit 15 by the adder 17, and the shading correction operation is completed.

If the result exceeds the dynamic range of the image data, it is clipped by the clip circuit 18.

Here, the bit configuration of the white correction data converted by the calculation LUT 14 will be described.

Assuming, for example, that the image data has an 8-bit configuration, the data configuration after the white reference data memory 9 has the same 8-bit configuration as the image data.

In other words, the white correction data, which is the result of the operation of (Equation 2), may be represented by 8 bits.

Therefore, assuming that the range of the data of the white reference original is a realistic value, 255 to 86 in decimal and F in hexadecimal.
If it is limited within the level of Fh to 56h, the range of the operation result of (Equation 2) becomes 1.0 to 2.9651, and the integer part takes only two kinds of values of 1 or 2, so it is an alternative. Representation with one bit is possible.

The remaining 7 bits can be used for the decimal part. That is, as shown in FIG.
B) for the integer part and the remaining 7 bits for the decimal part.

Of course, one bit representing the integer part does not have to be always at the MSB, and may be at any bit position.

As a result, the white correction data, which is the calculation result of (Equation 2), can be represented by an 8-bit irregular configuration.

As a condition, as described above, the data of the white reference original must be within the range of FFh to 56h.
In other words, the density control of the white reference document and the gain adjustment of the signal amplifier 4 in FIG.

The shading calculation accuracy when the decimal part is represented by 7 bits, but the least significant bit when the decimal part is represented by 8 bits in the conventional method is omitted in the bit configuration of the present invention. is there.

That is, when the least significant bit of the conventional reciprocal data has a value of 1/256 of the original image, and the image data has an 8-bit configuration and the maximum value 255 is input,
255/256 = 0.9961.

That is, when the decimal part is represented by 7 bits, an arithmetic error of less than one gradation at the maximum occurs as compared with the conventional case,
It is easily possible to judge it as a practical range.

Conversely, if there are a total of 15 bits of white reference data and a circuit configuration of 8 bits for the integer part and 7 bits for the fractional part in the conventional method, and 8 bits while maintaining the same calculation accuracy according to the present invention. It can be realized with the circuit configuration of FIG.

Referring again to the operation LUT 14 of FIG. 1, it has been described that the integer part of the white correction data is input to the bit shift circuit 15.

That is, since the integer part takes only 1 or 2, the image input data can be selected to be multiplied or doubled, and the operation of the integer part can be performed only by shifting one bit to the left.

That is, the calculation of the shading correction in the present invention is performed by selecting the left one bit shift for the integer part of the white correction data, and the decimal part is calculated by a small multiplier having an 8-bit configuration. It is realized by adding.

The case where the data has an 8-bit configuration has been described as an example, but the same applies to an image data having an arbitrary bit configuration other than 8-bit.

As means for converting the white reference data into white correction data, in addition to the method using the calculation LUT 14, the central processing means (hereinafter referred to as CPU) of the image reading apparatus is stored in the white reference data memory 9 as shown in FIG. 19 is configured so that data can be directly accessed, and the white reference data memory 9
Is read by the CPU 19, and the calculation of (Equation 2) is performed by the CPU
There is a method of storing the data again in the white reference data memory 9 after the calculation in step 19 is performed.

By using the reference data structure of the present invention, 8-bit white reference data is temporarily stored using an 8-bit width memory, converted by the CPU, and then 8-bit white correction data is stored in the same memory. Can be written to.

In this case, since the calculation LUT 14 can be omitted,
Further simplification of the circuit and cost reduction can be realized.

[0061]

As described in detail above, according to the present invention, the white correction data necessary for the shading correction calculation is formed into an irregular bit configuration, and the number of data bits itself is reduced.
Compared with conventional shading correction means, the amount of calculation can be reduced, a large amount of memory is not required, and a small-scale multiplier can be used, so that the circuit can be simplified and the cost can be reduced. .

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of shading correction according to the present invention.

FIG. 2 is a diagram illustrating a bit configuration of white correction data according to the present invention.

FIG. 3 is a block diagram showing another embodiment of the present invention.

FIG. 4 is a block diagram showing a configuration of a conventional shading correction.

FIG. 5 is a detailed block diagram of a conventional shading correction.

[Explanation of symbols]

 Reference Signs List 1 light source lamp 2 original 3 one-dimensional image sensor 4 signal amplifier 5 A / D converter 6 dark offset removing circuit 7 black reference data memory 8 shading correction circuit 9 white reference data memory 10 reference data generation circuit 11 data converter 12 multiplier 13 Clip Circuit 14 Operation LUT 15 Bit Shift Circuit 16 Multiplier 17 Adder 18 Clip Circuit 19 CPU

Claims (5)

[Claims] 1. A storage means for storing data obtained by reading a predetermined white reference document prior to reading an image by a one-dimensional image sensor having a plurality of light receiving elements, and processing and processing the read data of the white reference document. Conversion means for converting the integer part of the operation value into a bit configuration expressing one bit, and correction means for performing shading correction using a multiplier on a document image read based on the correction data converted by the conversion means An image reading apparatus comprising: 2. The image reading apparatus according to claim 1, wherein the calculation of the shading correction is divided into an integer part and a decimal part of the correction data, and the calculation is performed by different methods. 3. A lookup table is used as conversion means for processing read data of a predetermined white reference original and converting it into a bit configuration in which an integer part of the calculated value is represented by one bit. The image reading device according to claim 1. 4. A conversion means for processing read data of a white reference original and converting it into a bit configuration in which an integer part of the operation value is represented by 1 bit, wherein the conversion is performed by a CPU. 2. The image reading device according to 1. 5. The image reading apparatus according to claim 2, wherein the arithmetic operation between the integer part of the correction data and the original image is performed by bit shifting the original image data.