JPH10313403A - Still image pickup device, color copying device and display device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a high resolution image by providing a storage means which stores a prediction coefficient of an estimation equation for still image signal conversion in each class and performing an estimation operation based on a read prediction coefficient and a segmented still image signal. SOLUTION: An area segmenting circuit 12 segments image data and supplies the segmented image data (predictive tap) to an estimation arithmetic circuit 16. A class code generation circuit 14 performs an operation based on compressed pattern data from an ADRC circuit 13, detects a class to which a block belongs and supplies a class code showing the class to a ROM table 15. The circuit 16 calculates image data of high resolution by performing an operation according to the prediction tap supplied from the circuit 12 and a prediction coefficient supplied from the table 15, and outputs it through an output terminal. With this, a waveform that is closer to an actual wave is reproduced, and image data that has an excellent image quality and high resolution is outputted.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a still image pickup device for outputting a high-resolution image, a color copying device, and a display device.

[0002]

In order to transmit video, audio, data, and the like with a limited transmission bandwidth and to record them on a recording medium having a limited capacity, digital video and audio are nowadays used. Compression and digital transmission technologies have evolved.

However, when compressing / decompressing image data to store a large amount of image data, the decompressed image data does not return to the original complete image data.
The resolution may be reduced and the image quality may be degraded. At this time, it is conceivable to reduce the number of pixels in order to reduce the capacity of the image data, but the image data with the reduced number of pixels has low resolution and poor image quality and cannot be restored to the original high resolution. Was.

In the case where pixel number conversion processing is performed to increase the number of pixels of image data, linear interpolation processing is conventionally simply performed based on the original data. No data could be obtained.

The present invention has been made in view of such circumstances, and has as its object to provide a still image pickup device, a color copying device, and a display device capable of outputting a high-resolution image.

[0006]

In order to solve the above-mentioned problems, a still image pickup apparatus according to the present invention comprises: an image pickup means for generating a still image signal corresponding to light picked up by a subject; A first storage unit, a second storage unit for storing a prediction coefficient of an estimation formula for converting a still image signal for each class, and a still image signal read out from the first storage unit, The extracted still image signal is compressed to generate a compressed data pattern, a class code is generated based on the compressed data pattern, and a prediction coefficient corresponding to the generated class code is read out from the second storage unit and read out. Signal conversion processing means for performing an estimation operation based on the obtained prediction coefficient and the cut-out still image signal and outputting a converted still image signal.

The color copying apparatus according to the present invention includes a still image reading means for reading a still image and generating three primary color signals, and a table showing prediction coefficients of an estimation formula for converting the three primary color signals for each class code. Storage means for storing;
The three primary color signals are cut out, the cut out three primary color signals are compressed to generate a compressed data pattern, a class code is generated based on the compressed data pattern, and a prediction coefficient corresponding to the generated class code is read out from the storage means. A signal conversion processing unit for performing an estimation operation based on the read prediction coefficients and the cut-out three primary color signals and outputting a converted color difference signal; and a printing unit for printing a color still image based on the color difference signal. Things.

[0008] A display device according to the present invention comprises a storage means for storing, for each class code, a prediction coefficient of an estimation formula for converting the number of pixels of an image signal, an input image signal, and an output image signal. To generate a compressed data pattern, generate a class code based on the compressed data pattern, read a prediction coefficient corresponding to the generated class code, and, based on the read prediction coefficient and the cut-out image signal. Signal conversion processing means for performing an estimation operation and outputting an image signal having undergone the pixel number conversion processing, and display means for displaying an image based on the image signal having undergone the pixel number conversion processing.

[0009]

Embodiments of the present invention will be described below with reference to the drawings. A still image pickup device, a color copying device, and a display device according to the present invention use the image data conversion device shown in FIG. First, the image data conversion device will be described.

As shown in FIG. 1, the image data conversion device includes area extracting circuits 11 and 12 for extracting image data.
And an ADRC (Adaptive Dynamic Range Codin) that compresses the extracted image data to generate a compressed data pattern.
g) a circuit 13, a class code generating circuit 14 for generating a class code to which the clipped image data belongs, a ROM table 15 in which prediction coefficients of the estimation formula are stored for each class, And an estimation operation circuit 16 for performing an estimation operation based on the image data.

An area extracting circuit 11 extracts image data supplied from an input terminal for each predetermined area, and supplies the image data to an ADRC circuit 13. Here, it is assumed that the area extracting circuit 11 extracts, for example, image data of six samples. On the other hand, the area extracting circuit 12 extracts image data in the same manner as the area extracting circuit 11 and supplies the extracted image data (hereinafter, referred to as a prediction tap) to the estimation operation circuit 16.

The ADRC circuit 13 forms pattern compressed data by performing an operation of compressing the cut-out image data of each region from, for example, 8 bits to 2 bits, and converts the pattern compressed data into a class code generation circuit. 14.

The ADRC circuit 13 usually has a VTR (Vide
o Tape quantizer performs adaptive quantization developed for high-efficiency coding. Here, since a local pattern of the signal level can be efficiently represented by a short word length,
It is used to generate codes for classifying signal patterns.

[0014] More specifically, when attempting to classification of the image data of six 8-bit example, it shall be classified into enormous number of classes of 2 ^48. Although it is ideal in terms of grasping the characteristics of the signal waveform, it is not practical because of the heavy load on the circuit. Therefore, the ADRC circuit 13
Classification is performed based on the pattern compression data generated in step (1). For example, if 1-bit quantization is performed on six image data, the six image data can be represented by six bits, and can be classified into 2 ⁶ = 64 classes.

Here, assuming that the dynamic range in the area is DR, the bit allocation is n, the data level of the pixels in the area is L, and the quantization code is Q, the ADRC circuit 13 calculates the area according to the following equation (1). Maximum value MAX and minimum value M
Quantization is performed by equally dividing the interval between IN and IN by a designated bit length.

[0016]

(Equation 1)

The ADRC circuit 13 compresses each image data extracted by the area extracting circuit 11 into two bits. Here, the number of samples of the cut-out image data is set to, for example, six, and the compressed image data is set to q1 to q6, respectively.

The class code generation circuit 14 calculates the equation (2) based on the pattern compression data from the ADRC circuit 13, detects the class to which the block belongs, and stores the class code class indicating the class in the ROM. Supply to the table 15.

[0019]

(Equation 2)

This class code class indicates an address read from the ROM table 15.

The ROM table 15 stores prediction coefficients of a linear estimation formula for calculating high-resolution image data for each class. The method of creating the prediction coefficients stored in the ROM table 15 will be described later.
From the ROM table 15, from the address indicated by the class code class, wn (c
lass) (n = 1 to 6) is read out. This prediction coefficient is supplied to the estimation operation circuit 16.

Based on the prediction tap supplied from the region extracting circuit 12 and the prediction coefficient wn supplied from the ROM table 15, a high-resolution image corresponding to the input image data is estimated. Calculate the data.

More specifically, the estimation calculation circuit 16 calculates a prediction coefficient w based on the prediction tap supplied from the region extraction circuit 12 and the prediction coefficient supplied from the ROM table 15.
Based on n (n = 1 to 6), the following equation (3)
The high-resolution image data is calculated by performing the calculation shown in (1) and is output via the output terminal.

[0024]

(Equation 3)

As described above, the image data conversion apparatus stores the prediction coefficients for estimating the high-resolution image data in the ROM table 15 and reads out the input image data and the input image data from the ROM table 15. The above-described image data can be output by performing an estimation operation based on the predicted coefficient thus obtained. That is, since the image data conversion device performs an estimation operation based on a prediction coefficient obtained by learning described later from actual image data,
By reproducing a waveform closer to the actual one, it is possible to output high-resolution image data with good image quality.

Next, a prediction coefficient generation device for creating (learning) prediction coefficients stored in the ROM table 15 will be described with reference to FIG. The prediction coefficient generation device stores prediction coefficients in the ROM table 15 by using image data of high resolution and good image quality.

As shown in FIG. 2, the prediction coefficient generator 2 is a low-pass filter (hereinafter, referred to as LPF).
21 and area extracting circuits 22 and 23 for extracting image data
An ADRC circuit 24 for compressing the extracted image data to generate a compressed data pattern, a class code generating circuit 25 for generating a class code based on the compressed data pattern, and a normal equation for forming a normal equation for each class code A circuit 26, a prediction coefficient determination circuit 27 that determines a prediction coefficient for each class code, and a ROM table 28 that stores the determined prediction coefficient.

In order to obtain prediction coefficients by learning, it is necessary to first generate compressed image data from already known high-quality image data. The LPF 21 removes the high-band component of the image data input via the input terminal and supplies the same to the region extracting circuits 22 and 23. The area cutout circuit 22 cuts out image data from the LPF 21 for each predetermined area. That is, the region extraction circuit 22
The same function as that of the region extracting circuits 11 and 12 described above is performed.
4 The region extraction circuit 23 is
The image data from the LPF 21 is cut out and supplied to the normal equation circuit 26.

As described above, the ADRC circuit 24 performs an operation of compressing all or a part of the data in each area from, for example, 8-bit image data to 2-bit image data to form pattern compression data. Then, the pattern compression data is supplied to the class code generation circuit 25.
Note that the ADRC circuit 24 is the same as the ADRC circuit 13 described above.

The class code generation circuit 25 is the same as the class code generation circuit 14 described above.
The class code is determined based on the pattern compression data supplied from the RC circuit 24. Thus, the class code generation circuit 25 generates a class to which the block belongs, and supplies a class code indicating the class to the normal equation circuit 26.

The normal equation circuit 26 receives each class code class supplied from the class code generation circuit 25, image data x1, x2,..., Xn supplied from the area extracting circuit 23 for each class code class. Using the high-quality image data y supplied from the terminal, a normal equation described later is established.

Here, for explanation of the normal equation circuit 26, learning of a conversion formula from a plurality of high-quality image data to normal image data and signal conversion using the prediction formula will be described. In the following, for the sake of explanation, a case will be described in which learning is generalized and prediction is performed using n samples. The levels of the image data are x1, x2,..., And xn, respectively, and the quantized data resulting from the p-bit ADRC processing is q1,.

At this time, as described above, when the levels of the image data are x1, x2,..., Xn and the levels of the high-quality image data are y, the prediction coefficients w1, w2,. An n-tap linear estimation equation is set using ‥ and wn. This is shown in equation (4). Before learning, wn is an undetermined coefficient.

[0034]

(Equation 4)

The learning is performed on a plurality of image data for each class. When the number of data samples is m, the following equation (5) is set according to the equation (4).

[0036]

(Equation 5)

If m> n, the prediction coefficients w1, ‥‥ wn
Is not uniquely determined, the element of the error vector e is defined by the following equation (6), and a prediction coefficient that minimizes the equation (7) is obtained. This is a so-called least squares solution.

[0038]

(Equation 6)

[0039]

(Equation 7)

Here, the partial differential coefficient according to wn in equation (7) is obtained. What is necessary is just to find each wn (n = 1 to 6) so that the following equation (8) is set to "0".

[0041]

(Equation 8)

In the following, Xi is calculated as shown in equations (9) and (10).
When j and Yi are defined,

[0043]

(Equation 9)

[0044]

(Equation 10)

Equation (8) can be rewritten to equation (11) using a matrix.

[0046]

[Equation 11]

This equation is generally called a normal equation. Here, n = 6.

After the input of all the learning data is completed, the normal equation circuit 26 adds
The normal equation shown in the equation (10) is established, and this data is supplied to the prediction coefficient determination circuit 27.

The prediction coefficient determination circuit 27 solves the normal equation using a general matrix solution such as a sweeping method for each wn, and calculates a prediction coefficient for each class. In other words, the above equation (11) is rewritten as equation (12), and X · W = Y (12) Equation (13) is obtained by a matrix solution method such as a sweeping-out method. ,
The determinant W of the prediction coefficient is calculated for each class code.

W = X ⁻¹ · Y (13) Then, the prediction coefficient determination circuit 30 writes the calculated prediction coefficient in the ROM table 28.

As a result of such learning, the ROM table 28 stores, for each class, a prediction coefficient for estimating high-quality image data y, which can be statistically approximated to the true value. . The table stored in the ROM table 28 is the ROM table 15 used in the image data conversion device 1 of the present invention, as described above.
With this processing, the learning of the prediction coefficient for creating the high-quality image data from the normal image data is completed by the linear estimation formula.

As described above, the prediction coefficient generation device 2
Can learn the deterioration of the image quality caused by the image compression based on the image data before and after performing the image compression, thereby generating a prediction coefficient for reproducing the high-resolution image data.

In the description of the present embodiment, ADR is used as a pattern generating means for generating a compressed data pattern.
Although the C circuits 13 and 24 are provided, this is only an example, and any information compression means that can represent a signal waveform pattern with a small number of classes can be freely provided. For example, DPCM ( Of course, compression means such as predictive coding) and VQ (vector quantization) may be used.

In this embodiment, the area dividing circuit 2
Although the area division is performed by the 2 and area division circuit 23, the present invention is not limited to this, and it is a matter of course that these area divisions may be performed by one circuit.

Next, a digital still picture camera device (hereinafter, referred to as a still camera) using the video data converter 1 will be described.

The still camera 30, as shown in FIG.
A CCD image sensor 31 for capturing an image of an object, an A / D converter 32 for digitizing an image signal, a memory 33 for storing image data, and a transmission which is an interface for transmitting image data to, for example, a so-called personal computer. Unit 34, a resolution creation circuit 35 to which the video data converter 1 is applied, a D / A converter 36 for converting image data into an analog signal, and an LCD (Liquid Crystal Displ) for displaying a subject image based on an image signal.
ay) 37.

When a shutter button (not shown) is pressed, the CCD image sensor 31 generates an image signal of one still image based on the image pickup light of the subject, and removes reset noise by correlated double sampling processing. , And supplies the image signal to the A / D converter 32. The A / D converter 32 digitizes the image signal and converts it into, for example, 8-bit image data.

The memory 33 stores image data supplied from the A / D converter 32 as a still image.
The memory 31 is not limited to the memory 31 as long as it can store image data, and may be, for example, an optical disk or a magnetic disk. The transmission unit 34 converts the image data read from the memory 33 so as to conform to a predetermined interface, and transmits the converted image data to a personal computer by, for example, infrared light or a cable.

The resolution creation circuit 35 is a circuit to which the image data converter 1 is applied. Therefore, the resolution creation circuit 35 converts the image data read from the memory 33 into high-resolution image data and supplies the high-resolution image data to the D / A converter 36.

Here, in order to generate the prediction coefficients stored in the ROM table 15, the LPF 21 may be provided in the prediction coefficient generation device 2 as described above. Since the image data is totally blurred when the LPF processing is performed on the image data, the prediction coefficient generation device 2 generates a prediction coefficient for predicting image data before the image is blurred by learning the blur. can do. When the number of pixels of the CCD image sensor 31 is small, a thinning-out processing filter for thinning out the number of pixels of high-resolution image data may be provided instead of the LPF 21. This makes it possible to generate a prediction coefficient for predicting high-resolution image data even when the resolution is deteriorated due to a decrease in the number of pixels.

The D / A converter 36 converts the high-resolution image data from the resolution creating circuit 35 into an analog signal and converts it into an image signal, and supplies the image signal to the LCD 37. Accordingly, a still image with high resolution and good image quality is displayed on the LCD 37. The image data output from the resolution creation circuit 35 may be transmitted to a personal computer or the like via the transmission unit 34.

As described above, the still camera 30 includes:
Even if an image is blurred or degraded by applying the image data conversion device 1, if a prediction coefficient generated by learning in advance is stored in a ROM table, a high level can be obtained based on the prediction coefficient. A still image with good image quality at a resolution can be generated.

As shown in FIG. 4, an image compression circuit 38a may be provided between the A / D converter 32 and the memory 33, and an image expansion circuit 38b may be provided between the memory 33 and the resolution creation circuit 35. . Such a still camera 30A
Can store more image data of still images in the memory 33 than in the still camera 30.

At this time, the prediction coefficient generation device 2
By providing an image compression / decompression circuit in place of 21, it is possible to learn that the image quality is degraded without completely returning to the original image data even if the compression / decompression processing is performed. That is, the prediction coefficient generation device 2 can learn the deterioration of the image quality by the image compression / decompression circuit, and generate the prediction coefficient for predicting the high-quality image data.

Note that a decoder / resolution creation circuit 39 may be provided as shown in FIG. 5 instead of the image expansion circuit 38b and the resolution creation circuit 35 of the still camera 30A. The still camera 30 </ b> B achieves high resolution while decoding image data read from the memory 33. At this time, in the prediction coefficient generation device 2, LP
If an image compression circuit is provided instead of F21, a prediction coefficient for achieving high resolution while decoding is generated.

Next, a color copying machine (copier) using the video data converter 1 will be described.

As shown in FIG. 6, the color copying apparatus 40 includes a scanner 41 for reading a color image such as a photograph and outputting an image signal, and an A / A for digitizing the image signal.
D converter 42 and memory 43 for storing image data
A class classification adaptive color conversion circuit 44 for outputting image data with good color resolution; a model-specific control table circuit 45 for switching ROM tables; a D / A converter 46 for converting image data into analog data; An ink control circuit 47 for controlling the output of ink and an output unit 48 are provided.

The scanner 41 reads a color still image such as a photograph, generates image signals R, G, and B of red (R), green (G), and blue (B), and supplies the image signals to the A / D converter 42. Supply. The A / D converter 42 converts the image signals R, G,
Image data R, G, and B obtained by digitizing B are supplied to a class classification adaptive color conversion circuit 44 via a memory 43.

The class classification adaptive color conversion circuit 44 is obtained by applying the image data conversion device 1 described above. The classification adaptive color conversion circuit 44 converts the image data R, G, B into image data Y, Cr, Cb by performing a conversion process by the image data conversion device 1.

Normally, when converting image data R, G, B into image data Y, Cr, Cb, a matrix conversion process is performed. Here, if the determinant of the image data R, G, and B is I, the determinant of the image data Y, Cr, and Cb is J, and the determinant of the matrix conversion is A, the following expression (14) is established.

J = A · I (14) Therefore, in the prediction coefficient conversion device 2, instead of the LPF 21, an inverse matrix conversion processing circuit that performs a conversion process of A ⁻¹ May be provided. At this time, if high-quality image data Y, Cr, Cb is supplied to the input terminal of the prediction coefficient conversion device 2, a prediction coefficient for predicting the image data Y, Cr, Cb is written in the ROM table 28. .

The class classification adaptive color conversion circuit 44 is provided with a plurality of ROM tables 15. The model-specific control table circuit 45 stores the ROM
The table 15 can be switched.

The classification adaptive color conversion circuit 44 supplies the image data Y, Cr, Cb to the D / A converter 46. The D / A converter 46 converts the image data Y, Cr,
The Cb is converted into an analog signal and supplied to the ink control circuit 47.
The ink control circuit 47 controls the output of the color ink in the output unit 48 to print a color image with good color resolution on paper.

The memory 43 also stores external image data via the receiving unit 49 and the decoder 50. The image data read from the memory 43 is transmitted via the encoder 51 and the transmission unit 52.

As described above, the color copying apparatus 40
Uses a class classification adaptive color conversion circuit 44 instead of the image signal matrix conversion processing circuit to prevent the color resolution from deteriorating due to the matrix conversion processing, and to provide a color still image with high image quality and good color resolution. Can be copied.

It is needless to say that an image data conversion device can be applied to a facsimile as well as the color copying device 40.

Next, a computer using the video data converter 1 will be described. The computer 6
0, as shown in FIG. 7, a camera input terminal 61 to which an image signal from a video camera is input, and a video capture 62 for performing a process of capturing the image signal for a computer.
, A motherboard 63, a graphic accelerator 64 for converting the number of dots of image data, and a display 65 for displaying an image based on the image data.

The video capture 62 digitizes the image signal supplied via the camera input terminal 61, captures it as computer image data, and supplies it to the motherboard 63.

The motherboard 63 includes a network board 67 and a CD-ROM in addition to the video capture 62.
(Compact Disc-Read Only Memory), which is connected to a CD-ROM drive 68 for reading out image data recorded on the memory and other peripheral devices (peripherals). The motherboard 63 supplies image data from the video capture 62, image data supplied via the network 66 and the network board 67, or image data from the CD-ROM drive 68 to the graphic accelerator 64.

The graphic accelerator 64 is obtained by applying the image data converter 1 described above. The graphic accelerator 64 is, for example, 640 dots × 48
Image data of a VGA (Video Graphics Array) of 0 dots is converted into 800 dots × 600 dots of SVGA (Super
Video Graphics Array).

Here, the graphic accelerator 64
The VGA image data is stored in the ROM table of
A prediction coefficient for predicting GA image data must be stored. To learn this prediction coefficient,
In the prediction coefficient generation device 2 described above, it is necessary to use a down-convert filter for converting image data from SVGA to VGA instead of the LPF 21. Then, when SVGA video data is input to the input terminal of the prediction coefficient generation device 2, a prediction coefficient capable of predicting SVGA image data from VGA image data is recorded in the ROM table 28.

The graphic accelerator 64 is an SV
GA image data is supplied to the display 65. Therefore, the display 65 has the above 800 dots × 600.
The image of the Dodd is displayed.

As described above, in the computer, when applying the image data conversion apparatus 1 to perform the conversion process of the dot of the image data, the prediction coefficients generated by learning in advance are stored in the ROM table. If stored, image data with high resolution and good image quality can be obtained based on the prediction coefficients.

In this embodiment, the case where the graphic accelerator 64 converts image data from VGA to SVGA has been described as an example.
Dot x 768 Dot XGA (eXtended Graphics Ar
ray) SXGA from 1280 dots x 1024 dots
(Super eXtended Graphics Array), and the reverse conversion is also possible.

Further, the present invention is not limited to the above-described embodiment, and various design changes are possible within the scope of the technical idea described in the claims. Needless to say.

[0086]

As described above in detail, according to the still image pickup apparatus of the present invention, even when the image is blurred or deteriorated, the prediction coefficient generated by learning in advance is used as the second coefficient.
In this case, a high-resolution and high-quality still image based on the prediction coefficient can be generated by the signal conversion processing unit.

According to the color copying apparatus of the present invention, 3
The primary color signal is cut out, the cut out three primary color signals are compressed to generate a compressed data pattern, a class code is generated based on the compressed data pattern, and the prediction coefficient of an estimation formula for converting the three primary color signals is determined for each class. And outputs a prediction coefficient of the generated class code, performs an estimation operation based on the prediction coefficient and the cut-out three primary color signals, and outputs a converted color difference signal. Deterioration of resolution can be prevented, and a color still image with high image quality and good color resolution can be copied.

According to the display device of the present invention, when the conversion process of the number of dots of the image signal is performed, the prediction coefficient generated by learning in advance is stored in the storage means. , It is possible to obtain an image signal with high resolution and good image quality.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a specific configuration of an image data conversion device.

FIG. 2 is a block diagram illustrating a specific configuration of a prediction coefficient generation device.

FIG. 3 is a block diagram illustrating a configuration of a still camera according to the present invention.

FIG. 4 is a block diagram showing a configuration of another still camera.

FIG. 5 is a block diagram showing a configuration of another still camera.

FIG. 6 is a block diagram illustrating a configuration of a color copying apparatus according to the present invention.

FIG. 7 is a block diagram illustrating a configuration of a computer to which the display device according to the present invention is applied.

[Explanation of symbols]

11, 12 area extraction circuit, 13 ADRC circuit, 14
Class code generation circuit, 15 ROM table, 16
Estimation calculation circuit, 31 CCD image sensor, 33
Memory, 35 resolution creation circuit, 37 LCD, 38a
Image compression circuit, 38b Image expansion circuit, 41 scanner, 44 class classification adaptive color conversion circuit, 47 ink control circuit, 64 graphic accelerator, 65 display

Claims (9)

[Claims] An imaging unit that generates a still image signal corresponding to imaging light of a subject; a first storage unit that stores the still image signal; and a prediction coefficient of an estimation formula for converting the still image signal. A second storage unit for storing each class, a still image signal read out from the first storage unit, a compressed data pattern generated by compressing the extracted still image signal, And a prediction coefficient corresponding to the generated class code is read from the second storage means, and an estimation operation is performed based on the read prediction coefficient and the cut-out still image signal. And a signal conversion processing means for outputting a converted still image signal. 2. The still image pickup apparatus according to claim 1, further comprising display means for displaying a still image based on the converted still image signal. 3. The image processing apparatus according to claim 1, further comprising: a compression unit that compresses a still image signal generated by the imaging unit and supplies the still image signal to the storage unit; and an expansion unit that expands a still image signal read from the storage unit. The still image pickup device according to claim 1, wherein: 4. The apparatus according to claim 2, wherein said second storage means stores, for each class, a prediction coefficient of an estimation formula for converting the compressed / expanded still image signal into an original still image signal. 3. The still image pickup device according to 3. 5. The apparatus according to claim 1, wherein said second storage means stores, for each class, a prediction coefficient of an estimation formula for converting a still image signal having a coarse number of pixels into a high-resolution still image signal. Item 7. The still image pickup device according to Item 1. 6. A still image reading means for reading a still image to generate three primary color signals, a storage means for storing a table showing prediction coefficients of an estimation formula for converting the three primary color signals for each class code, The three primary color signals are cut out, the cut out three primary color signals are compressed to generate a compressed data pattern, a class code is generated based on the compressed data pattern, and a prediction coefficient corresponding to the generated class code is stored in the storage means. Signal conversion processing means for performing an estimation operation based on the read prediction coefficients and the cut-out three primary color signals and outputting a converted color difference signal; and printing a color still image based on the color difference signal. A color copying apparatus including a printing unit. 7. A table switching control unit, wherein the storage unit stores a plurality of tables indicating prediction coefficients of an estimation formula for converting three primary color signals for each class code, and the table switching control unit includes 7. The color copying apparatus according to claim 6, wherein control is performed to switch to one of the tables stored in the storage unit. 8. A storage means for storing, for each class code, a prediction coefficient of an estimation expression for converting the number of pixels of an image signal; A pattern is generated, a class code is generated based on the compressed data pattern, a prediction coefficient corresponding to the generated class code is read, and an estimation operation is performed based on the read prediction coefficient and the cut-out image signal. And a display unit for displaying an image based on the image signal subjected to the pixel number conversion processing. 9. The method according to claim 8, wherein said storage means stores, for each class code, a prediction coefficient of an estimation formula for converting a low-resolution image signal into a high-resolution image signal.
The display device according to the above.