JP2000350223A - Pixel data processor and pixel data processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a pixel data processor and a pixel data processing method which can reproduce an image high in quality. SOLUTION: An image interpolation part 100 includes a feature extraction part 1, a system selection part 2, a 1st system part 3, a 2nd system part 4 and a selector 5. The feature extraction part 1 extracts features of input data of a Bayer array. A system selection part 2 outputs a select signal for selecting the 1st system part 3 or 2nd system part 4 according to the extracted features. When the image data of the Bayer array are symmetrical, the interpolating process of the 2nd system 4 is selected. In other cases, the interpolating processing of the 1st system part 3 is selected. The selector 5 outputs interpolation data outputted from the 1st system 3 or 2nd system 4. Consequently, the optimum interpolating processing can be selected according to the features of the image data.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a pixel data processing apparatus and a pixel data processing, and more particularly, to a pixel data processing apparatus and a pixel data reproducing apparatus for reproducing a high-quality image from input pixel color data. It relates to a data processing method.

[0002]

2. Description of the Related Art With the spread of digital still cameras, imaging devices suitable for digital processing have been developed.
Among these image sensors, Bayer-type primary color CC
D image sensors have good color signal sensitivity and color reproducibility and are used in many digital devices.

[0003]

The Bayer type primary color CCD has red (R), blue (B) and green (G) filters alternately arranged.
Blue data and green data cannot be obtained for a pixel in which a red filter is arranged. Therefore, when a Bayer-type primary color CCD is used, it is necessary to interpolate missing data. For this reason, digital devices such as a conventional digital still camera are provided with an interpolation processing circuit that generates missing color data using only pixels of the same color as the interpolation target color, and reproduces an image. .

[0004] However, it is difficult to reproduce a high-quality image only with such a simple interpolation processing circuit. On the other hand, in consumer applications such as digital still cameras, high resolution and high-speed processing, and furthermore, realization with a microprocessor (small size) are essential.

Accordingly, the present invention has been made to solve such a problem, and an object of the present invention is to provide a pixel data processing apparatus and a pixel data processing apparatus which can obtain high-resolution pixel data at high speed with a small-scale circuit. An object of the present invention is to provide a pixel data processing method for a pixel data processing device.

[0006]

According to a first aspect of the present invention, there is provided a pixel data processing apparatus for processing a plurality of color data output from a plurality of color filters arranged for a plurality of pixels. A feature extraction unit that extracts features of a plurality of color data, an interpolation unit that performs mutually different processes for interpolating the color data to the interpolation target pixel in parallel, and a feature extracted by the feature extraction unit. A selector for selectively outputting any one of a plurality of interpolation processing results output by the interpolation unit, wherein the interpolation unit converts the color data based on a change amount of the color data of the pixel in a peripheral area of the interpolation target pixel. A first interpolation processing circuit that performs interpolation, and color data based on a ratio of color data of the same type as the color data to be interpolated and color data of a pixel of the same type as the interpolation target pixel in a peripheral area of the interpolation target pixel. And a second interpolation circuit for between.

According to a second aspect of the present invention, there is provided the pixel data processing apparatus according to the first aspect, wherein the feature extracting unit determines that the color data of the pixel in the peripheral area of the pixel to be interpolated is The selector detects whether or not there is symmetry. If there is symmetry, the selector selectively outputs the output of the second interpolation processing circuit.

A pixel data processing apparatus according to a third aspect is the pixel data processing apparatus according to the second aspect, wherein the plurality of pixels are arranged in a matrix in the vertical direction and the horizontal direction, and A first evaluation unit for detecting symmetry in the direction and a second evaluation unit for detecting symmetry in the horizontal direction, integrating the results of the first and second evaluation units to form a first interpolation processing circuit Alternatively, the method further includes a method selection unit that outputs a selection signal for selecting one of the second interpolation processing circuits, wherein the selector is configured to output one of the first interpolation processing circuit and the second interpolation processing circuit in accordance with the selection signal. One output is selectively output.

A pixel data processing apparatus according to a fourth aspect is the pixel data processing apparatus according to the second aspect, wherein the plurality of pixels are arranged in a matrix in the vertical and horizontal directions, and A first evaluation unit that detects the symmetry in the direction, a second evaluation unit that detects the symmetry in the horizontal direction, and a third evaluation unit that detects the symmetry in the oblique direction; And a method selection unit that integrates the results of the second evaluation unit and the third evaluation unit and outputs a selection signal for selecting one of the first interpolation processing circuit and the second interpolation processing circuit, Selectively outputs the output of one of the first interpolation processing circuit and the second interpolation processing circuit in accordance with the selection signal.

A pixel data processing device according to a fifth aspect is the pixel data processing device according to the second aspect, wherein the plurality of color filters are arranged according to a Bayer array, and each of the plurality of color data is a red data. , Blue data, or green data, and the feature extraction unit detects the symmetry of the green data.

A pixel data processing apparatus according to a sixth aspect is the pixel data processing apparatus according to the second aspect, wherein the feature extraction unit detects the symmetry of the same kind of color data as the color data of the pixel to be interpolated. .

A pixel data processing method according to claim 7 is a pixel data processing method for processing a plurality of color data output from a plurality of types of color filters for each of a plurality of pixels to reproduce an image. An interpolation step of performing a plurality of different interpolation processes for interpolating color data on the interpolation target pixel in parallel; a feature extraction step of extracting a plurality of color data features in synchronization with the interpolation step; A selecting step of selectively outputting any one of the results of the plurality of interpolation processes according to the feature that has been performed,
The plurality of interpolation processes include first interpolation for interpolating color data based on a change amount of color data of a pixel in a peripheral area of the pixel to be interpolated.
An interpolation process and a second interpolation process of interpolating color data based on a ratio of color data of the same type as the color data to be interpolated and color data of the same type of pixel as the interpolation target pixel in a peripheral area of the interpolation target pixel. Including.

According to an eighth aspect of the present invention, there is provided the pixel data processing method according to the seventh aspect, wherein the feature extracting step comprises the steps of: To determine whether or not there is symmetry with respect to
The result of the interpolation process is selectively output.

A pixel data processing method according to a ninth aspect is the pixel data processing method according to the eighth aspect, wherein the plurality of pixels are arranged in a matrix in the vertical and horizontal directions, and the feature extracting step includes An evaluation step of detecting the symmetry in the direction and the symmetry in the horizontal direction, and integrating a result of the evaluation step to output a selection signal for selecting one of a first interpolation process and a second interpolation process And selecting one of the first interpolation processing and the second interpolation processing according to the selection signal.

According to a tenth aspect of the present invention, there is provided a pixel data processing method comprising:
9. The pixel data processing method according to claim 8, wherein the plurality of pixels are arranged in a matrix in a vertical direction and a horizontal direction, and the feature extraction unit detects symmetry in the vertical, horizontal, and oblique directions. And a method selection step of integrating a result of the evaluation step and outputting a selection signal for selecting either the first interpolation processing or the second interpolation processing, wherein the selection step is performed according to the selection signal. Thus, either one of the first interpolation processing and the second interpolation processing is selectively output.

A pixel data processing method according to claim 11 is
9. The pixel data processing method according to claim 8, wherein the plurality of color filters are arranged according to a Bayer array, and each of the plurality of color data is any one of red data, blue data, and green data, and the feature extraction is performed. The step detects the symmetry of the green data.

According to a twelfth aspect of the present invention, there is provided a pixel data processing method comprising:
9. The pixel data processing method according to claim 8, wherein the feature extracting step detects a symmetry of the same type of color data as the color data of the interpolation target pixel.

[0018]

[First Embodiment] A pixel data processing apparatus according to a first embodiment of the present invention will be described.
The pixel data processing device according to the first embodiment of the present invention
Based on the characteristics of the input color data, a process for interpolating other color data to each pixel to which specific color data is given is performed.

A digital still camera to which the pixel data processing device according to the first embodiment of the present invention is applied will be described with reference to FIG. In the following, the same components are denoted by the same reference numerals and symbols, and description thereof will be omitted.

FIG. 1 shows a digital still camera 1 to which the pixel data processing device according to the first embodiment of the present invention is applied.
000 is a block diagram showing the outline of the configuration of FIG. Referring to FIG. 1, a digital still camera 1000 includes an optical system 10.
2. CCD 104, peripheral circuit 106, data processing unit 11
2, a memory 114, a camera control unit 116, and an input / output interface circuit 118.

The optical system 102 includes an image pickup lens inside. The camera control unit 116 controls a photographing operation and a data processing operation of the optical system. The optical signal obtained by the optical system 102 is converted into an electric signal by the CCD 104. CCD10
Reference numeral 4 denotes a Bayer type primary color filter. The peripheral circuit 106 converts the output of the CCD 104 into digital data and outputs the digital data to the data processing unit 112.

The data processing unit 112 includes an image processing unit 108
And a compression / expansion unit 110. The image processing unit 108 includes an image interpolation unit that generates missing color data (generates interpolation data). The compression / expansion unit 110 performs expansion processing for compressing image data or expanding the compressed image data. The image memory 114 stores image data. The input / output interface circuit 118 performs control for receiving photographing conditions and the like from the outside and extracting image data subjected to predetermined processing in the form of a memory card (not shown).

Next, the image interpolation unit 100 included in the image processing unit 108 will be described. FIG. 2 is a diagram for describing an outline of a configuration of the image interpolation unit 100 according to Embodiment 1 of the present invention. Referring to FIG. 2, image interpolation unit 100
Are the feature extracting unit 1, the method selecting unit 2, the first method unit 3, the second
It includes a method unit 4 and a selector 5.

The feature extracting unit 1 extracts features of the input image data. The method selecting unit 2 outputs a selection signal for selecting one of the first method unit 3 and the second method unit 4 based on the extraction result of the feature extracting unit 1. Each of the first method unit 3 and the second method unit 4 performs interpolation processing different from each other to generate interpolation data.
The selector 5 selectively outputs the interpolation data of one of the first method unit 3 and the second method unit 4 according to the selection signal output from the method selection unit 2.

FIG. 3 is a diagram for explaining the characteristics of the Bayer type primary color filter included in the CCD 104. R is a color filter corresponding to red, G is green, and B is a color filter corresponding to blue, each of which corresponds to one pixel. Filters of the same color are arranged every other pixel in both the vertical and horizontal directions (Bayer arrangement). The CCD 104 obtains specific color information of three primary colors (R, G, B) for each pixel. Of (R, G, B), a pixel having R data is called an R pixel, a pixel having G data is called a G pixel, and a pixel having B data is called a B pixel.

Referring to FIG. 2, image interpolating section 100 receives image data (input data) having specific color information for each pixel. FIG. 4 is a diagram illustrating an example of the input pattern. The image interpolation unit 100 receives the input pattern P1 shown in FIG. 4A or the input pattern P2 shown in FIG. In the input patterns P1 and P2, the symbol R is R
A pixel and a symbol G correspond to a G pixel, and a symbol B corresponds to a B pixel. Note that the case where the symbol R and the symbol B are interchanged with each other is also included.

In the first embodiment, the first method section 3 and the second method section 4 perform processing for providing G data (providing interpolation data G) to the R and B pixels. The feature extraction unit 1 detects the presence or absence of symmetry of the G data around the pixel to be interpolated. In order to reduce the processing time, the two types of interpolation circuits (the first method unit 3 and the second method unit 4) and the feature extraction unit 1 are operated in parallel to perform the feature extraction and the method selection. It is configured so that interpolation processing is executed.

It should be noted that interpolation of B data for R pixels, B
Interpolation of R data for pixels and R for G pixels
Interpolation of data and B data will be described in an embodiment described later.

Next, the feature extracting unit 1 according to the first embodiment of the present invention will be described with reference to FIG. FIG. 5 is a diagram for describing an outline of a configuration of the feature extracting unit 1 according to the first embodiment of the present invention. Referring to FIG. 5, feature extracting section 1 includes a pixel extracting section 6 for extracting pixels, a counter 7 for performing switching corresponding to input patterns P1 and P2, and an evaluating section for evaluating the extracted pixels. Including.
The evaluation section includes a first evaluation section 8a, a second evaluation section 8b, and a third evaluation section 8c for evaluating horizontal symmetry, a fourth evaluation section 8d for evaluating vertical symmetry, and a fourth evaluation section 8d. It comprises a fifth evaluation unit 8e and a sixth evaluation unit 8f. In addition,
In the following example, the range from which the feature is extracted is a range of 5 × 5 pixels.

The pixel extracting section 6 extracts G pixels from the input pattern P1 or P2. Note that the input pattern P
Since 1 and P2 are alternately input, the pixel extraction unit 6
The pixels to be extracted are changed based on the output of the 1-bit counter (counter 7).

A case where an input pattern P1 is received as input data will be described. For the input pattern P1, the pixel extraction unit 6 outputs G pixels, that is, pixels G11, G
13, G15, G31, G33, G35, G51, G5
Extract 3 and G55. The first evaluation unit 8a to the third evaluation unit 8c check the presence or absence of symmetry of the data arranged in the horizontal direction. Specifically, the first evaluation unit 8a determines (G11, G1
(3, G15), the following expression (2) for the second evaluation unit 8b for (G31, G33, G35), and the following expression (2) for the third evaluation unit 8c (G51, G53, G55). Check whether 3) is satisfied.

G11-G15 <th1, (G11 + G15) -G13> th2 (1) G31-G35 <th1, (G31 + G35) -G33> th2 (2) G51-G55 <th1, (G51 + G55) -G53> th2 (3) where th1 and th2 are threshold values, and th1 <th
It is determined experimentally by an arbitrary number in which 2 holds. Considering the size of the circuit, only the second evaluation unit 8b may be arranged.

Similarly, the fourth evaluation unit 8d to the sixth evaluation unit 8f
Checks the symmetry of the data arranged in the vertical direction. Specifically, the fourth evaluation unit 8d outputs (G11, G31,
G51) is given by the following equation (4).
13, G33, G53), and the sixth evaluation unit 8f checks whether the following expression (6) holds for (G15, G35, G55).

G11-G51 <th1, (G11 + G51) -G31> th2 (4) G13-G53 <th1, (G13 + G53) -G33> th2 (5) G15-G55 <th1, (G15 + G55) -G35> th2 (6) In consideration of the size of the circuit, only the fifth evaluation unit 8e may be arranged. The feature extraction unit 1 outputs the results (true / false) of the first evaluation unit 8a to the sixth evaluation unit 8f as outputs A1 to A
6

Next, the input pattern P2 is used as input data.
A description will be given of a case in which the information is received. For the input pattern P2, the pixel extraction unit 6 determines that the pixels G12 and G
14, G21, G23, G25, G32, G34, G4
1, G43, G45, G52 and G54 are extracted.

The first evaluator 8a and the second evaluator 8b check whether or not the data arranged in the horizontal direction is symmetric.
Specifically, the first evaluation unit 8a determines (G21, G23, G2
The following expression (1b) is applied to the fifth evaluation unit 5b.
1, G43, G45) to determine whether the following equation (2b) holds.

G21-G25 <th1, (G21 + G25) -G23> th2 (1b) G41-G45 <th1, (G41 + G45) -G43> th2 (2b) The fourth evaluation unit 8d to the fifth evaluation unit 8e The data arranged in the vertical direction is checked for symmetry. In particular,
The fourth evaluation unit 8d calculates the following equation (4b) for (G12, G32, G52), and the fifth evaluation unit 8e calculates (G14, G34,
G54) is checked to determine whether or not the following expression (5b) holds.

G12-G52 <th1, (G12 + G52) -G32> th2 ... (4b) G14-G54 <th1, (G14 + G54) -G34> th2 ... (5b) The feature extraction unit 1 includes the first evaluation units 8a to 8th. The results (true / false) of the second evaluation unit 8b and the fourth evaluation unit 8d to the fifth evaluation unit 8e are output as A1 to A2 and A4 to A5. At this time, the third evaluation unit 8c
The sixth evaluation unit 8f is not particularly considered.

FIG. 6 is a diagram for explaining an outline of the configuration of the method selecting section 2 according to the first embodiment of the present invention. Referring to FIG. 6, scheme selecting section 2 includes a horizontal symmetry determining circuit 9, a vertical symmetry determining circuit 10, an OR circuit 11, and a counter 12. The horizontal symmetry determining circuit 9
First evaluation unit 8a, second evaluation unit 8b, and third evaluation unit 8c
Of the G data in the horizontal direction is determined based on the respective outputs A1 to A3. The vertical symmetry determination circuit 10 determines the vertical symmetry of the G data based on the outputs A4 to A6 of the fourth evaluation unit 8d, the fifth evaluation unit 8e, and the sixth evaluation unit 8f. The horizontal symmetry determining circuit 9 and the vertical symmetry determining circuit 10
Switches the processing target (input pattern P1, P2) based on the output 0/1 of the 1-bit counter (counter 12).

The horizontal symmetry determining circuit 9 outputs true for the input pattern P1 if all of the outputs A1 to A3 are true, and outputs false otherwise. For the input pattern P2, It outputs true if the outputs A1 and A2 are true, and outputs false otherwise.

The vertical symmetry determining circuit 10 outputs true for the input pattern P1 if all of the outputs A4 to A6 are true, and outputs false otherwise. Outputs true if the outputs A4 and A5 are true, otherwise outputs false.

The OR circuit 11 performs the second method section 4 when one of the output of the horizontal symmetry determining circuit 9 and the output of the vertical symmetry determining circuit 10 is true, and the OR circuit 11 when the output is false. A selection signal for selecting the first method unit 3 is output. That is, when the G data has symmetry in either the horizontal direction or the vertical direction, the second method unit 4 is selected, and otherwise, the interpolation processing of the first method unit 3 is specified. I do.

Next, the first method unit 3 and the fourth method unit 4 according to the first embodiment of the present invention will be described. FIG.
FIG. 9 is a diagram for describing an interpolation process. In FIG. 7, the arrangement of pixels is represented by the X axis in the horizontal direction and the Y axis in the vertical direction, and the X coordinate (i−2) to (i + 2) and the Y coordinate (j
-2) to (j + 2) are displayed in 25 pixels. Pixels or data at coordinates (K, L) are represented by colors G, R,
The symbols G (K, L), R (K, L), B
Expressed as (K, L).

Referring to FIG. 7, a method for obtaining interpolation data G for pixel R (i, j) will be described. In addition,
By replacing the R pixel and the B pixel, the interpolation data G of the B pixel is calculated in the same procedure.

The amount of change in R data in the vertical direction centering on the pixel R (i, j) to be interpolated is ΔV,
The change amount of the data is assumed to be ΔH.

ΔV = {R (i, j−2) + R (i, j + 2)} − R (i, j) (7) ΔH = {R (i−2, j) + R (i + 2, j)} −R (i, j) (8) In the first method unit 3, assuming that the rate of change (difference) of color change is equal, ΔV or ΔH is regarded as a prediction error, and the following equations (9) to (10) ) Are calculated.

Gv (i, j) = １ ／ · {G (i, j + 1) + G (i, j−1)} − １ ／ · ΔV (9) gh (i, j) = １ ／ · {G (i−1, j) + G (i + 1, j)} − １ ／ · ΔH (10) In the first method unit 3, when the correlation in the vertical direction is strong (△ H>
ΔV), and outputs the interpolation data gv shown in Expression (9), and otherwise outputs the interpolation data gh shown in Expression (10).

The second method section 4 calculates interpolation data on the assumption that the rate of color change (the rate of low frequency components) is equal. First, a low frequency component represented by the following equation is calculated.

G _LPF 3v = １ ／ · (G2v + G4v) (11a) R _LPF 3v = １ ／ · R1v + ／ · R3 + ／ · R5v (12a) G _LPF 3h = １ ／ · (G2h + G4h) .. (11b) R _LPF 3h = １ ／ · R1h + ／ · R3 + ／ · R5h (12b) Here, R3 is R data of an R pixel to be interpolated.
G2v and G4v represent G data of a G pixel vertically adjacent to the pixel R3, and R1v and R5v represent pixels G2v and
The R data of the R pixel vertically adjacent to G4v is shown. Specifically, in FIG. 7, R3 is R (i,
j), G2v is G (i, j + 1), G4v is G
(I, j-1), R1v is R (i, j + 2), R5
v corresponds to R (i, j-2), respectively.

Further, G2h and G4h represent the G data of the G pixel adjacent to the pixel R3 in the horizontal direction by R1h, R4
5h indicates the R data of the R pixel horizontally adjacent to the pixels G2h and G4h. Specifically, in FIG. 7, G2h is G (i−1, j) and G4h is G (i +
1,1), R1h is R (i-2, j), R5h is R
(I + 2, j) respectively.

The correlation between the G and R in the coordinate (i, j) is assumed to be equal to the correlation between the low frequency component G _LPF and a low-frequency component R _{L PF.} The second method unit 4 calculates the expressions (11a) to (12)
a) is used to convert the interpolation data of Expression (13) into Expression (11b)
(12b) is used to calculate the interpolation data of the equation (14).

[0052] gv (i, j) = R (i, j) × G LPF 3v / R LPF 3v ... (13) gh (i, j) = R (i, j) × G LPF 3h / R LPF 3h ... (14) When the correlation in the vertical direction is strong (ΔH> Δ
V) outputs the interpolation data gv of Expression (13), and otherwise outputs the interpolation data gh of Expression (14).

Referring to FIG. 2, when G data has symmetry according to the selection signal, selector 5 replaces the interpolation data of second method unit 4 taking the low-frequency component into account with the other data. In this case, the interpolation data of the first method section 3 is output.

Here, an example of the configuration of the method section (first method section 3 and second method section 4) for performing the above-described interpolation processing will be described with reference to FIGS.

FIG. 8 is a diagram for explaining a configuration example of the first method section 3 according to the first embodiment of the present invention.
Referring to FIG. 8, first method unit 3 includes a G data calculation circuit 5
0, a comparison / determination circuit 51 and a selector 52 are included. The G data calculation circuit 50 receives the input data,
The interpolation data gv and gh corresponding to (10) are calculated.
The comparison determination circuit 51 compares the magnitude relationship between the displacement ΔH in the horizontal direction and the displacement ΔV in the vertical direction, and calculates the interpolation data gv, g
h is to be output. Selector 52
Outputs one of the interpolation data gv and gh as the interpolation data G based on the determination result of the comparison determination circuit 51. Note that the G data calculation circuit 50 may be configured to calculate one of gv and gh based on the result of the comparison determination circuit 51.

FIG. 9 is a diagram for describing a configuration example of second method section 4 according to Embodiment 1 of the present invention.
Referring to FIG. 9, the second scheme unit 4 includes a G data calculation circuit 5
3. It includes a comparison determination circuit 54 and a selector 52. The G data calculation circuit 53 receives the input data, and
The interpolation data gv and gh corresponding to (14) are calculated.
The comparison determination circuit 54 compares the magnitude relation between the displacement ΔH in the horizontal direction and the displacement ΔV in the vertical direction, and calculates the interpolation data gv, g
h is to be output. Selector 52
Outputs one of the interpolation data gv and gh as the interpolation data G based on the determination result of the comparison determination circuit 54. It should be noted that the G data calculation circuit 53 may be configured to calculate one of gv and gh based on the result of the comparison determination circuit 54.

As described above, by selecting the optimal interpolation processing according to the characteristics of the image, a high-quality image can be obtained. In addition, the first method unit 3 and the second method unit 4
Are performed in parallel, only the time required for the selection operation by the selector 5 is increased as compared with the conventional system. Therefore, it is possible to guarantee almost the same processing speed as the conventional system.

[Second Embodiment] In a second embodiment of the present invention, the symmetry in the horizontal direction, the vertical direction, and the oblique direction is detected for pixels located around the pixel to be interpolated.
FIG. 10 shows an image interpolation unit 2 according to the second embodiment of the present invention.
FIG. 9 is a diagram for explaining the configuration of the 00. The image interpolating unit 200 shown in FIG. 10 differs from the image interpolating unit 100 shown in FIG. 2 in that a feature extracting unit 21 that extracts features in the horizontal, vertical, and oblique directions is used instead of the feature extracting unit 1. The point is that a method selecting unit 22 is provided instead of the unit 2.

FIG. 11 is a diagram for describing a configuration of feature extracting section 21 according to the second embodiment of the present invention. Referring to FIG. 11, feature extracting section 21 includes a pixel extracting section 6, a counter 7, and an evaluating section for evaluating the extracted pixels. The evaluation section includes a first evaluation section 8a, a second evaluation section 8b, and a third evaluation section 8c for evaluating horizontal symmetry, a fourth evaluation section 8d for evaluating vertical symmetry, and a fourth evaluation section 8d. A fifth evaluation unit 8e and a sixth evaluation unit 8f, and a seventh evaluation unit 8g, an eighth evaluation unit 8h, a ninth evaluation unit 8i, a tenth evaluation unit 8j, and an eleventh evaluation unit for evaluating symmetry in an oblique direction. 8k and a twelfth evaluation unit 8l.

The pixel extraction unit 6, the counter 7, and the first to sixth evaluation units 8a to 8f are as described in the first embodiment. The seventh to twelfth evaluation units 8g to 8l for evaluating the symmetry in the oblique direction will be described.

A case where an input pattern P1 is received as input data will be described. The seventh evaluation unit 8g calculates (G1
1, G33, G55), and the eighth evaluation unit 8h checks whether the following expression (16) holds for (G15, G33, G51).

G11−G55 <th1, (G11 + G55) −G33> th2 (15) G15−G51 <th1, (G15 + G51) −G33> th2 (16) where th1 and th2 are thresholds, and th1 <Th
It is determined experimentally by an arbitrary number in which 2 holds. The feature extraction unit 21 outputs the results (true / false) of the first evaluation unit 8a to the eighth evaluation unit 8h as A1 to A8. At this time, the ninth evaluation unit 8i to the twelfth evaluation unit 8l are not considered.

Next, a case where the input data corresponds to the input pattern P2 will be described. The ninth evaluation unit 8i calculates (G2
1, G32, G43, G54),
The tenth evaluation unit 8j calculates (G12, G23, G34, G4
Expression (18) is obtained for 5), and the eleventh evaluation unit 8k calculates (G1
4, G23, G32, G41),
The twelfth evaluation unit 8l includes (G25, G34, G43, G5
It is checked whether Expression (20) holds for 2).

G21-G54 <th1, G32-G43 <th1, (G21 + G54)-(G32-G43)> th3 (17) G12-G45 <th1, G23-G34 <th1, (G12 + G45)-(G23-G34) )> Th3 (18) G14-G41 <th1, G23-G32 <th1, (G14 + G41)-(G23-G32)> th3 (19) G25-G52 <th1, G34-G43 <th1, (G25 + G52)- (G34-G43)> th3 (20) where th1 and th3 are threshold values, and th1 <th
It is determined experimentally by an arbitrary number in which 3 holds. The feature extraction unit 21 outputs the results (true / false) of the first to sixth evaluation units 8a to 8f and the ninth and twelfth evaluation units 8l to 8l as A1 to A6 and A9 to A12. At this time, the seventh evaluation unit 8g to the eighth evaluation unit 8h are not particularly considered.

FIG. 12 is a diagram for describing an outline of the configuration of the method selecting section 22 according to the second embodiment of the present invention. Referring to FIG. 12, method selection unit 22 includes horizontal symmetry determining circuit 9, vertical symmetry determining circuit 10, counter 12, oblique symmetry determining circuit 26, and OR circuit 2.
7 inclusive. The horizontal symmetry determining circuit 9, the vertical symmetry determining circuit 10, and the counter 12 are described in the first embodiment.
As described in the above.

The oblique direction symmetry determining circuit 26 evaluates the symmetry in the oblique direction from the seventh evaluator 8g to the twelfth evaluator 8l.
The outputs A7 to A12 of the above are received and the symmetry in the oblique direction is determined. At this time, a 1-bit counter (counter 1
Based on the output 0/1 of 2), the processing target (input pattern P
1, P2).

For example, if the input data is input pattern P1
If the output A7 is true or the output A8 is true, true is output; otherwise, false is output. When the input data corresponds to the input pattern P2, true is output when both outputs A9 and A10 are true, or true when both outputs A11 and A12 are true, and false otherwise.

If any one of the output of the horizontal symmetry determining circuit 9, the output of the vertical symmetry determining circuit 10, and the output of the oblique symmetry determining circuit 26 is true, the OR circuit 27 outputs A selection signal for selecting the second method unit 4 and the first method unit 3 in other cases is output. That is, if there is symmetry in any of the horizontal direction, the vertical direction, or the oblique direction, the second method unit 4 is selected. If no symmetry is detected in any of them, the interpolation processing of the first method unit 3 is performed. Specifies to apply.

Thus, the selector 5 shown in FIG.
When there is symmetry in any of the horizontal direction, the vertical direction, or the oblique direction, the interpolation data output from the second method unit 4 is
In other cases, the interpolation data output from the first method section 3 is selectively output.

With this configuration, it is possible to more accurately extract the features (symmetry) of the image, and thus it is possible to obtain a higher quality image.

[Embodiment 3] In Embodiment 3 of the present invention, horizontal and vertical symmetry is detected using pixels having the same type of color data as the pixel to be interpolated.

FIG. 13 is a diagram for explaining the outline of the configuration of the image interpolation unit 300 according to the third embodiment of the present invention. Referring to FIG. 13, image interpolating section 300 includes feature extracting section 31, method selecting section 32, first method section 33, second method section 34, and selector 5.

The feature extraction unit 31 extracts pixels having the same type of color data (R, G, B) as the pixel to be interpolated, and checks whether there is symmetry in the horizontal and vertical directions. Specifically, an input pattern P3 shown in FIG. 14 is received as input, and the features of X pixels are extracted. Note that X shown in FIG. 14 is a pixel to be interpolated, and corresponds to any one of the R pixel, the G pixel, and the B pixel.

FIG. 15 is a diagram for describing an outline of the configuration of feature extracting section 31 according to the third embodiment of the present invention. Referring to FIG. 15, feature extraction unit 31 includes pixel extraction unit 35, first evaluation unit 8a, second evaluation unit 8b, third evaluation unit 8c, fourth evaluation unit 8d, fifth evaluation unit 8e, and sixth evaluation unit 8e. An evaluation unit 8f is included. The pixel extraction unit 35 includes pixels X11, X13, X15, X31, X33, X35, and X shown in FIG.
Extract 51, X53 and X55.

The first evaluation unit 8a calculates (X11, X13, X1
The following expression (21) is applied to 5), and the second evaluation unit 8b calculates (X3
1, X33, X35), and the third evaluation unit 8c checks whether the following expression (23) holds for (X51, X53, X55).

X11-X15 <th1, (X11 + X15) -X13> th2 (21) X31-X35 <th1, (X31 + X35) -X33> th2 (22) X51-X55 <th1, (X51 + X55) -X53> th2 (23) where th1 and th2 are threshold values, and th1 <th
It is determined experimentally by an arbitrary number in which 2 holds. In consideration of the size of the circuit, it may be configured only with the second evaluation unit 8b.

The fourth evaluation section 8d calculates (X11, X31, X5
The following expression (24) is applied to the first evaluation unit 1), and the fifth evaluation unit 8e calculates (X1
3, X33, X53), and the sixth evaluation unit 8f checks whether the following expression (26) holds for (X15, X35, X55).

X11-X51 <th1, (X11 + X51) -X31> th2 (24) X13-X53 <th1, (X13 + X53) -X33> th2 (25) X15-X55 <th1, (X15 + X55) -X35> th2 (26) In addition, only the fifth evaluation unit 8e may be arranged in consideration of the size of the circuit. The feature extraction unit 31 outputs the result (true / false) of the first evaluation unit 8a to the sixth evaluation unit 8f as an output A1.
A6.

FIG. 16 is a diagram for describing an outline of the configuration of system selecting section 32 according to the third embodiment of the present invention. Referring to FIG. 16, scheme selecting section 32 includes a horizontal symmetry determining circuit 9, a vertical symmetry determining circuit 10, and an OR circuit 11. The horizontal symmetry determining circuit 9
First evaluation unit 8a, second evaluation unit 8b, and third evaluation unit 8c
X arranged in the horizontal direction based on the respective outputs A1 to A3
The symmetry of the pixel data is determined. The vertical symmetry determination circuit 10 determines the symmetry of the data of the X pixels arranged in the vertical direction based on the outputs A4 to A6 of the fourth evaluation unit 8d, the fifth evaluation unit 8e, and the sixth evaluation unit 8f.

The OR circuit 11 outputs the result of the second method unit 34 when either the output of the horizontal symmetry judging circuit 9 or the output of the vertical symmetry judging circuit 10 is true. A selection signal for selecting the first method unit 33 is output. In other words, when there is symmetry in either the horizontal direction or the vertical direction, it instructs to apply the interpolation processing of the second method unit 34, and otherwise applies the interpolation processing of the first method unit 33.

Next, an explanation will be given of the interpolation processing of the first method unit 33 and the second method unit 34 according to the third embodiment of the present invention. The process for providing the interpolation data G to each of the R pixel and the B pixel is as described in the first embodiment. In the following, a description will be given of a process of providing interpolation data R and B to G pixels, obtaining interpolation data R to B pixels, and obtaining interpolation data B to R pixels.

The R and B pixels differ from the G pixels in that the position of the referenceable pixel changes depending on the position of the pixel to be interpolated. Therefore, the interpolation formula must be changed according to the position of the pixel to be interpolated. For this reason, when obtaining the interpolation data R and B, the reference pixel range is limited to 3 × 3, and the interpolated data (interpolation data G) at the same position as the color data of the color to be interpolated within that range. Is used to determine the relationship between the color data. Based on these relationships, color data at the interpolation position is calculated from the G data at the interpolation position. In the following equations, G represents input data or interpolation data.

The processing of the first method section 33 will be described.
FIG. 17 is a conceptual diagram for explaining the interpolation process. FIG. 17A shows the coordinates (i,
j) is a G pixel, and FIG. 17B is a diagram in which the coordinates (i, j) are B pixels.
Cases corresponding to pixels are shown.

Referring to FIG. 17A, G pixels include pixels of colors to be interpolated in the vertical and horizontal directions.
Assuming that the rate of change (difference) of color change is equal, Equation (2)
7) to (28) hold.

R (i, j) −G (i, j) ≒ R (i−1, j) −G (i−1, j) ≒ R (i + 1, j) −G (i + 1, j) ( 27) B (i, j) -G (i, j) ≒ B (i, j-1) -G (i, j-1) ≒ B (i, j + 1) -G (i, j + 1) (28) Therefore, the interpolation data R and B in G (i, j) are calculated using equations (29) to (30). R (i, j) = G (i, j) + ／ · {R (i−1, j) −G (i−1, j) + R (i + 1, j) −G (i + 1, j)} (29) B (i, j) = G (i, j) + ／ · B (i, j−1) −G (i, j−1) + B (i, j + 1) −B (i, j + 1) (30) Referring to FIG. 17B, the B pixel (or R pixel)
There are pixels of colors to be interpolated in the oblique direction, up and down, and left and right directions. Therefore, equation (31) holds.

R (i, j) −G (i, j) ≒ R (i−1, j−1) −G (i−1, j−1) ≒ R (i + 1, j−1) −G ( i + 1, j-1) ≒ R (i-1, j + 1) -G (i-1, j + 1) ≒ R (i + 1, j + 1) -G (i + 1, j + 1) (31) Therefore, B (i, j) Of the interpolation data R in the equation (3)
(If the coordinates (i, j) are R pixels, R in equations (31) to (32) is replaced with the corresponding B).

R (i, j) = G (i, j) + ／ · ｛R (i−1, j−1) −G (i−1, j−1) + R (i + 1, j−1) −G (i + 1, j−1) + R (i−1, j + 1) −G (i−1, j + 1) + R (i + 1, j + 1) −G (i + 1, j + 1)} (32) The processing of the method unit 34 will be described. Assuming that the ratios of the color changes are equal (the ratios of the low-frequency components are equal), the equations (33) to (3) in FIG.
4) holds, and with respect to FIG. 17B, the equation (35) holds.

R (i, j) / G (i, j) ≒ R (i−1, j) / G (i−1, j) ≒ R (i + 1, j) / G (i + 1, j) ( 33) B (i, j) / G (i, j) ≒ B (i, j−1) / G (i, j−1) ≒ B (i, j + 1) / G (i, j + 1) (34) ) R (i, j) / G (i, j) ≒ R (i−1, j−1) / G (i−1, j−1) ≒ R (i + 1, j−1) / G (i + 1, j-1) ≒ R (i-1, j + 1) / G (i-1, j + 1) ≒ R (i + 1, j + 1) / G (i + 1, j + 1) (35) Therefore, G ( The interpolation data R and B of (i, j) are calculated using equations (36) to (37). The interpolation data R of B (i, j) in FIG.
Is calculated using equation (38). Note that the coordinates (i, j)
Is an R pixel, R in equations (35) and (38)
To the corresponding B.

R (i, j) = G (i, j) × {R (i−1, j) + R (i + 1, j)} / G (i + 1, j) + G (i−1, j) ( 36) B (i, j) = G (i, j) × {B (i, j-1) + B (i, j + 1)} / {G (i, j + 1) + G (i, j-1)} (37) R (i, j) = G (i, j) × ｛R (i−1, j−1) + R (i + 1, j−1) + R (i−1, j + 1) + R (i + 1, j + 1) ｝ / G (i−1, j + 1) + G (i + 1, j−1) + G (i−1, j + 1) + G (i + 1, j + 1) (38) Here, the system unit according to Embodiment 3 of the present invention. FIG. 18 shows an example of the configuration of the (first method unit 33 and the second method unit 34).
This will be described with reference to FIG.

FIG. 18 is a diagram for describing a configuration example of first method section 33 according to Embodiment 3 of the present invention. Referring to FIG. 18, first method unit 33 includes a pixel determination circuit 60, a G data calculation circuit 55, and R / B data calculation circuits 56 and 57. The pixel determination circuit 60 determines whether the input pixel is a G pixel, an R pixel, or a B pixel. If it is a G pixel, the R / B data calculation circuit 56
Is an R pixel or a B pixel, the R / B data calculation circuit 57 and the G data calculation circuit 55 are executed.

The R / B data calculation circuit 56 uses the equation (29)
As described in (30), the R data at the G pixel is calculated based on the G data and the R data, and the B data at the G pixel is calculated based on the G data and the B data.

The G data calculation circuit 55 includes the G data calculation circuit 50, the comparison / determination circuit 51, and the selector 52 described in the first embodiment, and calculates G data of an R pixel or a B pixel.

The R / B data calculation circuit 57 uses the equation (32)
As described in the above, B based on R data and G data
The R data of the pixel is calculated based on the B data and the G data.
The B data of the pixel is calculated.

FIG. 19 is a diagram for illustrating a configuration of second method section 34 according to the third embodiment of the present invention. Second
The method unit 34 includes a pixel determination circuit 60, a G data calculation circuit 6
5, including R / B data calculation circuits 66 and 67. If it is a G pixel based on the determination of the pixel determination circuit 60, R / B
If the data calculation circuit 66 is an R pixel or a B pixel,
R / B data calculation circuit 67 and G data calculation circuit 65
Execute

The R / B data calculation circuit 66 uses the equation (36)
As described in (37), the R data at the G pixel is calculated based on the G data and the R data, and the B data at the G pixel is calculated based on the G data and the B data.

The G data calculation circuit 65 includes the G data calculation circuit 53, the comparison determination circuit 54, and the selector 52 described in the first embodiment, and calculates G data of an R pixel or a B pixel.

The R / B data calculation circuit 67 uses the equation (38)
As described in the above, B based on R data and G data
The R data of the pixel is calculated based on the B data and the G data.
The B data of the pixel is calculated.

The selector 5 shown in FIG.
For each of the pixels, when the selection signal output from the method selecting unit 32 is true (there is data symmetry), the interpolation data of the second method unit 34 is used. Output data.

In the third embodiment of the present invention, since there is only one type of input pattern, no counter is required.
Therefore, a high-quality image can be obtained with a simple circuit configuration as in the first embodiment.

[Fourth Embodiment] In a fourth embodiment of the present invention, the symmetry in the vertical, horizontal, and oblique directions is detected using pixels having the same type of color data as the pixel to be interpolated. .

FIG. 20 is a diagram for describing a configuration of image interpolating section 400 according to the fourth embodiment. An image interpolating unit 400 shown in FIG. 20 includes a feature extracting unit 41 that extracts horizontal, vertical, and oblique features instead of the feature extracting unit 31 shown in FIG. Prepare.

FIG. 21 is a diagram for describing an outline of the configuration of feature extracting section 41 according to Embodiment 4 of the present invention. Referring to FIG. 21, feature extraction unit 41 includes pixel extraction unit 35, first evaluation unit 8a, second evaluation unit 8b, third evaluation unit 8c, fourth evaluation unit 8d, fifth evaluation unit 8e, and sixth evaluation unit 8e. An evaluation unit 8f, a seventh evaluation unit 8g, and an eighth evaluation unit 8h are included.

As described in the third embodiment, the first evaluation unit 8a to the sixth evaluation unit 8f (X11, X13, X1
5), (X31, X33, X35), (X51, X5
3, X55), (X11, X31, X51), (X1
3, X33, X53) and (X15, X35, X5
The presence or absence of symmetry in 5) is evaluated.

The seventh evaluation unit 8g and the eighth evaluation unit 8h
Evaluate the symmetry in the oblique direction. Specifically, the seventh
The evaluation unit 8g checks whether or not the following equation (39) holds for (X11, X33, X55). Further, the eighth evaluation unit 8h checks whether or not the following expression (40) holds for (X15, X33, X51).

X11−X55 <th1, (X11 + X55) −X33> th2 (39) X15−X51 <th1, (X15 + X51) −X33> th2 (40) where th1 and th2 are thresholds and th1 > Th
It is determined experimentally by an arbitrary number in which 2 holds. The feature extraction unit 41 outputs the results (true / false) of the first evaluation unit 8a to the eighth evaluation unit 8h as A1 to A8.

FIG. 22 is a diagram for describing an outline of the configuration of system selecting section 42 according to the fourth embodiment of the present invention. The method selection unit 42 includes a horizontal symmetry determination circuit 9, a vertical symmetry determination circuit 10, and a diagonal symmetry determination circuit 4.
6 and an OR circuit 27. The oblique direction symmetry determination circuit 46 outputs the output A7 of the seventh evaluation unit 8g or the eighth evaluation unit 8
If the output A8 of h is true, it outputs true; otherwise, it outputs false. The OR circuit 27 outputs true if any one of the output of the horizontal symmetry determining circuit 9, the output of the vertical symmetry determining circuit 10, and the output of the oblique symmetry determining circuit 46 is true, and If false, output false. O
When the selection signal output from the R circuit 27 is true, the selector 5 shown in FIG. 20 selects the interpolation data of the second method unit 34, and otherwise selects the interpolation data of the first method unit 33. Output.

As described above, by adopting the configuration of the fourth embodiment of the present invention, it is possible to perform an optimal interpolation process for all the pixels, so that a high quality image can be obtained. . Further, since a counter is not required, the circuit configuration can be simplified.

The first to fourth embodiments can be realized by either hardware or software.

The embodiments disclosed this time are to be considered in all respects as illustrative and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description of the embodiments, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

[0110]

As described above, claims 1 and 2 are as described above.
According to the pixel data processing device according to the above, for a plurality of pixels each having specific color data, by configuring to select the interpolation data according to the characteristics of the color data,
It is possible to reproduce a highly accurate image. In addition, it is possible to satisfy high speed by performing the interpolation processing in parallel.

A pixel data processing apparatus according to a third aspect is the pixel data processing apparatus according to the second aspect, wherein the pixel data processing apparatus extracts, in particular, the symmetry of the color data in the horizontal and vertical directions,
Thereby, the correction data is selected. As a result, when symmetry is recognized, correction data based on the rate of color change can be selectively output, and in other cases, correction data based on the rate of color change can be selectively output, so that appropriate interpolation can be performed. Become.

A pixel data processing apparatus according to a fourth aspect is the pixel data processing apparatus according to the second aspect, and in particular, extracts the feature of the horizontal, vertical, and oblique symmetry of the color data, thereby extracting Select the correction data. As a result, correction data based on the rate of color change can be selectively output when symmetry is recognized, and correction data based on the rate of color change can be selectively output otherwise. Becomes

A pixel data processing apparatus according to a fifth aspect is the pixel data processing apparatus according to the second aspect, and extracts the characteristic of the symmetry of the G data. Since the G data having the largest number of data is used, highly accurate feature extraction becomes possible.

A pixel data processing apparatus according to a sixth aspect is the pixel data processing apparatus according to the second aspect, wherein the pixel data processing apparatus extracts the feature of the symmetry of color data of the same color as the pixel to be interpolated. This makes it possible to easily extract the symmetry for all the color data obtained via the Bayer array CCD.

Further, according to the pixel data processing method according to the seventh and eighth aspects, interpolation data is selected for a plurality of pixels each having specific color data according to the characteristics of the color data. It is possible to reproduce a highly accurate image. In addition, it is possible to satisfy high speed by performing the interpolation processing in parallel.

A pixel data processing method according to a ninth aspect is the pixel data processing method according to the eighth aspect, in which the color data is extracted as a feature of the horizontal and vertical symmetry.
Thereby, the correction data is selected. As a result, when symmetry is recognized, correction data based on the rate of color change can be selectively output, and in other cases, correction data based on the rate of color change can be selectively output, so that appropriate interpolation can be performed. Become.

According to a tenth aspect of the present invention, there is provided a pixel data processing method comprising:
9. The pixel data processing method according to claim 8, wherein in particular, the symmetry of the color data in the horizontal direction, the vertical direction, and the oblique direction is extracted as a feature, and the correction data is selected. As a result, correction data based on the rate of color change can be selectively output when symmetry is recognized, and correction data based on the rate of color change can be selectively output otherwise. Becomes

The pixel data processing method according to claim 11 is
9. The pixel data processing method according to claim 8, wherein the symmetry of the G data is extracted as a feature. G with the largest number of data
Since data is used, highly accurate feature extraction is possible.

According to a twelfth aspect of the present invention, there is provided a pixel data processing method comprising:
9. The pixel data processing method according to claim 8, wherein the feature is that the symmetry of color data of the same color as the pixel to be interpolated is extracted as a feature.
This makes it possible to easily extract the symmetry for all the color data obtained via the Bayer array CCD.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an outline of a configuration of a digital still camera 1000 to which a pixel data processing device and a pixel data processing method according to Embodiment 1 of the present invention are applied.

FIG. 2 is an image interpolation unit 1 according to the first embodiment of the present invention.
FIG. 9 is a diagram for explaining the configuration of the 00.

FIG. 3 is a diagram illustrating characteristics of a Bayer type primary color filter included in a CCD 104;

FIG. 4 is a diagram for explaining an input pattern according to the first embodiment of the present invention.

FIG. 5 is a feature extraction unit 1 according to the first embodiment of the present invention.
FIG. 2 is a diagram for describing an outline of the configuration of FIG.

FIG. 6 is a diagram illustrating a method selection unit 2 according to the first embodiment of the present invention.
FIG. 2 is a diagram for describing an outline of the configuration of FIG.

FIG. 7 is a diagram illustrating an interpolation process.

FIG. 8 shows a first method unit 3 according to the first embodiment of the present invention.
FIG. 3 is a diagram for describing a configuration example of FIG.

FIG. 9 is a diagram illustrating a second method unit 4 according to the first embodiment of the present invention.
FIG. 3 is a diagram for describing a configuration example of FIG.

FIG. 10 is a diagram for describing an outline of a configuration of an image interpolation unit 200 according to Embodiment 2 of the present invention.

FIG. 11 is a diagram for describing an outline of a configuration of a feature extraction unit 21 according to the second embodiment of the present invention.

FIG. 12 is a diagram for describing an outline of a configuration of a method selection unit 22 according to the second embodiment of the present invention.

FIG. 13 is a diagram for describing an outline of a configuration of an image interpolation unit 300 according to Embodiment 3 of the present invention.

FIG. 14 is a diagram for describing an input pattern according to Embodiment 3 of the present invention.

FIG. 15 is a diagram for describing an outline of a configuration of a feature extracting unit 31 according to Embodiment 3 of the present invention.

FIG. 16 is a diagram for describing an outline of a configuration of a method selection unit 32 according to Embodiment 3 of the present invention.

FIG. 17 is a conceptual diagram for describing an interpolation process.

FIG. 18 is a diagram for describing a configuration example of a first method unit 33 according to Embodiment 3 of the present invention.

FIG. 19 is a diagram for describing a configuration example of a second method unit according to Embodiment 3 of the present invention.

FIG. 20 is a diagram for describing an outline of a configuration of an image interpolation unit 400 according to Embodiment 4 of the present invention.

FIG. 21 is a diagram for describing an outline of a configuration of a feature extraction unit 41 according to Embodiment 4 of the present invention.

FIG. 22 is a diagram for describing an outline of a configuration of a scheme selection unit according to Embodiment 4 of the present invention.

[Explanation of symbols]

102 optical system, 104 CCD, 106 peripheral circuit,
112 data processing unit, 116 camera control unit, 118
I / O interface circuit, 114 memory, 100
~ 400 Image interpolation unit, 1,21,31,41 Feature extraction unit, 2,22,32,42 Method selection unit, 3,33
1st method part, 4,34 2nd method part, 5 selectors, 8a
８8f Evaluation unit, 7 counter, 6,35 image extraction unit, 1000 digital still camera.

Claims (12)

[Claims] 1. A pixel data processing device for processing a plurality of color data output from a plurality of color filters arranged for a plurality of pixels, wherein the feature extraction unit extracts features of the plurality of color data. And an interpolation unit that performs mutually different processes for interpolating the color data on the interpolation target pixel in parallel, based on the feature extracted by the feature extraction unit,
A selector that selectively outputs any one of a plurality of interpolation processing results output by the interpolation unit, wherein the interpolation unit is configured to perform a process based on a change amount of color data of a pixel in a peripheral region of the interpolation target pixel. A first interpolation processing circuit that interpolates the color data; and a ratio of color data of the same type as the color data to be interpolated to color data of the same type of pixel as the interpolation target pixel in a peripheral area of the interpolation target pixel. A second interpolation processing circuit for interpolating the color data based on the color data. 2. The feature extraction unit detects whether color data of a pixel in a peripheral area of the pixel to be interpolated has symmetry with respect to the pixel to be interpolated. If there is, the second
2. The pixel data processing device according to claim 1, wherein the output of the interpolation processing circuit is selectively output. 3. The plurality of pixels are arranged in a matrix in a vertical direction and a horizontal direction. The feature extraction unit includes: a first evaluation unit that detects the symmetry in the vertical direction; and the symmetry in the horizontal direction. And a second estimator for detecting the first and second interpolators, and integrating the results of the first evaluator and the second evaluator to select either the first interpolation circuit or the second interpolation circuit. And a selector for selectively outputting one of the outputs of the first interpolation processing circuit and the second interpolation processing circuit according to the selection signal. The pixel data processing device according to claim 2. 4. The method according to claim 1, wherein the plurality of pixels are arranged in a matrix in a vertical direction and a horizontal direction, the feature extraction unit includes: a first evaluation unit that detects the symmetry in a vertical direction; And a third estimator for detecting the symmetry in the oblique direction, integrating the results of the first evaluator, the second evaluator, and the third evaluator, A selector for outputting a selection signal for selecting one of the first interpolation processing circuit and the second interpolation processing circuit, wherein the selector performs the first interpolation processing in accordance with the selection signal. 3. The pixel data processing device according to claim 2, wherein the output of one of the circuit and the second interpolation processing circuit is selectively output. 5. The color filter according to claim 1, wherein the plurality of color filters are arranged according to a Bayer array, and each of the plurality of color data is one of red data, blue data, and green data. 3. The pixel data processing device according to claim 2, which detects data symmetry. 6. The pixel data processing device according to claim 2, wherein the feature extraction unit detects the symmetry of the same type of color data as the color data of the pixel to be interpolated. 7. A pixel data processing method for reproducing an image by processing a plurality of color data output from a plurality of color filters for each of a plurality of pixels, wherein the color data is interpolated to a pixel to be interpolated. An interpolation step of performing a plurality of different interpolation processes in parallel for each other; a feature extraction step of extracting a feature of a plurality of color data in synchronization with the interpolation step; And a selecting step of selectively outputting any one of the results of the interpolation processing. The plurality of interpolation processings are performed based on a change amount of color data of a pixel in a peripheral area of the interpolation target pixel. A first interpolation process of interpolating the color data of the same type as the color data to be interpolated and the color data of the same type of pixel as the interpolation target pixel in the peripheral area of the interpolation target pixel. And a second interpolation process of interpolating the color data based on the ratio, the pixel data processing method. 8. The feature extracting step detects whether or not color data of a pixel located in a peripheral area of the interpolation target pixel is symmetric with respect to the interpolation target pixel. The pixel data processing method according to claim 7, wherein when the symmetry is present, a result of the second interpolation processing is selectively output. 9. The method according to claim 9, wherein the plurality of pixels are arranged in a matrix in a vertical direction and a horizontal direction, and the feature extracting step includes an evaluation step of detecting the symmetry in a vertical direction and the symmetry in a horizontal direction; A method selection step of integrating a result of the evaluation step and outputting a selection signal for selecting one of the first interpolation processing and the second interpolation processing, wherein the selection step includes: And selectively outputs one of the outputs of the first interpolation process and the second interpolation process.
The pixel data processing method according to claim 8. 10. The method according to claim 10, wherein the plurality of pixels are arranged in a matrix in a vertical direction and a horizontal direction, wherein the feature extracting unit detects the symmetry in a vertical direction, a horizontal direction, and an oblique direction; Integrating a result of the step and outputting a selection signal for selecting one of the first interpolation processing and the second interpolation processing, wherein the selection step is performed according to the selection signal. Selectively outputting either one of the first interpolation processing and the second interpolation processing,
The pixel data processing method according to claim 8. 11. A plurality of types of color filters are arranged according to a Bayer array, wherein each of the plurality of color data is one of red data, blue data, and green data, and the feature extracting step includes: 9. The pixel data processing method according to claim 8, wherein a symmetry of the green data is detected. 12. The pixel data processing method according to claim 8, wherein the feature extracting step detects a symmetry of color data of the same type as the color data of the pixel to be interpolated.