JP2007184721A - Solid-state imaging apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a single chip color camera using a color solid-state imaging element, that can attain high image quality. <P>SOLUTION: For example, a digital video signal corresponding to an output of the color solid-state imaging element 11 is fed to an image pattern detection circuit 41c via a 4H line memory 41b. The image pattern detection circuit 41c groups digital video signals of light receiving regions each comprising 5×5 pixels, calculates an average of signals of each group, rearranges the signal averages between the groups depending on the magnitude correlation of signal levels, extracts a difference of the signal levels of the averages between the groups, so as to detect an image pattern, and respectively outputs an EDGE signal and a FLAT signal depending on the image pattern to an RGB synchronization circuit 41k and a false contour automatic suppression circuit 41f. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a solid-state image pickup device such as a video camera or a digital camera using a color solid-state image pickup device, and more particularly to a single-plate color camera provided with a signal processing circuit for color separation.

  In recent years, in a so-called single-plate color camera using only one solid-state image sensor, specific color filters such as red (R), green (G), blue ( A color filter B) is provided. Usually, a pixel provided with an R color filter is an R pixel (red pixel), a pixel provided with a G color filter is G pixel (green pixel), and a pixel provided with a B color filter is B pixel (blue pixel). Pixel). There are various color filter arrays, and various signal processing methods for color separation corresponding to various color filter arrays have been devised.

  In a typical primary color mosaic type color filter array called a primary color Bayer array, for example, a plurality of basic blocks having a unit of 2 pixels in the horizontal direction × 2 pixels in the vertical direction (2 × 2 pixels) as a unit (1 unit) Are arranged. Each basic block has a structure in which two green pixels are arranged on one diagonal and a red pixel and a blue pixel are arranged on the other diagonal. As a signal processing method corresponding to this primary color mosaic type color filter array, for example, it has been common to obtain RGB signals synchronized (color separation) by simple interpolation processing, for example. That is, in the case of simple interpolation processing, in color separation processing for separating R signal, G signal, and B signal for each pixel from an image signal (analog video signal) output from a solid-state imaging device, for example, light reception of 3 × 3 pixels If the pixel of interest at the center of the region is an R pixel, the G signal of the corresponding pixel of interest is obtained by calculating the average value of the signals of the four G pixels adjacent to the pixel of interest. The B signal is obtained by obtaining an average value of signals of four B pixels adjacent to the target pixel.

  However, in such color separation processing by simple interpolation processing, when a bright pixel and a dark pixel are adjacent to each other on a solid-state imaging device, the border between the bright pixel and the dark pixel, that is, the edge portion of the image. There is a problem that a so-called color false signal that is different from the original color is generated.

  As a solution to this problem, a color solid-state imaging device having a signal processing circuit for color separation processing that can effectively suppress color false signals without increasing the circuit scale has already been proposed. (For example, refer to Patent Document 1). However, in the preceding example, in the signal processing for color separation processing, signals of other color pixels are used for calculation, and the calculation formulas in the signal processing are all in the screen (imaging area). Therefore, even though the color false signal can be suppressed, it cannot be erased.

As described above, conventionally, although the generation of color false signals can be controlled, further improvement in image quality has been demanded.
JP 2000-50290 A

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a solid-state imaging device capable of further improving the image quality and achieving higher image quality.

  According to one aspect of the present invention, each color of R, G, B is determined from a solid-state image sensor in which R, G, B color filters are arranged for each pixel in accordance with the primary color Bayer arrangement, and the output of the solid-state image sensor. A signal processing circuit that generates a signal for each pixel, and the signal processing circuit divides the image signals of the light receiving area of m × n pixels obtained from the output of the solid-state imaging device into a plurality of groups, and each group unit On the basis of the result of the calculation, means for obtaining an average video signal having the same color component, means for calculating a magnitude relationship between different groups of the average video signal and a signal level difference between groups, Means for detecting an image pattern in a light receiving region of m × n pixels, and a method for selecting an interpolation signal for RGB synchronization and a contour correction signal for performing contour enhancement according to the detection result A solid-state imaging device is provided.

  With the above configuration, it is possible to generate not only color false signals but also high-quality YUV signals with few false contour signals. As a result, the image quality can be further improved, and solid-state imaging that can achieve higher image quality Equipment can be provided.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, it should be noted that the drawings are schematic and dimensional ratios and the like are different from actual ones. Moreover, it is a matter of course that portions having different dimensional relationships and ratios are included between the drawings.

[First Embodiment]
FIG. 1 shows a basic configuration of a single-plate color camera (solid-state imaging device) using a color solid-state imaging device according to the first embodiment of the present invention. Here, a case where the signal processing circuit of the single-plate color camera is configured using a pattern extraction type camera signal processing circuit is shown. Further, an example will be described in which the size of the light receiving region to be subjected to signal processing is 5 × 5 pixels (m × n pixels).

  That is, in the color solid-state imaging device 11, color filters are arranged in a primary color Bayer arrangement in each pixel arranged in a matrix. For example, as shown in FIG. 2, the primary color Bayer arrangement is such that, on the screen corresponding to the imaging area, the G filter and the R filter are repeatedly arranged in a cycle of two pixels on each odd line, and the B filter and the G filter are arranged on each even line. Is a primary color mosaic type color filter array having a repetitive structure in which 2 × 2 pixels are one unit (basic block). Among the pixels of the color solid-state imaging device 11, R pixels (red pixels) are arranged with R filters, G pixels (green pixels) are arranged with G filters, and B pixels are arranged with B filters. This is called a pixel (blue pixel).

  In the case of the present embodiment, the color solid-state imaging device 11 is provided with, for example, color filters on photodiodes arranged in a matrix that form each pixel, and accumulates in each photodiode via each color filter. Each of the signal charges thus formed is configured by a CMOS (Complementary MOS) type image sensor that is read out by a signal scanning circuit including a plurality of MOS (Metal Oxide Semiconductor) transistors.

  For example, as shown in FIG. 1, an A / D (analog / digital) converter 21 is connected to the color solid-state imaging device 11. The color solid-state imaging device 11 and the A / D converter 21 are driven by a clock & timing generator 31. That is, under the control of the clock & timing generator 31, an optical image incident on the color solid-state image pickup device 11 having a color filter is photoelectrically converted there to be converted into an analog video signal (image signal). It is converted into a digital video signal by the D converter 21.

  A signal processing circuit (pattern extraction type camera signal processing circuit) 41 is connected to the A / D converter 21. In the present embodiment, the signal processing circuit 41 includes a pre-γ (gamma) correction circuit 41a, a 4H line memory 41b, an image pattern detection circuit 41c, a horizontal contour extraction circuit 41d, a vertical contour extraction circuit 41e, and a false contour automatic suppression circuit. 41f, an inverse γ correction circuit 41g, color correction circuits (1H, 2H, 3H) 41h, 41i, 41j, and an RGB synchronization circuit 41k.

  That is, the digital video signal output from the A / D converter 21 passes through a pre-γ correction circuit 41a for amplitude compression, and is a simultaneous signal of a plurality of lines (in this example, 5 lines from 0H to 4H). Is input to the 4H line memory 41b. The 4H line memory 41b is configured using, for example, an SRAM (Static Random Access Memory) capable of storing 4H (4 × 1 horizontal period) of the digital video signal. For example, the digital video signal in which the high luminance portion is nonlinearly compressed by the pre-γ correction circuit 41a has a smaller number of bits than the digital video signal output from the A / D converter 21, so the 4H line memory 41b is reduced in size. Can be

  Here, the simultaneous signal generated by the 4H line memory 41b corresponds to a digital video signal of each line (0H to 4H) in the light receiving region RR of 5 × 5 pixels, for example, as shown in FIG. .

  In FIG. 1, simultaneous signals for 5 lines (0H to 4H) output from the 4H line memory 41b are simultaneously input to an image pattern detection circuit 41c for performing image pattern detection. In the image pattern detection circuit 41c, an EDGE signal (interpolation signal) and false contour suppression, which will be described later, are used based on the digital video signals of the lines 0H to 4H in the 5 × 5 pixel light receiving region RR. A FLAT signal (contour correction signal) for performing contour emphasis by is generated. For example, grouping of digital video signals, calculation of signal addition average value (average video signal) for each group, rearrangement of signal addition average values according to the magnitude relationship of signal levels between groups, and signal addition average value The signal level difference extraction between the groups and the image pattern detection are performed, and an EDGE signal and a FLAT signal corresponding to the edge portion of the image (the luminance change point of the image pattern in the light receiving region RR of 5 × 5 pixels) Is selected. Details of the signal processing in the image pattern detection circuit 41c will be described later.

  Among the simultaneous signals for 5 lines (0H to 4H) output from the 4H line memory 41b, the digital image signals of the lines 1H, 2H, and 3H in the light receiving region RR of 5 × 5 pixels are simultaneously extracted with the horizontal contour. Is input to the horizontal contour extraction circuit 41d. In addition, the digital video signals of the lines 0H, 2H, and 4H in the 5 × 5 pixel light receiving region RR are simultaneously input to the vertical contour extraction circuit 41e that performs vertical contour extraction. Further, the digital video signals of the lines 0H to 4H in the light receiving region RR of 5 × 5 pixels simultaneously obtain signals 1H ′, 2H ′, and 3H ′ for three lines for RGB synchronization (color separation). Is input to the inverse γ correction circuit 41g.

  The horizontal contour extraction circuit 41d generates an HDTL signal based on the digital video signals of the lines 1H, 2H, and 3H in the light receiving region RR of 5 × 5 pixels, and outputs it to the false contour automatic suppression circuit 41f. The vertical contour extraction circuit 41e generates a VDTL signal based on the digital video signals of the lines 0H, 2H, and 4H in the light receiving region RR of 5 × 5 pixels, and outputs it to the false contour automatic suppression circuit 41f.

  The false contour automatic suppression circuit 41f, to which the HDTL signal from the horizontal contour extraction circuit 41d and the VDTL signal from the vertical contour extraction circuit 41e are supplied, corrects the contour correction in the subsequent stage according to the FLAT signal output from the image pattern detection circuit 41c. A signal DTL for controlling the processing is generated and output to a YUV generation / contour addition circuit 71 that generates a YUV signal and performs contour addition.

  On the other hand, the inverse γ correction circuit 41g is provided to enable accurate color correction processing in the subsequent stage, and can thereby perform ideal γ correction on an ideal RGB signal. It becomes. That is, the digital video signals of the lines 0H, 1H, and 2H output from the inverse γ correction circuit 41g pass through the color correction circuit (1H) 41h that performs color correction processing, and the signal 1H ′ having an ideal RGB component. Become. The digital video signals of the lines 1H, 2H, and 3H output from the inverse γ correction circuit 41g pass through a color correction circuit (2H) 41i that performs color correction processing, and a signal 2H ′ having an ideal RGB component. Become. The digital video signals of the lines 2H, 3H, and 4H output from the inverse γ correction circuit 41g pass through a color correction circuit (3H) 41j that performs color correction processing, and a signal 3H ′ having an ideal RGB component. Become.

  The signals 1H ', 2H', 3H 'output from the color correction circuits 41h, 41i, 41j are input to the RGB synchronization circuit 41k. The RGB synchronization circuit 41k receives the EDGE signal output from the image pattern detection circuit 41c. In the RGB synchronization circuit 41k, interpolation processing according to the EDGE signal output from the image pattern detection circuit 41c is performed on the signals 1H ′, 2H ′, and 3H ′ output from the color correction circuits 41h, 41i, and 41j. Thus, synchronized R, G, and B signals are obtained.

  Ideal R, G, and B signals output from the RGB synchronization circuit 41k are sent to a band-limited LPF (Low Pass Filter) 51, and high-frequency components are cut to become signals RL, GL, and BL. . Signals RL, GL, and BL output from the band-limited LPF 51 are sent to the γ correction circuit 61, where γ correction is performed. The signals R ′, G ′, and B ′ after the γ correction are output to the YUV generation / contour addition circuit 71.

  The YUV generation / contour addition circuit 71 uses the international standard Y from the R ′, G ′, B ′ signals output from the γ correction circuit 61 and the DTL signal output from the false contour automatic suppression circuit 41f. A (luminance) signal and a U / V (color) signal are generated. For example, the luminance signal Y is generated and output by adding the low luminance signal YL generated from the R ′, G ′, and B ′ signals and the DTL signal. Also, the color difference signals (R−Y, B−Y) generated based on the R ′, G ′, and B ′ signals are converted into U and V signals and output.

  However, in this embodiment, when the image pattern is detected by the image pattern detection circuit 41c, the brightness (darkness) of each pixel of the 5 × 5 pixel light receiving region RR is almost uniform, that is, the image If the edge portion (luminance change point of the image pattern in the 5 × 5 pixel light receiving region RR) cannot be detected, the contour correction processing (DUV signal addition in the YUV generation / contour addition circuit 71) is not executed.

  The signal processing in the image pattern detection circuit 41c, which is a characteristic part of the invention, according to this embodiment will be described in detail below. First, a light receiving region that is a target of signal processing for detecting an image pattern will be described.

  FIG. 3 shows an example of the light receiving region RR as a signal processing target. This spatially shows a pixel arrangement (signal output) for 5 × 5 pixels out of a pixel arrangement for one screen of the color solid-state imaging device 11 having a primary color mosaic type color filter (primary color Bayer arrangement). Is. In the case of this example, the center R pixel R22 is a pixel of interest when performing color separation processing by simple interpolation processing.

  Here, a color separation process by a simple interpolation process will be described using the 5 × 5 pixel light receiving region RR shown in FIG. 3 as an example. The RGB signal of the pixel of interest synchronized by the simple interpolation process is obtained by the following arithmetic expression.

R22 = R22
G22 = (Gb12 + Gr21 + Gr23 + Gb32) / 4
B22 = (B11 + B13 + B31 + B33) / 4
That is, the R signal of the R pixel R22 that is the target pixel is equivalent to the signal output of the R pixel R22. On the other hand, the G signal of the target pixel (G pixel G22) corresponding to the R pixel R22 is obtained by obtaining the average value of the signal outputs of the four G pixels Gb12, Gr21, Gr23, and Gb32 adjacent to the target pixel. . Similarly, the B signal of the target pixel (B pixel B22) corresponding to the R pixel R22 is obtained by calculating the average value of the signal outputs of the four B pixels B11, B13, B31, and B33 adjacent to the target pixel. It is done. In the case of simple interpolation processing, as described above, the signal output (digital video signal) of 3 × 3 pixels in the central portion in the light receiving region RR of 5 × 5 pixels is used.

  However, in this simple interpolation process, for example, assuming a bright vertical object passing through the R pixels R02, R22, and R42, the signal output of the G pixel Gb12, the R pixel R22, and the G pixel Gb32 increases. The signal outputs of the B pixels B11 and B13, the G pixels Gr21 and Gr23, and the B pixels B31 and B33 become small, and the result of the calculation by the above formula (RGB signal) and the actual synchronized RGB ideal value There will be a gap.

Calculation result: R22>G22> B22
RGB ideal value: R22 ≒ G22 ≒ B22
Such a color signal generated based on the RGB signal having a gap with the RGB ideal value is called a so-called color false signal, and causes a color that does not actually exist to appear on the photographed image. .

  FIG. 4 shows an example of grouping of digital video signals (color pixel signal outputs) for image pattern detection by the image pattern detection circuit 41c in order to suppress the color false signals described above. . In this example, the digital video signal of the light receiving region RR of 5 × 5 pixels is divided into three groups (H1, H2, H3) in the vertical direction.

  The digital video signals (signal addition average values) averaged in the vertical direction in each of the groups H1 to H3 are obtained by the following arithmetic expressions, respectively.

H1 = (R00 + 2 × (Gb10 + R20 + Gb30) + R40) / 8
H2 = (R02 + 2 × (Gb12 + R22 + Gb32) + R42) / 8
H3 = (R04 + 2 × (Gb14 + R24 + Gb34) + R44) / 8
In the case of this example, the signal component of the average value according to the arithmetic expression is (R + Gb) / 2.

  The signal addition average values as the calculation results of the groups H1 to H3 are rearranged according to the magnitude relationship of the signal levels, and the respective differences are obtained, thereby detecting the image pattern, that is, in which region, It is suitable for use in discriminating what kind of amplitude the vertical line-shaped subject (the edge portion of the image) is.

  In general, the luminance signal Y mainly includes the G component as represented by 0.3R + 0.59G + 0.11B. Therefore, the luminance change point of the image pattern corresponding to the edge portion of the image can also be detected from the following arithmetic expression using only the G pixel as the signal component.

H1 = (Gb10 + Gb30) / 2
H2 = (Gb12 + Gb32) / 2
H3 = (Gb14 + Gb34) / 2
In the case of this example, the signal component of the average value according to the arithmetic expression is only Gb.

  FIG. 5 shows another example of grouping of digital video signals (color pixel signal outputs) for image pattern detection by the image pattern detection circuit 41c in order to suppress the color false signal described above. It is. In this example, the digital video signal of the light receiving region RR of 5 × 5 pixels is divided into three groups (V1, V2, V3) in the horizontal direction.

  The digital video signals (signal addition average values) averaged in the horizontal direction for each of the groups V1 to V3 are obtained by the following arithmetic expressions.

V1 = (R00 + 2 × (Gr01 + R02 + Gr03) + R04) / 8
Or
= (Gr01 + Gr03) / 2
V2 = (R20 + 2 × (Gr21 + R22 + Gr23) + R24) / 8
Or
= (Gr21 + Gr23) / 2
V3 = (R40 + 2 × (Gr41 + R42 + Gr43) + R44) / 8
Or
= (Gr41 + Gr43) / 2
The signal addition average values as the calculation results of the groups V1 to V3 are rearranged according to the magnitude relationship of the signal levels, and the respective differences are obtained, thereby detecting the image pattern, that is, in which region, This is suitable for determining what amplitude the horizontal line-shaped subject (the edge portion of the image) is.

  FIG. 6 shows another example of grouping of digital video signals (color pixel signal outputs) for image pattern detection by the image pattern detection circuit 41c in order to suppress the color false signal described above. It is. In the case of this example, the digital video signal of the light receiving region RR of 5 × 5 pixels is divided into four groups (OB1, OB2, OB3, OB4) in the diagonally upward direction or four in the diagonally upward direction. It is divided into groups (OB11, OB12, OB13, OB14).

  The digital video signals (signal addition average values) averaged in the diagonally upward direction of each group OB1 to OB4 are obtained by the following arithmetic expressions, respectively.

OB1 = (Gr01 + Gb10) / 2
OB2 = (Gr03 + Gb12 + Gr21 + Gb30) / 4
OB3 = (Gb14 + Gr23 + Gb32 + Gr41) / 4
OB4 = (Gb34 + Gr43) / 2
Further, the digital video signals (signal addition average values) averaged in the diagonally upward direction of each group OB11 to OB14 are obtained by the following arithmetic expressions.

OB11 = (Gr03 + Gb14) / 2
OB12 = (Gr01 + Gb12 + Gr23 + Gb34) / 4
OB13 = (Gb10 + Gr21 + Gb32 + Gr43) / 4
OB14 = (Gb30 + Gr41) / 2
The signal addition average value that is the calculation result of the groups OB1 to OB4 or the signal addition average value that is the calculation result of the groups OB11 to OB14 is rearranged separately according to the magnitude relationship of the signal levels, and each difference Therefore, it is suitable for detecting an image pattern, that is, for determining which region has an oblique line-shaped subject (an edge portion of an image).

  In order to accurately rearrange the signal addition average values according to the magnitude relationship of the signal levels, each of the groups H1 to H3, the groups V1 to V3, the groups OB1 to OB4, and the groups OB11 to OB12 Each signal component must be the same.

  In particular, if the result of rearrangement of the signal addition average values and the difference between the four groups H1 to H3, V1 to V3, OB1 to OB4, and OB11 to OB12 are determined in combination (overall), More accurate image pattern detection is possible. For example, it is possible to detect a more accurate image pattern by determining that the subject is a vertical line only when H1≈H2 << H3 and V1≈V2≈V3.

  Further, when the rearrangement of the signal addition average value of 25 (5 × 5) pixels and the difference thereof are obtained without grouping, the circuit for detecting the image pattern becomes complicated, and the circuit scale also increases. Increase. That is, as in the case of the present embodiment, vertical / horizontal / diagonal groups of digital video signals are created, and the signal addition average values are rearranged according to the magnitude relationship of the signal levels, and the respective signal addition average values are It is easier to obtain the difference, and the detection of the image pattern by a small circuit can be realized.

  FIG. 7 shows another example of the light receiving region RR ′ as a signal processing target. Here, a case where the G pixel G23 is arranged in the center in the 5 × 5 pixel light receiving region RR ′ is shown as an example. Also, this is because the digital video signal of the light receiving region RR ′ of 5 × 5 pixels is divided into three groups (OB1 ′, OB2 ′, OB3 ′) in the upward-right diagonal direction, or 3 in the upward-left diagonal direction. This is an example in the case of being divided into groups (OB11 ′, OB12 ′, OB13 ′).

  That is, the digital video signals (signal addition average values) averaged in the diagonally upward direction of each group OB1 'to OB3' are obtained by the following arithmetic expressions.

OB1 ′ = (Gr03 + 2 × Gb12 + Gr21) / 4
OB2 ′ = (Gr05 + 2 × (Gb14 + Gr23 + Gb32) + Gr41) / 8
OB3 ′ = (Gr25 + 2 × Gb34 + Gr43) / 4
Further, the digital video signals (signal addition average values) averaged in the diagonally upward direction of each group OB11 ′ to OB13 ′ are obtained by the following arithmetic expressions.

OB11 ′ = (Gr03 + 2 × Gb14 + Gr25) / 4
OB12 ′ = (Gr01 + 2 × (Gb12 + Gr23 + Gb34) + Gr45) / 8
OB13 ′ = (Gr21 + 2 × Gb32 + Gr43) / 4
In the case of this example, the grouping is different from the case of FIG. 6, but the detection of the image pattern is more reliable when the luminance change point is detected using the signal component including the G pixel. At that time, it is preferable to switch the diagonal group for each pixel arrangement.

  Depending on how the groups are divided, that is, how to obtain the signal addition average value, there is a case where it is not necessary to switch the diagonal group for each pixel arrangement.

  FIG. 8 shows a digital video signal (color) for detecting an image pattern in the image pattern detection circuit 41c in order to suppress the color false signal described above in the 5 × 5 pixel light receiving region RR shown in FIG. Another example of grouping of pixel signal outputs) will be described. In the case of this example, the digital video signals in the 5 × 5 pixel light receiving region RR are divided into three groups (OB1 ″, OB2 ″, OB3 ″) in the diagonally upward direction. However, each group (OB1 ″, OB2 ″, OB3 ″) is not switched for each pixel arrangement.

  The digital video signals (signal addition average values) averaged in the diagonally upward direction of each group OB1 ″, OB2 ″, and OB3 ″ are obtained by the following arithmetic expressions.

OB1 '' = (Gr01 + Gb10 + (R02 + 2 × B11 + R20) / 2) / 4
OB2 ″ = (Gr03 + Gb12 + Gr21 + Gb30 + Gb14 + Gr23 + Gb32 + Gr41 + R04 + 2 × (B13 + R22 + B31) + R40) / 16
OB3 ″ = ((R24 + 2 × B33 + R42) / 2 + Gb34 + Gr43) / 4
As described above, in the image pattern detection circuit 41c, vertical, horizontal, and diagonal groups are formed from digital video signals in the same 5 × 5 pixel light receiving region RR (RR ′), and the signal addition average value of each group is signaled. The image pattern is detected by rearranging according to the level relationship and obtaining the difference between the respective signal addition average values. Further, an interpolation signal (EDGE signal) for RGB synchronization and a contour correction signal (FLAT signal) for performing contour enhancement are appropriately selected according to the image pattern.

  FIG. 9 shows all arithmetic expressions (9 types) for interpolation processing of the B pixel B22 that can be selected according to the image pattern. In the case of this example, there are no arithmetic expressions corresponding to the interpolation signal selected according to the detected image pattern other than the nine types shown below.

B22 = (B11 + B13 + B31 + B33) / 4
Or, B22 = (B11 + B13) / 2
Or, B22 = (B11 + B31) / 2
Or, B22 = (B13 + B33) / 2
Or, B22 = (B31 + B33) / 2
Or B22 = B11
Or, B22 = B13
Or B22 = B31
Or B22 = B33
That is, in the RGB synchronization circuit 41k, interpolation is performed on the signals 1H ′, 2H ′, and 3H ′ output from the color correction circuits 41h, 41i, and 41j according to the EDGE signal output from the image pattern detection circuit 41c. As processing, color separation processing corresponding to any of nine arithmetic expressions is performed, whereby synchronized R, G, B signals are obtained.

  For the G pixel G22, there are nine arithmetic expressions corresponding to the interpolation signals selected according to the detected image pattern as shown below.

G22 = (Gb12 + Gr21 + Gr23 + Gb32) / 4
Or, G22 = (Gb12 + Gr21) / 2
Or, G22 = (Gb12 + Gr23) / 2
Or, G22 = (Gr21 + Gb32) / 2
Or, G22 = (Gr23 + Gb32) / 2
Or, G22 = Gb12
Or, G22 = Gr21
Or G22 = Gr23
Or, G22 = Gb32
As described above, as the interpolation process, the optimum color separation process (interpolation calculation) is performed according to the result of the image pattern detection, so that the RGB ideal value can be obtained even for a bright subject such as a vertical line. An RGB signal close to can be generated. Therefore, it is possible to suppress the color false signal, and it is possible to prevent a color that does not actually exist on the captured image from appearing.

  Next, an operation (processing flow) related to detection of an image pattern in the image pattern detection circuit 41c described above will be described. Here, in order to simplify the description, an example will be described in which a simple method is used in which the subject is discriminated by grouping in the vertical direction and the horizontal direction without performing grouping in the oblique direction.

(1) First, as shown in FIG. 4, the signal addition average value of three groups (H1 to H3) in the vertical direction is obtained. In this example, the following formula is a solution (signal addition average value) obtained by the previous calculation formula.

H1 = R00 + 2 × (Gb10 + R20 + Gb30) + R40
H2 = R02 + 2 × (Gb12 + R22 + Gb32) + R42
H3 = R04 + 2 × (Gb14 + R24 + Gb34) + R44
(2) Next, the magnitude relationship between the signal levels of the signal addition average values of the groups (H1 to H3) is obtained, and the determination results (HR, HL) are output according to the following conditions.

H3≈H2≈H1 (H1 and H2 are approximate values, and H2 and H3 are also approximate values)
... HL = 0, HR = 0
H3 ≠ H2≈H1 (H1 and H2 are approximate values, but H2 and H3 are different)
... HL = 0, HR = 1
H3≈H2 ≠ H1 (H1 and H2 are different, but H2 and H3 are approximate values)
... HL = 1, HR = 0
H3 ≠ H2 ≠ H1 (H1 and H2 are different, and H2 and H3 are also different)
... HL = 0, HR = 0
In the above determination, whether or not the signal addition average value of each of the groups H1, H2, and H3 is an approximate value can be obtained from the following arithmetic expression using α as an arbitrary coefficient, for example.

(H2−α) ≦ H1 and H1 ≦ (H2 + α) ... H1 and H2 are approximate values H3 <(H2α) or (H2 + α) <H3 ... H2 and H3 are different (3) Next, as shown in FIG. Then, the signal addition average value of the three groups (V1 to V3) in the horizontal direction is obtained. In the case of this example, the following formula is a solution (signal addition average value) obtained by the previous calculation formula.

V1 = R00 + 2 × (Gr01 + R02 + Gr03) + R04
V2 = R20 + 2 × (Gr21 + R22 + Gr23) + R24
V3 = R40 + 2 × (Gr41 + R42 + Gr43) + R44
(4) Next, the magnitude relation of the signal level of the signal addition average value of each group (V1 to V3) is obtained, and the determination result (VT, VB) is output according to the following conditions.

V3≈V2≈V1 (V1 and V2 are approximate values, and V2 and V3 are also approximate values)
... VT = 0, VB = 0
V3 ≠ V2≈V1 (V1 and V2 are approximate values, but V2 and V3 are different)
... VT = 0, VB = 1
V3≈V2 ≠ V1 (V1 and V2 are different, but V2 and V3 are approximate values)
... VT = 1, VB = 0
V3 ≠ V2 ≠ V1 (V1 and V2 are different, and V2 and V3 are also different)
... VT = 0, VB = 0
(5) Next, the subject is discriminated from four (4 bits) judgment results HL, HR, VT, and VB, and one interpolation signal is selected.

    When HL = 0, HR = 0, VT = 0, and VB = 0, the luminance change point cannot be detected, and the conventional simple interpolation process (4 pixel average) ) Is selected.

G22 = (Gb12 + Gr21 + Gr23 + Gb32) / 4
B22 = (B11 + B13 + B31 + B33) / 4
When HL = 0, HR = 0, VT = 0, and VB = 1, the edge portion of the image is detected from the lower side of the light receiving region RR of 5 × 5 pixels. An interpolation signal is selected.

G22 = (2 × Gb12 + Gr21 + Gr23) / 4
B22 = (B11 + B13) / 2
When HL = 0, HR = 0, VT = 1, and VB = 0, it is assumed that the edge portion of the image is detected from the upper side of the light receiving region RR of 5 × 5 pixels, and interpolation is performed to synchronize using the following arithmetic expression A signal is selected.

G22 = (Gr21 + Gr23 + 2 × Gb32) / 4
B22 = (B31 + B33) / 2
When HL = 0, HR = 1, VT = 0, and VB = 0, it is assumed that the edge portion of the image is detected from the right side of the light receiving region RR of 5 × 5 pixels, and interpolation is performed to synchronize using the following arithmetic expression A signal is selected.

G22 = (Gb12 + 2 × Gr21 + Gb32) / 4
B22 = (B11 + B31) / 2
When HL = 0, HR = 1, VT = 0, and VB = 1, it is assumed that the edge portion of the image is detected from the lower right side of the 5 × 5 pixel light receiving region RR and is synchronized by the following arithmetic expression. Are selected.

G22 = (Gb12 + Gr21) / 2
B22 = B11
When HL = 0, HR = 1, VT = 1, and VB = 0, it is assumed that the edge of the image is detected from the upper right side of the 5 × 5 pixel light receiving region RR. An interpolation signal is selected.

G22 = (Gr21 + Gb32) / 2
B22 = B31
When HL = 1, HR = 0, VT = 0, and VB = 0, it is assumed that the edge portion of the image is detected from the left side of the light receiving region RR of 5 × 5 pixels, and interpolation is performed to synchronize using the following arithmetic expression A signal is selected.

G22 = (Gb12 + 2 × Gr23 + Gb32) / 4
B22 = (B13 + B33) / 2
When HL = 1, HR = 0, VT = 0, and VB = 1, it is assumed that the edge portion of the image is detected from the lower left side of the 5 × 5 pixel light receiving region RR. An interpolation signal is selected.

G22 = (Gb12 + Gr23) / 2
B22 = B13
When HL = 1, HR = 0, VT = 1, and VB = 0, the edge portion of the image is detected from the upper left side of the 5 × 5 pixel light receiving region RR. An interpolation signal is selected.

G22 = (Gr23 + Gb32) / 2
B22 = B33
In the present embodiment, there are no combinations other than the above nine combinations of the determination results HL, HR, VT, and VB.

  Although the detailed description is omitted, the selection of the FLAT signal according to the edge portion of the image (the luminance change point of the image pattern in the 5 × 5 pixel light receiving region RR) is performed in the same manner.

  As described above, the digital video signals in the light receiving area of m × n pixels are divided into a plurality of groups, the signal addition average values of each group are rearranged according to the magnitude relationship of the signal levels, and each signal addition average value An image pattern is detected based on the result of obtaining the difference between them, and an interpolation signal and an outline correction signal for RGB synchronization according to the image pattern are selected. This makes it possible to generate not only color false signals but also high-quality YUV signals with few false contour signals. As a result, in the single-plate color camera, the image quality can be further improved and the image quality can be further improved.

  In addition, according to the present embodiment, as the digital video signal in the light receiving area of m × n pixels, the digital video signal in which the high luminance portion is nonlinearly compressed is used, so the 4H line memory can be reduced in size. It is.

  In particular, since one line memory is also used for RGB synchronization and contour extraction, the signal processing circuit can be configured with a minimum line memory.

  In addition, when detecting an image pattern, vertical / horizontal / diagonal groups of digital video signals are formed, so that the image pattern detection circuit can be simplified and downsized.

  In the present embodiment, the case where the light receiving area is 5 × 5 pixels has been described as an example, but the present invention is not limited to this.

  In addition, the present invention is not limited to the above (each) embodiment, and various modifications can be made without departing from the scope of the invention in the implementation stage. Further, the above (each) embodiment includes various stages of the invention, and various inventions can be extracted by appropriately combining a plurality of disclosed constituent elements. For example, even if several constituent requirements are deleted from all the constituent requirements shown in the (each) embodiment, the problem (at least one) described in the column of the problem to be solved by the invention can be solved. When the effect (at least one of the effects) described in the “Effect” column is obtained, a configuration from which the constituent requirements are deleted can be extracted as an invention.

1 is a block diagram showing a basic configuration of a single-plate color camera according to a first embodiment of the present invention. The figure which shows the primary color Bayer arrangement | sequence which is a color filter arrangement | sequence of a color solid-state image sensor concerning this embodiment. The figure which shows the pixel arrangement | sequence of the light reception area | region of 5x5 pixel concerning this embodiment. The figure which shows an example of grouping of the digital video signal of the light reception area | region of 5x5 pixel concerning this embodiment. The figure which shows another example of grouping of the digital video signal of the light reception area | region of 5x5 pixel concerning this embodiment. The figure which shows another example of grouping of the digital video signal of the light reception area | region of 5x5 pixel concerning this embodiment. The figure which shows another example of the light reception area | region of 5x5 pixel concerning this embodiment. The figure which shows another example of grouping of the digital video signal of the light reception area | region of 5x5 pixel concerning this embodiment. The figure shown in order to demonstrate the relationship between the light reception area | region of 5 * 5 pixel concerning this embodiment, and the arithmetic expression which can be selected.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 11 ... Color solid-state image sensor, 21 ... A / D converter, 31 ... Clock & timing generator, 41 ... Signal processing circuit, 41a ... Pre-gamma correction circuit, 41b ... 4H line memory, 41c ... Image pattern detection circuit, 41d ... Horizontal contour extraction circuit, 41e ... Vertical contour extraction circuit, 41f ... False contour automatic suppression circuit, 41g ... Reverse gamma correction circuit, 41h, 41i, 41j ... Color correction circuit, 41k ... RGB synchronization circuit, 51 ... Band-limited LPF 61... Γ correction circuit, 71... YUV generation / contour addition circuit.

Claims (5)

A solid-state imaging device in which R, G, and B color filters are arranged for each pixel according to a primary color Bayer arrangement;
A signal processing circuit that generates R, G, and B color signals for each pixel from the output of the solid-state imaging device;
The signal processing circuit includes:
Means for dividing the video signals of the light receiving area of m × n pixels obtained from the output of the solid-state imaging device into a plurality of groups, and obtaining an average video signal having the same color component in each group unit;
Means for calculating a magnitude relationship between different groups of the average video signal and a difference in signal level between the groups;
Means for detecting an image pattern in the light receiving area of the m × n pixels based on the result of the calculation;
A solid-state imaging device comprising: an interpolation signal for performing RGB synchronization and a contour correction signal for performing contour enhancement according to the detection result.   2. The solid-state imaging device according to claim 1, wherein a signal obtained by performing nonlinear compression processing on an output of the solid-state imaging device is used as the video signal of the light receiving region of the m × n pixels.   2. The solid-state imaging device according to claim 1, wherein the average video signal is always obtained by the same grouping and the same arithmetic expression regardless of the arrangement of the color filters arranged for each pixel.   The video signal in the light receiving region of m × n pixels includes a plurality of groups corresponding to the vertical pixel columns, a plurality of groups corresponding to the horizontal pixel columns, or a plurality of pixels corresponding to the diagonal pixel columns. The solid-state imaging device according to claim 1, wherein the solid-state imaging device is divided into groups.   2. The solid-state imaging device according to claim 1, wherein the detection of the image pattern is to extract a luminance change point in a light receiving region of the m × n pixels.

Images