JP2002010273A - Imaging device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain good images by the way being more easy in signal processing and less degraded in image quality than prior art at least in the case of monochrome imaging while keeping a high frame rate through the addition-readout. SOLUTION: In a color-imaging mode, a addition-readout drive selection control means 201 controls a drive of an imaging element 105 through CCD driver 106 so that an identical color addition of two pixels is performed in a vertical direction, and in a monochrome-imaging mode, an adjacent pixel addition of two pixels is performed in a vertical direction. Thus, the use of the addition-readout method in the monochrome-imaging mode different from in the color-imaging mode eliminates the need of considering a color generation, and further allows the addition inside element which reduces the number of signal processing and provides an output image signal best suited to improve the quality of monochromatic image.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image pickup apparatus using a color image pickup device such as a CCD, and more particularly to an image pickup device configured to obtain an image pickup signal by using an addition drive of the image pickup device.

[0002]

2. Description of the Related Art Conventionally, a single-chip color image pickup apparatus using a Bayer array as a color array has been widely used. The Bayer array handled here is a periodic array in which two 2 × 2 pixels are arranged as a unit array. Two of the four pixels in the unit array have the same color, and two pixels of the same color are arranged diagonally. is there. FIG. 4 shows an example of a typical RGB Bayer arrangement.

[0003] A component color signal (R, G, B or Y, R-
Y, BY) are generated in various signal processing systems.
A typical example is shown below.

In the following, the coordinates of an original pixel point and a point of interest (generated pixel point) are represented by (i, j), and for simplicity, index functions eG, eR, eB for R, G, and B, respectively.
Will be introduced. As shown in the left and upper ends of FIG. 4, the lattice phase is shifted by a half pitch between the original pixel point and the generated pixel point. Therefore, even in the case of the coordinates having the same numerical value, the spatial positions are different by that amount. The index function eG is a function for extracting only a G component from a certain signal processing area, and is “1” for a G pixel area and “0” for another color pixel area. Becomes For example, R00, G0 in FIG.
When the index function eG is applied to four pixels of 1, G10, and B11, "0", "1", "1", and "0" are obtained. The same applies to eR and eB.

[0005] The high-frequency luminance signal HY (i, j) is shifted by one pixel at a time so as to include a shared pixel.
Pixel unit ｛(R00, G01, G10, B11), (G01, R02, B11, G12), (R0
2, G03, G12, B13), (G03, R04, B13, G14), (R04, G05, G14, B1
5),... Are generated using only the G included therein. HY (i, j) = Σ [2 × 2 (i, j)] eG × P (i ′, j ′) (1) where the symbol Σ [2 × 2 (i, j)] is the point of interest. P (i ′, j ′) means the sum (with the pixel coordinates as variables) of the adjacent 2 × 2 (original pixels) 4-pixel area of (i, j), and P (i ′, j ′) represents the coordinates (i ′, j). The signal output value of the original pixel in ') is shown. Accordingly, the high-frequency luminance signal HY (i, j) {= HY (0,0)} corresponding to (i, j) = (0,0) in FIG. G0 in B11
It is generated from the sum of 1 and G10 image signals.
Similarly, HY is calculated from the sum of the image signals of G01 and G12.
(0, 1) is HY from the sum of the image signals of G12 and G03.
(0,2) is generated.

The high-frequency luminance signal HY obtained in this manner is
Regarding (i, j), the high-frequency luminance signals HY (i−1, j), HY of the coordinates (of the generated pixel point) adjacent to each other, respectively.
(I, j-1), HY (i, j + 1), HY (i + 1,
j) and a difference signal from the luminance edge signal AP
(Aperture signal). AP (i, j) = 4HY (i, j)-{HY (i-1, j) + HY (i, j-1) + HY (i, j + 1) + HY (i + 1, j) } (2) In addition, as for the low-frequency primary color signals LR, LG, and LB, an area larger than the high-frequency luminance signal HY, for example, a point of interest (i,
The average value of 4 × 4 = 16 pixels around j), that is, the average value of the pixel signals of 8 pixels included in 16 pixels for G, and the corresponding 4 values included in 16 pixels for R and B
It is calculated using the average value of the pixel signals of the pixels. Using the same notation as above, LR (i, j) = {Σ [4 × 4 (i, j)] eR × P (i ′, j ′)} / 4 (3) R LG (i, j) = {Σ [4 × 4 (i, j)] eG × P (i ′, j ′)} / 8 (3) G LB (i, j) = {Σ [4 × 4 (i, j )] eB × P (i ′, j ′)} / 4 (3) B For the low-frequency luminance signal LY, a matrix operation of 0.3R + 0.59G + 0.11B is performed on the low-frequency primary color signals LR, LG, and LB. Thus, LY (i, j) = 0.3LR (i, j) + 0.59LG (i, j) + 0.11LB (i, j) (4)

[0007] A broadband luminance signal Y and a color difference signal R-
Y (BY) is Y (i, j) = LY (i, j) + k × AP (i, j) RY (i, j) = LR (i, j) -LY (i, j) BY (i, j) = LB (i, j) -LY (i, j) (5) k is a predetermined count. Equation (5)
R (i, j) = LR (i, j) + k × AP (i, j) G (i, j) = LG (i, j) + k × AP (i, j) B (i, j) = LB (i, j) + k × AP (i, j) (6) Equation (6) is completely equivalent to equation (5).

As described above, by adding the luminance edge signal (high-frequency luminance signal) AP obtained by the equations (1) and (2) to the low-frequency luminance signal LY obtained by the equation (4), a wide-band luminance signal Y is obtained. Can be obtained, but the color difference signals RY, BY
Is processed only in the low band, and as can be seen from equation (6), the low-band primary color signals LR,
Since the same (and thus achromatic) high-frequency signal (k × AP) is superimposed on the RGB and LB in RGB, the signal processing is less likely to cause false colors in the edge portion. (For example, if a (broadband) luminance signal is calculated as it is from the information of only the Bayer unit 4 pixels without performing such band-by-band processing and a simple matrix operation is performed, a false color is remarkably generated at an edge portion. In addition, since the high-frequency luminance signal HY is obtained only from G according to the equation (1), the jagged luminance signal due to the color edge portion is suppressed.

That is, if equation (1) is rewritten, HY (i, j) = G (i, j + 1) + G (i + 1, j)... When i + j is an even number, G (i, j) + G (i + 1, j + 1)... I + j is an odd number, and the information of the original pixel G is spread over four adjacent generation points (for example, in the case of G12, (i, j) = ( 0,
1), (0,2), (1,1). (4 generation points of (1,2)) The contribution to one generation point is always ２ and is averaged with the information of the adjacent original pixel point of the same color G, so that the jaggedness is not noticeable due to the smoothing effect. In other words, since the original pixel point of the same color G is always adjacent to the original pixel G (for example, G01, G03, G21, G23 for G12), the two adjacent original pixel G From the signal.

To supplement this point, the high-frequency luminance signal HY is
Obtaining only from RB means discarding RB information, which leads to a corresponding decrease in SN. However, if RB is added, the luminance signal is strongly jagged due to the color edge portion. Not something. In other words, when the high-frequency luminance signal HY is generated by adding all four pixels in the Bayer unit array, that is, instead of Expression (1), HY (i, j) = Σ [2 × 2 (i, j) ] × P (i ′, j ′)… (1) ′, and if this is written down, HY (i, j) = R (i, j) + G (i, j + 1) + G ( i + 1, j) + B (i + 1, j + 1)… When both i and j are even = G (i, j) + R (i, j + 1) + B (i + 1, j ) + G (i + 1, j + 1)… when i is even and j is odd = G (i, j) + B (i, j + 1) + R (i + 1, j) + G ( i + 1, j + 1) ... i is odd and j is even = B (i, j) + G (i, j + 1) + G (i + 1, j) + R (i + 1, j + 1)... i and j are both odd numbers, and the original pixel information of R or B spreads to four adjacent generation points with a contribution of 1. That is, R or B
Since there is no adjacent original pixel of the same color at any of the four generation points, the information is exactly the same for each of the four generation points (information of only one pixel is directly adjacent to the four generation points). 4 generation points). Therefore, for example, at a red edge portion that does not include G and B information, a blue edge portion that does not include R and G information, or a red and blue edge portion, the pixel density is substantially the same as that in which the pixel density is reduced to ４. In this state, apparently four pixels are actually only one pixel, so that the above-mentioned jaggedness occurs. In particular, since this occurs locally, it was even more noticeable than a uniform reduction in resolution of the entire image.

[0011] Under such circumstances, the processing using only G is used even if the SN is reduced. With the above-described processing, it is possible to generate a high-resolution image having the same number of pixels (pixel density) as the number of pixels of the image sensor (pixel density for the subject).

On the other hand, in recent years, the number of pixels of an image sensor has been increased, and recording of the same number of pixels as the number of pixels of an image sensor is not always appropriate when recording an image. That is, recently, an image sensor using a 4 million pixel image sensor has been supplied for consumer use.
It is expected that a device having 10 million pixels will be realized. With the increase in the number of pixels (higher pixel density), if a recorded image is generated with the same number of recording pixels as the number of pixels of the image sensor, the actual use environment of a personal computer or the like is increased due to an increase in the file size of the recorded image. In this case, the image quality (S / N, S / N, S / N) is reduced due to the principle of the image sensor itself such as an overload, a decrease in the frame rate, and a shortage of the accumulated photocharge per pixel of the image sensor.
(Dynamic range) is deteriorated.

In this respect, there is known a technique for improving the sensitivity and the frame rate by setting a recording mode in which pixel addition is particularly performed. In particular, even when a color image pickup device is used, a reading method is devised to improve the sensitivity and the frame rate. The present applicant has also applied for a pixel addition reading technique for performing addition color imaging by performing addition (Japanese Patent Application No. 2000-173947). According to the present technology, by using both the same-color two-pixel addition inside the imaging device and the same-color two-pixel addition outside the device, the same-color four-pixel addition processing is performed on the four-million-pixel image signal. Is generated. As a result, a sufficient sensitivity and a dynamic range can be secured, and the recording image quality can be improved. Note that at least with respect to color imaging, the above-described conventional signal processing is directly applied to a signal after pixel addition.

[0014]

However, the above-described conventional signal processing (Equations (1) to (5)) is based on the premise that the overall image quality at the time of color imaging is optimized. In addition to the above, the resolution was considerably low in addition to the decrease in SN or the pseudo signal (jagged color edge portion) as described above. That is, the signal after the addition of the same color four pixels is subjected to the conventional signal processing (formula (1) to
Since (5)) is performed, considering the distribution of the two Gs added in Expression (1) on the original image sensor, this spreads to a 4 × 4 = 16 pixel area. Will be. That is, as shown in FIG. 5, in the same-color four-pixel addition, the same-color addition between the odd-numbered lines when reading out the odd-numbered fields and the same-color addition between the even-numbered lines when reading out the even-numbered fields result in 2 × 4 × 4 = 16 pixel areas. × 2 four pixel signals (G1 + G2 + G5 + G6, R1
+ R2 + R3 + R4, B1 + B2 + B3 + B4, G3 +
G4 + G7 + G8) is generated. When equation (1) is applied to these four pixel signals, all G pixels in a 4 × 4 = 16 pixel area, that is, G1 + G2 + G5 + G6 and G3 + G
The addition of 4 + G7 + G8 is performed, and as a result, the number of luminance sampling points is one in 4 × 4 = 16 pixel areas, so that the resolution of the luminance signal is reduced.

The resolution itself is 1/4 pixel number (for example, 1
When compared with an image sensor of (00000 pixels), it is not inferior because it corresponds to a 2 × 2 pixel area of the image sensor, but high pixels such as 4 million pixels, 6 million pixels, and 10 million pixels It cannot be said that the image density is sufficiently utilized in the image pickup device, and it can be considered as a factor of image quality deterioration.

As described above, in the conventional pixel addition readout method, when photographing in black and white using a luminance signal as a recording image signal (the object is not limited to black and white but is a general color subject, In (2), since addition reading for color imaging is used as it is, a reduction in resolution and complicated signal processing are required. That is, when performing black-and-white shooting (hereinafter, referred to as monochrome imaging), the luminance signal Y of Expression (5) is used.
Recording (or outputting to the outside) only (i, j) achieves the prima facie purpose, but it is necessary to perform signal processing relating to color information that is originally useless. There was a problem that it could not be obtained.

The present invention has been made in view of the above-mentioned circumstances, and at the time of monochrome imaging, at least at the time of monochrome imaging, is simpler signal processing and has no image quality deterioration, while maintaining a high frame rate by addition reading. It is an object to provide an imaging device capable of obtaining a good image.

[0018]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, the present invention provides a color image sensor having a Bayer array, driving means for driving the image sensor to read out an image signal, and processing the image signal. An image signal generating means capable of selectively generating a color image signal and a monochrome image signal of a predetermined format, and selecting one of a color mode in which the image generated by the image signal generating means is a color mode and a monochrome mode in which the image is monochrome An image pickup apparatus comprising: a mode control unit for designating the first readout drive when the mode designated by the mode control unit is the color mode; Is characterized in that it is configured to perform a second read drive different from the first read drive.

In the image pickup apparatus using the color image pickup device of the Bayer arrangement, a color image pickup mode and a monochrome image pickup mode are prepared, and the read driving method is switched between the color image pickup mode and the monochrome image pickup mode. . By using different addition readouts in the color imaging mode and the monochrome imaging mode in this way, in the monochrome imaging mode, it is possible to perform optimal pixel addition for improving the image quality of a monochrome image without considering color generation. Therefore, it is possible to obtain a good monochrome image with simple signal processing and no image quality deterioration compared to the related art while maintaining a high frame rate by addition reading.

In this case, especially in the color imaging mode, a first addition reading is performed in which the unit pixel signals corresponding to the image receiving pixels of the same color are added together inside the color imaging device, and the first addition reading is performed. It is preferable to execute a second addition readout in which the unit pixel signals corresponding to the adjacent image receiving pixels are read out while being added together inside the color image sensor.

As described above, in the color imaging mode, by performing the same addition and reading as in the related art in consideration of the color, the occurrence of the jaggedness is prevented. By performing the addition, it is possible to suppress the spread of the pixel area on the original image sensor corresponding to the luminance sampling point, and to sufficiently improve the resolution, as compared with the case of performing the same color addition.

Further, in this case, the addition readout drive in the first and second readout drive is performed in such a manner that the charge transfer from the vertical transfer path to the horizontal transfer path of the image pickup device is performed for 2n pixels (2n transfer) in each horizontal blanking period. (2: n is a natural number). 2n addition driving is used to transfer charges. Further, when the charges are transferred from the photoelectric conversion storage unit to the vertical transfer path prior to the charge transfer, the first read drive is performed by A plurality of transfer gate pulses for transferring the charges are output a plurality of times with the timing shifted for each phase, and a predetermined pixel (a predetermined transfer unit) is output during the plurality of output of the transfer gate pulses.
The same-color adjacent charge transfer, in which the pixel signal charges of the same color are rearranged on the vertical transfer path so as to be continuous for at least 2n pixels by accompanying the vertical transfer path driving for the second reading. It is preferable that the driving does not perform the same color adjacent charge transfer.

As a result, the same-color pixel addition and the adjacent pixel addition can be easily realized even for a progressive scanning type image pickup device.

Further, the present invention provides a color image pickup device having a Bayer array, driving means for driving the image pickup device to read out an image pickup signal, and an image pickup device having a pixel density smaller than that of the image pickup device. A pixel density matching unit that generates unit pixel information of a pixel density corresponding to the pixel density of the generated image; and selectively one of a color image signal and a monochrome image signal of a predetermined format based on an output of the pixel density matching unit. An image pickup apparatus comprising: an image signal generation unit capable of generating a color image; and a mode control unit that selectively designates a color mode in which an image generated by the image signal generation unit is monochrome and a monochrome mode in which the image is monochrome. When the mode specified by the mode control unit is a color mode, the pixel density matching unit adds pixel signal information of the same color. To generate unit pixel information for generating the color image signal in the image signal generating unit, and when the mode specified by the mode control unit is a monochrome mode, add pixel signal information of all colors. In this case, the image signal generating means generates unit pixel information for generating the monochrome image signal.

With this configuration, optimal signal processing can be realized when performing monochrome imaging using a color imaging device.
Since a monochrome image signal having a low pixel density can be generated with a high frame rate, a high resolution, no need to generate unnecessary pixel points, and no generation of jaggedness due to color edges, a large number of 4 million pixels, 6 million pixels or more can be generated. Pixel density 1/4 from signal obtained by pixelated color image sensor
Therefore, it is possible to realize optimal signal processing when generating a monochrome image, and to improve the image quality of a recorded image with a small number of signal processings.

The Bayer arrangement used in the present invention is RGB.
A Bayer arrangement is preferred.

Thus, each of the above-described functions can be obtained while securing high color reproducibility during color imaging.

[0028]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows a configuration of an imaging device according to an embodiment of the present invention. Here, a case where the present invention is implemented as a digital camera will be described as an example.

In the figure, reference numeral 101 denotes an imaging lens system including various lenses; 102, a lens driving mechanism for driving the lens system 101; 103, an exposure control mechanism for controlling the aperture of the lens system 101; An optical filter for outer cut, 105 is a CCD color image sensor having a Bayer array color filter, and 106 is an image sensor 1
05, a CCD driver for driving the CCD, 107, a pre-processing circuit including a gain control amplifier, an A / D converter, etc., 108, a digital processing circuit for performing component signal generation processing and other various digital processing, and 109, a card An interface 110 is a memory card, and 111 is an LCD image display system.

The color image sensor 105 has a plurality of photoelectric conversion elements as image receiving pixels. For example, a progressive scan (progressive scan) in a Bayer array of 4 million pixels, for example, is performed.
A type or the like is used.

In the figure, reference numeral 112 denotes a system controller (CPU) for overall control of each unit, 113 denotes an operation switch system including various switches, and 114 denotes an operation display for displaying an operation state and a mode state. System, 11
Reference numeral 5 denotes a lens driver for controlling the lens driving mechanism 102; 116, a strobe as a light emitting unit; 117, an exposure control driver for controlling the strobe 116;
Reference numeral 8 denotes a nonvolatile memory (EEPROM) for storing various setting information and the like.

In the digital camera of the present embodiment,
The system controller 112 performs overall control, and in particular controls the driving of the CCD color image pickup device 105 by the CCD driver 106 to perform exposure (charge accumulation) and signal reading, and the signal is transferred to the pre-processing circuit 10.
7 to generate a recording image signal by taking it into the digital process circuit 108 through the card interface 1
09 to the memory card 110.

In the digital camera of the present embodiment,
The imaging / recording of the subject is performed by using the addition imaging using the pixel addition reading as a so-called default imaging mode. That is, there is no non-addition mode, and vertical two-pixel addition in the element is always performed by two-line addition driving (outputting a vertical transfer V clock of two transfer units in the H blanking period). As a result, two pixel unit pixel signals are added for each vertical transfer path on the horizontal transfer path of the image sensor 105 and read as one line of pixel signals. Then, by adding two horizontal pixels outside the element to the read signal in the digital process circuit 108, an image of １ ／ pixel number of the image sensor 105 is recorded. Therefore, the frame rate is improved to twice that of the standard (non-addition) driving.

The imaging mode includes a "color imaging mode" for obtaining a color image signal as an output image signal for recording.
And a “monochrome imaging mode” for obtaining a monochrome image signal as an output image signal for recording.
“Color imaging mode” and “monochrome imaging mode” can be switched by a manual operation of the operation switch system 113 by the user.

In both the "color image pickup mode" and the "monochrome image pickup mode", four-pixel addition (two pixels in both the horizontal and vertical directions) of the combined use of the inside and outside elements is performed. Different addition reading is performed between the time and the “monochrome imaging mode”.
To realize this, the system controller 112 is provided with an addition read drive switching control unit 201. The addition read drive switching control unit 201 includes a CCD
This is for controlling the driving of the CCD color image pickup device 105 using the driver 106. In the “color image pickup mode”, the CCD driver 106 controls the image pickup device 105 so that the same color addition of two pixels is performed in the vertical direction. In the “monochrome imaging mode”, C is controlled so that two adjacent pixels are added in the vertical direction.
The driving of the image sensor 105 by the CD driver 106 is controlled.

As described above, by using a different addition reading method between the “color imaging mode” and the “monochrome imaging mode”, it is not necessary to consider the color generation in the “monochrome imaging mode”, and the number of signal processes is reduced. Less,
In addition, in-element addition for obtaining an output image signal optimal for improving the image quality of a monochrome image can be performed. Hereinafter, specific operations in the “color imaging mode” and the “monochrome imaging mode” will be described.

(Color Imaging Mode) In the color imaging mode, the same color pixel addition of two vertical lines inside the element and two horizontal lines outside the element are performed.
The combined use of the same-color pixel addition of the pixels results in the same four-pixel addition processing for each color as described with reference to FIG. The signal after the addition, that is, the conventional Bayer signal processing shown in the above equations (1) to (5) is performed on the Bayer data of 1/4 pixel number, and the obtained component color image signal is
A recording process including compression is performed as necessary, and recorded on a medium.

In this embodiment, since the image pickup device 105 is of a progressive scan type, the vertical drive of two lines in the device is different from that shown in FIG. . Figure 2 shows this situation.
Shown in

FIG. 2 shows the relationship between the arrangement of pixels in a specific column (vertical direction) of the image sensor 105 and the arrangement of corresponding pixel signals in the vertical transfer path. In addition, TG indicates a transfer gate (or a charge transfer pulse for controlling the transfer gate) provided between the charge storage unit including the photoelectric conversion element and the vertical transfer path. The TG is divided into four, TG1 to TG4, and is periodically arranged in that order. One column of pixels can be divided into a plurality of pixels and transferred from the charge storage unit to the vertical transfer path.

1. First, by outputting charge transfer pulses corresponding to TG1 and TG2, as shown in the figure,
TG1, TG2 from each pixel column of 4 pixels in the vertical direction
Are transferred to the vertical transfer path (R0, G0, R2, G2, R4, G4, R6, G
6).

2. Driving the vertical transfer path for one clock,
The pixel signal in the vertical transfer path is shifted by one pixel. As a result, the positions of Gs (for example, G0 and G1) that originally had a nested structure with R interposed therebetween are aligned so as to be adjacent to each other. 3. By outputting the charge transfer pulse corresponding to TG4, T is output from each pixel column of four pixels in the vertical direction.
Only the signal of one pixel corresponding to G4 is transferred to the vertical transfer path (G1, G3, G5, G7). As a result, R0, G0, G1, R2, G
2, G3, R4, G4, G5, R6, G6, G7.

4. Driving the vertical transfer path for two clocks,
The pixel signal in the vertical transfer path is shifted by two pixels. Thereby, the positions of R (for example, R0 and R1) are aligned so that they are adjacent to each other. 5. By outputting the charge transfer pulse corresponding to TG3, only the signal of one pixel corresponding to TG3 is transferred to the vertical transfer path from each pixel row of four pixels in the vertical direction (R1, R3, R5, R7). . As a result, in the vertical transfer path, as shown in FIG.
Pixels will be arranged adjacently. For R,
Although the arrangement is inverted upside down with respect to the original arrangement of pixels, there is no problem because the addition processing is performed.

6. Vertical transfer of two transfer units is performed during the H blanking period, and normal transfer is performed on the horizontal transfer path, thereby performing 2n-line addition reading between pixels of the same color. Then, after the signal is read, two pixels are added every other pixel by digital operation as in the conventional case.

(Monochrome Imaging Mode) FIG.
The relationship between the Bayer array of the 10,000 pixel CCD image sensor 105 and the pixel points (pixel generation points) for obtaining a monochrome recorded image of 1 million pixels is shown. In the monochrome imaging mode, as shown by the thick line in FIG.
Pixel point Y at one rate for every × 2 Bayer unit pixel array
(Shown by a black circle) is generated. That is, the pixel point (Y) is an adjacent 2 × 2 4 having no shared pixel (no overlap).
Generated in pixel units. That is, assuming that the coordinates of the generated pixel point of interest are (i, j), the following luminance signal Y is generated only at the points of i = 2m (even number) and j = 2n (even number).

Y (i, j) = Σ [2 × 2 (i, j)] P (i ′, j ′) = R (i, j) + G (i, j + 1) + G (i + 1, j) + B (i + 1, j + 1) (7) Here, the symbol Σ [2 × 2 (i, j)] is a 2 × 2 ( It means the sum (with the pixel coordinates as variables) for the four pixel area (of the original pixels), and P (i ', j') indicates the signal output value of the original pixels at the coordinates (i ', j'). .

That is, R00, G01, G10, B11
A luminance signal Y (0,0) is generated by adding the four pixels of
Similarly, the luminance signal Y (0, 2) is obtained by adding four pixels of R02, G03, G12, and B13 to R04, G05, and G1.
The luminance signal Y (0, 4) is obtained by adding four pixels of B4 and B15.
However, a luminance signal of a corresponding pixel point is generated by adding all four pixels for each Bayer unit pixel array.

The luminance image signal obtained in this way is subjected to a recording process including compression as necessary as a monochrome image signal, and is recorded on a medium. The Bayer signal processing as in the equations (1) to (5) is unnecessary. (However, gamma and Kn
It is a matter of course that general signal processing such as ee and setup is appropriately performed as needed. Of the four-pixel addition, addition of two pixels in the vertical direction is performed by in-element addition driving, and addition of two pixels in the horizontal direction is performed by digital operation by the digital process circuit 108. In the intra-element addition drive in the monochrome imaging mode,
The pixel order permutation as in FIG. 2 is not performed. Specifically, for two vertical pixels (two lines), simple two-line addition driving is performed by simultaneous TG transfer of all lines, and adjacent two pixels are added by digital operation after signal reading.

As described above, by using the addition readout dedicated to the monochrome imaging mode, it is possible to perform the addition between the pixels optimal for improving the image quality of the monochrome image.
While maintaining the high frame rate by the addition reading, it is possible to obtain a good monochrome image with simple signal processing and no image quality deterioration compared to the related art.

For example, as compared with the case where the conventional processing is performed (that is, the luminance signal in the color mode), the pixel area on the image sensor corresponding to one pixel of the generated image is 2
× 2 4 pixel area, extremely high resolution ・ Signal processing is simple and no waste ・ The uniformity of the image is high because all generated pixels have the same structure (one pixel of R and B is generated) (Because it corresponds to only one pixel, there is no jaggedness at the color edge portion.)

Various other embodiments are conceivable.

For example, in this embodiment, the addition in the horizontal direction is performed by an external digital operation. However, if it becomes possible by further improving the driving of the image pickup device, the addition in the horizontal direction is also performed in the device. Is preferred. Further, the signal processing of the present embodiment is not limited to four-pixel addition, but can be extended to arbitrary (2n) ² addition, so that the pixel density becomes 1 / 2n ² (1 / 2n both horizontally and vertically). Can be applied to Here, n is an integer of 1 or more. That is, in this embodiment, the pixel density is 1/4.
Although the signal processing in the case of generating the image of (1) is exemplified, the present invention can also be applied to the signal processing for image generation of the pixel density of 1/16, the pixel density of 1/36,. In either case,
The addition of at least two lines (preferably 2n lines) in the vertical direction is performed in the element.

In addition, at the time of 2n addition in the horizontal direction in (2n) ² addition, arbitrary weighting is performed on the signal P (i, j) of each pixel according to the distance from the point of interest and the color of the pixel. Is also good.

[0053]

As described above, according to the present invention, while maintaining the high frame rate by the addition reading, the signal processing is simpler than the conventional processing in the monochrome mode, and the pseudo signal is more than the conventional one. It is possible to obtain a homogeneous, high-resolution, and good image with less image quality.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an imaging device according to an embodiment of the present invention.

FIG. 2 is an exemplary view for explaining a vertical pixel permutation type two-line addition drive performed in a color imaging mode according to the embodiment;

FIG. 3 is an exemplary view for explaining a pixel generation point and an adjacent four-pixel addition process in a monochrome imaging mode according to the embodiment;

FIG. 4 is a diagram for explaining conventional pixel generation points and Bayer signal processing.

FIG. 5 is a diagram for explaining a conventional four-pixel addition process.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 101 ... Lens system 102 ... Lens drive mechanism 103 ... Exposure control mechanism 104 ... Filter 105 ... CCD color image sensor 106 ... CCD driver 107 ... Preprocess circuit 108 ... Digital process circuit 109 ... Card interface 110 ... Memory card 111 ... LCD image display System 112: System controller (CPU) 201: Addition read drive switching control unit

Claims (5)

[Claims] 1. A color image pickup device having a Bayer array, driving means for driving the image pickup device to read out an image pickup signal, and processing the image pickup signal to selectively generate a color image signal and a monochrome image signal in a predetermined format. An image pickup apparatus comprising: a possible image signal generation unit; and a mode control unit that selectively designates one of a color mode in which an image generated by the image signal generation unit is in color and a monochrome mode in which the image is monochrome. The drive means performs first readout drive when the mode specified by the mode control means is the color mode, and performs second readout drive different from the first readout drive when the mode specified by the mode control means is the monochrome mode. An imaging apparatus characterized in that the imaging apparatus is configured to perform the operation. 2. A first read-out drive is a same-color addition read-out drive for reading out while adding electric charges of the same-color pixels in the image pickup device. A second read-out drive adds electric charges of adjacent pixels in the image pickup device. 2. The imaging apparatus according to claim 1, wherein adjacent addition read driving is performed while reading. 3. The addition readout drive in the first and second readout drives is such that, at the time of charge transfer from the vertical transfer path to the horizontal transfer path of the image sensor, 2n pixels (2n transfer units: n is a natural number), and 2n addition driving is used to transfer electric charges by an amount equal to (n is a natural number). Further, when electric charges are transferred from the photoelectric conversion storage section to the vertical transfer path prior to the above-described electric charge transfer, the first read driving is performed by The transfer gate pulse for charge transfer is output a plurality of times with the timing shifted for each phase, and vertical transfer is performed by driving a vertical transfer path for a predetermined pixel (a predetermined transfer unit) between the plurality of output of the transfer gate pulse. And performing the same-color adjacent charge transfer that rearranges pixel signal charges of the same color on the road so as to be continuous for at least 2n pixels. Moving the imaging device according to claim 2, characterized in that is not conducted the same color adjacent of charge transfer. 4. A color image pickup device having a Bayer array, driving means for driving the image pickup device to read out an image pickup signal, and when the pixel density of the image to be generated is smaller than the pixel density of the image pickup device, A pixel density matching unit that generates unit pixel information of a pixel density corresponding to the pixel density; and a color image signal or a monochrome image signal of a predetermined format can be selectively generated based on an output of the pixel density matching unit. An image pickup apparatus comprising: an image signal generation unit; and a mode control unit that selectively designates a color mode in which an image generated by the image signal generation unit is a color mode and a monochrome mode in which the image is monochrome. The density matching unit adds the pixel signal information of the same color when the mode specified by the mode control unit is the color mode. The image signal generation unit generates unit pixel information for generating the color image signal, and when the mode specified by the mode control unit is a monochrome mode, adds pixel signal information of all colors to generate an image. An imaging apparatus for generating unit pixel information for generating the monochrome image signal in a signal generation unit. 5. The image pickup apparatus according to claim 1, wherein said Bayer array is an RGB Bayer array.