JPS63306790A - Solid-state image pickup element - Google Patents

Abstract

PURPOSE:To obtain a titled element for color image whose average aperture ratio is improved by allocating the photoelectric conversion elements uniformly to (l) kinds of colors in the initialization, and modifying the allocation corresponding to the optimization of the picture element allocation to respective colors. CONSTITUTION:It is assumed that, under a system requirement, the color balance is set so that the required aperture ratio among the three colors of one unit is DR:DG:DB =40:25:60 in terms of percent. One unit is executed in accordance with an inequality I when the picture element allocation is optimized for respective colors. The subtraction in said inequality means the subtraction of a requested aperture ratio value at the minimum from a requested aperture ratio value at the maximum, (m) and (n) represent respectively the number of rows and that of columns constituting one unit (for example m=n=3). When prescribed values are substituted to the equation I, to be satisfied with equalization, the number of picture elements in the color corresponding to the minimum requested aperture ratio i.e. G is decreased by one, while the picture elements in the color corresponding to the maximum requested aperture ratio i.e. B, is increased by one piece. Then the relative ratios of the aperture ratios for respective colors after modifying the picture element allocation is re-calculated, in order to optimize the picture element allocation.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a solid-state image pickup device for color images that can provide high-sensitivity output.

(Prior art) It is well known (for example) to use an MO8 type solid-state image sensor as a measurement cell.

Recent color printers classify the color original dA (color negative film, etc.) into multiple scenes, and perform color correction by adjusting the insertion amount of the color correction filter into the optical path according to each scene, resulting in good color balance. I am creating beautiful printed photos. For this scene classification, a solid-state image sensor for color images is used, and the three-color density of each point of the color original image is measured before printing.

Force 2 - When reading an image, it is well known that there are two types: the Tokoura three-plate type, which uses three image sensors, and the so-called single-plate type, which uses one image sensor. The latter is
On the light-receiving surface of the image sensor, nine color filters are attached corresponding to each pixel, with blue (B), green (G), and red (R) arranged in a mosaic pattern. It is widely used because of its low cost compared to the plate type.

Although the terms "image sensor" and "solid-state imaging device" are originally synonymous, here, for convenience, only the semiconductor device will be referred to as an "image sensor", and the device including the microfilter will be referred to as a "solid-state imaging device".

The inventors first defined m83 adjacent pixels in the light receiving section as one unit, divided the pixels KR, G, and B within one unit in parentheses into nine equal parts, and added the signals of the same color pixels. proposed a single-chip solid-state image sensor that outputs images (%Global Application No. 61-071299).

However, in such a single-chip solid-state image sensor, the aperture ratios of R, G, and H are appropriately set according to the photometry condition, light source balance, and sensor sensitivity balance.
The relative aperture ratio of GB is R:G:B=40%:50
%: It was set to 6096. Therefore, in this case, the simple average aperture ratio per unit was 50g6.

(Problems to be Solved by the Invention) However, as a measuring sensor, it may be inconvenient if the color balance of the three colors is significantly disrupted due to S/stem side requirements. For example, the relative aperture ratio of RGB is R: G
: B=40%: 25%: This is the case where you want to set it to 60%. In this case, the color distribution of pixels within l unit is R, G, as in the above-mentioned configuration.

If B is distributed equally, the simple average aperture ratio per unit will drop to 4296, and the sensitivity will be lowered. That is, the system requirement is a change in color balance, and a decrease in sensitivity is not desirable.

An object of the present invention was to provide a solid-state image sensor for color images that can improve the average aperture ratio by optimizing the color distribution of pixels within the unit according to the system specifications. It's about doing.

(Means for Solving the Problems) In order to achieve the above object, a solid-state imaging device of the present invention shows the following. That is, adjacent m rows and n columns (m
, n>z, and tn = n = an integer other than 2) as one unit, and one color of the color filter (l is 3
(an integer greater than or equal to 1. D1, D
2,..., when arranging pixels according to Dl,
M&X (D, , D1, D2, ...? D
ρ−Min (Dl * D2 s..., Dρ≦
The number of pixels for each color is incremented so as to satisfy the relationship 2.times.(D, +D2+...Dr)/mn.

In a preferred embodiment of the present invention, the photoelectric conversion elements arranged in a matrix are grouped into one unit of mxn, for example 3x3, that is, nine photoelectric conversion elements.

These nine photoelectric conversion elements are attached with filters of two colors, for example, one of the three primary colors (for each), for example, R, G, and B. The colors of the filters combined with the photoelectric conversion element may be arranged in any manner within one unit. However, the color distribution of, for example, nine photoelectric conversion elements is uniformly distributed to, for example, three colors in the initial setting, and the distribution is changed according to optimization of pixel distribution to each color.

Note that if the pixels cannot be distributed evenly to each color in the initial settings, they are distributed approximately evenly. It is necessary to add signal charges of the same color within the l unit, but in order to avoid adding them using a special adder circuit after signal readout, the solid-state image sensor of the present invention When reading signals from pixels, signal charges accumulated in photoelectric conversion elements of the same color are added and extracted.

Various methods can be considered for this addition, one of which is to provide each unit with a number of vertical MOS switches corresponding to the number of colors, and to connect a photoelectric conversion element of the corresponding color to the source of each vertical MOS switch. Just connect. A horizontal line of a corresponding color is connected to the gate of this vertical MOS switch, and a vertical line of a corresponding color is connected to the drain. In this way, for example, when configuring 91 units of 3 x 3 photoelectric conversion elements using three colors of R, G, and B and combining three of each, one signal line is placed in each column and row of the matrix. Just do it.

This has one fewer signal line than a structure in which a vertical MOS switch is provided for each photoelectric conversion element.

(Example) Embodiments of the present invention will be described below with reference to the drawings.

In FIGS. 1 and 2, photoelectric conversion elements arranged in 3 rows and 3 columns, which constitute a part of the light receiving section, are shown as 10 units.

It is shown that R, G, and B color filters are respectively arranged on these nine photoelectric conversion elements corresponding to one pixel 1t'C. Further, each number shown in small size below the symbols R, G, and B indicates the aperture ratio of each pixel (<96).

Now, due to system requirements, the ratio of the required aperture ratios of the three colors in one unit is DR:DG:D8=4o96:25%
The case where the one color balance is set to 60% will be described.

In addition, in the above, D, , DB indicate the required aperture ratio of the pixel corresponding to each color.

FIG. 2 shows a conventional example in which each color of R, G, and B is equally allocated to three pixels each, and has the color arrangement shown in the figure, with 3 rows x 3 columns as one unit. Each pixel is R, G
, H, respectively, are set to predetermined required aperture ratios.

Next, FIG. 1 shows one unit in which the pixel allocation to each color is optimized based on FIG. 2. This optimization is performed using the following method.

However, in the above formula, Max(DR, DB, DB)
-Min (DR, D1, DB) means subtracting the minimum required aperture ratio value from the maximum required aperture ratio value, and m and n are the numbers of rows and columns that make up l unit. ,
That is, in this embodiment, m=n=3 is substituted.

By substituting each predetermined value into the above equation (1), if the equality relationship in equation (1) is established, the pixel of the color corresponding to the minimum required aperture ratio, that is, 9 pixels of color G in this embodiment is decreased by one, and one pixel of the color corresponding to the maximum required aperture ratio, that is, one pixel of one color B, is increased. Then, the relative ratio of the aperture ratio of each color after changing the element distribution is recalculated.

What is the pixel distribution after the change?

R:B:G=3a:2b:4e (2).

In the above equation (2), a, b, and e represent the correction values of the required aperture ratio of 9 per pixel corresponding to each color after changing the pixel allocation.℃
Ori.

DR 3a = X mn DR+D0+DB, a=0.96, b=Q, 90
, e=1.08 are obtained, respectively. Then, the aperture ratio after changing each color is, from the viewpoint of device manufacturing, assuming that the upper limit of the aperture ratio of one pixel is, for example, 6096 for color B, the aperture ratio of the ordinary pixel is as follows.

What is the aperture ratio of the G pixel?

The aperture ratio of B's 11ii element is.

D, = DB = 60 (%) are obtained respectively.

Next, using the aperture ratio after the above change, the equality relationship in the above equation (1) is confirmed again. In this case, if the equality relationship in equation (1) remains unchanged, that is, if the right term is eclipsed, then
Furthermore, the color of the minimum aperture ratio is reduced by one pixel.

Then, the color with the maximum aperture ratio is multiplied by one pixel, and the aperture ratio after changing the pixel number g is calculated again based on the above method. Then, the above method is repeated until the left term of equation (1)≦eclipse. In this embodiment, the equal sign in equation (1) above changes by changing the pixel number allocation once.

This means that the distribution of the number of pixels per pixel has been optimized. Therefore, in this embodiment, the optimization of the pixel number allocation to each color is changed to R: 3 pixels, G: 2 pixels, and B: 4 pixels, and the ratio of the aperture ratio per pixel is R.A.
This is obtained by setting G:B=53%:50%:60%. With this configuration, the average aperture ratio per unit is 42% in the conventional example shown in FIG.

The device of the present invention has an improvement of about 143 times to 5596.

Moreover, the ratio of the required aperture ratio of 9 per l unit is the system requirement DR:DG:DB=4096:25%:609
6 has been achieved.

Next, in order to obtain the preferable color arrangement of the present invention shown in FIG. 1 from the conventional example shown in FIG. It's a translation. At this time, it is necessary to perform color allocation according to the weighting of one color signal. Therefore, one method is to change G in the 1st column and 3rd row to B, and then swap the 3rd row of the 1st and 3rd columns to obtain the pixel array shown in Figure 1. .

Note that the pixel arrangement of the present invention is not limited to the embodiment shown in FIG. 1 (for example, the generally known Bayer arrangement or striped arrangement may be used).

(Effects of the Invention) As described above, according to the solid-state image sensor of the present invention,
By optimizing the color distribution of each pixel filter color in each unit in accordance with the system specifications, the average aperture ratio can be improved and sensitivity can be increased.

[Brief explanation of the drawing]

FIG. 1 shows a preferred embodiment of a color filter applied to the solid-state image sensor of the present invention, showing the one-color arrangement and aperture ratio, and FIG. 2 shows a conventional example used for comparison with FIG. FIG. 1...pixel, 10...l Unit agent Patent attorney (8107) Kiyotaka Sasaki (and 3 others)

Claims (1)

[Scope of Claims] In a solid-state image sensor for color images in which mosaic color filters are arranged corresponding to pixels on photoelectric conversion elements arranged in a matrix, adjacent m rows and n columns (m, n
>2 and m=n=an integer other than 2) as one unit, and l colors of the color filter (l is an integer of 3 or more) are each required aperture ratio D_1, D_2, ..., D_
When arranging pixels according to l, Max (D_1, D_2, ..., D_l) - Min
(D_1, D_2, ..., D_l)≦2×(D_1
A solid-state image sensor characterized in that the number of pixels is allocated to each color so as to satisfy the relationship: +D_2+...D_l)/mn.