JPH0287785A - Image pickup device - Google Patents

Abstract

PURPOSE:To expand the dynamic range by moving a stored charge of each picture element in a photodetection section to a vertical transfer section for plural number of times of different time intervals in the case of reading the charge from a solid-state image pickup element and adding the charges on the vertical transfer section. CONSTITUTION:The photodetection section 2 in the solid-state image pickup element 1 generates an electric charge in response to the quantity of photodetection. A solid- state image pickup element drive means outputs a pulse for signal read within one vertical scanning period for plural different time intervals. Let A be an exposure period and B be a readout period, then after the charge of the picture element corresponding to phiV-V4 is moved to the vertical transfer section 3 at the start of the period A, the charge is thrown away by reverse transfer. Then the charge is moved from the photodetection section 2 to the vertical transfer section 3 separately by 5 times (C1, C2, C3, C4, C5) and the charges added on the vertical transfer section 3 are read out for the period B. The timing of the charge movement for 5 times is varied to vary the brightness-output voltage characteristic. The charge from the vertical transfer section 3 is transferred to a horizontal transfer section 4. Thus, the range of a wide brightness is converted in this way.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to an MS device, and more particularly to an imaging device suitable for using a solid-state imaging device.

(Background of the Invention) Solid-state image carriers are known to have a much narrower dynamic range (latitude) with respect to brightness than photographic film. For this reason, when performing a scene with a large difference in brightness level, for example, an outdoor positive view on a sunny day, high brightness areas often appear white, and conversely, low brightness areas often appear black.

In a normal solid-state image sensor (COD, MOS, etc.), the amount of incident light and the generated charge have a linear relationship (γ-1). This situation is shown in FIG.

In this case, the dynamic range of the signal is narrow, and high-brightness portions quickly become saturated. For this reason, the luminance part a of the output image was blown out in white, making it impossible to distinguish the contrast in the subject.

In order to solve these drawbacks, it was proposed in 1978 that solid-state image sensors be given knee characteristics.
National Conference of the Television Society of Japan, pp. 43-44 rccD
Knee characteristic control of image sensor V". The applicant of the present invention also proposed the knee characteristics of COD in Japanese Patent Application No. 1987-8739.

This knee characteristic widens the dynamic range of luminance by providing a knee characteristic to the relationship between the amount of incident light and the generated charge, as shown in FIG.

(Problem to be Solved by the Invention) The above-mentioned knee characteristics widen the incident dynamic range, but it is possible to accurately control the amount of accumulated charge in the light receiving section or the transfer voltage between the light receiving section and the transfer section, and to ensure that there is no variation between each pixel. There is no problem and it is fast! It is extremely difficult to do D.

The present invention has been made in view of the above-mentioned problems, and its purpose is to provide a decoy that can fully utilize the solid-state R@hydrogen dynamic range and obtain photoelectric conversion output over a wide brightness range. The aim is to realize image purchasing.

(Means for Solving the Problems) The present invention that solves the above problems includes: a light receiving section that converts incident light from the outside into charges; a vertical transfer section that transfers the charges from the light receiving section in a vertical direction; f! A solid body 11 equipped with a horizontal transfer section that transfers the charge of the direct transfer section in the horizontal direction! The solid-state image sensor has an i-image element and a solid-state image sensor drive circuit that supplies a drive pulse to the solid-state image sensor, and the solid-state image sensor drive means includes: 1
Pulses for reading signals at multiple different time intervals are output during the vertical scanning period, and the charge generated in the light receiving section is divided into multiple different time intervals and transferred to the vertical transfer section. The feature is that it is configured to add and read by [.

(Function) When reading the charge from the solid-state image element, the bulk charge (4) of the light receiving part of each pixel is transferred to the vertical transfer part in several different time intervals, and then added up on the vertical transfer part. One after another?l
,.

<Examples> Examples of the present invention will be described in detail below with reference to the drawings.

FIG. 1 is a block diagram showing the structure of a main part of an embodiment of the present invention. In this figure, numeral 1 indicates a solid-state image sensor, and the case of an interline COD will be described as an example.

2 is a light receiving section that generates charges according to the amount of received light, 3 is a vertical transfer section that vertically transfers the charges generated in the light receiving section, and 4 is! I! This is a horizontal transfer section that horizontally transfers charges from the direct transfer section.

Here, 3×4 pixels are shown.

Further, it is assumed that the potentials of the light receiving section 2 and the vertical transfer section 3 of this solid state 1liI image element 1 are as shown in FIG. 2 or 3. That is, it is assumed that the transfer section can store five times as much charge as the light receiving section.

FIG. 4 is a time chart showing drive pulses when driving this solid-state image sensor 1. In FIG. Note that the solid-state image sensor drive circuit that generates this drive pulse can be realized by making some changes to a known circuit, and the detailed circuit configuration will be omitted.

The operation of the embodiment of the present invention will be described below with reference to FIGS. 1 to 4.

If △ in Fig. 4 is the period of exposure and BI'fii output, then at the beginning of ∆, the charges of pixels corresponding to φv1 to φv4 are transferred to the vertical transfer section, and then the charges are discarded by reverse transfer. . During the attack, charge is transferred from the light receiving section 2 to the vertical transfer section 3 in five steps (C1, C2° C3, C4, C5).
What has been brightened on the vertical transfer section 3 is read out during period B.

By changing the timing of these five charge transfers,
The relationship between the amount of light and the voltage of the output signal changes.

Next, the relationship between the amount of light and the voltage of the output signal will be explained.

FIG. 5 is a characteristic diagram showing the relationship between the charge accumulation time and the amount of accumulated Ff1 charge. At tl+..., t5 in this figure, charges are transferred from the light receiving section to the vertical transfer section. Moreover, QO is the amount of saturation charge of the light receiving section 2. Here, solid lines, broken lines, etc. indicate how charges accumulate at each luminance. The brightness ratio here doubles in the order of solid line Δ, broken line B, . . . . Also, here CI =02=03=04=
The case of 4xC5 is shown.

The output voltage of each pixel is proportional to the charge m of the pixel at each time point tl+j2+jl-j4°t5, which is shown in FIG. This figure shows the relationship between luminance 1 and output voltage V. Here, (-^ is the luminance at the characteristic shown in Fig. 5, Ls is the luminance at the characteristic B of Fig. 5, Lc is the luminance at the characteristic C of Fig. 5, and Lo is the luminance at the characteristic B of Fig. 5.
This is the brightness for the characteristics shown in the figure. In addition, the characteristic (■) is the characteristic when conventional exposure is performed, and (■) is the characteristic when the exposure < C shown in FIG.
1=C2=C3-C4=4XC5), and (2) is the characteristic when exposure is performed as C1=C2=C3=C4=8XC5.

As is clear from ■ and ■ in this figure, by performing the exposure shown in Figure 5, the output is compressed in the high n degree range < LA to Lc, which is approximately three times that of normal exposure. The dynamic range of 19.

Furthermore, when exposure is performed with CI = C2 = C3 = C4 = 8XC5, the output is compressed in the high brightness region of LA-LD, and the dynamic range becomes about 7 times.

By the way, what is important in the above operation is the ratio of 01 to C5, and the absolute value is not important. For example, the F of the lens
When the value is 2.8 and when the F value is 4, the shape of the characteristics shown in FIG. 6 remains the same, but the brightness doubles. Also, C1
If the time is halved while the ratio of ~C5 remains unchanged, the shape of the characteristic shown in FIG. 6 remains unchanged, but the brightness doubles.

As an example, 1. -1o=12-1. -13-jz=t
+ j3-1/250 seconds, t5-t4 =1/100
I'll explain in 0 seconds. The bright a L c pixel is jo ”-
j+, t+ "t, 4, t2~t3, t3
The signal becomes saturated in the first 1/1000 second of each period from ~t4 (FIG. 5C), and the remaining time is meaningless as an exposure time. Since saturation does not occur in the minimum 1/1000 second (tn to t5), it is meaningful as an exposure time. Therefore, the actual exposure time for this pixel is 5X1/1000
= 1/200 second. Similarly, a pixel with luminance La is 1o
-1,.

The signal is saturated in the first 1/500 seconds of each period of t, ~j2+, t2-13, and 13-t4 (Fig. 5B), and the remaining time is meaningless as an exposure time. ! 1/ after Ii
Since saturation does not occur in 1000 seconds (t4~js),
It has meaning in terms of exposure time. Therefore, the actual exposure time for this pixel is 4×11500+1/1000=
It becomes 1/111 second. Similarly, the pixel with luminance LA is tO""jl + il ~ t2. t2-13,1.
The signal becomes saturated in the first 1/250 seconds of each period from t4 to t4 (FIG. 5 △), and the remaining time is meaningless as an exposure time. Since saturation does not occur at the minimum 1/1000 second (jn~ts), it is meaningful as an exposure time. Therefore, the actual exposure time for this pixel is 4x1/250+1
/1000=1159 seconds. For pixels darker than luminance rJIt-A, all exposure times are valid, so similarly 115
It will be 9 seconds. In other words, the dynamic range is expanded by making the exposure time of dark pixels a little longer and the exposure time of bright pixels a little shorter.

Figure 7 shows C1-02=4XC3=4XC4=16XC
5 is a characteristic diagram showing the case of No. 5; FIG. In this figure, the brightness is Δ,
B, ..., E are each doubled. FIG. 8 shows the relationship between luminance and output voltage when such exposure is performed.

Figure 8 (2) shows the characteristics when normal exposure is performed.

With this characteristic, good output can be obtained in dark areas, but
When 1lll[ becomes a little large, it becomes saturated.

Therefore, the dynamic range is extremely narrow. Also, ■ is the characteristic when normal exposure is performed by adjusting the aperture or shutter speed. This will expand the dynamic range to some extent, but the output will be small in dark areas and the S/N will be low.
Therefore, in this case as well, a satisfactory result cannot be obtained. The method shown in (2) is the method explained in FIG. With this characteristic, a sufficient output can be obtained in the dark area, and there is no saturation in the In degree part Tb. The slope of the middle part is between the high brightness part and the low brightness part, and an easy-to-see image can be obtained.

Figure 9 shows C1-4XC2-4XC3-4XC4=1
It is a characteristic diagram showing the case of 6xC5. In this figure, the brightness is doubled for A, B, . . . , E, respectively. FIG. 10 shows the relationship between luminance and output voltage when such exposure is performed. The brightness and output characteristics are bent at two places as in the case of Figure 8, but
The bending point and slope are slightly different.

In this way, by changing the combination of C1 to C5,
Brightness-output voltage characteristics can be freely designed. Further, in the above example, the types of exposure time are 3f4 types, but it is also possible to use 4 types or 5 types. Furthermore, the number of times of exposure and 'R charge movement is not limited to five times, but can be increased or decreased depending on the ratio of the capacitances of the light receiving section and the transfer section.

It should be noted that, although field transfer is not possible in a normal COD, frame transfer is also possible in a cCD that has separate vertical transfer units for odd and even fields.

The above explanation is for shooting still images, but in the case of a CCD or FIT-COD that has separate vertical transfer sections for the odd and even fields,
It can also be applied to videos.

FIG. 11 is a block diagram showing the structure of Ffl--COD. Components J3 in this figure that are the same as those in FIG. 1 are given the same numbers and their explanations will be omitted. The difference from the COD shown in FIG. 1 is that it has a memory section 5 that stores charges corresponding to each pixel.

FIG. 12 shows pulses for driving this FIT-COD in field accumulation mode. In other words, the light receiving part! Charges are transferred between the backward transfer sections multiple times, and after exposure is completed, the charges on the vertical transfer section are transferred to the memory section at high speed without being read out from the memory section at the video rate. While the memory section is reading data, the light receiving section is exposing the next field. Here, in order to output it as a video signal, the condition that C1+C2-toC3+C4+C5≦1/60 seconds is added, but other than that, it is the same as in the case of a still image. In this case, usually CI +C2+C3+C4+C3
= 1/60 seconds, but in the case of a COD that allows the signal charge to be swept out midway (for example, in the case of a COD that can discard the charge to the A-bar 70-drain (OFD), C1 + 02 + C30C4+ 05 < 1 /
It can also be set to 60 seconds.

Further, FIG. 13 shows pulses for driving this Fl'r-COD with the frame accumulation mode. Here, charges are first swept out by reverse transfer.

Then, charges are transferred between the light receiving section and the vertical transfer section multiple times, and after exposure is completed, the charges on the vertical transfer section are transferred at high speed to the memory section and read out from the memory section at a video rate. While the memory section is reading data, the light receiving section is exposing the next field. 01 for each field
The same applies to performing exposure for +C20C30C41-C5≦1/60 seconds. Furthermore, in the FIT-COD in which charge can be discarded to the OFD, a pulse to o i: ca is applied instead of the reverse transfer pulse train.

As an example of the frame accumulation mode, the case shown in FIG. 15 can be considered. In this case, there is no reverse transfer sweep, and during the period when exposure, accumulation, and charge transfer are being performed for the signal of one field, the signals of the other field do not move, so exposure and accumulation can continue during that time. become. Therefore, in this case, C1≧1/60 seconds, C1+C
The condition is 2+C3+C4+C3=1/3o seconds.

Another possible example of the history of frame accumulation modes is as shown in FIG. This is an example using a CCD that can dump charge into an OFD. Since the charge can be discarded to the OFD, the conditions as in the case of FIG. 15 are not applied.

If it is possible to control such that only one field is discarded when discharging charge to the OFD, drive with a pulse as shown in Fig. 17, and the conditions are C1 + C2 + C3 + C4 + C
5≦1/30 seconds only. However, if the signals of the fields on both sides are discarded when the charge is discarded to the OFD, it is necessary to consider the influence thereof. That is, the first
The case shown in Figure 6 is fine, but the case shown in Figure 18 is C.
The exposure/accumulation time of No. 3 is C3'.

In COD, which has separate vertical transfer sections for odd and even fields, each field can move between the light receiving section and the vertical transfer section at completely arbitrary timing, so under the condition of CI + C2 + C3 + 04 + 05 = 1/30 seconds. Now you can freely set C1 to C5. C.I.
102 + 03 +04 + C5 < 1/30 second knee r
- Kill 71) 1 No LL, depends on whether the charge can be discarded to the OFD.

Note that since the characteristics are determined by the ratio of 01 to C5, by changing the overall exposure time by m while keeping the ratio fixed, it is possible to shift the brightness range where m shadows are possible.

Further, although the optimum brightness-output voltage characteristic curve may differ depending on the content of the image, this can be easily dealt with by changing the timing of the pulse.

(Effects of the Invention) As explained in detail above, in the present invention, the charges generated in the light receiving section are read out in multiple times according to the capacities of the light receiving section and the vertical transfer section, and are added on the vertical transfer section. and read it out. Therefore, the output of the solid-state image sensor does not saturate even in a high-brightness region, and the dynamic range is widened. Therefore, it is possible to realize an imaging device that can fully utilize the dynamic range of the solid-state Hl element image and obtain photoelectric conversion output over a wide brightness range.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention, FIGS. 2 and 3 are explanatory diagrams showing the potential of a solid-state llIl elementary image, FIG. 4 is a time chart showing drive pulses, and FIG.
Figures 1 to 10 are characteristic diagrams of the solid-state imaging device of the present invention.
Figure 1 is a configuration diagram showing the configuration of FIT-COD, Figures 12 to 18 are time charts showing other examples of drive pulses, and Figures 19 and 20 are characteristic diagrams of conventional solid-state image devices. be. 11...Solid image carrier 2...Light receiving section 3...Vertical transfer section 4...Horizontal transfer section Patent applicant Konica Co., Ltd. Representative Patent attorney Fuji Ijima 1 person Figure 1 Figure 4 φH1φH2 Figure 2 Figure 3 Vsync ψV21~φV24 Figure 12 Figure 13 φ27~~24 Yusho 14 Figure 15

Claims (1)

[Scope of Claims] A light receiving section that converts incident light from the outside into charges, a vertical transfer section that transfers the charges from the light receiving section in the vertical direction, and a horizontal transfer section that transfers the charges from the vertical transfer section in the horizontal direction. a solid-state image sensor, and a solid-state image sensor driving means for supplying a drive pulse to the solid-state image sensor, and the solid-state image sensor driving means reads signals at a plurality of different time intervals within one vertical scanning period. The device is characterized by a structure in which the charge generated in the light receiving section is transferred to the vertical transfer section several times at different time intervals, and the charges are added up on the vertical transfer section and read out. imaging device.