JPH04152770A - Signal read processing system for solid-state image pickup element - Google Patents

Abstract

PURPOSE:To obtain a signal suppressing an FPN at a dark state stably by reading a 1st half signal for one horizontal scanning period, resetting a stored charge, reading the signal again for a latter half, matching the timing of both read signals at an external circuit and applying subtraction processing to the signal without no photo charge from a signal in response to the photo charge. CONSTITUTION:A signal in response to a photo charge by one horizontal picture element string is read for a first half for one horizontal scanning period by applying high speed horizontal scanning to a solid-state image pickup element 1 and the stored charge is subject to high speed horizontal scanning again to read a signal by a same horizontal picture element while the stored charge is reset simultaneously or being reset for the latter half. A memory control section 12 is subject to time axis conversion by controlling a phase of a memory read pulse, the result is D/A-converted by D/A converters 9, 10, resulting in obtaining a signal B by a photo charge and a signal C at a dark state returned to the usual time axis. Then the signal B by the photo charge and the signal C at the dark state are inputted to a subtractor 11, in which the signal C is subtracted by the signal B by the photo charge to obtain a video signal D in which the FPN at dark state is suppressed is obtained.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention provides a solid-state imaging device that can follow changes in dark fixed pattern noise caused by temperature changes and can always stably suppress dark fixed pattern noise. This invention relates to a signal readout processing method for an element.

(Prior Art) Conventionally, static induction transistors (Statistical
c Induction Transistor:
(abbreviated as SIT) and charge modulation element (Cha
rge ModulationDevice: CM
In devices that reproduce images using an image sensor that uses an internally amplified photoelectric conversion element (abbreviated as FPN) as a unit pixel, it is necessary to cancel fixed pattern noise (hereinafter abbreviated as FPN) specific to the image sensor. A storage means such as a frame memory is provided in the storage means, the FPN is stored for each pixel in the storage means, and the FP corresponding to the pixel is determined from the image information obtained from each pixel of the image sensor.
An FPN suppression means that obtains an image signal by subtracting N is required.

FIG. 5 is a schematic diagram of a conventional FPN suppressing means used in an image reproducing apparatus using such an internally amplified image sensor. The image signal output from the image sensor 102 is amplified by the amplifier circuit 103 and
The data is converted into digital data by the D converter 104. This digital data becomes dark FPN data of each pixel when light is blocked by a shutter 101 placed in front of the image sensor 102. At this time, switch 1
05 is connected to the a side, and this FPN data is sequentially stored in the frame memory 106. After the accumulation of l frame period is completed, switch 105 is connected to side b, and shutter 1 is turned on.
When 01 is opened, the image signal based on the photoelectric charge sent to the A/D converter 104 after the release is A/D converted, and is repeatedly subtracted from the FPN data stored in the frame memory 106 by the subtracter 107. , FPN are input to the D/A converter 10B as data with canceled data, and the data is D/A converted and output as an output video signal.

Further, FIG. 6 is a block diagram showing another example of the configuration of the conventional FPN suppressing means. This configuration example is a bright signal line memo consisting of a capacitance column in the solid-state image sensor 111.
and a dark signal line memo January 13, and light receiving section 114.
The image signals generated by the photocharges of one horizontal pixel column are simultaneously transferred and held in the capacitor column of the bright signal line memory corresponding to each pixel during the horizontal blanking period. After transferring the signal based on the photocharge to the bright signal line memory 112, the same horizontal pixel column is reset 7) at the same time within the same horizontal blanking period, and the dark signal (FPN) is handled for each pixel. The dark signal line memory 113 simultaneously transfers and holds the dark signal to the capacitor column of the line memory 113. Next, during the horizontal scanning period, signals are simultaneously read out sequentially from both the bright signal line memory 112 and the dark signal line memo 13, and the external subtracter 115 subtracts the dark signal from the signal due to the photocharge to calculate the dark FPN. It is designed to output a suppressed video signal.

The problem that r invention tries to solve! ! By the way, the conventional FPN suppressing means shown in FIG.
Since the video signal is output by subtracting N,
When the dark FPN changes due to temperature changes or the like, there is a problem in that it is not possible to follow this change and it is not always possible to stably suppress the FPN.

Furthermore, in the conventional FPN suppression means shown in FIG.
Since two line memories must be placed within the solid-state image sensor, the structure of the solid-state image sensor becomes complicated, the chip area increases, and the number of output signal lines from the solid-state image sensor becomes two. be. This becomes a fatal problem in applications where chip area and mounting space are limited, for example, when the chip is mounted on the tip of an endoscope. Furthermore, variations in the capacitor columns corresponding to each pixel constituting the line memory generate new variations in signal level, that is, new FPN, so there is a problem that the FPN suppression effect as a whole does not become large.

The present invention has been made in order to solve the above-mentioned problems of the FPN suppressing means of the conventional internally amplified solid-state image sensor, and prevents the increase in chip area due to line memory, and prevents changes in dark FPN due to temperature changes. It is an object of the present invention to provide a signal readout processing method for an internally amplified solid-state image pickup device that can follow FPN in the dark and stably suppress FPN at all times.

[Means and effects for solving the problem] In order to solve the above problems, the present invention provides a signal readout method for a solid-state image sensor in which an internal amplification type pre-conversion element is used as a unit pixel and the unit pixels are arranged in a matrix. In the processing method, after reading out the signal of one horizontal pixel column corresponding to the photocharge in the first half of one horizontal scanning period, immediately after or while resetting the accumulated charge of the one horizontal pixel column, the second half of one horizontal scanning period is read out. Then, read out the signal in the state where there is no photocharge for each horizontal pixel column again, synchronize the timing of both readout signals with an external circuit, and then subtract the signal in the state where there is no photocharge from the signal corresponding to the photocharge. It is something to do.

In the signal readout processing method configured in this way,
Every time a signal corresponding to a photocharge in one horizontal pixel column is read out, the signal of the same horizontal pixel column immediately after without photocharge is read out and subtracted, and a signal with FPN removed is output. . Therefore, the dark FPN is suppressed in real time, and even if the dark FPN changes due to temperature changes, etc.
A signal that immediately follows the change and always stably suppresses the dark FPN can be obtained.

According to this unique signal readout processing method, there is no need to provide a line memory in the solid-state image sensor, so it is possible to prevent an increase in the chip area of the element due to the addition of a line memory, and to prevent new FPNs from occurring due to variations in the capacity of the line memory. Ru.

〔Example〕

Next, an example will be described. FIG. 1 is a block diagram of a signal readout processing device for explaining a first embodiment of the signal readout processing method according to the present invention. In the figure, l is a solid-state image sensor, which includes a light receiving section 2 configured by arranging internally amplified pixels such as SIT and CMD in a matrix, and a vertical section for selecting horizontal pixel rows of the light receiving section 2. It is composed of a scanning shift register 3 and a horizontal scanning shift register 5 that selects a vertical signal line of the light receiving section 2 via a horizontal scanning switch 4. Then, by high-speed horizontal scanning, a signal corresponding to the photocharges for one horizontal pixel column is read out in the first half of one horizontal scanning period, and in the second half of the horizontal scanning period, immediately after or after resetting the accumulated charges of one horizontal pixel column at the same time. However, it is configured to perform high-speed scanning again to read out signals for the same horizontal pixel column.

Reference numeral 6 denotes an amplifier, which amplifies the signal read out via the horizontal scanning switch 4 of the solid-state image sensor 1, and the signal amplified by the amplifier 6 is A/D converted by an A/D converter 7-. and is taken into the memory 8. Reference numerals 9 and 10 denote D/A converters for D/A converting a signal corresponding to a photocharge taken into the memory 8 and a signal in a state without photocharge, respectively; Each signal D/A converted in step 10 is input to a subtracter 11, where it is subtracted and output as an output video signal. Note that 12 is memory 8
This is a memory control unit that controls the operation of the memory.

Next, the operation of the signal readout processing device configured as described above will be explained. FIG. 2 is a timing chart showing signal waveforms of each part to explain this operation. First, the solid-state image sensor 1 is scanned horizontally at high speed, and a signal corresponding to photocharges for one horizontal pixel column is read out in the first half of one horizontal scanning period.

After reading out a signal corresponding to the photocharges for a horizontal pixel column, in the latter half of one horizontal scanning period, immediately after or while resetting the accumulated charges of the one horizontal pixel column,
High-speed horizontal scanning is performed again to read out signals for the same horizontal pixel column, which corresponds to reading out dark signals.

Internal amplification type image sensors generate FPN due to variations in the characteristics of the transistors that amplify the accumulated charge, but reading out the dark signal described above is based on the dark FP when the dark current is set to zero.
This also means reading out N (hereinafter used in this sense). Dark current has recently been suppressed to a very low level and can be ignored under normal solid-state imaging device usage conditions, so dark FPN is mostly caused by variations in transistor characteristics. . Dark FPN
exists at a constant level regardless of the amount of light, and is offset-wise superimposed on the signal corresponding to the first read photocharge.

As shown by the signal waveform A in FIG. 2, the photoelectric charge signals and the dark signal read out by the high-speed horizontal scanning are time-axis compressed and the H
It is read out as a line sequential signal A every /2. Note that FIG. 2 shows horizontal pixel column signals of m-1 rows, 1m rows, and 1m+1 rows. The line sequential signal A read out in this manner is amplified by an amplifier 6, converted into a digital signal by an A/D converter 7, and taken into a memory 8.

Next, time axis conversion is performed by controlling the phase of the memory read pulse by the memory control unit 12, and a time axis expansion operation is performed to return the time axis compressed signal due to the photocharge and the dark time signal to the normal time axis. At the same time, the two are read out from the memory 8 with their temporal phases aligned, and the respective D/A converters 9. A signal B and a dark signal C based on the photocharges are obtained by D/A conversion using lO and returned to the normal time axis. Note that the signals in the (m-1) row and m row in the photocharge signal B and the dark signal C returned to the normal time axis are (m-1)', respectively.
It is shown as m'. Next, the signal B due to the photocharge and the dark signal C are input to the subtracter 11, and the signal B due to the photocharge is input to the subtracter 11.
By further subtracting the dark signal C, a video signal with suppressed dark FPN can be obtained.

Note that in this embodiment, the signal read from the memory is
I have shown the one that performs calculation processing after conversion,
It is clear that similar effects can be obtained by performing D/A conversion after arithmetic processing.

FIG. 3 is a block configuration diagram showing a signal readout processing device for explaining a second embodiment of the present invention, and the same or equivalent members as shown in FIG. 1 are given the same reference numerals. It is shown as follows. FIG. 4 is a timing chart showing signal waveforms of various parts for explaining the operation of this second embodiment.

In this embodiment, similarly to the first embodiment, the time axis is compressed by high-speed horizontal scanning, and the signal due to the photocharge and the dark signal, which are read out as a line sequential signal A every H/2, are compressed on the time axis. Before conversion, subtraction processing is performed using an H/2 delay circuit 13 and a subtracter 11. That is, as in the first embodiment, the read line sequential signal A is delayed through the H/2 delay circuit 13 to form an H/2 delay line sequential signal.

Then, by inputting this H/2 delay delay line sequential signal type sequential signal A to the subtracter 11 and subtracting the H/2 delay delay line sequential signal type sequential signal A, the dark FPN is An FPN suppression signal -B-A) containing a signal due to the photocharge from which is subtracted is obtained.

This FPN suppression suppression signal-/D converter 7 A/D converts'
and import it into memory 8. Next, by controlling the phase of the memory read pulse by the memory control unit 12,
FPN Suppression Suppression Signal Only the signal due to the photocharge in the second half of the horizontal scanning period is read out from the memory 8 with the time axis compression returned to normal. Then, by D/A converting this signal with the D/A converter 14, an output video signal with suppressed dark FPN can be obtained.

In this embodiment, since the input signal to the A/D converter 7 does not include the dark FPN, there is an advantage that the dynamic range of the A/D converter 7 can be used effectively.

In the first and second embodiments described above, the time axis is compressed by high-speed horizontal scanning to obtain the line-sequential signal A, but since the number of horizontal pixels is originally small, the horizontal signal can be obtained even without high-speed horizontal scanning. For solid-state image sensors with a period shorter than H/2, the same horizontal pixel row is scanned twice at the normal scanning speed.
By reading the signal once, it corresponds to the line sequential signal A shown in FIGS. 2 and 4! a sequential signal can be obtained.

In this case, the time axis expansion operation at the time of memory read in each of the above embodiments is simply unnecessary, and the first and second
The FPN suppression processing method of the embodiment can be applied.

Furthermore, in the first and second embodiments described above, high-speed scanning is used to read out signals from the solid-state image sensor, but it is known that internally amplified solid-state image sensors have little deterioration in their output characteristics even with high-speed scanning. (Refer to Proceedings of the 1989 Electrical and Information Society Federation Conference 17-2 Amplified solid-state image sensor) For example, in a CMD image sensor, for a single boat readout frequency of 12 to 14 MHz for NTSC use,
It is said that high-speed scanning up to 60MHz is possible (Television Society Technical Report, Study on high-speed driving of CMD ITE
JTechnical Report νo1. IL
No. 28), there is no technical problem in the high-speed scanning in each of the above embodiments.

Further, in each of the above embodiments, an A/D converter.

Memory, D/A converter, etc. are configured as external circuits, but the video camera is digitized and functions.

Improving performance is common, and parts for these applications can be shared, so there is no cost problem when implementing it.

[Effects of the Invention] As described above based on the embodiments, according to the present invention, the dark FPN is suppressed in real time, and even if the dark FPN changes due to temperature changes, the change can be immediately followed. , a signal with stable dark FPN suppression can be obtained.

Furthermore, since there is no need to provide a line memory within the solid-state image sensing device, it is possible to prevent an increase in the chip area of the device due to the addition of a line memory, and to prevent new FPN from occurring due to variations in the capacity of the line memory.

[Brief explanation of drawings]

FIG. 1 is a block configuration diagram of a signal readout processing device for explaining a first embodiment of the signal readout processing method according to the present invention, and FIG. 2 is a timing chart showing signal waveforms for explaining its operation. , FIG. 3 is a block configuration diagram of a signal readout processing device for explaining the second embodiment,
FIG. 4 is a timing chart showing signal waveforms for explaining its operation, FIG. 5 is a block diagram showing a configuration example of a conventional FPN suppressing means, and FIG. 6 is a timing chart showing a conventional FPN suppressing means.
FIG. 7 is a block configuration diagram showing another example of the configuration of the suppression means. In the figure, 1 is a solid-state image sensor, 2 is a light receiving section, 3 is a vertical scanning shift register, 4 is a horizontal scanning switch, 5 is a horizontal scanning shift register, 6 is an amplifier, 7 is an A/D converter,
8 is a memory, 9.10 is a D/A converter, 11 is a subtracter,
12 is a memory control unit, 13 is an H/2 delay circuit, and 14 is a D
/A converter is shown. Patent applicant: Olympus Optical Industry Co., Ltd. Figure 1 (m-1)'m' Figure 5 Figure 6 (111 Solid-state image sensor

Claims (1)

[Claims] 1. In a signal readout processing method for a solid-state image sensor in which an internally amplified photoelectric conversion element is used as a unit pixel and the unit pixels are arranged in matrix, After reading out the signal of one horizontal pixel column, immediately after or while resetting the accumulated charge of the one horizontal pixel column, in the latter half of one horizontal scanning period, the one horizontal pixel column is read again in a state where there is no photocharge. A signal readout processing method for a solid-state image sensor characterized by reading out a signal, aligning the timing of both readout signals using an external circuit, and then subtracting a signal in a state where there is no photocharge from a signal corresponding to the photocharge. 2. The readout of the signal corresponding to the photocharge of one horizontal pixel row and the readout of the signal in the absence of photocharge are performed by compressing the time axis by high-speed scanning, and are performed before or after the subtraction process by an external circuit. 2. The signal readout processing method for a solid-state image sensor according to claim 1, wherein a time axis extension operation is performed.