KR101639385B1 - Image sensor - Google Patents

Abstract

The image sensor has multiple exposure times to acquire a high-sensitivity image. The image sensor includes a pixel array including a plurality of pixels arrayed in a matrix and a driving signal providing section for applying driving signals for driving each pixel of the pixel array. In the pixel array, two or more kinds of pixels whose sensitivity are different from each other are arranged in one row. The different kinds of pixels are configured to receive a transfer signal applied to a transfer transistor for transferring light charges accumulated in the photodiode from different signal lines to the floating node. The driving signal supply unit applies different transmission timings to the pixels included in one row at different timings so that two or more different kinds of pixels have different light exposure time.

Image sensor, multiple exposure, binning, resolution, pixel

Description

Image sensor

More particularly, to an image sensor for acquiring a high-sensitivity image.

Currently, researches are actively conducted to increase the sensitivity characteristics of image sensors. Among various methods, there is a method of arranging white or transparent pixels in addition to the existing R, G, and B pixels to have high sensitivity characteristics. Gray pixel information or white pixel information, and outputs gray information having a sensitivity characteristic of about 3 to 5 times that of the conventional R, G, and B pixels, and color information is generated by interpolating information of neighboring R, G, and B pixels I will.

The sensor of this structure has a problem in selecting a proper exposure time because the sensitivity difference between the R, G, and B pixels and the high sensitivity pixel is very remarkable. If the exposure time is adjusted according to the R, G, and B pixel output values and the exposure time is relatively long, the high-sensitivity pixels become saturated and the image information is lost, and exposure to the high- When the time is adjusted, the exposure time is relatively short, and the amount of light that can be received by the R, G, and B pixels becomes small. As a result, the SNR (Signal-to-Noise Ratio) of the image is remarkably deteriorated.

On the other hand, Binning technique is widely used for high sensitivity sensors. The binning method basically combines the values of adjacent pixels to obtain a high-quality, high-sensitivity image. The binning operation can be realized by combining the signal value of the adjacent pixel with the accumulated charge in the pixel or by combining the analog output value of the pixel in the column circuit. Or binning may be performed in the digital image processing unit. This conventional binning technique can increase the sensitivity effectively but it has a disadvantage that the spatial resolution is reduced because it is commonly applied to the entire image.

There is provided an image sensor structure for differently controlling the exposure time for each kind of pixel when there are pixels having two or more different sensitivities.

There is provided an image sensor structure for performing local binning while differently controlling the exposure time for each kind of pixel when there are two or more pixels having different sensitivities.

An image sensor according to one aspect includes a pixel array including a plurality of pixels arrayed in a matrix and a drive signal providing section for applying a drive signal for driving each pixel of the pixel array. In the pixel array, two or more kinds of pixels having sensitivity different from each other are arranged in one row. The different kinds of pixels are configured to receive a transfer signal applied to a transfer transistor for transferring light charges accumulated in the photodiode from different signal lines to the floating node. The driving signal supply unit applies different transmission timings to the pixels included in one row at different timings so that two or more different kinds of pixels have different light exposure time.

When two or more pixels having different sensitivities are present, the exposure time may be adjusted differently for each type of pixel to improve the sensitivity without loss of image information or SNR. Also, only a pixel region that can not acquire image information due to its low illuminance acquires image information at sacrificing spatial resolution, and other pixel regions can obtain a high-resolution and high-quality image without sacrificing spatial resolution.

Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings. In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear. In addition, the terms described below are defined in consideration of the functions of the present invention, which may vary depending on the intention of the user, the operator, or the custom. Therefore, the definition should be based on the contents throughout this specification.

1 is a view showing an example of a sensor structure for multiple exposure.

FIG. 1 shows an example of an 8 × 8 array sensor structure having two types of exposure time. The image sensor 100 may include a pixel array 110 and a driving signal providing unit 120, a column lead-out circuit 130, and a column decoder 140.

The pixel array 110 converts a light signal into an analog signal and includes a plurality of unit pixels arranged in a matrix form.

The driving signal providing unit 120 selects a row line of the pixel array 110 and generates driving signals for driving each pixel of the pixel array 110 and applies the driving signals to each pixel.

The column lead-out circuit 130 reads a signal from the pixel array 110.

The column decoder 140 decodes the address signals in the column direction and sequentially outputs the signals read out from the column readout circuit 230. Here, the column indicates the direction in which the pixels are read, and does not mean an array in the vertical direction.

The pixel array 110 is configured such that two or more kinds of pixels having different sensitivities are arranged in one row. In Fig. 1, it is assumed that a pixel displayed in gray represents a low-sensitivity pixel and a white pixel is a high-sensitivity pixel. The low-sensitivity pixel may be a color pixel, and the high-sensitivity pixel may be a transparent pixel. High-sensitivity pixels saturate in a shorter time than low-sensitivity pixels.

The general image sensor uses one transmission signal TX for controlling the transfer transistor, one for each row, to perform the operation of reading out or resetting the photodiode signal. In contrast, pixels of different types in the pixel array 110 are configured to receive transmission signals applied to a transmission transistor that transfers light charges accumulated in the photodiodes from different signal lines to the floating node. That is, one row includes signal lines for applying two or more transmission signals TX.

The driving signal supply unit 120 is configured to apply different transmission timings to pixels included in one row so that two or more different kinds of pixels have different light exposure times. In Fig. 1, two signals TX1 and TX2 are arranged per one row. In order to have more kinds of exposure times, lines may be arranged in which TX signals are transmitted for a number of different exposure times.

In the column readout circuit 130, column readers 131 to 138 are arranged for each column of the pixel array 110. The column readers 131 to 138 usually have an analog circuit and an analog to digital converter (ADC) for removing noise. For example, when a pixel signal of one row enters the input of each of the column readers 131 to 138, the noise is removed first through the noise elimination circuit, and then the digital signal is converted into a digital value through the ADC. do. The operation up to this point is column parallel operation. Then, when a column decoded signal sequentially applied to each column is input, the corresponding pixel output is output to the sensor output one by one.

2 is a table showing the operation of pixels according to the sensor structure of FIG.

It is assumed that the operation of the image sensor 100 starts from the first row, and the operation of the row 8 is the last. It is assumed that the color pixel (low-sensitivity pixel) has an exposure time of 4 line readout time and that the exposure time of the transparent pixel (high-sensitivity pixel) is 2 line lead-out time. Here, the line lead-out time has the same meaning as the row lead-out time for reading the pixel value from the pixels included in one row.

When the image sensor 100 starts to operate, it first starts a reset operation which empties the photodiode before the start of the integration of the color pixels. In the image sensor 100, the operation of each pixel is the same as the operation of a typical four-transistor CMOS image sensor, except that there is a separate transmission line to be input to each of the transmission pixels of the color pixel and the transparent pixel in one row .

The gate signal TX12 of the transfer transistor included in the color pixels of the first row is "ON" for resetting the photodiode included in the color pixels of the first row in the time "1 ". The color pixels of the reset first row are exposed and light begins to intergrate.

At time "2 ", the gate signal TX22 of the transfer transistor included in the color pixels of the second row is" ON "for resetting the photodiode included in the color pixels of the second row. The color pixels of the second row that are reset start the integration.

At time "3 ", the gate signal (TX32) signal of the transfer transistors of the color pixels of the third row is" ON "for resetting the photodiode included in the color pixels of the third row. Further, at time "3 ", the gate signal TX11 of the transfer transistors in the first row is" ON "for resetting the photodiode included in the transparent pixels in the first row. Thus, after the photodiodes included in the color pixels of the third row and the transparent pixels of the first row are reset, the integration starts.

At time "4 ", the gate signal TX42 of the transfer transistors of the color pixels of the fourth row is" ON "for resetting the photodiode contained in the color pixels of the fourth row, The gate signal TX21 of the transfer transistors of the transparent pixels of the second row is "ON" to reset the photodiodes included in the second row of transparent pixels, The photodiodes are reset. Then, the color pixels of the fourth row and the transparent pixels of the second row begin their integration.

At time "5 ", the gate signal TX52 of the transfer transistors of the color pixels of the fifth row is" ON "for resetting the photodiode contained in the color pixels of the fifth row, The gate signal TX31 of the transfer transistors in the third row is "ON ". The color pixels of the reset fifth row and the transparent pixels of the third row start the integration. On the other hand, at this time, in order to lead out all the pixels of the first row, the TX11 and TX12 signals of the first row are all turned ON and the pixel values of all the pixels of the first row are read out.

Since the color pixels in the first row are reset at time 1 and read at time 5, the exposure time is the 4-line readout time, and the transparent pixels in the first row start to reset at time 3, So that the exposure time becomes the 2-line lead-out time. In the same way, when the operation is finished to the time "12 ", the values of all the pixels are integrated for the exposure time suited to them, and the pixel value can be read.

FIGS. 3A to 3F are diagrams showing the operation of all the pixels in the second row, the transparent pixels in the fourth row, and the color pixels in the sixth row at the time "6" time in the operation of FIG.

3A and 3B, for all the pixels in the second row, the RST signal is turned ON in the RST READ period, and the reset signal of the pixel is read. Here, the reset signal resets the floating diffusion FD to the operating voltage Vdd during the intergration time during which the light enters the photodiode PD, thereby resetting the floating diffusion FD to the reset voltage V _R , And means a signal corresponding to the reset voltage V _R of the floating diffusion.

Next, in the TX period, the gate signal TX22 of the transfer transistor T of the color pixel and the gate signal TX21 of the transfer transistor T of the transparent pixel are turned ON and the SEL signal is turned ON , The photoelectric charge accumulated in the photodiode PD is transferred to the floating diffusion FD and outputted as the signal voltage V _sig through the photoelectric transferring source follower transistor SF. When a signal corresponding to the signal voltage V _sig is an image signal, a reset voltage (V _R ) reset signal and a signal voltage V (V) are reset to remove a reset noise, a fixed pattern noise _sig ) is used as a final video signal. The amount of charge accumulated during different times for each pixel type is read through the column bus.

3C and 3D, in the reset period, the gate signal TX41 of the transfer transistor T of the transparent pixels of the fourth row is turned ON and the gate signal RST of the reset transistor R is also Quot; ON ", so that the charge in the photodiode of the transparent pixels is discharged toward the driving voltage Vdd, and the photodiode is reset.

3E and 3F, in the reset period, the gate signal TX62 of the transfer transistor T of the color pixels of the sixth row is turned ON and the gate signal RST of the reset transistor R is also " ON ", so that the charge in the photodiode of the color pixels is discharged toward the driving voltage Vdd, and the photodiode is reset.

All the remaining pixels continue the photocharge accumulation operation without changing the external control signal, that is, the gate signal of the transfer transistor, the gate signal RST of the reset transistor, and the gate signal SEL of the selection transistor.

4 is a view showing another example of the sensor structure for multiple exposure.

FIG. 4 shows an example of an 8 × 8 array sensor structure having three types of exposure time. The image sensor 400 includes a pixel array 410 including a plurality of pixels arrayed in a matrix, a driving signal supplier 420 for applying driving signals for driving each pixel of the pixel array, a column leadout circuit 430, And a column decoder 440.

In FIG. 4, the pixel array 410 includes A type, B type, and C type pixels. The sensitivity assumes that the C type pixel is the highest, the A type pixel is the lowest, and the B type pixel has the intermediate sensitivity between the C type pixel and the A type pixel. The A-type pixel, the B-type pixel, and the C-type pixel included in one row in the pixel array 410 transmit the transfer signal applied to the transfer transistor for transferring the light charge accumulated in the photodiode from the different signal line to the floating node And receive. In Fig. 4, signal lines for applying three transmission signals are included in one row.

The driving signal supplying unit 120 supplies driving signals to the pixels included in one row at different timings so that different three kinds of pixels have different light exposure times.

The column lead-out circuit 430 and the column decoder 440 perform the same functions as the components of the image sensor shown in FIG. 1 of the same name, so that detailed description is omitted.

5 is a table showing the operation of pixels according to the sensor structure of FIG.

It is assumed that the operation of the image sensor 100 starts from the first row, and the operation of the row 8 is the last. It is assumed that the exposure time of the A type pixel is the 3 line readout time, the exposure time of the B type pixel is the 2 line readout time, and the exposure time of the C type pixel is one line lead-out time.

When the image sensor 100 starts to operate, it first starts a reset operation which empties the photodiode before the start of the integration of the color pixels. At time "1 ", the TX111 signal of the first row is turned ON to reset the photodiodes of the A type pixels of the first row. The integration begins for the A type pixels in the reset first row.

At time "2 ", the TX121 signal becomes" ON "for the photodiode reset of the A type pixels of the second row. Further, at time "2 ", the TX112 signal of the first row becomes" ON "in order to reset the photodiodes of the B type pixels of the first row. The initialization of the A type pixels of the reset second row and the B type pixels of the first row are started.

At time "3 ", the TX131 signal becomes" ON "for the photodiode reset of the A type pixels of the third row. Further, at time "3 ", the TX122 signal of the second row becomes" ON "in order to reset the photodiodes of the B type pixels of the second row. At time "3 ", the TX113 signal of the first row is turned" ON " in order to reset the photodiodes of the C type pixels of the first row. The A type pixels of the reset third row, the B type pixels of the second row, and the C type pixels of the first row are started.

At time "4 ", the TX 141 signal is" ON "for photodiode reset of the A type pixels of the fourth row. At time "4 ", the TX132 signal of the second row is turned" ON " in order to reset the photodiodes of the B type pixels of the third row. At time "4 ", the TX123 signal of the second row is turned" ON " in order to reset the photodiodes of the C type pixels of the second row. The A type pixels of the reset fourth row, the B type pixels of the third row, and the C type pixels of the second row are started.

On the other hand, at time "4 ", the TX111, TX112 and TX113 signals of the first row are all turned" ON " and the pixel value of each pixel is read out to lead out all the pixels of the first row.

Since the A type pixels in the first row start to reset at time 1 and read at time 4, the exposure time becomes 3 line readout time, the B type pixels of the second row start reset at time 2, The exposure time is the 2-line readout time, the C-type pixels of the third row start to reset at time 3, and are read at time 4, so that the exposure time becomes one line lead-out time. In this way, when the operation is completed to the time "11 ", the values of all the pixels are integrated for the exposure time suited to itself, and the pixel value corresponding to the accumulated amount of light charge during the exposure time can be read .

6 is a diagram illustrating a concept of a local binning method according to an exemplary embodiment.

A local binning method according to an embodiment acquires image information at a sacrifice of spatial resolution only in a pixel region in which image information can not be acquired due to low illumination, and other pixel regions acquire image information in a high resolution high- It is a way to obtain images.

In order to perform local binning only for the low-luminance pixel region, it is necessary to determine which pixel is a low-luminance pixel region. The determination criterion is that the output value of each pixel is checked every frame, and if the output value is too small to obtain a high-quality image, it is determined as a low-illuminance pixel.

Here, the frame means an image in which the pixel values of the pixel array are output as a whole. Also, the low-illuminance means that the amount of received light is smaller than a predetermined threshold value. For example, when the output value of a predetermined pixel has an output value that is lower than a value that is less than or equal to a predetermined pixel value than the output value of the surrounding pixel region or the output value of the surrounding pixel region, it can be determined as a low-luminance region.

Based on the result of the determination, a low-illuminance map is created in which low-illuminance pixels are recorded. Then, when the image of the next frame is outputted, when each pixel is displayed as a low-illuminance pixel in the previous frame, binning is performed, the low-illuminance pixel area is determined again based on the output image, and the determination result is updated to the low illuminance map . So that the binning operation in the next frame again can be controlled according to the updated low-illuminance map. This operation is repeatedly performed every frame.

On the other hand, the low-illuminance map need not be generated by determining whether the pixel value of all the pixels of the pixel array is smaller than a predetermined threshold value in pixel units. That is, an image is generated at a low resolution using pixel values of some pixels in the pixel array, and a low-illuminance low-resolution map is generated using the generated image. Then, a high- Binning can be performed.

Referring to FIG. 6, a first image frame output by the light incident on the pixel array 1 arranged in a matrix form is acquired (10), low-illuminance region information is generated in the first image frame, Area information is recorded (20).

A second image frame is obtained (30), a binarization is performed on an area corresponding to the low-illuminance area determined from the first image frame with respect to the second image frame, and a second image frame where local binning is performed only for the low- (40). Based on the obtained second image frame, a low-illuminance area to be applied to the next input image frame is determined and recorded (50).

A third image frame is acquired (60), a binning is performed for an area corresponding to the low-illuminance area determined from the second image frame with respect to the third image frame, and a third input image subjected to local binning only for the low- (70). Based on the obtained third input image, a low-illuminance area to be applied to the next input image frame is determined and recorded (80). A local binning technique in accordance with one embodiment includes both column lead-out circuits or methods of locally binning in a digital domain.

Conventional global binning increases sensitivity but has a large loss of spatial resolution, whereas local binning according to one embodiment results in loss of spatial resolution only in low-light areas and no loss in general light areas. In addition, since only the sensitivity of the low-illuminance region is increased, an effect of expanding the dynamic range to the low-illuminance region can be obtained.

 7 is a view showing a configuration of a column reading section according to an embodiment.

First, referring to FIG. 1, the column lead-out circuit 130 may further perform an operation for local binning in addition to reading a signal from the pixel array 110. The column lead-out circuit 130 generates low-luminance area information indicating a low-luminance area for the input image frame, and generates a low-luminance area for the next image frame, which is determined in accordance with the low- So that the operation is performed.

The low-illuminance region information may be stored in a plurality of unit pixels as low-illuminance information indicating whether a video signal of a previous video frame corresponding to all or a part of a plurality of unit pixels indicates low light intensity. The low-luminance region information does not need to be generated for each unit pixel, and thus can be generated as a pixel value of some unit pixel. Here, the low-illuminance information means information to be recorded in a pixel as information indicating whether one pixel is low-illuminance, and the low-illuminance region information means information indicating which region is low-illuminated in one frame. To this end, each of the column readers 131 to 138 further performs an operation of reading out low-illuminance information from each pixel and storing low-illuminance information in each pixel, in addition to reading out a video signal of a normal pixel.

The low-illuminance information indicating whether the corresponding pixel of the previous image frame is low-light is first applied to the binarizing determiner 711. [ The low-illuminance information indicating whether the corresponding pixel of the previous frame is low-illuminance may be stored in the corresponding pixel, or may be stored in a separate memory inside or outside the image sensor 100. [ Accordingly, the binarizing unit 711 receives low-illuminance information from the corresponding pixel or from a separate memory, and determines that the low-illuminance information is to perform a binning operation when the low-illuminance information indicates that the corresponding pixel is low- It is determined that the binning operation is not performed. According to one embodiment, the binning determination unit 711 may receive the image signals (or pixel values) from the pixel array 110 and perform binning directly in the analog domain.

When the pixel array 110 uses transparent pixels and color pixels to obtain a high-sensitivity image signal, the binning operation can be performed between pixels of the same color as the surrounding columns. Here, the binning operation of the pixels of the same color includes the binning of the transparent pixels of the neighboring columns. For this, the binarizing unit 711 can receive an image signal output from a pixel of the same color in a peripheral column when a predetermined column pixel is a pixel in a low-illuminance region, and perform an analog binning operation. The binning determination unit 711 may be implemented in various manners as long as it can determine whether or not to perform the binning operation according to the output values of the pixels of the corresponding columns.

The binning is performed in the binning determining unit 711, or the pixel signal not performed is transmitted to the noise removing unit 712. The noise removing unit 712 performs CDS (Correlated Double Sampling). More specifically, the noise remover 712 removes noise by performing a difference between a reset signal resulting from the reset of the pixel and a video signal according to the amount of charge generated in accordance with the photocharge accumulation in the pixel, To the ADC unit 313. Here, the reset of the pixel is distinguished from the reset of the photodiode described above, and means the operation of outputting the reset voltage of the floating node to remove the FPN. The image signal transmitted from the ADC unit 713 is digitally converted, and the digitally converted signal is stored in the latch 714.

The low-illuminance determination unit 715 determines whether the video signal read out based on the video signal stored in the latch 714 is low-illuminance, and the low-illuminance information update unit 716 updates the low-illuminance information using the determination result. The low-illuminance information can be stored in the analog memory of the pixel from which the video signal is read. For example, the low-illuminance determination unit 715 can record a "High" voltage when the pixel is a low-illuminance pixel and a "Low" voltage when the pixel is a general illuminated pixel. The above operation is done column-parallel.

Although the low-illuminance information is stored in the pixels, the low-illuminance information may be stored in a separate frame memory (not shown), which may be included in the image sensor 100, And may be stored in a memory (not shown) for utilization.

8 is a diagram illustrating a pixel structure capable of analog local binning according to one embodiment.

The pixel 800 includes a photodiode PD, a transfer transistor T, a floating diffusion node capacitor C _FD , a reset transistor R, a source follower transistor SF and a selection transistor S.

The photodiode PD receives light to generate a photocharge.

The transfer transistor T is connected to the photodiode PD and transfers the photocharge to the floating diffusion node FD in accordance with the signal TX.

The floating diffusion node capacitor (C _FD ) accumulates the light charge transferred through the transfer transistor (T). The low-illuminance information according to one embodiment may be stored in a floating diffusion node capacitor (C _FD ) of a floating diffusion (FD) node. The floating diffusion node capacitor C _FD may be formed by adding a capacitor to the FD node or may use a parasitic capacitor of the FD node. The use of parasitic capacitors eliminates the need for additional hardware.

The reset transistor R discharges the photoelectric charge accumulated in the floating diffusion node FD.

The source follower transistor SF amplifies the _photocharge in the floating diffusion node capacitor C _FD to a voltage signal.

The selection transistor S selectively outputs the output voltage in accordance with the selection signal SEL.

The pixel 400 according to one embodiment differs from the general 4Tr structure in that the drain node of the RST transistor is connected to the RST_col signal line without being connected to the power supply (V _DD ). The voltage of the line RST_col varies according to the operation order. RST_col line has a voltage applied to the reset of the photodiode (for example, V _DD), the voltage (V _HIGH) or a voltage representing the low-illuminance information indicating that it is not a low luminance (V _LOW) that corresponds to the low-illuminance information indicating that the low-illuminance is .

The low-intensity information indicating whether the value of the corresponding unit pixel in the previous image frame is low is temporarily stored in the floating diffusion node capacitor C _FD at a time different from the time when the photoelectric charge of the photodiode PD is transferred, Can be output at a time different from the time at which the video signal is output.

FIG. 9 is a timing diagram illustrating an operation method of a pixel capable of analog local binning of FIG. 8. FIG.

Unlike general sensor operation, the pixel operation method according to one embodiment requires a C _FD READ section for reading low-illuminance information and a C _FD WRITE section for recording low-illuminance information.

It is assumed that information indicating whether the image signal of the pixel 800 in the previous image frame is low is stored in the FD capacitor C _FD . By the signal RST of the reset transistor (R) to "ON" and read the reset signal of the pixel 800, the old and the select transistor (S) to "ON" in the C _FD READ intervals stored in the FD capacitor (C _FD) The low-illuminance information indicating whether the video signal of the pixel 400 in the image frame is low-illuminance is read.

RST READ, Signal READ, and CDS & ADC sections are performed in the same manner as normal image sensor operation. Specifically, RST_col is set to "ON" and the gate signal RST of the reset transistor R is turned "ON" to output a reset signal. Then, in the signal READ period, the gate signal TX of the transfer transistor T is turned "ON ", and a video signal due to the photoelectric charge accumulated by the light is outputted.

In the CDS & ACD interval, the binarization information is received by the binning determination unit 711 of the column reading unit 131. If the pixel is a low-illuminance pixel, an analog binning operation is performed using the image signal read from peripheral pixels of the same row. To the noise removing unit 712 without binning. The noise eliminator 712 removes noise by using the reset signal and the video signal, converts the noise-eliminated video signal into the digital signal by the ADC 713, and stores the digitally-converted video signal in the latch 714 . The digitally converted image signal is output by the column decoder 240.

The digitally converted image signal is input to the low-illuminance information determination unit 715, and it is determined whether or not the image signal is low-illuminance. The low-illuminance information updating unit 716 updates the low-illuminance information according to whether the low-illuminance is determined, and records the updated low-illuminance information in the pixel analog memory as the RST_col signal in the C _FD WRITE period. The signal RST is also "ON" in the C _FD WRITE period, and the RST_col signal is transmitted to the floating diffusion node FD. The signal level of the signal RST_col in the C _FD WRITE period can be adjusted depending on whether the read video signal is low level. The above operation is repeated.

That is, the low-illuminance information indicating whether the video signal for the previous image frame stored in the floating diffusion node FD is low-level is obtained before the reset transistor R is turned on and the reset signal is outputted in the state that the transfer transistor T is off After the transfer transistor T and the selection transistor S are turned on to output a video signal corresponding to the accumulation of the photocharge of the photodiode with respect to the current video frame, the video signal for the current video frame is read out from the floating diffusion node, Low-illuminance information indicative of whether or not low-illuminance is recorded.

Figs. 10A to 10F are diagrams for explaining the operations of all the pixels in the second row, the transparent pixels in the fourth row, and the color pixels in the sixth row at time "6" Fig.

In the local binning operation, since the pixels have the low-illuminance information of the corresponding pixel of the previous frame in the FD node, even if the reset transistor and the transfer transistor of the pixel are turned on to emit the light charge accumulated in the photodiode before the integration period, And operates to maintain the low-illuminance information of the previous frame stored in the FD node. To this end, the low-illuminance information is read and temporarily stored before the photodiode reset of the pixel, and the low-illuminance information temporarily stored after the photodiode reset of the pixel is stored again in the FD node. Here, in order to temporarily store the low-illuminance information, a storage means such as a capacitor included in the binning determination unit 711 of the column reading unit 131 in Fig. 7 may be used.

Also, as described with reference to Fig. 9, when the signal corresponding to the accumulated light charge during the pixel's integration period is read out, the low-illuminance information of the previous frame of the FD node is read out before the signal is read out, After the signal is read out, the operation of storing the low-illuminance information of the current frame in the corresponding pixels is performed. After the reset and the readout are completed, the operation of writing the C _FD information again is necessary. Other operations are the same as for a general 4T-based pixel.

10A and 10B are operation timing diagrams of pixels and pixels showing a pixel lead-out operation for all the pixels in the second row.

During the C _FD READ0 period, the SEL signal is turned "ON ", and the low-luminance information of the previous frame stored in each of the color pixels and transparent pixels in the second row is output. When the SEL signal is "ON" in the RST READ period, the RST signal is turned "ON" and the reset signal of the pixel is read. In the TX period, the gate signal TX21 of the color pixel of the transfer transistor T is turned ON and the SEL signal is held "ON " to convert the photocharge accumulated in the photodiode PD into a floating diffusion FD, and is output as the signal voltage V _sig corresponding to the image signal via the photoelectric transferring source follower transistor SF. The operation of the CDS & ACD section is as described with reference to FIG. With respect to the transparent pixels of the second row, the same operation as the color pixels of the second row is performed except that the gate signal TX22 of the transfer transistor T is turned "ON " in the TX period.

10C and 10D are operation timing diagrams of pixels and pixels showing a photodiode reset operation of pixels for transparent pixels in the fourth row.

Referring to FIGS. 10C and 10D, in the C _FD READ 2 period before the start of the CDS & ADC period of the readout operation of the pixels included in the second row, the SEL signal is turned ON, The low-illuminance information of the stored previous frame is output and temporarily stored. Then, in the fourth row of the transparent pixels, the TX41 signal is turned ON, the RST signal and the RST_col signal are also turned ON, the charge in the photodiode is discharged toward the drive voltage RST_col, and the photodiode is reset . Then, the low-luminance information of the previous frame temporarily stored in the C _FD WRITE 2 period is written back to the capacitor (C _FD ) of the FD node.

10E and 10F are operation timing diagrams of pixels and pixels showing a photodiode reset operation of a pixel for color pixels in a sixth row.

10E and 10F, in the C _FD READ 2 period before the start of the CDS & ADC period of the readout operation of the pixels included in the second row, the SEL signal is turned "ON " The low-illuminance information of the stored previous frame is output and temporarily stored. Then, in the sixth row of the color pixels, the TX62 signal is turned ON, the RST signal and the RST_col signal are also turned ON, the charge in the photodiode is discharged toward the drive voltage RST_col, and the photodiode is reset . Then, the low-luminance information of the previous frame temporarily stored in the C _FD WRITE 2 period is written back to the capacitor (C _FD ) of the FD node.

All the remaining pixels continue the optical accumulation operation without changing the external control signals TX1, TX2, RST, and SEL.

Figure 11 is a timing diagram illustrating the operation of pixels included in the first row for multiple exposure and local binning in the image sensor structure of Figure 1;

FIG. 11 is a diagram showing the relationship between the color pixels of the first row and the pixels of the transparent pixels when the exposure time of the color pixel is 2 line lead-out time and the exposure time of the transparent pixel is 1 line lead- Fig. 5 is a timing chart showing the operation until the diode is reset, accumulates photo charges during the integration period, and all the pixels in the first row are read out.

The phototiode reset operation of the color pixels is performed in the "1" period, the photodiode reset operation of the transparent pixels is performed in the "2" period, And a signal corresponding to the light charge accumulated in the transparent pixels during the integration period.

With this operating method, the high-sensitivity pixel and the low-sensitivity pixel can have an exposure time optimized for each pixel type and can perform local binning. Such multiple exposure and local binning operations are applied to a sensor composed of several kinds of pixels having different sensitivity characteristics, so that a high-quality image can be obtained.

One aspect of the present invention may be embodied as computer readable code on a computer readable recording medium. The code and code segments implementing the above program can be easily deduced by a computer programmer in the field. A computer-readable recording medium includes all kinds of recording apparatuses in which data that can be read by a computer system is stored. Examples of the computer-readable recording medium include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical disk, and the like. The computer-readable recording medium may also be distributed over a networked computer system and stored and executed in computer readable code in a distributed manner.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the invention. Therefore, the scope of the present invention should not be limited to the above-described embodiments, but should be construed to include various embodiments within the scope of the claims.

1 is a view showing an example of a sensor structure for multiple exposure.

2 is a table showing the operation of pixels according to the sensor structure of FIG.

FIGS. 3A to 3F are diagrams showing the operation of all the pixels in the second row, the transparent pixels in the fourth row, and the color pixels in the sixth row at the time "6" time in the operation of FIG.

4 is a view showing another example of the sensor structure for multiple exposure.

5 is a table showing the operation of pixels according to the sensor structure of FIG.

6 is a diagram illustrating a concept of a local binning method according to an exemplary embodiment.

7 is a view showing a configuration of a column reading section according to an embodiment.

8 is a diagram illustrating a pixel structure capable of analog local binning according to one embodiment.

FIG. 9 is a timing diagram illustrating an operation method of a pixel capable of analog local binning of FIG. 8. FIG.

Figs. 10A to 10F are diagrams for explaining the operations of all the pixels in the second row, the transparent pixels in the fourth row, and the color pixels in the sixth row at time "6" Fig.

Figure 11 is a timing diagram illustrating the operation of pixels included in the first row for multiple exposure and local binning in the image sensor structure of Figure 1;

Claims (13)

A pixel array including a plurality of pixels arranged in a matrix; A driving signal supplier for applying a driving signal for driving each pixel of the pixel array; And And a column readout circuit configured for each column of the plurality of pixels, the column readout circuit receiving a video frame from the pixel array, Two or more kinds of pixels having different sensitivities are arranged in one row in the pixel array and the different kinds of pixels are connected to a transfer transistor for transferring light charges accumulated in the photodiode from different signal lines to a floating node And to receive the transmission signal, Wherein the drive signal provider comprises: By applying transmission signals received from the different signal lines at different timings to pixels included in one row, the different two or more kinds of pixels have different light exposure time, The column lead-out circuit generates low-illuminance area information representing a low-illuminance area for the input image frame. When the next image frame input in the next frame period is input, And an in-binning operation is performed. The method according to claim 1, Wherein the different kinds of pixels include transparent pixels and color pixels. 3. The method of claim 2, Wherein the drive signal provider comprises: And applies the transmission signal so that the light exposure time of the transparent pixel is shorter than the light exposure time of the color pixel when the transmission signal applied from the different signal line is applied at a different timing. delete The method according to claim 1, The low-illuminance region information for the image frame is information indicating whether a video signal of a previous video frame corresponding to all or a part of the plurality of pixels indicates low light intensity. In each pixel, And the low-illuminance information is read out from the plurality of pixels by the column reading unit. The method according to claim 1, Wherein the column lead-out circuit performs a local binning operation on the low-luminance region in an analog domain using an image signal for the next image frame. The method according to claim 1, The column reading unit includes: Determining whether a video signal of another pixel should be binarized with respect to the video signal output from the pixel based on the low-illuminance information read from the pixel, and determining whether to binarize the video signal read from another pixel of the same row A binning determining unit for binning the image using the image; A noise removing unit for removing noise from the video signal; An analog digital converter for converting the noise-eliminated video signal into a digital signal; A latch for storing the video signal converted into the digital signal; A low-illuminance determination unit for determining whether a video signal output from the pixel indicates low illumination using the video signal converted into the digital signal; And And a low-illuminance information updating unit for updating the previously stored low-illuminance information according to whether the low-illuminance is judged. 8. The method of claim 7, Wherein the updated low-illuminance information is stored in the pixel, and is used by the binning determination unit to determine whether or not to bin the image signal of the next image frame in the next image frame processing period. The method according to claim 1, Wherein the pixel comprises: A photodiode for receiving light to generate photocharge; A transfer transistor connected to the photodiode and for transferring the photocharge; A floating diffusion node through which the photocharge is transferred through the transfer transistor; A reset transistor for discharging the photoelectric charge accumulated in the floating diffusion node; A source follower transistor for amplifying and converting the photocharge in the floating diffusion node into a voltage signal; And And a selection transistor for selectively outputting a signal of the pixel, Wherein the drain terminal of the reset transistor is connected to a signal line that delivers low illumination information indicating whether the value output from the photodiode in the column readout in the previous frame is low. 10. The method of claim 9, Wherein the low-illuminance information is stored in a capacitor of the floating diffusion node. 10. The method of claim 9, The low-illuminance information indicating whether a video signal for a previous image frame stored in the floating diffusion node is low-level is generated by turning on the reset transistor and the transfer transistor to emit light accumulated in the photodiode before the start of the integration period Is held in the photodiode. 12. The method of claim 11, In order to maintain the low-illuminance information in the photodiode after the operation of emptying the photo-charges accumulated in the photodiode before the start of the integration period, the low-illuminance information is read from the floating diffusion node before the start of the integration period And, Wherein the readout low-illuminance information is written back to the floating diffusion node at a predetermined time while the integration is started and light charges are accumulated in the photodiode. 12. The method of claim 11, The low-illuminance information indicating whether the image signal for the previous image frame stored in the floating diffusion node is low-level is read from the floating diffusion node before the reset transistor is turned on and the reset signal is output in a state where the transfer transistor is off , And the selection transistor is turned on to output the image signal corresponding to the photocharge accumulation of the photodiode with respect to the current image frame, the low-illuminance information indicating whether the image signal for the current image frame is low- Is written to the floating diffusion node.

Images