KR19980041905A - How to increase the dynamic range of an active pixel sensor cell - Google Patents

Abstract

The dynamic range of an image cell is obtained by reading the image cell several times during each integration period and obtaining the read multiple values. When the last read out of a plurality of readings indicates that the cell is saturated, the readout value before saturation is used to extrapolate the final value, which would be the appropriate value that the image cell would have if the image cell had not saturated Loses.

Description

How to increase the dynamic range of an active pixel sensor cell

The present invention relates to a method of increasing the dynamic range of an active pixel sensor cell, particularly an active pixel sensor cell.

Charge-coupled devices (CCDs) have been a key part of conventional imaging circuits for converting light energy pixels into electrical signals that represent the intensity of light energy. In general, a charge coupled device uses a photonic gate to convert light energy into an electrical charge, and a series of electrodes to transfer the charge collected at the gate to the output sense node.

Although charge-coupled devices have many advantages, including high sensitivity and fill-factor, charge-coupled devices have many drawbacks as well. Among the drawbacks of limited read rate and dynamic range limitations, the most noteworthy is the difficulty in implementing a charge coupled device with a CMOS-based microprocessor.

To address these limitations of imaging circuits based on charge coupled devices, recent imaging circuits use active pixel sensor cells to convert light energy pixels into electrical signals. Conventional photodiodes with active pixel sensor cells are typically a combination of a number of active transistors that provide for the formation of electrical signals as well as the functions of amplification, readout control, and reset adjustment.

1 is an example of a conventional CMOS active pixel sensor cell 10. 1, the cell 10 includes a photodiode 12 coupled between a first intermediate node N _IM1 and ground, a reset transistor (not shown) coupled between the power supply node N _PS and the first intermediate node N _IM1 14).

In addition, the cell 10 also includes a buffer transistor 16 and a row-select transistor 18. 1, the buffer transistor 16 has a drain connected to the power supply node N _PS , a source connected to the second intermediate node N _IM2 , and a gate connected to the first intermediate node N _IM1 , The row select transistor 18 is connected between the second intermediate node N _IM2 and the output node N _O.

The operation of the active pixel sensor cell 10 is performed in the following three steps: the cell 10 is reset from the integration period of the previous stage; the image integration step in which the light energy is collected and converted into an electrical signal; Read signal reading phase.

Figures 2A-2B are timing diagrams illustrating the reset, image integration, and reading steps associated with the cell 10. As shown in FIGS. 1 and 2A-2B, the reset step begins by pulsing the reset voltage V _RT at the gate of the reset transistor 14 at time t ₁ . The reset voltage V _RT activates the reset transistor 14 to raise the voltage of the photodiode 12 and the buffer transistor 16 to the initial integrated voltage. In turn, the voltage on the source of the buffer transistor 16 is raised to a voltage below the threshold voltage by the source-follower action of the buffer transistor 16, rather than the initial integral voltage of the buffer transistor 16. [

Subsequently, a numerical value of the initial integration voltage smaller than the threshold voltage of the buffer transistor 16 is read by pulsing the gate of the row selection transistor 18 to the row selection voltage V _RS at time t2. The row select voltage V _RS activates the row select transistor 18, which causes the voltage on the source of the buffer transistor 16 to appear on the source of the row select transistor 18. The voltage on the source of the row select transistor 18 is detected by a conventional detection circuit and then stored as a reset voltage.

Next, during the integration period, the photon energy in the form of a photon will strike the photodiode 12, thus creating a number of electron-hole pairs. The photodiode 12 is designed to limit recombination between newly formed electron-hole pairs. As a result, the holes created by the light are attracted to the ground terminal of the photodiode 12, while electrons created by the light are attracted to the cathode terminal of the photodiode 12, The voltage on the photodiode 12 is lowered.

After the image integration period, the final integrated voltage on cell 10 is read by pulsing the gate of row select transistor 18 to row select voltage V _RS at time t ₃ . At this time, the final integrated voltage on the photodiode 12, which is less than the threshold voltage of the buffer transistor 16, is present at the drain of the row select transistor 18. [ As a result, when the row selection transistor 18 is activated, the voltage of the drain of the row selection transistor 18 appears at the source of the row selection transistor 18, where the voltage is detected and stored as a reading value Loses.

Thus, at the end of the integration period, a collected photon number value representing the number of photons absorbed by the photodiode 12 during the image integration period is calculated by subtracting the reset value taken at the beginning of the integration period from the readout value taken at the end of the integration period Can be determined.

However, one problem with the active pixel sensor cell 10 is that the imaging system using the array of active pixel sensor cells has a limited dynamic range. Conventionally, the dynamic range is defined as the maximum number of photons that a cell can collect during an integration period without saturating the cell, i. E., Without exceeding the capacity of the cell, and the maximum number of photons that can be detected over the noise layer, Has been defined by a minimum number of photons.

The effect of the limited dynamic range is most pronounced in images that include bright light and low intensity light. In this situation, if the integration period of the array is such that no information of the bright light region is lost, that is, the number of photons collected by the cells exposed to the bright light does not exceed the capacity of the cell during the integration period , Most of the information in the low luminous intensity region of the cell exposed to low luminous intensity will be lost, resulting in a black image. Because the collected photons can not be identified across the noise floor.

On the other hand, if the integration interval of the array is increased to capture information of a low luminosity region, i.e., the number of photons collected by the cells exposed to low luminosity can be detected over the entire noise layer, A significant portion of the information will be lost, resulting in a white image. This is because the number of photons collected by the cells exposed to bright light will far exceed the capacity of these cells.

One approach to solving the dynamic range problem is to use a non-integral active pixel sensor cell with a non-linear load device, such as a MOSFET diode, in weak inversion to obtain logarithmic response. However, this approach has many drawbacks.

First, the noise in the non-integral cell is much larger than the noise in the conventional integrating cell such as the cell 10 in Fig. In conventional integration cells, the effect of any noise events is averaged over the integration period, while the effect of any noise events in the non-integration cell produces substantial distortion. Secondly, the precise nonlinear transfer function of this type of device must be carefully calibrated to avoid changes between cell and cell and due to temperature variations.

Another approach to the problem of dynamic range used in CCD systems is to integrate twice: once with short exposure and once with long exposure. Information of the bright light region is stored during short exposure, while information of the low light region is discarded. Thus, during long exposures, information in the low light region is stored while information in the bright light region is discarded.

The information from the two exposures is then combined to form a composite image. However, the disadvantage of this approach is that the resulting image is formed by combining the image data for two different time intervals.

Thus, there is a need for an active pixel sensor cell having a substantially increased dynamic range in order to successfully capture sources of bright light and low light areas in the same image.

Conventionally, the dynamic range of an image cell in an imaging system is determined by the maximum number of photons that the cell can collect during the integration interval (without exceeding the cell's capacity), and the maximum number of photons that can be detected over the noise layer , The cell has been defined by the minimum number of photons that can be collected during the integration interval.

However, in the method of the present invention, the dynamic range of the image cell is significantly extended by reading the cell several times to obtain a large number of read values during each integration period. At the end of the integration interval, the imaging system determines whether the last read out of a number of readings indicates that the cell is saturated. When the last read out of the plurality of readings indicates that the cell has saturated, the readings taken before the image cell saturates can be obtained by extrapolating the last number that the image cell would have had if the image cell had not saturated .

1 is a schematic diagram illustrating a conventional active pixel sensor cell 10;

FIGS. 2A-2B are timing diagrams illustrating reset, image integration, and read steps for the cell 10. FIG.

3 is a schematic diagram illustrating a single cell imaging system 100 in accordance with the present invention.

Figures 4A-4B are timing diagrams illustrating operation of the imaging system 100 in accordance with the method of the present invention.

5 is a diagram illustrating the operation of image control and storage device 110 in accordance with the present invention.

6 is a schematic diagram illustrating an imaging system 200 implementing an imaging system 100 having an array of active pixel sensor cells 10;

FIG. 7 is a schematic illustration showing an imaging system 300 with two cells according to the method of the first variant of the present invention.

8A-8D are timing diagrams illustrating the operation of the imaging system 300 in accordance with the method of the first variant.

9 is a schematic diagram illustrating an imaging system 400 implementing an imaging system 300 having an array of active pixel sensor cells 10.

[Description of Reference Numerals]

10: active pixel sensor cell 12: photodiode

14: reset transistor 16: buffer transistor

18: Row select transistor 100: Single image system

110: Image control device and storage device

A method of increasing the dynamic range of an image cell in accordance with the present invention includes resetting the image cell for the first time and reading the image cell to obtain the first reset value after the image cell is reset for the first time. After the image cell is read to obtain the initial reset value, the image cell is read many times to obtain a corresponding multiple readout value. Then, the image cell is reset a second time, and then the image cell is read out to obtain a second reset value after being reset a second time. Next, the step of determining whether the last read out of the plurality of readings indicates that the image cell is saturated continues in the present invention.

According to the method of the present invention, an image cell is read many times before the image cell is saturated before the image cell is read to obtain a second reset value after the image cell is read to obtain the initial reset value.

Determining which of the readings is taken before the image cell is saturated if the last read value indicates that the image cell is saturated; and, if the last read value indicates that the image cell is saturated, The step of determining the numerical value is continued in the present invention. Next, the accumulated photon number values are calculated by calculating the difference between the final reading value and the initial reset value or the second reset value.

On the other hand, if the last read value indicates that the image cell is saturated, the method of the present invention as described above may include the steps of determining which of the read values was taken before the image cell was saturated, Lt; RTI ID = 0.0 > a < / RTI > first reset value or a second reset value. Next, by extrapolating the data values, the aggregated photon number values are determined.

In the method of the first modification of the present invention, the dynamic range of the image system having the image cell and the reference cell is determined by resetting the image cell for the first time and resetting the image cell to obtain the first reset value after the image cell is reset for the first time. Is increased by reading. Then, after the image cell is read to obtain the initial reset value, the image cell is read many times correspondingly to obtain a large number of read values. After the image cell is read many times, the image cell is reset a second time, and then read to obtain a second reset value.

In addition, the reference cell is reset several times, and the reference cell is reset each time the video cell is read. Also, the reference cell is read correspondingly several times in order to obtain a plurality of reset values, so that the reference cell is read every time the reference cell is reset. Each read value has a corresponding reset value. The step of determining whether the last read out of the plurality of readings indicates that the image cell is saturated continues in the present invention.

According to the method of the first modification of the present invention, before the image cell is read to obtain the first reset value and then after the image cell is read to obtain the second reset value, It is read many times.

Determining which one of the readings is taken before the image cell is saturated when the last read value indicates that the image cell is saturated, and comparing the read values with the readout values indicating that the image cell is not saturated, The step of forming a plurality of data values is determined in the present invention. Next, the collected photon number values are determined by extrapolating from the data values.

BRIEF DESCRIPTION OF THE DRAWINGS The features and advantages of the present invention will become better understood with reference to the following detailed description and the accompanying drawings, which illustrate an exemplary embodiment using the teachings of the present invention.

FIG. 3 shows a schematic diagram of a single imaging system 100 in accordance with the present invention. As shown in FIG. 3, the imaging system 100 includes the active pixel sensor cell 10 of FIG. 1 with a storage device 110 and an image control device that controls the operation of the cell 10. (Other active pixel sensor cells that are described in the context of the cell 10 of Fig. 1 in the present invention but have non-destructive readings may also be used.)

4A-4B are timing diagrams illustrating operation of the imaging system 100 in accordance with the method of the present invention. As shown in FIGS. 4A-4B, the method of the present invention applies a row reset voltage V _RT to the gate of the reset transistor 14 at time t ₁ , as described with respect to FIGS. 2A-2B, a row select voltage to the gate of the row select transistor 18 at t ₂ is started by applying a V _RS.

Thus, as in Figures 2A-2B, the cell 10 of the imaging system 100 is first reset, and then the reset value, which represents the initial integral voltage of the photodiode 12 less than the threshold voltage drop of the buffer transistor 16, ≪ / RTI >

Next, according to the present invention, Once the cell 10 is read to obtain the reset level, the cell 10 is read multiple times during the remainder of the integral term starts and ends at time t ₇ at time t _2.

The length of the integration interval is defined by a scan rate of about 30 mS in video applications whereas the length of the integration interval in still picture applications is defined by the lowest light level or the highest light level being captured . The longer integration periods cause more and more photons to gather from the dim light sources, which in turn causes more and more of these collected photons to exceed the noise layer.

Further, according to the method of the present invention, the cell 10 must be read at least twice, preferably three times, before the cell 10 is saturated when exposed to the brightest light.

As shown in the embodiment of FIG. 4B, cell 10 is read five more times during the remainder of the integration interval by applying row select voltage V _RS to times t ₃ , t ₄ , t ₅ , t ₆ and t ₇ . Row selection voltage V _RS, the line as cause the voltage of buffer transistor 16, a small photodiode 12 than the threshold voltage drop, time _{t 3, t 4, t 5} , t 6 , and the output voltage at t ₇ V _O so that it appears to the source of the selection transistor 18, the output voltage V _O at the source are detected by the image control device and the storage device 110 and stored.

In regard to timing, the time t _2, t elapsed time between _{3, t 4, t 5,} t 6 and t ₇ are set equal in duration, or, or alternatively may be set otherwise, to have a different duration. Thus, for example, the time between times t ₂ and t ₃ and the time between times t ₃ and t ₄ may be equal to 6 mS, or may be different, respectively, between 1 μS and 10 μS. Other durations are needed to obtain a minimum number of read times before the cell 10 is saturated.

Next, in accordance with the present invention, once the integration interval is over, the image control and storage device 110 compares the read value taken at time t ₇ with the saturation value of the cell to determine whether the cell is saturated at time t ₇ . If the time read value taken at t ₇ has indicated that it is not saturated and the cell 10, then device 110 hours from the taken reset value at time t ₂ to determine a combined photon numbers of cells 10 t ₇ Lt; / RTI > Thus, the collected photon number indicates how many photons have been collected by the cell 10 during the integration interval.

On the other hand, if the time t is read out value taken at ₇ indicates that the cell 10 is a saturated surface, then device 110 at time t _3, t _4, t which one cell from among the taken in read value from ₅ with t ₆ ( 10) has been taken before it is saturated. The device 110 then uses the readout value taken before the cell 10 saturates to indicate the appropriate reading that the cell 10 would have had if the cell 10 had not saturated at time t ₇ The readout value is extrapolated. Next, the accumulated photon number for cell 10 is calculated by subtracting the extrapolated read value from the reset value taken at time t ₂ .

5 is a diagram illustrating operation of the image control and storage device 110 in accordance with the present invention. 5, line L1 represents the number of photons collected by the cell 10 when the cell 10 is exposed to bright light, while line L2 represents the number of photons collected by the cell 10, Represents the number of photons collected by the cell 10 when exposed.

Likewise, line L3 represents the number collected by cell 10 when cell 10 is exposed to moderate low light while line L4 represents cell 10 when cell 10 is exposed to low light, The number of photons collected by.

As described above, when the cell 10 is exposed to, or line shown in L4 as very low light when exposed to low light, moderate As shown in the line L3, the cell 10 is not saturated at the time t ₇ Do not. As a result, the accumulated photon number of the cell 10 under these conditions is determined by subtracting the value obtained at time t ₇ from the reset value obtained at time t ₂ .

By the way, when the cell 10 is exposed to bright light, as indicated by the line L1 in Fig. 5, the cell 10 can be read only three times before the cell is saturated. Thus, as indicated by line L2, when the cell 10 is exposed to medium bright light, the cell 10 can be read only four times before the cell is saturated. Once saturated, the values read from the cell 10 remain essentially constant.

Thus, the readout value taken at time t ₇ indicates that the cell 10 is saturated when the cell 10 is exposed to bright or moderate bright light. By the way, in bright light, the device 110 is able to determine from the readings taken at times t ₃ , t ₄ and t ₅ , before the cell 10 is saturated, the cell (s) at time t ₇ Extrapolates the appropriate reading value that the readout data would have.

Similarly, in the light of the medium, the apparatus 110 includes a cell 10 is shown, the time t _3, t _4, t from the taken in read figures ₅ and t _6, the line L2 'before saturation extrapolates as the cell 10, the appropriate read levels have had at time t _7.

The change in voltage appearing from incoming photons is nonlinear, among other things, due to the change in capacitance. As a result, the greater the number of readings that can be taken before the cell 10 is saturated, the more accurate the extrapolated result will be. Thus, the accuracy of line L2 'is more accurate than the accuracy of line L1'.

Instead of using unsaturated readings to extrapolate the last readout value that is subtracted from the reset value, in an alternative approach, each non-saturated readout may include a series of data values that are used to extrapolate the appropriate aggregated photon- Can be subtracted from the reset value to obtain.

For example, if the readings taken at times t ₃ , t ₄ and t ₅ are the only readings that indicate that the cell 10 is not saturated, then these values are subtracted from the reset value taken at time t ₂ , Data values. The three data values are then used to extrapolate the collected photon number values that the cell 10 would have had if the cell 10 had not saturated at time t ₇ .

Thus, an advantage of the present invention is that by extrapolating the read or collected photon number values for the cell 10 when the readout value taken at the end of the integration period indicates that the cell 10 is saturated, Can be substantially increased.

In addition to using the reset value taken at time t ₂ , a reset value taken at the beginning of the next integration interval may also be used. Returning to Figures 4A-4B, after the cell 10 is read at time t ₇ , the end of the first integration period, the cell 10 is reset again at time t ₈ to start the second integration period. After the cell 10 is reset by the second at time t _8, the row selection voltage V _RS is applied again at time t ₉ will determine a second reset value.

Instead of subtracting the read value taken at time t ₇ from the reset value taken at time t ₂ , which is a common case in the prior art, when the read value taken at time t ₇ indicates that the cell 10 is not saturated, the read value is taken at time t ₇ is subtracted from the reset value taken at time t _9.

The advantage of subtracting the read value taken at time t ₇ from the reset value taken at time t ₉ is that the time between when the reset value is taken and when the read values are taken is minimized. By reducing the time between when the reset value is taken and when the read values are taken, the effect of the 1 / f noise originating from the buffer transistor 16 can be substantially reduced. The effect of the 1 / f noise on the reset transistor 14 is described by Richard Merrill on September 10, 1996, for a method for operating an active pixel sensor cell that reduces noise in optical information extracted from a cell As discussed in U. S. Patent Application Serial No. 08 / 707,933, which is incorporated herein by reference.

In addition, when the readout value taken at time t ₇ indicates that the cell 10 is saturated and the cell 110 has used the readings values taken before the cell 10 saturates at time t ₇ when extrapolating the read value indicating the proper value read out, by subtracting the read value extrapolated from the reset value taken at time t ₉ the photon numbers instead of time t ₂ for the combined cell 10 can also be calculated.

In another approach, unsaturated readings may be subtracted from the first reset value, the second reset value, or both reset values to obtain a series of data sets that are used to extrapolate the collected photon quantities .

For example, if the readings taken at times t ₃ , t _4, and t ₅ are the only readings indicating that the cell 10 is not saturated, then these values may be used to generate three data values, may be subtracted from the first reset value taken at t ₂ , all subtracted from the second reset value taken at time t ₉ , or some from the first reset value and some subtracted from the second reset value . The three data values are then, if the cell 10 is used for the time t Had not saturated at ₇ to extrapolate proper collected photons value eoteul the cell 10 held at the time t _7.

FIG. 6 shows a schematic diagram of an imaging system 200 implementing an imaging system 100 having an array of active pixel sensor cells 10. As shown in FIG. 6, the imaging system 200 uses both a series of reset voltages V _RT1 -V _RTn and a series of row selection voltages V _RS1 -V _RSn corresponding to the row number of cells in the array.

In addition, the imaging system 200 includes a series of output lines V _OUT1 -V _OUTn corresponding to the number of columns in the array. In operation, the imaging system 200 first operates in parallel on one row of cells and then sequentially in each of the cells.

FIG. 7 shows a schematic diagram of an image two-cell system 300 according to a first variant method of the present invention. As shown in FIG. 7, the imaging system 300 includes an image cell C1 configured as the active pixel sensor cell 10 of FIG. 1, a reference cell C2 comprised of the active pixel sensor cell 10 of FIG. 1, And an image control device and a storage device (110). (Other active pixel sensor cells having non-destructive readout results may also be used, although the first modified method describes cell 10 of FIG. 1, as described above).

Figures 8A-8D show timing diagrams illustrating operation of the imaging system 300 in accordance with the first variant method. Also as shown in 8A-8D, a modified example In the first method begins by applying a row reset voltage V _RT1 to the gate of the reset transistor 14 of cell C1 at time t _1, followed by the row select voltage from the time t ₂ V _RS1 is applied. As before, the row reset voltage V _RT1 at time t ₁ resets the photodiode 12 to the initial integrated voltage, while the row select voltage V _RS1 at time t ₂ reads the first reset value.

Once the cell C1 is read to obtain the first reset value, the cell C1 is read again many times during the remaining interval of the integration interval, beginning at time t2 and ending at time t7. As above, cell C1 should be read at least three times, at least twice, before cell C1 is saturated when exposed to the brightest light.

As shown in the example of FIG. 8B, cell C1 is read more than five times during the remainder of the integration interval by applying row select voltage V _RS1 at times t ₃ , t ₄ , t ₅ , t ₆ and t ₇ . As described above, the time required between times t ₂ , t ₃ , t ₄ , t ₅ , t _6, and t ₇ can be set to have a constant duration or alternatively can be set to have different durations .

Row selection voltage V _RS1 is the buffer transistor little photodiode 12 voltage two hours _{t 3, t 4, t 5} , t 6 , and the row selection transistor 18 at t ₇ on a sub-threshold voltage drop of the 16 So that the output voltage V _O appears at the source. The voltage V _O at this source is detected and stored as a corresponding number of readings by the image controller and storage device 110.

Further, according to the present invention, the reference cell C2 is also reset and read many times corresponding to the number of times that the image cell C1 is read after reset at time t1. Thus, as shown in Figures 8C-8D, the cell C2 is reset by applying a reset voltage V _RT2 at times t ₈ , t ₉ , t ₁₀ , t ₁₁ and t ₁₂ , and at times t ₁₃ , t ₁₄ , t ₁₅ , t ₁₆ and t ₁₇ by applying the row selection voltage V _RS2 .

The row select voltage V _RS2 is the voltage of the row select transistor 18 at times t ₁₃ , t ₁₄ , t ₁₅ , t ₁₆ and t ₁₇ , which is less than the threshold voltage drop of the buffer transistor 16, And appear as an output voltage V _{O at the} source. The voltage V _O at this source is detected and stored as a corresponding plurality of reset values by the image controller and storage device 110.

Once the integration period is over, the image control and storage device 110 determines whether the read value of cell C1 taken at time t7 indicates that cell C1 is saturated. If taking the read value is taken at time t7 indicates that the cell C1 is full, then the device 110 is time t _3, t _4, t _5, and among the taken in read value at t ₆ before which one cell 10 is saturated To decide whether to lose. The device 110 then subtracts the non-saturated readings from the corresponding reset values to determine a series of data values.

For example, if the readings taken at times t ₃ , t _4, and t ₅ represent non-saturating values, then these values are obtained from the reset values taken at times t ₁₃ , t ₁₄ and t ₁₅ to obtain data values, respectively Is subtracted. Once the data points have been determined, the device 110, if in the cell C1 the time t ₇ Had this is not in saturation prior to time t ₇ from the data value representing the photons were able to collect the cell C1, the extrapolated proper collected photons figures do.

The advantage of using the reference cell and subtracting each non-saturated read value of the corresponding cell C1 from the reset value of the cell C2 is that the time between each read and reset value can be minimized, It is possible to further reduce the effects of 1 / f noise on the values.

On the other hand, if the readout value taken at time t ₇ indicates that cell C 1 is not saturated, then the device 110 determines the time from the reset value taken at time t ₁₇ to determine the aggregated photon- subtracts the read value taken at t ₇ .

Time In addition to subtracting the read value taken at time t ₇ from the taken reset value at time t ₁₇ when t is the cell C1 is a non-saturated _7, the read value is taken at time t ₇ from the taken reset value from the start of the next integration interval It can be subtracted instead.

After also the Returning to 8A-8B, cell C1 is read out from the first integral term kkeutin time t _7, the cell C1 is reset again at the time t ₁₈ to start the second integration interval. Cell C1 is applied this time in t ₁₈ after both the second reset, the row selection voltage V _RS1 from this time t ₁₉ again is determined for the second reset value for the cell C1.

According to the present invention, rather than subtracting the read value taken at time t ₇ from the reset value taken at time t ₁₇ , the read value taken at time t ₇ may be subtracted from the reset value taken at time t ₁₉ . Alternatively, the read value taken at time t ₇ may also be subtracted from the reset value taken at time t ₂ .

The advantage of subtracting the read value taken at time t ₇ from the reset value taken at time t ₁₉ is that the difference between the reset values of cell C 1 and cell C 2 can be eliminated when cell 10 is non-saturating.

FIG. 9 shows a schematic diagram of an imaging system 400 implementing an imaging system 300 having an array of active pixel sensor cells 10. As shown in FIG. 9, the imaging system 400 differs from the imaging system 200 only in that one row of cells in the array is represented as a row of reference cells C2 and thus is not used for image acquisition.

In operation, the imaging system 400 operates on the cells C1 in the first row of image cells, while operating on the cells C2 in the row of reference cells, and then on the cells C1 Lt; RTI ID = 0.0 > C2 < / RTI > in the rows of reference cells.

Thus, when the device 110 is operating on the cells in the first row, the read values of the cells C11, C12, C13 and C14 are subtracted from the reset values of the cells R11, R12, R13 and R14, respectively, 110 are operating in the cells in the second row, the read values of cells C21, C22, C23 and C24 are subtracted from the reset values of cells R11, R12, R13 and R14, respectively.

It should be understood that various modifications to the embodiments of the invention described herein may be applied in practicing the invention. For example, instead of using one row of reference cells, if a reference reset value read from a single reference cell is used for all of the cells C1 in the row being operated, a single reference Cells can also be used.

Also, instead of resetting and reading the reference cell (or cells) C2 each time the video cell C1 is read, the reference cell (or cells) C2 can be reset and read once or more. In this case, the read values may be subtracted from any combination of a first reset value, a second reset value, and a reference reset value.

For example, time t _3, t _4, and the read value is taken at t ₅ are, respectively, the first reset value, a reference reset level, and the two may be subtracted from the first reset value or, time t ₃ and taken read value from t ₄ Can be subtracted from the first reset value and the read value taken at time t ₅ can be subtracted from the reference reset value. Similarly, time t ₃ and t ₄ are taken from a read value can be subtracted from the reset reference value, and at the same time, the read value is taken at time t ₅ can be subtracted from the second reset levels.

It is therefore to be understood that within the scope of the appended claims, the appended claims define the scope of the invention and that methods and structures within the scope of such claims and their equivalents may be included within the scope of the invention .

By the method of the present invention as described above, the dynamic range of the image cell can be significantly extended by reading the cell several times to obtain multiple readings during each integration period.

Claims (36)

Resetting the video cell, Reading the image cell to obtain a reset value after the image cell is reset; Reading the image cell correspondingly several times to obtain a plurality of readings after the image cell is read to obtain a reset value, Determining whether the last read out of the plurality of readings indicates that the image cell is saturated. ≪ Desc / Clms Page number 16 > 16. A method for increasing the dynamic range of an imaging system having an image cell. The method of claim 1, Characterized in that the image cell is read out several times before the image cell is read out to obtain a reset value. The method according to claim 1, If the last read value indicates that the image cell is saturated, determining which of the read values was taken before the image cell was saturated, Determining a final read value by extrapolating from readings that indicate that the image cell is not saturated, Further comprising the step of determining a collected photon number value by calculating a difference between a final reading value and a reset value. The method according to claim 1, Determining which of the readout values was taken before the image cell was saturated if the last readout value indicates that the image cell is saturated, Forming a plurality of data values by determining a difference between reset values and read values indicating that the image cell is not saturated, ≪ / RTI > further comprising the step of determining an aggregated photon number value by extrapolating from the data values. Resetting the video cell for the first time, Reading an image cell to obtain a first reset value after the image cell is reset for the first time, Reading the image cell correspondingly several times to obtain a plurality of readings after the image cell is read to obtain a first reset value, Resetting the image cell a second time after the image cell is read many times, Reading an image cell to obtain a second reset value after the image cell is reset a second time, Determining whether the last read out of the plurality of readings indicates that the image cell is saturated. The method of claim 5, Characterized in that the image cell is read out several times before the image cell is saturated before the image cell is read to obtain a second reset value after the image cell is read to obtain the first reset value. 6. The method of claim 5, If the last read value indicates that the image cell is saturated, determining which of the read values was taken before the image cell was saturated, Determining the final read value by extrapolating from the read value indicating that the image cell is not saturated, Further comprising the step of determining a collected photon number value by calculating a difference between a final read value and a first reset value. 6. The method of claim 5, If the last read value indicates that the image cell is saturated, determining which of the read values was taken before the image cell was saturated, Determining the final read value by extrapolating from the read values indicating that the image cell is not saturated, Further comprising the step of determining a collected photon number value by determining a difference between a final read value and a second reset value. 6. The method of claim 5, If the last read value indicates that the image cell is saturated, determining which of the read values was taken before the image cell was saturated, Forming a plurality of data values by determining a difference between readings and a first reset value indicating that the image cell is not saturated, ≪ / RTI > further comprising the step of determining an aggregated photon number value by extrapolating from the data values. 6. The method of claim 5, If the last read value indicates that the image cell is saturated, determining which of the read values was taken before the image cell was saturated, Forming a plurality of data values by determining a difference between read values and a second reset value indicating that the image cell is not saturated, ≪ / RTI > further comprising the step of determining an aggregated photon number value by extrapolating from the data values. 6. The method of claim 5, If the last read value indicates that the image cell is saturated, determining which of the read values was taken before the image cell was saturated, Determining a difference between a read value indicating that the image cell is not saturated and a first reset value and a read value indicating that the image cell is not saturated and a second reset value, ≪ / RTI > further comprising the step of determining an aggregated photon number value by extrapolating from the data values. 6. The method of claim 5, Forming a collected photon number value by determining a difference between a last read value and a second reset value among a plurality of read values when the last read out value among the plurality of read out values indicates that the image cell is not saturated ≪ / RTI > Resetting the video cell, Reading the image cell correspondingly several times to obtain a plurality of read values after the image cell is reset, Resetting the reference cell many times so that the reference cell is reset every time the video cell is read; Reading the reference cell correspondingly many times to obtain a plurality of readings so that the reference cell is read whenever the reference cell is reset and each readout value has a corresponding reset value, Determining whether the last read out of the plurality of readings indicates that the image cell is saturated. ≪ Desc / Clms Page number 15 > 18. A method for increasing the dynamic range of an imaging system having an image cell and a reference cell. 14. The method of claim 13, Characterized in that the image cells are read out several times before the image cells are saturated after being reset. 14. The method of claim 13, Further comprising the step of forming a collected photon number value by determining a difference between a last read value and a last read value among a plurality of read values, when the last read value among the plurality of read values indicates that the image cell is not saturated ≪ / RTI > 14. The method of claim 13, If the last read value indicates that the image cell is saturated, determining which of the read values was taken before the image cell was saturated, Forming a plurality of data values by determining a difference between readout values and corresponding reset values indicating that the image cell is not saturated, ≪ / RTI > further comprising the step of determining a collected photon number value by extrapolating the data values. 14. The method of claim 13, Characterized in that the video cell is reset immediately after reading. 18. The apparatus of claim 17, And the reference cell is read immediately after reset. Resetting the video cell for the first time, Reading the image cell to obtain a first reset value after the first time the image cell is reset; Reading the image cell correspondingly several times to obtain a plurality of read values after the image cell is read to obtain a first reset value, Resetting the image cell a second time after the image cell is read many times, Reading an image cell to obtain a second reset value after the second reset of the image cell, Resetting the reference cell many times so that the reference cell is reset every time the video cell is read; Reading the reference cell correspondingly many times to obtain a plurality of reset values so that the reference cell is read every time the reference cell is reset and each read value has a corresponding reset value, Determining whether the last read out of the plurality of readings indicates that the image cell is saturated. ≪ Desc / Clms Page number 15 > 18. A method for increasing the dynamic range of an imaging system having an image cell and a reference cell. 20. The method of claim 19, Characterized in that the image cell is read out several times before the image cell is saturated after it has been read in order to obtain the first reset value. 20. The method of claim 19, Forming a collected photon number value by determining a difference between a last read value and a first reset value among a plurality of read values, when the last read value among the plurality of read values indicates that the image cell is not saturated ≪ / RTI > 20. The method of claim 19, Forming a collected photon number value by determining a difference between a last read value and a second reset value among a plurality of read values when the last read out value of the plurality of readings indicates that the image cell is not saturated ≪ / RTI > 20. The method of claim 19, Characterized in that the video cell is reset immediately after reading. 24. The apparatus of claim 23, Characterized in that the reference cell is read immediately after reset. Resetting the video cell, Reading the image cell correspondingly several times to obtain a plurality of read values after the image cell is reset; Resetting the reference cell once after the video cell is reset; Reading the reference cell once to obtain a reset value so that the reference cell is read whenever the reference cell is reset, Determining whether the last read out of the plurality of readings indicates that the image cell is saturated. ≪ Desc / Clms Page number 15 > 18. A method for increasing the dynamic range of an imaging system having an image cell and a reference cell. 26. The method of claim 25, Characterized in that the image cells are read out several times before the image cells are saturated after being reset. 26. The method of claim 25, Further comprising the step of forming a collected photon number value by determining a difference between a last read value and a reset value among a plurality of read values, when the last read value among the plurality of read values indicates that the image cell is not saturated Lt; / RTI > 26. The method of claim 25, If the last read value indicates that the image cell is saturated, determining which of the read values was taken before the image cell was saturated, Forming a plurality of data values by determining a difference between read values and reset values that indicate that an image cell is not saturated, ≪ / RTI > further comprising the step of determining a collected photon number value by extrapolating the data values. Resetting the video cell for the first time, Reading an image cell to obtain a first reset value after the image cell is reset for the first time, Reading the image cell correspondingly several times to obtain a plurality of read values after the image cell is read to obtain a first reset value, Resetting the image cell a second time after the image cell is read many times, Reading an image cell to obtain a second reset value after the second reset of the image cell, Resetting the reference cell once so that the reference cell can be reset after the video cell is read to obtain a first reset value; Reading the reference cell once to obtain a reference reset value so that the reference cell is read whenever the reference cell is reset, Determining whether the last read out of the plurality of readings indicates that the image cell is saturated. ≪ Desc / Clms Page number 15 > 18. A method for increasing the dynamic range of an imaging system having an image cell and a reference cell. 30. The method of claim 29, Characterized in that the image cell is read many times before it is saturated after the image cell is read to obtain the first reset value. 30. The method of claim 29, Forming a collected photon number value by determining a difference between a last read value and a first reset value among a plurality of read values, when the last read value among the plurality of read values indicates that the image cell is not saturated ≪ / RTI > 30. The method of claim 29, Forming a collected photon number value by determining a difference between a last read value and a second reset value among a plurality of read values when the last read out value of the plurality of readings indicates that the image cell is not saturated ≪ / RTI > 30. The method of claim 29, If the last read value indicates that the image cell is saturated, determining which of the read values was taken before the image cell was saturated, By determining the difference between the readings of the first group and the first reset value indicating that the image cell is not saturated and the read value of the second group indicating that the image cell is not saturated and the reference reset value Forming a plurality of data values, ≪ / RTI > further comprising the step of determining an aggregated photon number value by extrapolating from the data values. 30. The method of claim 29, If the last read value indicates that the image cell is saturated, determining which of the read values was taken before the image cell was saturated, By determining the difference between the read values of the first group and the second reset value indicating that the image cell is not saturated and the read values of the second group indicating that the image cell is not saturated and the reference reset value Forming a plurality of data values, ≪ / RTI > further comprising the step of determining an aggregated photon number value by extrapolating from the data values. 30. The method of claim 29, If the last cell indicates that the image cell is saturated, determining which of the read values was taken before the image cell was saturated, Between the readings of the first group and the first reset value indicating that the image cell is not saturated and between the readings of the second group and the second reset value indicating that the image cell is not saturated, Forming a plurality of data values by determining a difference between a readout value of a second group and a reference reset value, ≪ / RTI > further comprising the step of determining an aggregated photon number value by extrapolating from the data values. 30. The apparatus of claim 29, Characterized in that the reference cell is read immediately after reset.