Adaptive correlation multiple sampling

Through adaptive related multi-sampling technology, the reading circuit calculation method of the image sensor is dynamically adjusted according to the signal size, which solves the problem of insufficient noise processing in the prior art, and achieves a lower readout period and a higher signal-to-noise ratio.

CN119031268APending Publication Date: 2024-11-26OMNIVISION TECHNOLOGIES INC
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202410636511.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2023-05-23
Filing Date
2024-05-22
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

Existing image sensors have insufficient noise reduction in the readout circuit noise processing, especially photon shot noise dominates under high signal conditions, and multi-sampling technology increases costs and reduces frame rates, which cannot effectively reduce low-frequency noise such as 1/f Noise and random telegram signals.

Method used

Adaptive related multi-sampling (CMS) technology is adopted, and the calculation method is dynamically adjusted according to the signal size through the combination of comparator, sampling circuit and counter. CMS calculation is performed when small signals is performed, non-CMS calculation is performed when large signals are performed, and slope signals are used. The difference and product calculations are performed with the reset signal to reduce noise.

Benefits of technology

The readout cycle is reduced, the noise is reduced, especially the random noise is effectively reduced under low signal conditions, the signal-to-noise ratio is improved, and unnecessary CMS calculations are avoided under high signal conditions, and the performance of the image sensor is improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119031268A_ABST
    Figure CN119031268A_ABST
Patent Text Reader

Abstract

The invention relates to adaptive correlation multiple sampling. A readout circuit includes a comparator having a first input coupled to receive a ramp signal from a ramp generator and a second input coupled to receive an analog image data signal from one of a plurality of bit lines. The comparator is configured to generate a comparator output in response to a comparison of the ramp signal and the analog image data signal. A sampling circuit has a first input coupled to receive a sampling control signal and a second input coupled to receive the comparator output. The sampling circuit is configured to generate a sampled output. A counter has a first input coupled to receive a counter control signal and a second input coupled to receive one of the comparator output and a signal from the sampling circuit. The readout circuit is configured to perform a correlated multi-sampling CMS calculation or a non-CMS calculation in response to the sampling output.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates generally to image sensors, and particularly, but not exclusively, to high dynamic range (HDR) complementary metal oxide semiconductor (CMOS) image sensors. Background Art

[0002] Image sensors have become ubiquitous and are now widely used in digital cameras, cell phones, security cameras, and in medical, automotive, and other applications. As image sensors are integrated into a wider range of electronic devices, it is desirable to enhance their functionality, performance metrics, and the like in as many ways as possible (e.g., resolution, power consumption, dynamic range), through device architecture design and image acquisition processing. The technology used to manufacture image sensors continues to advance rapidly. For example, the demand for higher resolution and lower power consumption has driven further miniaturization and integration of these devices.

[0003] A typical image sensor operates in response to image light incident on the image sensor from an external scene. The image sensor includes a pixel array having photosensitive elements (e.g., photodiodes) that absorb a portion of the incident image light and generate image charge when absorbing the image light. The image charge generated by the pixel light can be measured as an analog output image signal on a column bit line, which varies as a function of the incident image light. In other words, the amount of image charge generated is proportional to the intensity of the image light, and the image light is read out from the column bit line as an analog image signal and converted into a digital value to generate a digital image (e.g., image data) representing the external scene. The analog image signals on the bit lines are coupled to a readout circuit that includes an input stage with an analog / digital conversion (ADC) circuit to convert those analog image signals from the pixel array into digital image signals. Summary of the invention

[0004] On the one hand, the present application provides a readout circuit for an image sensor, comprising: a comparator having a first input coupled to receive a ramp signal from a ramp generator and a second input coupled to receive an analog image data signal from one of a plurality of bit lines, wherein the comparator is configured to generate a comparator output in response to a comparison of the ramp signal and the analog image data signal; a sampling circuit having a first input coupled to receive a sampling control signal and a second input coupled to receive the comparator output, wherein the sampling circuit is configured to generate a sampling output; and a counter having a first input coupled to receive a counter control signal and a second input coupled to receive the comparator output and one of a signal from the sampling circuit, wherein the counter is configured to generate a counting code, wherein the readout circuit is configured to perform a correlated multiple sampling (CMS) calculation or a non-CMS calculation in response to the sampling output.

[0005] On the other hand, the present application provides a method of operating an image sensor, comprising: comparing a ramp signal from a ramp generator and an analog image data signal from one of a plurality of bit lines to generate a comparator output; sampling the comparator output with a sampling circuit in response to a sampling control signal to generate a sampling output; generating a counter code with a counter circuit in response to a counter control signal and one of the comparator output and a signal from the sampling circuit; reading out a plurality of reset signals during a plurality of reset signal readout cycles; reading out a plurality of image signals during a plurality of image signal readout cycles; and determining whether to perform correlated multiple sampling (CMS) calculation or non-CMS calculation in response to the sampling output.

[0006] On the other hand, the present application provides a method for performing adaptive correlated multi-sampling (CMS), comprising: reading out multiple reset signals and multiple image signals; performing CMS calculations on small signals, including: finding multiple CMS differences, wherein each CMS difference is a difference between one of the image signals and a corresponding one of the reset signals; finding multiple CMS products, wherein each CMS product is a product between one of the CMS differences and one of a plurality of CMS weights, wherein the sum of the CMS weights is 1; and finding the sum of the CMS products; performing non-CMS calculations on large signals, including: finding a non-CMS difference between a representative image signal and a representative reset signal, wherein the representative image signal is based only on a last one of the image signals; and finding a quotient between the non-CMS difference and a divisor. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Non-limiting and non-exhaustive embodiments of the present disclosure are described with reference to the following figures, wherein like reference numerals refer to like parts throughout the various views unless otherwise specified.

[0008] Figure 1 One example of an imaging system including a pixel array according to the teachings of the present disclosure is shown.

[0009] Figure 2A and 2B Schematic diagram showing two examples of readout circuits according to the teachings of the present disclosure.

[0010] Figure 3A and 3B A timing diagram of a readout cycle of an example readout circuit according to the teachings of the present disclosure is shown.

[0011] Figure 4 A schematic diagram illustrating another example of a readout circuit in accordance with the teachings of the present disclosure.

[0012] Figure 5A timing diagram of a readout cycle of another example readout circuit according to the teachings of the present disclosure is shown.

[0013] Figure 6 is a flow chart illustrating an example method of performing adaptive correlated multi-sampling (CMS) according to the teachings of the present disclosure.

[0014] Figure 7 is a flow chart illustrating an example method of performing CMS calculations according to the teachings of the present disclosure.

[0015] Figure 8 is a diagram of an example method for performing weighted adaptive CMS according to the teachings of the present disclosure.

[0016] Fig. 9 is a diagram of another example method of performing weighted adaptive CMS according to the teachings of the present disclosure.

[0017] Fig.10 is a flow chart illustrating an example method of performing non-CMS computations according to the teachings of the present disclosure.

[0018] In the several views of the drawings, corresponding reference characters indicate corresponding components. It will be appreciated by those skilled in the art that the elements in the figures are illustrated for simplicity and clarity and are not necessarily drawn to scale. For example, the dimensions of some elements in the figures may be exaggerated relative to other elements to help facilitate understanding of the various embodiments of the present disclosure. In addition, common but well-known elements that are useful or required in commercially feasible embodiments are generally not depicted in order to facilitate unobstructed viewing of these different embodiments of the present disclosure. DETAILED DESCRIPTION

[0019] An example of an imaging system with a pixel unit readout circuit is disclosed for performing adaptive correlated multiple sampling (CMS) and providing a reduced readout cycle. In the following description, numerous specific details are set forth to provide a comprehensive understanding of the examples. However, one skilled in the relevant art will recognize that the techniques described herein may be practiced without one or more of the specific details or with other methods, components, materials, etc. In other instances, well-known structures, materials, or operations are not shown or described in detail to avoid obscuring certain aspects.

[0020] References to "one example" or "one embodiment" in this specification mean that the particular features, structures, or characteristics described in conjunction with the example are included in at least one example of the present disclosure. Therefore, the phrases "in one example" or "in one embodiment" appearing in different places in this specification do not necessarily refer to the same example. In addition, in one or more examples, the particular features, structures, or characteristics may be combined in any suitable manner.

[0021] For ease of description, spatially relative terms (e.g., "below," "below," "above," "below," "above," "up," "top," "bottom," "left," "right," "center," "middle," and the like) may be used herein to describe the relationship of one element or feature to another (or more) element or feature, as shown in the figures. It should be understood that, in addition to the orientation depicted in the figures, spatially relative terms are also intended to cover different orientations of the device in use or operation. For example, if the device in the figure is rotated or flipped, then the element described as being "below" or "below" or "below" other elements or features will be oriented "above" other elements or features. Therefore, the exemplary terms "below" and "below" can cover both above and below orientations. The device can be oriented in other ways (rotated 90 degrees or in other orientations) and the spatially relative descriptors used in this article are interpreted accordingly. In addition, it should also be understood that when an element is referred to as being "between" two other elements, it can be the only element between the two other elements or there can also be one or more intervening elements.

[0022] In this specification, several technical terms are used. These terms adopt the common meaning in the art to which they belong, unless explicitly defined herein or otherwise clearly implied by the context of their use. It should be noted that in this document, element names and symbols are used interchangeably (e.g., Si and silicon), however, both have the same meaning.

[0023] As will be discussed, various examples of imaging systems with pixel unit readout circuits are disclosed for performing adaptive CMS and providing reduced readout cycles. CMS is a technique for reducing noise in analog / digital conversion (ADC) circuits by averaging correlated noise sources. CMS involves sampling the same analog signal multiple times at different times and correlating the samples to reduce noise, especially high-frequency noise. One of the advantages of CMS is reduced readout noise. Under dark conditions, when the signal is small, random noise from the readout circuit components dominates, so using CMS can be advantageous. However, under bright conditions where the signal is large, photon shot noise dominates, so using CMS may not improve the signal-to-noise ratio (SNR). CMS may also be unable to effectively reduce low-frequency noise, such as 1 / f noise and random telegraph signals (RTS). In addition, using CMS increases costs, such as increasing readout time and reducing frame rate.

[0024] In various examples, the readout circuit determines whether the signal is small or large based on the comparator output. In various examples, if the signal is determined to be small, then a CMS calculation is performed and if the signal is determined to be large, then a non-CMS calculation is performed. Adjusting the calculations performed based on the signal size retains the benefits of using CMS while reducing the readout cycle.

[0025] In various examples of the present disclosure, a readout circuit includes a comparator having a first input coupled to receive a ramp signal from a ramp generator and a second input coupled to receive an analog image data signal from one of a plurality of bit lines. The comparator is configured to generate a comparator output in response to a comparison of the ramp signal and the analog image data signal. The readout circuit also includes a sampling circuit having a first input coupled to receive a sampling control signal and a second input coupled to receive the comparator output. The sampling circuit is configured to generate a sampled output. The readout circuit further includes a counter having a first input coupled to receive a counter control signal and a second input coupled to receive the comparator output and one of the signals from the sampling circuit. The counter is configured to generate a count code. The readout circuit is configured to perform a correlated multiple sampling (CMS) calculation or a non-CMS calculation in response to the sampled output.

[0026] To illustrate, Figure 1 An example of an imaging system 100 having a readout circuit 106 according to the teachings of the present disclosure is shown. In particular, Figure 1 The example depicted in shows an imaging system 100 including a pixel array 102, bit lines 112, control circuitry 110, readout circuitry 106, and function logic 108. In one example, the pixel array 102 is a two-dimensional (2D) array including a plurality of pixel circuits 104 (e.g., P1, P2, ..., Pn) arranged in rows (e.g., R1 to Ry) and columns (e.g., C1 to Cx) to acquire image data of a person, place, object, etc., which can then be used to present an image of the person, place, object, etc.

[0027] In various examples, readout circuitry 106 may be configured to read out image signals via column bit lines 112. As will be discussed, in various examples, readout circuitry 106 may include an analog-to-digital converter (ADC) in accordance with the teachings of the present disclosure. In an example, the digital image data values ​​generated by the analog-to-digital converter in readout circuitry 106 may then be received by function logic 108. Function logic 108 may simply store the digital image data or even manipulate the digital image data by applying post-process image effects (e.g., cropping, rotating, removing red eye, adjusting brightness, adjusting contrast, etc.).

[0028] In one example, control circuit 110 is coupled to pixel array 102 to control the operation of a plurality of photodiodes in pixel array 102. For example, control circuit 110 may generate a rolling shutter or global shutter signal for controlling image acquisition. In other examples, image acquisition is synchronized with a lighting effect such as a flash.

[0029] In one example, the imaging system 100 may be included in a digital, cell phone, laptop computer, endoscope, security camera, or automotive imaging device, or the like. Additionally, the imaging system 100 may be coupled to other hardware, such as a processor (general purpose or otherwise), memory elements, outputs (USB ports, wireless transmitters, HDMI ports, etc.), lighting / strobes, electronic inputs (keyboards, touch screen displays, track pads, mice, microphones, etc.), and / or displays. The other hardware may transmit instructions to the imaging system 100, extract image data from the imaging system 100, or manipulate image data provided by the imaging system 100.

[0030] Figure 2A and 2B Schematic diagrams showing two examples of readout circuits 206 according to the teachings of the present disclosure. It should be understood that Figure 2A and 2B The readout circuit 206 may be included in Figure 1 1. An example of readout circuitry 106 in imaging system 100 is shown in FIG. 1, and similarly named and numbered elements described above are similarly coupled and function below.

[0031] refer to Figure 2A and 2B In both, the readout circuit 206 includes a comparator 216 having a first input coupled to receive a ramp signal VRAMP 214 from a ramp generator and a second input coupled to receive an analog image data signal from one of a plurality of bit lines 212. The comparator 216 is configured to generate a comparator output 230 in response to a comparison of the ramp signal VRAMP 214 and the analog image data signal from the bit line 212. The sampling circuit 220 has a first input coupled to receive a sampling control signal Samp (or Samp1) 240 and a second input coupled to receive the comparator output 230. The sampling circuit 220 is configured to generate a sampling output Samp_out 246. The counter 218 has a first input coupled to receive a counter control signal Count 244 and is configured to generate a count code Data 248.

[0032] exist Figure 2A , sampling circuit 220 includes a latch 222 having a first input coupled to receive Samp 240 and a second input coupled to receive comparator output 230. Latch 222 is configured to generate sample output 246. Counter 218 has a second input coupled to receive comparator output 230.

[0033] exist Figure 2B, the sampling circuit 220 includes a first latch 222 and a second latch 224. The first latch 222 has a first input coupled to receive a first sampling control signal Samp1 240 and a second input coupled to receive a comparator output 230. The first latch 222 is configured to generate a latch output 232. The second latch 224 has a first input coupled to receive a second sampling control signal Samp2 242 and a second input coupled to receive the latch output 232. The second latch 224 is configured to generate a sampling output 246. The counter 218 has a second input coupled to receive the latch output 232.

[0034] Each of latches 222 and 224 is configured to latch (ie, retain) the value of a second input in response to a first input. Figure 2A In , when Samp 240 is high, sample output 246 is equal to comparator output 230. In response to the falling edge of Samp 240, latch 222 latches comparator output 230 and remains equal to that value even after the falling edge of Samp 240 (i.e., when Samp 240 is low). Figure 2B , when Samp2 242 is high, the sample output 246 is equal to the latch output 232. In response to the falling edge of Samp2 242, the second latch 224 latches the latch output 232 and remains equal to the value even after the falling edge of Samp2 242 (i.e., when Samp2 242 is low). When Samp1 240 is high, the latch output 232 is equal to the comparator output 230. In response to the falling edge of Samp1 240, the first latch 222 latches the comparator output 230 and remains equal to the value even after the falling edge of Samp1 240 (i.e., when Samp1 240 is low). As will be described below with reference to Figure 3B In further detail, in various examples, depending on the timing of the ramp signal VRAMP 214 , the second latch 224 may not be necessary.

[0035] Figure 3A and 3B 1 shows a read cycle timing diagram of an example read circuit according to the teachings of the present disclosure. It should be understood that Figure 3A and 3B The timing diagram can be as follows Figure 2A and 2B , and similarly named and numbered elements described above are similarly coupled and function similarly below.

[0036] refer to Figure 3A and 3BIn both cases, the readout cycle includes reading out two pairs (m=2) of reset signals R1, R2 and image signals S1, S2. In other embodiments, a greater number of pairs (e.g., m=4) may be read out. Prior to signal readout, reset signal RST 351 is pulsed to reset the pixel cell being read out. During readout of reset signals R1, R2 and image signals S1, S2, counter control signal Count 344 remains on and ramp signal VRAMP 314 ramps. Between the last reset signal readout R2 and the first image signal readout S1, transfer transistor TX 352 coupled to the photodiode is pulsed to transfer image charge to the floating diffusion region and comparator (e.g., Figure 2A and 2B 216 shown in FIG.

[0037] exist Figure 3A 340 is turned on only during the first image signal readout S1. In the example shown, during the first image signal readout S1, the falling edge of the sampling control signal Samp 340 occurs before the falling edge of the counter control signal Count 344 because there may be a timing gap between the latch (e.g., latch 222) and the counter (e.g., counter 218). The falling edge of Samp 340 occurring before the falling edge of Count 344 ensures that the count code is accurate at the falling edge of Samp 340. In various examples, Count 344 is turned off at the end of the first image signal readout S1 using Samp 340.

[0038] exist Figure 3B In the example shown, the first sampling control signal Samp1 340 is turned on during the readout of all signals of R1, R2, S1, S2. Figure 3A For the reasons mentioned, during the first image signal readout S1, the falling edge of the first sampling control signal Samp1 340 occurs before the falling edge of the counter control signal Count 344. In various examples, Count 344 is turned off at the end of the first image signal readout S1 with Samp1 340. The second sampling control signal Samp2 342 is pulsed between the first and second image signal readouts S1, S2.

[0039] Reference again Figure 3A and 3BIn both cases, VRAMP 314 during the first image signal readout S1 is shorter than VRAMP 314 during the second or last image signal readout S2. This is because the first image signal S1 is used as a test case to determine whether the image signal is small or large. If the image signal is small, the bit line signal will intersect the relatively short VRAMP 314 during S1, causing the comparator to flip. The sampled output will latch the flipped comparator output to indicate to subsequent circuits that the image signal is small and CMS should be performed (e.g., using both image signal readouts S1 and S2). If the image signal is large, the bit line signal will not be able to intersect the relatively short VRAMP 314 during S1. The sampled output will latch the non-flipped comparator output to indicate to subsequent circuits that the image signal is large and CMS should not be performed (i.e., discard the first image signal readout S1 that is inaccurate due to insufficient comparator range, and only use the second image signal readout S2).

[0040] Using the first image signal S1 as a test case is advantageous in reducing the readout cycle. In conventional imaging systems that use CMS for all readouts regardless of signal size, VRAMP 314 during S1 is the same length as VRAMP 314 during S2 because there is no comparator range issue for any readout image signal. However, in an embodiment according to the teachings of the present disclosure, VRAMP 314 during S1 may be shorter than VRAMP 314 during S2 by a value that depends on what the desired threshold between small and large signals is.

[0041] For example, if the resolution of the output signal is 10 bits, the ADC count of the second image signal S2 must be approximately 2 times greater than the ADC count of the first and / or second reset signals R1, R2. 10 However, the ADC count of the first image signal S1 can be almost the same as the ADC count of the smaller reset signals R1, R2. Therefore, if the ADC count of S1 is reduced to 32 counts, the ADC count code of S1 is reduced by 2 compared to the conventional imaging system. 10 -32=992 counts. If the counting frequency is 1.8 GHz, the S1 readout period per frame can be reduced by 551 ns. In addition, due to the smaller voltage swing, the ramp settling time after S1 can also be reduced. For example, the ramp settling time for each readout period can be reduced by 130 ns, so the total readout period can be reduced by about 680 ns. Compared with a conventional imaging system using CMS, an imaging system with a 10-bit resolution using adaptive CMS according to an embodiment of the present technology reduces the readout period by approximately 10%. As the number of CMS sample pairs and / or the ADC resolution increases, the readout period decreases. For example, an imaging system with a 12-bit resolution using adaptive CMS according to an embodiment of the present technology reduces the readout period by approximately 20%.

[0042] In other embodiments, VRAMP 314 during the first image signal readout S1 is longer than the second or subsequent image signal readout S2. The second image signal readout S2 (and any and all subsequent image signal readouts) is used as a test case to determine whether the image signal is small or large. Although having a longer initial ramp results in a slight degradation in the noise performance of the second (and subsequent) image signal readouts, a simpler circuit implementation and circuit control may be used. For example, a second latch (e.g., second latch 224 shown in FIG. 2 ) may not be necessary and a single latch (e.g., first latch 222) may be sufficient to determine whether the signal is small or large (i.e., a test case) after the second or subsequent image signal readout.

[0043] Figure 4 Schematic diagram showing another example of a readout circuit 406 according to the teachings of the present disclosure. It should be understood that Figure 4 The readout circuit 406 may be included in Figure 1 1. An example of readout circuitry 106 in imaging system 100 is shown in FIG. 1, and similarly named and numbered elements described above are similarly coupled and function below.

[0044] The readout circuit 406 includes a comparator 416 having a first input coupled to receive a ramp signal VRAMP 414 from a ramp generator and a second input coupled to receive an analog image data signal from one of the plurality of bit lines 412. The comparator 416 is configured to generate a comparator output 430 in response to a comparison of the ramp signal VRAMP 414 and the analog image data signal from the bit line 412. The sampling circuit 420 has a first input coupled to receive a sampling control signal Samp 440 and a second input coupled to receive the comparator output 430. The sampling circuit 420 is configured to generate a sampling output Samp_out 446. The counter 418 has a first input coupled to receive a counter control signal Count 444, a second input coupled to receive Samp_out 446, and a third input coupled to receive the comparator output 430. The counter 418 is configured to generate a count code Data 448.

[0045] The sampling circuit 420 includes an AND gate 426, a first input of which is coupled to receive a sampling control signal 440 and is configured to generate an AND gate output 432. The sampling circuit 420 also includes a flip-flop 422, a first input of which is coupled to receive the AND gate output 432, a second input of which is coupled to receive the comparator output 430, and a third input of which is coupled to receive a flip-flop reset signal rst450. The flip-flop 422 is configured to generate a sampling output 446. The sampling circuit 420 further includes an inverter 428 coupled to receive the sampling output 446 and configured to generate an inverter output 434, which is coupled to a second input of the AND gate 426.

[0046] Sample output 446 is equal to comparator output 430, and AND gate 428 output is high. In response to the falling edge of Samp 440, flip-flop 422 latches comparator output 430 and remains equal to that value even after the falling edge of Samp 440 (i.e., when Samp440 is low). Due to inverter 428, once flip-flop 422 generates a high value for Samp_out 446, AND gate output 432 is a low value and thus flip-flop 422 continues to generate high values ​​in subsequent readouts regardless of Samp 440 and comparator output 430 until rst 450 resets flip-flop 422.

[0047] Figure 5 FIG. 2 shows a readout cycle timing diagram of another example readout circuit according to the teachings of the present disclosure. It should be understood that Figure 5 The timing diagram can be as follows Figure 4 0048] An example timing diagram of readout circuit 406 is shown in FIG. 406, and similarly named and numbered elements described above are similarly coupled and function below.

[0048] The readout cycle includes reading out four pairs (m=4) of reset signals R1, R2, R3, R4 and image signals S1, S2, S3, S4. In other embodiments, fewer or more pairs may be read out (e.g., m=3, m=8). Before the signals are read out, the reset signal RST 551 is pulsed to reset the pixel cell being read out. During the readout of the reset signals R1, R2, R3, R4 and the image signals S1, S2, S3, S4, the counter control signal Count 544 remains on and the ramp signal VRAMP 514 is ramped. Between the last reset signal readout R4 and the first image signal readout S1, the transfer transistor TX 552 coupled to the photodiode is pulsed to transfer the image charge to the floating diffusion region, which is read out to the comparator (e.g., via the bit line) via the bit line. Figure 4 The bit line 412 of the comparator 416 shown in FIG. 5 ) and the flip-flop reset signal rst 550 are also pulse modulated.

[0049] The sampling control signal Samp 540 is turned on during all image signal readouts except the last one (i.e., S1, S2, S3). In the example shown, the counter control signal Count 544 is turned off at the end of each of the first, second, and third image signal readouts S1, S2, S3. In various examples, due to the above description of Figure 3A For the reasons mentioned, during each of S1 , S2 , and S3 , the falling edge of the sampling control signal Samp 540 may occur before the falling edge of the counter control signal Count 544 .

[0050] The ramp events in VRAMP 514 during S1, S2, S3 are shorter than the ramp events in VRAMP 514 during S4. This is because S1, S2, S3 are used as test cases to determine whether the image signal is small or large. If the image signal is small, then the bit line signal will intersect the relatively short VRAMP 514 during S1, causing the comparator to flip. If the image signal is determined to be small in all three test cases, then the sample output will latch the flipped comparator output to indicate to subsequent circuits that the image signal is small and CMS should be performed (e.g., using all image signals to read out S1, S2, S3, S4). If the image signal is large, then during S1, S2, S3, the bit line signal will not be able to intersect the relatively short ramp events in VRAMP 514. If the image signal is determined to be large under at least one of the three test cases, the sampled output will indicate to subsequent circuitry that the image signal is large and that CMS should not be performed (i.e., discard the first, second, and third image signal readouts S1, S2, S3, which may be inaccurate due to insufficient comparator range, and use only the fourth image signal readout S4). Inverters and AND gates (e.g., Figure 4 The inverter 428 and AND gate 426 shown in FIG. 4 can be used to ensure that if any of the signals during S1, S2, and S3 are large, then the sampled output indication signal is large.

[0051] Using the first, second and third image signals S1, S2, S3 as test cases has the advantage of reducing the readout period. In conventional imaging systems, all readouts use CMS regardless of signal size, and VRAMP 514 during S1, S2, S3 is as long as VRAMP 514 during S4 because there is no comparator range issue for any of the image signals read out. However, in an embodiment according to the teachings of the present disclosure, VRAMP 514 during S1, S2, S3 can be shorter than VRAMP 514 during S4 by a value that depends on what the desired threshold between small and large signals is.

[0052] Figure 66 is a flow chart illustrating an example method 600 for performing adaptive correlated multi-sampling (CMS) according to the teachings of the present disclosure. It should be understood that Figure 6 The processing described in Figure 2A , 2B 4, such as readout circuits 206 and 406. In various examples, method 600 is used to reduce readout cycles.

[0053] At block 610, readout circuitry of an imaging system (e.g., readout circuitry 206 and 406) reads out a plurality of reset signals and a plurality of image signals (e.g., R1, R2, ..., S1, S2, ... Figure 3A , 3B 5). At block 620, the readout circuit determines whether the signal is small or large. In various examples, Figures 2A to 5 The readout circuit and timing diagram shown in are used to perform block 620. If the image signal is small at block 620, the readout circuit or another circuit (e.g., an arithmetic logic unit (ALU)) of the imaging system (the "circuit") continues to perform the CMS calculation at block 630. At block 632, the circuit finds a plurality of CMS differences, where each CMS difference is a difference between one of the image signals and a corresponding one of the reset signals. At block 634, the circuit finds a plurality of CMS products, where each CMS product is a product between one of the CMS differences and one of the plurality of weights. At block 636, the circuit finds the sum of the CMS products.

[0054] If the image signal is large at block 620, the circuit continues to perform non-CMS calculations at block 640. At block 642, the circuit finds the non-CMS difference between the representative image signal and the representative reset signal. The circuit then proceeds to block 644 or block 646. At block 644, the circuit applies random least significant bit (LSB) dithering. Block 644 may be repeated a predetermined number of times. At block 646, the circuit finds the quotient between the non-CMS difference and the divisor.

[0055] At block 644, random LSB dithering is applied to reduce quantization errors that occur during ADC. Random noise (LSBs of quantized samples are randomly increased or decreased by 1) is added to the signal. This random perturbation of the LSBs causes some quantization errors to be evenly distributed across the signal, helping to reduce the average quantization noise.

[0056] As follows about Figure 7 and Fig.10 In further detail, in other embodiments, the blocks may be performed in a different order, or the final result of a CMS or non-CMS calculation may be obtained via a different order of arithmetic operations.

[0057] Figure 77 is a flow chart illustrating an example method of performing a CMS calculation (block 730) according to the teachings of the present disclosure. It should be understood that Figure 7 The processing described in can be Figure 6 Examples of boxes 630, 632, 634 and 636 depicted in .

[0058] At block 732, the circuit finds a plurality of CMS differences. For example, if m=4, the CMS differences may be S1-R1, S2-R2, S3-R3, and S4-R4. Next, the circuit may proceed to block 734a or 734b, depending on which CMS calculation is preferred. At block 734a, the circuit finds a plurality of CMS products between one of the CMS differences and the multiplicative inverse of the number of image signals. For example, if m=4, the multiplicative inverse of 4 is 1 / 4. At block 736 following block 734a, the circuit finds the sum of the CMS products. For example, if m=4, then the sum of the CMS products may be (1 / 4)*(S1-R1)+(1 / 4)*(S2-R2)+(1 / 4)*(S3-R3)+(1 / 4)*(S4-R4), which simplifies to ((S1-R1)+(S2-R2)+(S3-R3)+(S4-R4)) / 4. In other words, the difference between the arithmetic mean of the image signal and the arithmetic mean of the reset signal is derived along the path through block 734a.

[0059] At block 734b, the circuit finds a plurality of CMS products between one of the CMS differences and one of a plurality of weights that sum to 1. For example, if m=4, the plurality of weights may be a1=0.3, a2=0.4, a3=0.1, and a4=0.2. At block 736 following block 734b, the circuit finds the sum of the CMS products. For example, if m=4, the sum of the CMS products may be a1*(S1–R1)+a2*(S2–R2)+a3*(S3–R3)+a4*(S4–R4). In other words, the differences between the differently weighted image signals and the differently weighted reset signals are derived along the path through block 734b.

[0060] In other embodiments, the CMS calculation may be performed in a different arithmetic order. For example, the reset signals may be added together one by one (e.g., R1, R1+R2, R1+R2+R3, R1+R2+R3+R4), then the sum of the reset signals may be flipped to its negative value (e.g., -(R1+R2+R3+R4)), then the image signals may be added one by one (e.g., S1-(R1+R2+R3+R4), S1+S2-(R1+R2+R3+R4), S1+S2+S3-(R1+R2+R3+R4), S1+S2+S3+S4-(R1+R2+R3+R4)), then the result may be divided by the number of image reset signal pairs (e.g., (S1+S2+S3+S4-(R1+R2+R3+R4)) / 4).

[0061] Figure 8 is a diagram of an example method of performing weighted adaptive CMS according to the teachings of the present disclosure. The x-axis represents the signal magnitude and the y-axis represents the weight function n. In various examples, if m=2, the output of the weighted adaptive CMS may be (S1–R1)*(0.5–n)+(S2–R2)*(0.5+n). When the signal is below a first predetermined threshold a, n=0, so the output is (S1–R1)*0.5+(S2–R2)*0.5, in other words, the difference between the arithmetic mean of the image signal and the arithmetic mean of the reset signal. When the signal is above a second predetermined threshold b, n=0.5, so the output is S2–R2, in other words, a correlated double sampling (CDS) calculation rather than a CMS calculation. If the signal is between the first and second predetermined thresholds a and b, for example n=0.25, the output is (S1–R1)*0.25+(S2–R2)*0.75, in other words, the difference between differently weighted image signals and differently weighted reset signals. The weight function n allows the circuit to adaptively perform CMS by applying average CMS, weighted CMS, and CDS depending on the signal size.

[0062] Fig. 9 is a diagram of another example method of performing weighted adaptive CMS according to the teachings of the present disclosure. Noise Optimized CMS (NOCMS) provides a mathematical method for noise optimization by iteratively pairing image signals and reset signals that are read closest to each other and weighting image signal and reset signal pairs that are read out closer to each other more heavily than image signal and reset signal pairs that are read out closer to each other (i.e., X1 is more than X2). N Weighted more heavily). In the example shown, m=5 and the pairs are S1-R5, S2-R4, S3-R3, S4-R2, and S5-R1. In one example, weights may be assigned such that the output is (S1-R5)*0.4+(S2-R4)*0.3+(S3-R3)*0.2+(S4-R2)*0.075+(S5-R1)*0.025.

[0063] Fig.10 1040 is a flow chart illustrating an example method of performing non-CMS calculations (block 1040) according to the teachings of the present disclosure. It should be understood that Fig.10 The processing described in can be Figure 6 Examples of boxes 640, 642, 644 and 646 depicted in .

[0064] The circuit may proceed to one of blocks 1042a, 1042b, and 1042c. At block 1042a, the circuit finds the non-CMS difference between the last of the image signals and one of the reset signals. For example, if m=2, the output may be S2-R2 or S2-R1. If applied Figure 6 In block 646 shown in FIG. , the divisor will be 1.

[0065] At block 1042b, the circuit finds the non-CMS difference between (1) the product between the last of the image signals and the number of reset signals and (2) the sum of the reset signals. Next, the circuit may apply random LSB dithering one or more times at block 1044, or proceed directly to block 1046b, where the circuit finds the quotient between the non-CMS difference and the number of reset signals. For example, if m=4, then the output may be (4*S4-(R1+R2+R3+R4)) / 4.

[0066] At block 1042c, the circuit finds a non-CMS difference between (1) the product between the last of the plurality of image signals and the sum of the plurality of non-CMS weights and (2) the sum of the plurality of non-CMS products between each of the reset signals and each of the non-CMS weights. The circuit may then apply random LSB dithering one or more times at block 1044, or proceed directly to block 1046c, where the circuit finds a quotient between the non-CMS difference and the sum of the plurality of non-CMS weights. For example, if m=4 and the weights are a1, a2, a3, and a4, the output may be (S4*(a1+a2+a3+a4)−(a4*R4+a3*R3+a2*R2+a1*R1) / (a1+a2+a3+a4)=S4−(a4*R4+a3*R3+a2*R2+a1*R1) / (a1+a2+a3+a4).

[0067] In other embodiments, the non-CMS calculations may be performed in a different arithmetic order. For example, the reset signals may be added together one by one (e.g., R1, R1+R2, R1+R2+R3, R1+R2+R3+R4), then the sum of the reset signals may be flipped to its negative value (e.g., -(R1+R2+R3+R4)), then the image signals may be added to match the number of reset signals (e.g., S1-(R1+R2+R3+R4), S1+S1-(R1+R2+R3+R4), S1+S1+S1-(R1+R2+R3+R4), S1+S1+S1+S1-(R1+R2+R3+R4)), then the result may be divided by the number of image reset signal pairs (e.g., (S1+S1+S1+S1-(R1+R2+R3+R4) / 4).

[0068] The above description of the illustrated examples of the present disclosure (including the description in the Abstract) is not intended to be exhaustive or to limit the present disclosure to the precise forms disclosed. Although specific examples of the present disclosure are described herein for illustrative purposes, those skilled in the relevant art will recognize that various modifications are possible within the scope of the present disclosure.

[0069] These modifications may be made to the present disclosure in light of the above detailed description. The terms used in the appended claims should not be interpreted as limiting the present disclosure to the specific examples disclosed in the specification. Instead, the scope of the present disclosure will be determined entirely by the appended claims, which will be interpreted in accordance with established principles of claim interpretation.

Claims

1. A readout circuit (206) for an image sensor, wherein include: a comparator (216) having a first input coupled to receive a ramp signal (214) from a ramp generator and a second input coupled to receive an analog image data signal from one of a plurality of bit lines (212), wherein the comparator is configured to generate a comparator output (230) in response to a comparison of the ramp signal and the analog image data signal; a sampling circuit (220) having a first input coupled to receive a sampling control signal (240) and a second input coupled to receive the comparator output (230), wherein the sampling circuit is configured to generate a sampled output (246); and a counter (218) having a first input coupled to receive a counter control signal (244) and a second input coupled to receive one of the comparator output (230) and a signal (232) from the sampling circuit, wherein the counter is configured to generate a count code (248), Wherein the readout circuit is configured to perform correlated multiple sampling calculations or uncorrelated multiple sampling calculations in response to the sample outputs.

2. The readout circuit according to claim 1, wherein the sampling circuit include: a latch (222) having a first input coupled to receive the sampling control signal (240) and a second input coupled to receive the comparator output (230), wherein the latch is configured to generate the sampling output (246), and wherein the second input of the counter (218) is coupled to receive the comparator output (230).

3. The readout circuit of claim 2 , wherein the sampling output (246) is equal to the comparator output (230) when the sampling control signal is high, wherein the sampling circuit is configured to latch the comparator output (230) in response to a falling edge of the sampling control signal (240), and wherein the sampling output (246) remains equal to the comparator output latched in the sampling circuit after the falling edge of the sampling control signal (240).

4. The readout circuit according to claim 1, wherein the sampling control signal is a first sampling control signal (240), wherein the sampling circuit (220) include: a first latch (222) having a first input coupled to receive the first sampling control signal and a second input coupled to receive the comparator output, wherein the first latch is configured to generate a latch output (232); and a second latch (224) having a first input coupled to receive a second sampling control signal (242) and a second input coupled to receive the latch output (232), wherein the second latch is configured to generate the sampling output (246), and Wherein the second input of the counter (218) is coupled to receive the latch output (232).

5. The readout circuit of claim 4 , wherein when the first sampling control signal (240) is high, the latch output (232) is equal to the comparator output, wherein the first latch is configured to latch the comparator output in response to a falling edge of the first sampling control signal, wherein after the falling edge of the first sampling control signal, the latch output remains equal to the comparator output, wherein when the second sampling control signal (242) is high, the sampling output (246) is equal to the latch output (232), wherein the second latch (224) is configured to latch the latch output (232) in response to a falling edge of the second sampling control signal (242), and wherein after the falling edge of the second sampling control signal (242), the sampling output remains equal to the latch output latched in the second latch (224).

6. The readout circuit according to claim 1, wherein the sampling circuit include: an AND gate (426) having a first input and a second input coupled to receive the sampling control signal (440), wherein the AND gate is configured to generate an AND gate output (432); a flip-flop (422) having a first input coupled to receive the AND gate output (432) and a second input coupled to receive the comparator output (430), wherein the flip-flop (422) is configured to generate the sampled output (446); and an inverter (428) coupled to receive the sampled output (446), wherein the inverter is configured to generate an inverter output (434), and wherein the second input of the AND gate (426) is coupled to receive the inverter output (434), and Wherein the second input of the counter (418) is coupled to receive the comparator output (430).

7. The readout circuit of claim 6, wherein when the AND gate (428) output is high, the sampled output (446) is equal to the comparator output (430), wherein the flip-flop (422) is configured to latch the comparator output (430) in response to a falling edge of the AND gate output (432), and wherein when the AND gate output (432) is low, the sampled output (446) remains equal to the comparator output latched in the flip-flop (422) after the falling edge of the AND gate output (432).

8. The readout circuit of claim 6, wherein the flip-flop has a third input coupled to receive a reset signal (450), and wherein the counter (418) has a third input coupled to receive the sampled output (446).

9. A method of operating an image sensor, wherein include: comparing a ramp signal from a ramp generator and an analog image data signal from one of a plurality of bit lines to generate a comparator output; In response to a sampling control signal, sampling the comparator output with a sampling circuit to generate a sampled output; generating a counter code with a counter circuit in response to a counter control signal and one of the comparator output and a signal from the sampling circuit; reading out a plurality of reset signals during a plurality of reset signal reading cycles; reading out a plurality of image signals during a plurality of image signal readout cycles; and A determination is made in response to the sample output whether to perform correlated multi-sampling computations or uncorrelated multi-sampling computations.

10. The method according to claim 9, wherein the sampling circuit include: a latch having a first input coupled to receive the sampling control signal and a second input coupled to receive the comparator output, wherein the latch is configured to generate the sampling output, and Wherein the second input of the counter is coupled to receive the comparator output.

11. The method according to claim 10, further comprising: include: configuring the sampling control signal to be high during a first one of the plurality of image signal readout cycles and low otherwise; and The counter control signal is configured to be high during the plurality of reset signal readout periods, the plurality of image signal readout periods, and a period immediately following the first one of the plurality of image signal readout periods, and to be low otherwise.

12. The method according to claim 9, wherein the sampling control signal is a first sampling control signal, wherein the sampling circuit include: a first latch having a first input coupled to receive the first sampling control signal and a second input coupled to receive the comparator output, wherein the first latch is configured to generate a latch output; and a second latch having a first input coupled to receive a second sampling control signal and a second input coupled to receive the latch output, wherein the second latch is configured to generate the sampling output, and Wherein the second input of the counter is coupled to receive the latch output.

13. The method according to claim 12, further comprising: include: configuring the first sampling control signal to be high during the plurality of reset signal readout periods and the plurality of image signal readout periods, and low in other cases; configuring the second sampling control signal to be high between a first and a second of the plurality of image signal readout cycles, and low in other cases; and The counter control signal is configured to be high during the plurality of reset signal readout periods, the plurality of image signal readout periods, and a period immediately after a first one of the plurality of image signal readout periods, and to be low otherwise.

14. The method according to claim 9, wherein the sampling circuit include: an AND gate having a first input and a second input coupled to receive the sampled control signal, wherein the AND gate is configured to generate an AND gate output; a flip-flop having a first input coupled to receive the AND gate output and a second input coupled to receive the comparator output, wherein the flip-flop is configured to generate the sampled output; and an inverter coupled to receive the sampled output, wherein the inverter is configured to generate an inverter output, and wherein the second input of the AND gate is coupled to receive the inverter output, and Wherein the second input of the counter is coupled to receive the comparator output.

15. The method according to claim 14, further comprising: include: configuring the sampling control signal to be high during the plurality of image signal readout cycles except for a last one of the plurality of image signal readout cycles and to be low otherwise; and The counter control signal is configured to be high during the plurality of reset signal readout periods and the plurality of image signal readout periods, and to be low in other cases.

16. A method (600) for performing adaptive correlated multisampling, wherein include: reading out a plurality of reset signals and a plurality of image signals; (610) Perform correlated multisampling calculations on small signals, including: (620, 630) Finding a plurality of correlated multisampled differences, wherein each correlated multisampled difference is a difference between one of the image signals and a corresponding one of the reset signals; (632) finding a plurality of correlated multisampling products, wherein each correlated multisampling product is a product between one of the correlated multisampling differences and one of a plurality of correlated multisampling weights, wherein the correlated multisampling weights sum to 1; (234) and Finding the sum of the correlated multi-sample products; (236) Perform uncorrelated multisampling calculations on large signals, including: (620), (640) finding an uncorrelated multi-sampled difference between a representative image signal and a representative reset signal, wherein the representative image signal is based only on a last one of the image signals; (642) and A quotient between the uncorrelated multisampled differences and a divisor is found (646).

17. The method of claim 16, wherein when performing the correlated multisampling calculation: Each of the correlated multi-sampling weights is equal to the multiplicative inverse of a number of image signals (734a).

18. The method of claim 16, wherein when performing the correlated multi-sampling calculation: A first one of the correlated multi-sampled differences is a difference between a first one of the image signals and a last one of the reset signals; A last one of the correlated multi-sampled differences is a difference between a last one of the image signals and a first one of the reset signals; and A first one of the correlated multi-sampling weights multiplied by the first one of the correlated multi-sampling differences is greater than a last one of the weights multiplied by the last one of the correlated multi-sampling differences (734b, FIG. 9).

19. The method of claim 16, wherein when performing the non-correlated multi-sampling calculations: The representative image signal is the last one of the image signals (1042a); The representative reset signal is one of the reset signals (1042a); and The divisor is 1 (1046a).

20. The method of claim 16, wherein when performing the non-correlated multi-sampling computation: The representative image signal is a product of the last one of the image signals and the number of reset signals (1042b); The representative reset signal is the sum of the reset signals (1042b); and The divisor is the number of the reset signals (1046b).

21. The method of claim 16, wherein when performing the non-correlated multi-sampling computation: The representative image signal is a product of the last of the plurality of image signals and a sum of a plurality of non-correlated multi-sampling weights (1042c); The representative reset signal is a sum of a plurality of uncorrelated multi-sampling products, wherein each of the uncorrelated multi-sampling products is a product between each of the reset signals and each of the uncorrelated multi-sampling weights (1042c); and The divisor is the sum of the non-correlated multi-sampling weights (1046c).

22. The method of claim 16, wherein when performing the non-correlated multi-sampling computation: The one of the uncorrelated multi-sampling weights multiplied by a first one of the reset signals is greater than the one of the uncorrelated multi-sampling weights multiplied by a last one of the reset signals.

23. The method of claim 22, wherein performing the non-correlated multi-sampling computation further comprises: Random least significant bit dithering is applied before finding the quotient (644 / 1044).

24. The method of claim 16, wherein the method is performed by readout circuitry of an imaging device.

25. The method of claim 16, wherein the correlated multisampling calculations and the uncorrelated multisampling calculations are performed by logic circuitry external to readout circuitry of an imaging device.

Citation Information

Patent Citations

  • Image sensor power supply rejection ratio noise reduction through ramp generator

    CN106060434A

  • Reading circuit of CMOS image sensor based on CMS

    CN111182246A

  • Arithmetic logic cell design in column analog-to-digital converter with shared gray code generator for correlated multiple sampling

    CN114979508A

  • Receiver and reception method

    JP2016010057A

  • Image sensor with shared gray code generator and parallel column arithmetic logic units

    US20210329185A1