Dynamic vision sensor configured with calibration event signal and method of operating the same

Abstract

A dynamic vision sensor comprises: a dynamic vision sensor pixel array, the dynamic vision sensor pixel array comprising an active area and an optical black area, the active area being configured to output a plurality of first signals corresponding to a plurality of first events indicating a change in light intensity, the optical black area being configured to output a plurality of second signals corresponding to a plurality of second events not indicating a change in light intensity; and a processing circuit, the processing circuit being configured to generate a plurality of event signals based on the plurality of first signals, to generate a plurality of noise signals based on the plurality of second signals, to calculate a value representing a real event among the plurality of first events based on the plurality of event signals and the plurality of noise signals, and to selectively output the plurality of event signals based on the value representing the real event and a threshold value.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the priority benefit of Korean Patent Application No. 10-2019-0036151 filed on March 28, 2019, in the Korean Intellectual Property Office, the disclosure of which is incorporated herein by reference in its entirety. Background Art

[0003] Some example embodiments described herein relate to dynamic vision sensors, and more particularly, to dynamic vision sensors configured to calibrate event signals by using an optical black zone.

[0004] Generally, image sensors can be classified as asynchronously operated image sensors or synchronously operated image sensors. A complementary metal oxide semiconductor (CMOS) image sensor is a typical example of a synchronously operated image sensor. A dynamic vision sensor is a typical example of an asynchronously operated image sensor.

[0005] The dynamic vision sensor detects the change of light intensity due to the movement of the object and generates an event. Therefore, unlike a general CMOS image sensor, when the object does not move (e.g., when the light intensity does not change), no information is provided to the user. In contrast, when the object moves, because the light intensity changes mainly at the boundary or outline of the object, an image associated with the silhouette of the object is provided to the user.

[0006] However, in some cases, in addition to the change in light intensity corresponding to the movement of the object, unexpected events may be output from the pixels of the dynamic vision sensor due to various factors. The unexpected event may distort the information provided to the user. In addition, the unexpected event may increase the workload of the processor that processes the event signal. Therefore, it would be desirable to efficiently remove the unexpected event. Summary of the invention

[0007] Some example embodiments provide apparatus and methods capable of determining whether an event occurring in an active region is from noise.

[0008] According to some example embodiments, a dynamic vision sensor includes: a dynamic vision sensor pixel array, the dynamic vision sensor pixel array including an active area and an optical black area, the active area is configured to output a plurality of first signals corresponding to a plurality of first events indicating a change in light intensity, and the optical black area is configured to output a plurality of second signals corresponding to a plurality of second events not indicating a change in light intensity; and a processing circuit, the processing circuit is configured to generate a plurality of event signals based on the plurality of first signals, generate a plurality of noise signals based on the plurality of second signals, calculate a value representing a real event among the plurality of first events based on the plurality of event signals and the plurality of noise signals, and selectively output the plurality of event signals based on the value representing the real event and a threshold.

[0009] According to some example embodiments, a dynamic vision sensor includes: a dynamic vision sensor pixel array, the dynamic vision sensor pixel array including a plurality of active areas and a plurality of optical black areas, each of the plurality of active areas including a plurality of first pixels, and each of the plurality of optical black areas including a plurality of second pixels; and a processing circuit configured to generate a plurality of event signals based on a plurality of first events detected by a plurality of first pixels of a first active area among the plurality of active areas, generate a plurality of noise signals based on a plurality of second events detected by a plurality of second pixels of a first optical black area among the plurality of optical black areas, determine a number of first events included in a plurality of first events based on the plurality of event signals, determine a number of second events included in a plurality of second events based on the plurality of noise signals, perform a subtraction operation on the number of first events and one of the number of second events or a reference reduction value to generate a value representing a real event among the plurality of first events, and selectively output a plurality of event signals based on the value representing the real event and a threshold.

[0010] According to some example embodiments, a method of operating a dynamic visual sensor including an active area and an optical black area includes: detecting a plurality of first events indicating a change in light intensity by the active area, detecting a plurality of second events not indicating a change in light intensity by the optical black area, generating a plurality of event signals based on the plurality of first events, generating a plurality of noise signals based on the plurality of second events, determining a true event among the plurality of first events by performing a subtraction operation based on the plurality of event signals and the plurality of noise signals, and selectively outputting the plurality of event signals based on the determined result and a threshold. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The above and other objects and features of some example embodiments will become apparent by describing in detail detailed examples of some example embodiments with reference to the accompanying drawings.

[0012] Figure 1The diagram illustrates an electronic device according to some example embodiments.

[0013] Figure 2A is a graph showing the change over time from Figure 1 The second area outputs a cumulative graph of events.

[0014] Figure 2B is a graph showing the Figure 1 A graph of the accumulation of events detected in the first region.

[0015] Figure 3 The diagram conceptually illustrates a method of calibrating an event signal according to some example embodiments.

[0016] Figure 4 is a diagram conceptually illustrating processing of a calibration event signal according to some example embodiments.

[0017] Figure 5 The diagram shows Figure 1 An example configuration of a dynamic vision sensor is shown in FIG.

[0018] Figure 6 The diagram shows Figure 5 A circuit diagram of an example configuration of a DVS pixel of a DVS pixel array.

[0019] Figure 7 The diagram shows Figure 6 Example configuration of the DVS pixel backend circuit.

[0020] Figure 8 The figure shows an example format of a signal output from the dynamic vision sensor of the present disclosure.

[0021] Fig. 9 is a flow chart illustrating a method of operation of a dynamic vision sensor according to some example embodiments.

[0022] Fig.10 The diagram shows Figure 5 An example configuration of an event signal calibrator is shown in FIG.

[0023] Fig.11 The diagram shows Fig.10 An example configuration of a noise value determiner is illustrated in .

[0024] Fig.12 The diagram shows Figure 5 An example configuration of an event signal calibrator is shown in FIG.

[0025] Fig.13 is a flowchart illustrating an operating method of an electronic device according to some example embodiments.

[0026] Fig.14The diagram illustrates an example configuration of an electronic system including a dynamic vision sensor according to some example embodiments. DETAILED DESCRIPTION

[0027] Below, some example embodiments may be described in detail and clearly to the extent that a person having ordinary skill in the art may easily implement the same.

[0028] Processing circuits will be used to implement the components described in the specific embodiments with reference to the terms "unit", "module", "block", "~device or ~component" and the like and the functional blocks illustrated in the figures. As used in the present disclosure, the term 'processing circuit' may refer to, for example, hardware (such as a logic circuit), a hardware / software combination (such as a processor that executes software), or a combination thereof. In some example embodiments, the software may be machine code, firmware, embedded code, and / or application software. For example, the processing circuit may more specifically include, but is not limited to, a central processing unit (CPU), an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA), a system on a chip (SoC), a programmable logic unit, a microprocessor, an application specific integrated circuit (ASIC), a processor, a computer, and the like. Additional hardware that may be used to implement the components and functional blocks includes electrical circuits, electronic circuits, integrated circuits, integrated circuit cores, pressure sensors, inertial sensors, microelectromechanical systems (MEMS), passive components, or a combination thereof.

[0029] Figure 1 An electronic device 100 is illustrated according to some example embodiments.

[0030] The electronic device 100 includes a dynamic vision sensor (DVS) 10 and / or an event signal processor 20. For example, the electronic device 100 may be implemented using one or more electronic devices such as a desktop computer, a laptop computer, a tablet computer, a smart phone, a wearable device, a smart speaker, a home security Internet of Things (IoT), a video game console, a workstation, a server, and / or an autonomous vehicle.

[0031] The dynamic vision sensor 10 can sense changes in light intensity. For example, the dynamic vision sensor 10 can detect an event in which the light intensity increases and / or an event in which the light intensity decreases. The dynamic vision sensor 10 can generate a signal based on the detected intensity change, can process the generated signal, and can provide the result of the processing to the user. In some example embodiments, the dynamic vision sensor 10 may include a DVS pixel array 110, an event detection circuit 120, and / or an event signal calibrator 130. According to some example embodiments, the operations described herein as being performed by the dynamic vision sensor 10, the event signal processor 20, the event detection circuit 120, and / or the event signal calibrator 130 may be performed by a processing circuit.

[0032] The DVS pixel array 110 may include a first area 112 and a second area 114. The first area 112 may detect an increase or decrease in light intensity as an event and may output a signal corresponding to the detected event. The second area 114 may not be affected by the change in light intensity. Therefore, the second area 114 may not output a signal corresponding to the change in light intensity. Alternatively, the second area 114 may output a signal based on noise (hereinafter collectively referred to as noise) occurring due to interference from light incident on the first area 112, thermal electrons generated according to the operation of the electronic device 100, any other factors from the outside, etc. For example, the first area 112 may be an active area, and the second area 114 may be an optical black area. For example, the pixels of the first area 112 and the pixels of the second area 114 may have the same or substantially the same structure.

[0033] The event detection circuit 120 may generate an event signal and / or a noise signal based on a signal from the DVS pixel array 110. In some example embodiments, the event signal may be based on a signal output from the first region 112. For example, the event signal may include information associated with the properties of a real event that occurs due to a change in light when an object moves (e.g., an increase or decrease in light intensity), and the time when the event occurs. In contrast, the noise signal may be based on a signal output from the second region 114. For example, the noise signal may include information associated with the properties of the noise and the time when the noise occurs.

[0034] The event signal calibrator 130 may receive the event signal ES and the noise signal NS. The event signal calibrator 130 may compare the number of event signals based on the events detected by the pixels of the first area 112 with a specific threshold value based on the noise signal NS. The event signal calibrator 130 may block event signals determined to have a relatively large noise ratio among the events detected by the pixels of the first area 112 based on the comparison result. The signal output from the event signal calibrator 130 may be provided to the event signal processor 20 as a calibrated event signal CES.

[0035] The noise output from the first area 112 can be efficiently filtered by the above operation performed by the event signal calibrator 130. As such, the amount of data to be processed by the event signal processor 20 can be reduced. Therefore, the performance of the electronic device 100 can be improved (for example, the event processing speed can be increased) and / or the power consumption can be reduced. The operation of the event signal calibrator 130 will be described in detail below.

[0036] The event signal processor 20 may generate an image to be provided to a user based on the signal output from the event signal calibrator 130. The event signal processor 20 may perform various processing on the signal output from the event signal calibrator 130. For example, as is well known, the event signal processor 20 may calibrate the time stamp values ​​of noise pixels, hot pixels, and / or dead pixels by using the time correlation between the time stamp values ​​of adjacent pixels of the dynamic vision sensor 10.

[0037] Figure 2A is a graph showing the change over time from Figure 1 The second region 114 outputs a cumulative graph of events.

[0038] As described above, the second region 114 (refer to Figure 1 ) is configured so that light incident from the outside is blocked. Thus, the second region 114 cannot directly detect a change in light intensity. Alternatively, the second region 114 may be configured such that light incident from the outside is blocked. Figure 1 ), outputs a signal according to interference of hot electrons, noise, etc. (eg, events) generated by the operation of the electronic device 100. That is, the event related to the first area 112 and the event related to the second area 114 may be different from each other. Figure 2A The example is illustrated in FIG. 1 as "n0" events output from 0 to t0. In some example embodiments, referring to 201 in the figure, the number of events output from the first region 112 may increase linearly over time.

[0039] Figure 2B is a graph showing the Figure 1 A graph of the accumulation of events detected in the first region 112 .

[0040] exist Figure 2B , the example is illustrated as "n1" events detected from 0 to t1 and (n2-n1) events additionally detected from t1 to t2. 202 in the figure may be associated with an event due to noise, rather than a real event occurring according to the movement of the object. In contrast, 203 in the figure may be associated with a real event detected as a change in light intensity. That is, the period from 0 to t1 may be a period in which the movement of the object is not detected, and the period from t1 to t2 may be a period in which the movement of the object is detected.

[0041] As the signal output from the first area 112 and determined to be noise is blocked based on the signal output from the second area 114, the reference Figure 1Some example embodiments are briefly described. This is based on the assumption that the first region 112 and the second region 114 have the same or similar characteristics. That is, it is assumed that the amount of noise output per unit time from the pixels of the first region 112 (e.g., the number of times an indication of detected noise is output) is the same or similar to the amount of noise output per unit time from the pixels of the second region 114. In addition, it is assumed that Figure 2A The distribution of 201 (profile) and Figure 2B The distribution of 202 is identical or nearly identical.

[0042] Figure 3 The conceptual diagram illustrates a method of calibrating an event signal according to some example embodiments. For better understanding, reference will be made to Figure 1 Give a description together.

[0043] The DVS pixel array 310 may include a first region 312 and a second region 314. The DVS pixel array 310 corresponds to Figure 1 The DVS pixel array 110. The first area 312 may be divided into a plurality of active areas each including a plurality of pixels. For example, the first area 312 may be divided into a first active area AR1 to a ninth active area AR9 (e.g., a first active area AR1, a second active area AR2, a third active area AR3, a fourth active area AR4, a fifth active area AR5, a sixth active area AR6, a seventh active area AR7, an eighth active area AR8, and a ninth active area AR9). As described above, the second area 314 may be divided into a first optical black area OB1 to a third optical black area OB3.

[0044] The first active area AR1 may include nine pixels. In some example embodiments, the first active area AR1 includes a pixel PX[1,1] at the first row and the first column and eight surrounding (e.g., associated) pixels. As described above, the third optical black area OB3 includes a pixel PX[9,12] at the ninth row and the twelfth column and eight surrounding pixels. However, the number of pixels included in each active area may be different from the above example. In addition, the second area 314 is illustrated as being adjacent to the right side of the first area 312, but may be placed at various positions of the DVS pixel array 310.

[0045] The event detection circuit 120 may generate an event signal based on a signal corresponding to an event (e.g., an event indication) output by each pixel of the first region 312. The event detection circuit 120 may generate a noise signal based on a signal corresponding to an event (e.g., an event indication) output by each pixel of the second region 314. The event signal may include information about the attributes of the event that occurred and the time when the event occurred.

[0046] The event signal calibrator 130 may count the number of events detected by the pixels of each active area during a unit time (e.g., duration). In some example embodiments, the event signal calibrator 130 may count event signals based on signals output from nine pixels of the first active area AR1 to generate a count signal AR ACC[1]. The counting may be performed equally or similarly for the remaining active areas AR2 to AR9.

[0047] In addition, the event signal calibrator 130 may count the number of noise signals based on the signal (e.g., event indicator) output from the pixel of each optical black area during the unit time. In some example embodiments, the event signal calibrator 130 may count the events (e.g., event indicator) output from the nine pixels of the first optical black area OB1 to generate the count signal OB ACC [1]. The count may be performed equally or similarly for the remaining optical black areas OB2 and OB3. Alternatively, the count may not be performed for the remaining optical black areas OB2 and OB3.

[0048] Regardless of the change in light intensity, count signals OB ACC[1] to OB ACC[3] may be generated based on signals output from the optical black areas OB1 to OB3; because the pixels of the first area 312 and the second area 314 have the same or substantially the same structure, noise corresponding to the count signals OB ACC[1] to OB ACC[3] may be considered to have been applied to the count signals AR ACC[1] to AR ACC[9]. In other words, the signal output from each optical black area corresponds to noise, and the signal output from each active area corresponds to noise and real events. According to some example embodiments, each first active area may have an equal number of pixels or a similar number of pixels as each optical black area.

[0049] According to some example embodiments, the event signal calibrator 130 may generate a reference reduction value from the count signals OB ACC [1] to OB ACC [3]. The reference reduction value may correspond to the number of signals from noise among the signals output from the pixels of each active area. In some example embodiments, the reference reduction value may be an average value of the count signals OB ACC [1] to OB ACC [3]. Alternatively, the reference reduction value may be a maximum (e.g., highest) value or a minimum (e.g., lowest) value of the count signals OB ACC [1] to OB ACC [3].

[0050] The event signal calibrator 130 may subtract a reference reduction value from each of the count signals AR ACC[1] to AR ACC[9]. The event signal calibrator 130 may compare the result value of the subtraction with a threshold value. When the result value of the subtraction is greater than the threshold value, the event output from the active area may be considered to be a real event (e.g., a true event). In contrast, when the result value of the subtraction is less than the threshold value, the event output from the active area may be considered to be an event mainly derived from noise (e.g., a false event). In this case, the event signal calibrator 130 may block signals related to false events (i.e., noise-based events) output from the active area. That is, because the blocked events (reference Figure 1 ), so the speed at which the event signal processor 20 processes events can be improved (e.g., increased).

[0051] Figure 4 is a diagram conceptually illustrating processing of a calibration event signal according to some example embodiments.

[0052] For convenience of description, it is assumed that there is a lack of movement of the object in the period from 0 to t1, the movement of the object in the period from t1 to t2 is relatively large, and the movement of the object in the period from t2 to t3 is relatively small.

[0053] During the period from 0 to t1, even if the movement of the object is not detected, the activity area (e.g., Figure 3 Any one of AR1 to AR9 of FIG. 1 outputs a signal corresponding to "n" events due to noise, etc. 401 in the figure shows that output events are accumulated over time.

[0054] During the period from t1 to t2, because a relatively large movement of an object is detected, the number of events can increase dramatically, and (n2-n1) events can be output from each active region. The events output from the active region during the period from t1 to t2 include real events from the movement of the object and events from noise. 402 in the figure shows that output events are accumulated over time.

[0055] In the period from t2 to t3, because relatively small movement of the object is detected, the number of events can increase smoothly, and (n3-n2) events can be output from the active area. The events output from the active area in the period from t1 to t3 include real events from the movement of the object and events from noise. 403 in the figure shows that the output events are accumulated over time.

[0056] Because noise continuously occurs in the period in which the real event corresponding to the movement of the object occurs and in the period in which the real event does not occur, the events from the noise can be accumulated linearly over time, and 404 in the figure shows the accumulation of events from the noise. That is, the slope of 401 in the figure can be the same as or similar to the slope of 404 in the figure.

[0057] The event signal calibrator 130 may calculate the difference between the number of true events (e.g., n3) and the number of false events (e.g., n4) occurring in each active region, and compare the calculated difference with a threshold value TH. For example, when the calculated difference is greater than the threshold value TH, the event signal output from the active region is considered to be a true event from the movement of the object. In contrast, when the calculated difference is less than the threshold value TH, the event signal output from the active region is considered to be an event from noise that is not associated with the movement of the object.

[0058] In some example embodiments, the threshold TH may be a pre-specified value. In some example embodiments, the threshold TH may be a value measured in real time. In this case, the threshold TH may be the number of false events (e.g., n4) occurring during a unit time (e.g., 0 to t3). In some example embodiments, the threshold TH may be a value variable by a user. For example, in an electronic device (e.g., Figure 1 In the case where the electronic device 100 is operated in a low power mode, the magnitude of the threshold TH can become relatively high. In this case, because the number of events to be processed is reduced, the speed at which the electronic device processes events can be further increased, and power consumption can be further reduced.

[0059] Figure 5 The diagram shows Figure 1 An example configuration of a dynamic vision sensor 10 is illustrated in FIG.

[0060] The dynamic vision sensor 500 may include a DVS pixel array 510, an event detection circuit 520, and / or an event signal calibrator 530. The event detection circuit 520 may include a column address event representation (AER) circuit 522, a row AER circuit 524, and / or an input / output circuit 526. According to some example embodiments, the operations described herein as being performed by the column AER circuit 522, the row AER circuit 524, and / or the input / output circuit 526 may be performed by a processing circuit.

[0061] The DVS pixel array 510 may include a plurality of pixels arranged along a plurality of rows and a plurality of columns in a matrix form. As described above, the DVS pixel array 510 may include a first region 512 (which includes a plurality of active regions) and a second region 514 (which includes a plurality of optical black regions). A pixel that detects an event among the plurality of pixels of the DVS pixel array 510 may output a signal (e.g., a column request) CR indicating the occurrence of an event of an increase or decrease in light intensity to the column AER circuit 522. Here, the column request CR may be associated with a real event detected in the first region 512 or may be associated with noise occurring in the first region 512 or the second region 514.

[0062] The column AER circuit 522 may output an affirmative signal ACK to the pixel that performs event output in response to the column request CR received from the pixel. The pixel that receives the affirmative signal ACK may output polarity information PI of the event to the row AER circuit 524. The polarity information PI may include information for identifying an event in which light intensity increases or decreases. The column AER circuit 522 may generate a column address C_ADDR of the pixel that detects the event based on the column request CR received from the pixel that detects the event.

[0063] The row AER circuit 524 may receive polarity information PI from the pixel in which the event is detected. The row AER circuit 524 may generate a timestamp TS including information about the time when the event occurs based on the polarity information PI. In some example embodiments, the timestamp TS may be generated by a time stamper provided in the row AER circuit 524. For example, the time stamp may be implemented by using a clock tick generated in units of several microseconds to tens of microseconds. The row AER circuit 524 may output a reset signal RST to the pixel in which the event occurs in response to the polarity information PI. The pixel in which the event occurs may be reset by the reset signal RST. In addition, the row AER circuit 524 may generate a row address R_ADDR of the pixel in which the event occurs.

[0064] The row AER circuit 524 may control the period in which the reset signal RST is generated. For example, in order to prevent the workload from increasing due to the occurrence of many events, or to reduce the amount of increase, the row AER circuit 524 may control the period in which the reset signal RST is generated so that the event does not occur during a specific period. That is, the row AER circuit 524 may control the refractory period for the occurrence of an event.

[0065] The input / output circuit 526 may generate an event signal ES or a noise signal NS based on the timestamp TS, the column address C_ADDR, the row address R_ADDR, and / or the polarity information PI. The input / output circuit 526 may add a header indicating the start of transmission of the signal in front of the event signal ES or the noise signal NS or add a tail indicating the end of transmission of the signal behind it.

[0066] Figure 6 The diagram shows Figure 5 A circuit diagram of an example configuration of a DVS pixel of a DVS pixel array.

[0067] The DVS pixel 600 is configured to detect a change in light intensity and output a signal corresponding to the detected intensity change. The DVS pixel 600 may include a photoreceptor 610 and a DVS pixel backend circuit 620. Figure 5 ) may have a DVS pixel with an active area of ​​the first region 512 (reference Figure 5 However, the DVS pixels in the optical black area may be implemented to not receive light.

[0068] The photoreceptor 610 may include a photoelectric conversion element PSD, a logarithmic amplifier LA, and / or a feedback transistor FB. The photoelectric conversion element PSD may generate photoelectrons in response to incident light. That is, the photoelectric conversion element PSD may convert an optical signal into an electrical signal to generate a photocurrent IP. For example, the photoelectric conversion element PSD may include a photodiode, a phototransistor, a pinned phototube, and / or any other similar element. The logarithmic amplifier LA may amplify a voltage corresponding to the photocurrent IP generated by the photoelectric conversion element PSD of the DVS pixel 600. The logarithmic amplifier LA may output a logarithmic voltage VLOG of a logarithmic scale. The feedback transistor FB may separate the photoreceptor 610 from the DVS pixel backend circuit 620.

[0069] The DVS pixel back-end circuit 620 may perform various processes on the logarithmic voltage VLOG. In some example embodiments, the DVS pixel back-end circuit 620 may amplify the logarithmic voltage VLOG, may determine a change in the intensity of light incident on the photoelectric conversion element PSD by comparing the amplified voltage with a reference voltage, and may output an event signal (e.g., an on event or a off event) corresponding to the determined result. After the DVS pixel back-end circuit 620 outputs the on event or the off event, the DVS pixel back-end circuit 620 may be reset by a reset signal RST. According to some example embodiments, the reference voltage may be a design parameter determined by empirical research.

[0070] Figure 7 The diagram shows Figure 6 Example configuration of the DVS pixel backend circuit.

[0071] The DVS pixel backend circuit 700 may include a differentiator 710, a comparator 720, and / or a readout circuit 730. According to some example embodiments, operations described herein as being performed by the DVS pixel backend circuit 700, the comparator 720, and / or the readout circuit 730 may be performed by a processing circuit.

[0072] The differentiator 710 may be configured to amplify the voltage VLOG to generate a voltage VDIFF. For example, the differentiator 710 may include capacitors C1 and C2, a differential amplifier DA, and a switch SW, and the switch SW may operate in response to a reset signal RST. For example, the capacitors C1 and C2 may store electrical energy generated by at least one photoelectric conversion element PSD. For example, the capacitances of the capacitors C1 and C2 may be appropriately selected in consideration of the shortest time (e.g., an inappropriate period) between two events that occur consecutively in one pixel. When the switch SW is turned on by the reset signal RST, the pixel may be initialized (or reset). The row AER circuit 524 (reference signal) may be used to initialize the pixel. Figure 5 ) receives the reset signal RST.

[0073] The comparator 720 may compare the level of the output voltage VDIFF of the differential amplifier DA with the level of the reference voltage Vref, and may determine whether the event occurring in the pixel is an on event or an off event. When an event of an increase in light intensity is detected, the comparator 720 may output a signal ON indicating that the detected event is an on event; when an event of a decrease in light intensity is detected, the comparator 720 may output a signal OFF indicating that the detected event is an off event. According to some example embodiments, the reference voltage Vref may be a design parameter determined through empirical research.

[0074] The readout circuit 730 may output information about an event occurring at a pixel. The information about the event output from the readout circuit 730 may include information (e.g., a bit) about whether the event occurring is an on event or an off event. The information about the event output from the readout circuit 730 may be referred to as polarity information PI (reference Figure 5 ). The polarity information PI may be provided to the row AER circuit 524.

[0075] exist Figure 7 The configuration of the pixel illustrated in FIG. 1 is an example, and some example embodiments may be applied to various configurations of a DVS pixel configured to determine the type of an event based on a result of detecting light intensity.

[0076] Figure 8The figure shows an example format of a signal output from the dynamic vision sensor of the present disclosure.

[0077] The output circuit 800 may correspond to Figure 5 The output circuit 800 can be based on the input / output circuit 526 from the first region 512 (reference Figure 5 ) output signal, from the column AER circuit 522 (reference Figure 5 ) output column address C_ADDR and / or from row AER circuit 524 (reference Figure 5 ) to generate the event signal ES. The output circuit 800 may generate the event signal ES based on the row address R_ADDR output from the second region 514 (reference Figure 5 ) output signal, from the column AER circuit 522 (reference Figure 5 ) output column address C_ADDR and / or from row AER circuit 524 (reference Figure 5 ) to generate the noise signal NS according to the row address R_ADDR outputted by the event signal ES. For example, the event signal ES and the noise signal NS may have the same format or similar formats.

[0078] Continue to refer Figure 8 The signal output from the output circuit 800 may include a timestamp TS, a column address C_ADDR, a row address R_ADDR and / or polarity information PI. The signal output may also include a header H indicating the start of transmission of the signal and a tail T indicating the end of transmission of the signal.

[0079] The time stamp TS may include information about the time when the event occurs. For example, the time stamp TS may be composed of 32 bits, but is not limited to 32 bits.

[0080] Each of the column address C_ADDR and the row address R_ADDR may be composed of 8 bits. 8 Row and 2 8 However, this is only an example, and the number of bits of the column address C_ADDR and the number of bits of the row address R_ADDR may be determined differently according to the number of pixels.

[0081] The polarity information PI may include information about a turn-on event and / or a turn-off event. In some example embodiments, the polarity information PI may be composed of a bit including information about whether a turn-on event occurs and a bit including information about whether a turn-off event occurs. For example, a bit "1" may indicate that an event occurs, while a bit "0" may indicate that an event does not occur. Therefore, an event in which both the bit indicating the turn-on event and the bit indicating the turn-off event are "0" may occur, but an event in which both the bit indicating the turn-on event and the bit indicating the turn-off event are "1" may not occur.

[0082] Fig. 9 is a flow chart illustrating a method of operation of a dynamic vision sensor according to some example embodiments.

[0083] For ease of description, in this example, it is assumed that the noise signal NS means a signal based on an event output from an optical black area, and the event signal ES means a signal based on an event output from an active area. Figure 5 to give a description.

[0084] In operation 910, the event signal calibrator 530 may count the number of noise signals NS output per unit time from the event detection circuit 520. The event signal ES and the noise signal NS have the same format or a similar format, but the event signal calibrator 530 may distinguish the event signal ES from the noise signal NS with reference to the row address R_ADDR and / or the column address C_ADDR.

[0085] In operation 920, the event signal calibrator 530 may count the number of event signals ES output from the event detection circuit 520 per unit time. Also, the event signal calibrator 530 may distinguish the event signal ES from the noise signal NS with reference to the row address R_ADDR and / or the column address C_ADDR.

[0086] In this example, the operation of counting the noise signal is described as being performed before the operation of counting the event signal, but some example embodiments are not limited thereto. For example, the operation of counting the event signal and the noise signal may be performed in parallel, at the same time, or in the same period, or the operation of counting the event signal may be performed before the operation of counting the noise signal.

[0087] In operation 930, the number of events that actually occurred (i.e., real events) may be calculated based on the number of noise signals NS and the number of event signals ES. As described above, since a signal from noise is included in the event signal ES, the number of noise signals NS based on events occurring in the optical black region is considered to have been applied to the event signal ES based on events occurring in the active region.

[0088] In operation 940, the number of real events may be compared with a threshold value TH. The number of real events may be a value obtained by subtracting the number of noise signals NS from the number of event signals ES. When the number of real events is greater than the threshold value TH, operation 950 is performed; when the number of real events is less than the threshold value TH, operation 960 is performed.

[0089] In operation 950, the event signal calibrator 530 may output the event signal ES output from the event detection circuit 520 as a calibrated event signal CES. Before processing by the event signal calibrator 530, the calibrated event signal CES may be the same as or similar to the event signal ES.

[0090] In operation 960, the event signal calibrator 530 may filter (or block) the event signal ES output from the event detection circuit 520. That is, before the processing of the event signal calibrator 530, the calibrated event signal CES may be different from the event signal ES. In detail, since the calibrated event signal CES is considered as an event not occurring (e.g., reflecting that the event does not occur), all bits of the polarity information PI (reference Figure 8 ) can be "0".

[0091] Fig.10 The diagram shows Figure 5 An example configuration of the event signal calibrator 530 is shown in FIG.

[0092] For ease of description, in this example, it is assumed that noise signals NS[1] to NS[m] mean signals based on events output from an optical black area, and event signals ES[1] to ES[n] mean signals based on events output from an active area.

[0093] The event signal calibrator 1000 may include a noise signal accumulator 1010, an event signal accumulator 1020, a noise value determiner 1030, a calibrator 1040, a comparator 1050, and / or a selector 1060. According to some example embodiments, the operations described herein as being performed by the noise signal accumulator 1010, the event signal accumulator 1020, the noise value determiner 1030, the calibrator 1040, the comparator 1050, and / or the selector 1060 may be performed by a processing circuit.

[0094] The noise signal accumulator 1010 may accumulate (eg, receive and / or store) noise signals received from the event detection circuit 520 (see Figure 5 ) among the noise signals outputted by the noise signal accumulator 1010 (e.g., noise signal NS[1], noise signal NS[2], noise signal NS[3] ... noise signal NS[m]) and the number of noise signals NS[1] to NS[m] accumulated thereby can be counted. For this purpose, the noise signal accumulator 1010 can include a buffer and / or a memory, and / or can be implemented using a buffer and / or a memory. The noise signal accumulator 1010 can output a counting signal OB ACC as a counting result (e.g., the noise signal accumulator 1010 can determine the number of noise signals included in the received noise signals NS[1] to NS[m]).

[0095] The event signal accumulator 1020 can accumulate (e.g., receive and / or store) event signals ES[1] to ES[n] (e.g., event signal ES[1], event signal ES[2], event signal ES[3] ... event signal ES[n]) among the event signals output from the event detection circuit 520 and can count the number of event signals ES[1] to ES[n] accumulated thereby. Similarly, the event signal accumulator 1020 can include a buffer and / or a memory, and / or can be implemented using a buffer and / or a memory. The event signal accumulator 1020 can output a counting signal AR ACC as a counting result (e.g., the event signal accumulator 1020 can determine the number of event signals included in the received event signals ES[1] to ES[n]).

[0096] In some example embodiments, the event signal calibrator 1000 may further include a component configured to classify the signal before the noise signal accumulator 1010 and the event signal accumulator 1020 process the signal. For example, the component may refer to the column address and / or row address included in each of the noise signal and / or the event signal to determine whether each event is from the first region 512 (reference Figure 5 ) or the second region 514 (reference Figure 5 ). Depending on the result of the determination, the noise signals NS[1] to NS[m] may be provided to the noise signal accumulator 1010, and / or the event signals ES[1] to ES[n] may be provided to the event signal accumulator 1020. According to some example embodiments, the operations described herein as being performed by components may be performed by a processing circuit.

[0097] The noise value determiner 1030 may select one of the count signal OB ACC and / or the reference reduction value. The signal OBACC and the reference reduction value may correspond to the noise output from each active area. In some example embodiments, the noise value determiner 1030 may output the reference reduction value in a first mode. For example, the reference reduction value may be a predetermined value. In some example embodiments, a register, a memory, etc. may store the reference reduction value. In some example embodiments, the noise value determiner 1030 may output the count signal OB ACC in a second mode. In some example embodiments, the mode control signal MOD may be used to change the operating mode of an electronic device that employs the event signal calibrator 1000 of the present disclosure.

[0098] The noise value determiner 1030 may determine a calibration period of the calibrator 1040. In some example embodiments, the calibration period of the calibrator 1040 may be the same as or different from the period in which the timestamp is generated. The operation of determining the calibration period of the calibrator 1040 will be described in detail below.

[0099] The calibrator 1040 may calculate the remaining real events other than false events among the events occurring in the active area based on the counting signals AR ACC and OB ACC. In other words, the calibrator 1040 may calculate the values ​​related to the real events among the event signals ES[1] to ES[n]. In some example embodiments, the calibrator 1040 may include a subtractor for performing a subtraction operation on the counting signals AR ACC and OB ACC (e.g., the calibrator 1040 may calculate the remaining real events by subtracting the number of noise signals from the number of event signals). The calibrator 1040 may output a signal AR ACC' as a result of the subtraction operation (e.g., the calibrator 1040 may output a value representing the remaining real events). According to some example embodiments, the calibrator 1040 may calculate the remaining real events other than false events among the events, which occur in each active area (e.g., Figure 3 in the active area AR1 to the active area AR9).

[0100] The comparator 1050 may compare the signal AR ACC' with a threshold value TH (e.g., the calibrator 1040 may determine whether the value representing the remaining real events is greater than or less than the threshold value). The comparator 1050 may output a signal for controlling the selector 1060 as a comparison result. When the signal AR ACC' is less than the threshold value TH, the event signals ES[1] to ES[n] may be considered to be noise (even though the event may actually have occurred). In this case, the comparator 1050 may output a signal that allows the selector 1060 to select "0". In contrast, when the signal AR ACC' is greater than the threshold value TH, the signal AR ACC' may output a signal that allows the selector 1060 to select the event signals ES[1] to ES[n]. In other words, the comparison result of the comparator 1050 may point to the event signals ES[1] to ES[n] or "0".

[0101] The selector 1060 may select the event signal ES[1] to ES[n] or "0" based on the comparison result from the comparator 1050 and may output the calibrated event signal CES as the selection result. When the calibrated event signal CES is "0", information indicating that there is no movement of an object in the active area from which the event corresponding to the event signal ES[1] to ES[n] is output may be provided to the user.

[0102] Although not illustrated in the drawings, the noise value determiner 1030 may further include a multiplier that makes it possible to easily perform the subtraction operation of the calibrator 1040. For example, it is assumed that the count signal OB ACC is 0.25. In this case, because the calibrator 1040 uses a decimal point to perform the subtraction operation, the operation may be complicated. In order to prevent the complicated operation, or to reduce the complexity of the operation, a multiplier may be additionally provided for the multiplication operation for converting the count signal OB ACC into an integer. For example, the multipliers may be respectively set on the paths through which the count signals OB ACC and AR ACC are transmitted to the calibrator 1040. According to some example embodiments, the operations described herein as being performed by the multiplier may be performed by a processing circuit.

[0103] Fig.11 The diagram shows Fig.10 An example configuration of the noise value determiner 1030 is illustrated in FIG.

[0104] The noise value determiner 1030 may include a first multiplexer 1110 , a counter 1120 , and / or a second multiplexer 1130 .

[0105] The first multiplexer 1110 may output one of the count signal OB ACC and the reference reduction value in response to the mode control signal MOD. The reference reduction value may be a fixed value, and the count signal OB ACC may be a value that varies with the optical black area (eg, Figure 3 According to some example embodiments, the noise value determiner 1030 may receive the mode control signal MOD from a source outside the electronic device 100 and / or from a processing circuit of the electronic device 100.

[0106] The counter 1120 may determine the reduction period of the calibrator 1040. For example, the counter 1120 may count the timestamp TS as many times as set by the reduction time interval, and may reset the count value to "0" after counting the timestamp TS as many times as set thereby. For example, when the reduction time interval is 3, whenever the timestamp TS is received, the counter 1120 may repeatedly perform a counting operation from 0 to 2. The count value may be input to the second multiplexer 1130. According to some example embodiments, the operations described herein as being performed by the first multiplexer 1110, the second multiplexer 1130, and / or the counter 1120 may be performed by a processing circuit.

[0107] The second multiplexer 1130 may select the output value of the first multiplexer 1110 or “0” based on the count value received from the counter 1120. For example, the second multiplexer 1130 may output “0” when the count value received from the counter 1120 is not “0” and may select the output value of the first multiplexer 1110 when the count value received from the counter 1120 is “0”.

[0108] The output value of the second multiplexer 1130 may determine the reduction period of the calibrator 1040. In detail, when the count value of the counter 1120 is not "0", because "0" selected by the second multiplexer 1130 is input to the calibrator 1040, the subtraction operation of the calibrator 1040 may not be substantially performed. Therefore, the input value AR ACC and the output value ARACC' of the calibrator 1040 may be the same or similar. In contrast, when the count value of the counter 1120 is "0", because the count signal OB ACC or the reference reduction value selected by the first multiplexer 1110 is input to the calibrator 1040 through the second multiplexer 1130, the subtraction operation of the calibrator 1040 is performed. Therefore, the input value AR ACC and the output value AR ACC' of the calibrator 1040 may be different. According to some example embodiments, the reduction time interval may be a design parameter determined by empirical research.

[0109] Fig.12 The diagram shows Figure 5 An example configuration of the event signal calibrator 530 is shown in FIG.

[0110] Fig.10 The example configuration is mostly similar to Fig.12 However, the event signal calibrator 1200 may differ from the event signal calibrator 1000 in that the event signal calibrator 1200 includes a classifier 1201, a plurality of noise signal accumulators (NSAs) 1210_1 to 1210_3 (e.g., NSA 1210_1, NSA 1210_2, and NSA 1210_3), and / or a plurality of layers 1205_1 to 1205_9 (e.g., layer 1205_1, layer 1205_2, ... layer 1205_9), each of which includes an event signal accumulator (e.g., event signal accumulator 1220_1), a calibrator (e.g., calibrator 1240_1), a comparator (e.g., comparator 1250_1), and / or a selector (e.g., selector 1260_1). According to some example embodiments, each of the plurality of noise signal accumulators 1210_1 to 1210_3 may be coupled to a plurality of noise signal accumulators 1210_1 to 1210_3. Fig.10According to some example embodiments, each event signal accumulator (e.g., event signal accumulator 1220_1), calibrator (e.g., calibrator 1240_1), comparator (e.g., comparator 1250_1), and / or selector (e.g., selector 1260_1) of the plurality of layers 1205_1 to 1205_9 may be the same as or similar to the noise signal accumulator 1010 described in association therewith. Fig.10 The event signal accumulator 1020, calibrator 1040, comparator 1050, and / or selector 1060 described in association are the same or similar. According to some example embodiments, the operations described herein as being performed by the classifier 1201 may be performed by a processing circuit.

[0111] Classifier 1201 can obtain information from event detection circuit 520 (refer to Figure 5 )(Detailed description from input / output circuit 526 (refer to Figure 5 )) receives the event signal ES and the noise signal NS. The classifier 1201 can refer to the row address and / or column address of the event signal ES and / or the noise signal NS to determine whether each event belongs to the active area AR1 to AR9 (reference Figure 3 ) and / or optically black areas OB1 to OB3 (reference Figure 3 ). The noise signal classified according to the determination result may be provided to the relevant noise signal accumulator. For example, the noise signals NS[1] to NS[m] corresponding to the events classified as generated in the first optical black area OB1 may be provided to the first noise signal accumulator 1210_1. As in the above description, the event signals classified according to the determination result may be provided to the relevant event signal accumulator. For example, the event signals ES[1] to ES[n] corresponding to the events classified as generated in the first active area AR1 may be provided to the first event signal accumulator 1220_1. The layers 1205_1 to 1205_9 may process the event signals ES[1] to ES[n] received from the active areas AR1 to AR9, respectively.

[0112] The noise signal accumulators 1210_1 to 1210_3 may generate count signals OB ACC by processing noise signals from events occurring in the optical black areas OB1 to OB3. The count signals OB ACC may be an average value, a maximum value (eg, a highest value), or a minimum value (eg, a lowest value) of the outputs of the noise signal accumulators 1210_1 to 1210_3.

[0113] The output value of the noise value determiner 1230 and the threshold TH may be input to the layers 1205_1 to 1205_9, respectively. The output values ​​processed by the layers 1205_1 to 1205_9, respectively, may be output as calibrated event signals CES[1] to CES[9] (e.g., calibrated event signal CES[1], calibrated event signal CES[2], ... calibrated event signal CES[9]).

[0114] Fig.13 is a flowchart illustrating an operating method of the electronic device 1300 according to some example embodiments.

[0115] In operation 1351, light whose intensity varies according to the moving object may be incident on the first area 1312. However, light whose intensity varies may not be incident on the second area 1314.

[0116] Even if the second area 1314 does not detect light whose intensity changes, the second area 1314 may detect a noise-based event, and the detected event may be provided to the event detection circuit 1320 (operation 1352). The event detection circuit 1320 may generate a noise signal NS based on the event received from the second area 1314 (operation 1353). Each noise signal may include the attributes of the event that occurred, information about the time when the event occurred (e.g., a timestamp), the address of the pixel in which the event was detected, etc. The event detection circuit 1320 may provide the generated noise signal NS to the event signal calibrator 1330.

[0117] Events detected in the first region 1312 may also be provided to the event detection circuit 1320 (operation 1354). The event detection circuit 1320 may generate an event signal ES based on the events received from the first region 1312 (operation 1355). Each event signal may include polarity information, a timestamp, an address of a pixel in which the event occurred, etc. The event detection circuit 1320 may provide the generated event signal ES to the event signal calibrator 1330.

[0118] The event signal calibrator 1330 may generate a count signal OB ACC based on the noise signal NS (operation 1356). The count signal OB ACC may indicate the number of events occurring in each optical black area per unit time. The event signal calibrator 1330 may generate a count signal AR ACC based on the event signal ES (operation 1357). The count signal AR ACC may indicate the number of events occurring in each active area per unit time.

[0119] The event signal calibrator 1330 may perform calibration by performing a subtraction operation on the count signals OB ACC and AR ACC (operation 1358 ). This is in accordance with the assumption that as many false events as the number of false events occurring in each optical black area occur in each active area.

[0120] The event signal calibrator 1330 may compare the number of real events obtained according to the result of the calibration with the threshold value TH (operation 1359). When the number of real events is less than the threshold value TH (no), because the number of real events is relatively small, the relevant active area may be considered as an area in which the event does not occur. Therefore, the event signal calibrator 1330 may output "0" as a calibrated event signal (operation 1360). When the number of real events is greater than the threshold value TH (yes), the event signal calibrator 1330 may output the event signal ES received from the event detection circuit 1320 as a calibrated event signal without modification (operation 1361).

[0121] The event signal processor 1340 may perform event signal processing based on the signal received from the event signal calibrator 1330 (operation 1363).

[0122] According to some example embodiments, events determined to be noise among events detected in a DVS pixel array can be efficiently removed. Therefore, the number of events (e.g., false events) processed by an event signal processor can be reduced, thereby increasing the speed at which the event signal processor processes events (e.g., true events). In addition, the power consumption of an electronic device employing the present disclosure can be reduced.

[0123] Fig.14 The diagram illustrates an example configuration of an electronic system including a dynamic vision sensor according to some example embodiments.

[0124] The electronic system 2000 may be implemented using a data processing device that can use or support an interface proposed by the Mobile Industry Processor Interface (MIPI) Alliance. For example, the electronic system 2000 may be implemented using one of electronic devices such as a digital camera, a video camera, a smart phone, a tablet device, and / or a wearable device (e.g., a smart watch or a smart bracelet).

[0125] The electronic system 2000 may include an application processor 2100, a display 2220, an image sensor 2230, and / or a dynamic vision sensor 2290. The application processor 2100 may include a DigRF master 2110, a display serial interface (DSI) host 2120, a camera serial interface (CSI) host 2130, and / or a physical layer 2140.

[0126] The DSI host 2120 may communicate with the DSI device 2225 of the display 2220 through the DSI. For example, an optical serializer SER may be implemented in the DSI host 2120. For example, an optical deserializer DES may be implemented in the DSI device 2225.

[0127] The CSI host 2130 may communicate with the CSI device 2235 of the image sensor 2230 through the CSI. For example, an optical deserializer DES may be implemented in the CSI host 2130. For example, an optical serializer SER may be implemented in the CSI device 2235.

[0128] The electronic system 2000 may further include a radio frequency (RF) chip 2240 for communicating with the application processor 2100. The RF chip 2240 may include a physical layer 2242, a DigRF slave device 2244, and / or an antenna 2246. For example, the physical layer 2242 of the RF chip 2240 and the physical layer 2140 of the application processor 2100 may exchange data with each other according to a DigRF interface supported by the MIPI Alliance.

[0129] The electronic system 2000 may further include a working memory 2250 and / or an embedded / card storage device 2255. The working memory 2250 and / or the embedded / card storage device 2255 may store data provided from the application processor 2100. In addition, the working memory 2250 and / or the embedded / card storage device 2255 may provide the data stored therein to the application processor 2100. For example, the working memory 2250 and / or the embedded / card storage device 2255 may store image data.

[0130] The working memory 2250 may temporarily store data processed or to be processed by the application processor 2100. The working memory 2250 may include a volatile memory such as a static random access memory (SRAM), a dynamic RAM (DRAM), and / or a synchronous DRAM (SDRAM), and / or a nonvolatile memory such as a flash memory, a phase change RAM (PRAM), a magneto-resistive RAM (MRAM), a resistive RAM (ReRAM), and / or a ferroelectric RAM (FRAM).

[0131] The embedded / card storage device 2255 can store data regardless of whether power is supplied. The embedded / card storage device 2255 may include one or more non-volatile memories, memory controllers, and / or buffers. For example, the embedded / card storage device 2255 may include at least one of non-volatile memories such as flash memory, PRAM, MRAM, ReRAM, and / or FRAM. For example, the embedded / card storage device 2255 may be a device such as a secure digital (SD) card and / or an embedded multimedia card (eMMC).

[0132] The electronic system 2000 may communicate with an external system through a communication module such as Worldwide Interoperability for Microwave Access (WiMAX) 2260, a Wireless Local Area Network (WLAN) 2262, and / or an Ultra Wide Band (UWB) 2264. Even though WiMAX 2260, WLAN 2262, and UWB 2264 are mentioned, the electronic system 2000 may further include various communication modules. According to some example embodiments, the communication module of the electronic system 2000 may transmit and / or receive an information signal and / or an image signal.

[0133] The electronic system 2000 may further include a speaker 2270 and / or a microphone 2275 (e.g., for processing voice information). The electronic system 2000 may further include a global positioning system (GPS) device 2280 for processing location information. The electronic system 2000 may include a dynamic visual sensor 2290 of the present disclosure. The dynamic visual sensor 2290 may be configured to calibrate an event signal based on an event output from an active area by using an event output from an optical black area.

[0134] A variety of semiconductor packages can be used to install circuits, chips and devices according to some example embodiments. For example, the following packages can be used to install circuits, chips and devices according to some example embodiments: stacked packages (PoP), ball grid arrays (BGA), chip scale packages (CSP), plastic lead chip carriers (PLCC), molded dual in-line packages (PDIP), die in waffle packages, die in wafer form, chip on board (COB), ceramic dual in-line packages (CERDIP), metric quad flat packages (MQFP), thin quad flat packages (TQFP), small outline integrated circuits (SOIC), shrink small outline packages (SSOP), thin small outline packages (TSOP), system-in-package (SIP), multi-chip packages (MCP), wafer-level manufacturing packages (WFP) and / or wafer-level processed stacked packages (WSP).

[0135] According to some example embodiments, an apparatus and method capable of determining whether an event occurring in an active area including a plurality of DVS pixels is from noise may be provided.

[0136] In detail, according to some example implementations, when an event occurring in an active area is determined to be from noise, the event thus determined may not be output, thereby improving the performance of a processor processing an event signal.

[0137] Furthermore, by preventing a false event signal from being output or reducing its output, power consumption of a processor that processes the event signal can be reduced.

[0138] While certain example embodiments have been described, it will be apparent to those of ordinary skill in the art that various changes and modifications can be made thereto without departing from the spirit and scope thereof as set forth in the following claims.

Claims

1. A dynamic vision sensor, comprising: a dynamic vision sensor pixel array comprising an active area and an optically black area, the active area being configured to output a plurality of first signals corresponding to a plurality of first events indicating a change in light intensity, and the optically black area being configured to output a plurality of second signals corresponding to a plurality of second events not indicating a change in light intensity; as well as The processing circuit is configured to: generating a plurality of event signals based on the plurality of first signals, generating a plurality of noise signals based on the plurality of second signals, calculating a value representing a true event among the plurality of first events by performing a subtraction operation based on the plurality of event signals and the plurality of noise signals, and The plurality of event signals are selectively output based on the values ​​representing real events and a threshold value.

2. The dynamic vision sensor according to claim 1, wherein each corresponding event signal among the plurality of event signals includes polarity information regarding an attribute of a corresponding first event among the plurality of first events, a timestamp indicating when the corresponding first event occurred, and an address of a pixel among a plurality of first pixels constituting the active area in which the corresponding first event was detected; and Each corresponding noise signal among the multiple noise signals includes polarity information related to the attributes of a corresponding second event among the multiple second events, a timestamp indicating when the corresponding second event occurred, and an address of a pixel among the multiple second pixels constituting the optical black area in which the corresponding second event was detected.

3. The dynamic vision sensor according to claim 1, wherein: The processing circuit is configured to: determining a number of first events included in the plurality of first events based on the plurality of event signals; determining a number of second events included in the plurality of second events based on the plurality of noise signals; as well as The value representing the real event is calculated by performing a subtraction operation on the number of the first events and one of the number of the second events or a reference reduction value corresponding to an average value, a maximum value or a minimum value of count values ​​of a signal output from an optical black area.

4. The dynamic vision sensor according to claim 3, wherein: The processing circuit is configured to select one of the number of the second events or the reference reduction value in response to a mode control signal.

5. The dynamic vision sensor according to claim 3, wherein: The processing circuit is configured to: classifying the plurality of event signals with reference to a plurality of first addresses of the active area at which the plurality of first events occur; as well as The plurality of noise signals are classified with reference to a plurality of second addresses of the optical black area where the plurality of second events occur.

6. The dynamic vision sensor according to claim 3, wherein: The processing circuit is configured to: outputting the plurality of event signals in response to the value representing a real event being greater than the threshold; as well as In response to the value representing a real event being less than the threshold, outputting the plurality of event signals is blocked.

7. The dynamic vision sensor according to claim 6, wherein: The threshold value is a predetermined value.

8. The dynamic vision sensor according to claim 6, wherein: The threshold value is a value of the number of the second events.

9. The dynamic vision sensor according to claim 1, wherein: A plurality of first pixels constituting the active area and a plurality of second pixels constituting the optical black area have the same structure.

10. The dynamic vision sensor according to claim 9, wherein Each first pixel among the plurality of first pixels and each second pixel among the plurality of second pixels includes, a photoelectric conversion element configured to generate photoelectrons in response to light, a logarithmic amplifier configured to amplify a voltage corresponding to a photocurrent generated by the photoelectrons to generate an amplified voltage; and The processing circuit is configured to generate information about the change in the light intensity based on the amplified voltage.

11. A dynamic vision sensor, comprising: A dynamic vision sensor pixel array, comprising a plurality of active areas and a plurality of optical black areas, each of the plurality of active areas comprising a plurality of first pixels, and each of the plurality of optical black areas comprising a plurality of second pixels; as well as The processing circuit is configured to: generating a plurality of event signals based on a plurality of first events detected by the plurality of first pixels of a first active area among the plurality of active areas, generating a plurality of noise signals based on a plurality of second events detected by the plurality of second pixels of a first optical black area among the plurality of optical black areas, determining the number of first events included in the plurality of first events based on the plurality of event signals, determining the number of second events included in the plurality of second events based on the plurality of noise signals, performing a subtraction operation on the number of the first events and one of the number of the second events or a reference reduction value to generate a value representing a real event among the plurality of first events, wherein the reference reduction value corresponds to an average value, a maximum value, or a minimum value of count values ​​of a signal output from an optical black area, and The plurality of event signals are selectively output based on the values ​​representing real events and a threshold value.

12. The dynamic vision sensor according to claim 11, wherein: The processing circuit is configured to: outputting the plurality of event signals in response to the value representing a real event being greater than the threshold; as well as In response to the value representing a real event being less than the threshold, outputting the plurality of event signals is blocked.

13. The dynamic vision sensor according to claim 11, wherein: The processing circuit is configured to select one of the number of the second events or the reference reduction value in response to a mode control signal.

14. The dynamic vision sensor according to claim 11, wherein: The number of the plurality of first pixels of the first active area is the same as the number of the plurality of second pixels of the first optical black area.

15. The dynamic vision sensor according to claim 11, wherein: The threshold value is a value of the number of the second events.

16. A method of operating a dynamic vision sensor comprising an active area and an optically black area, the method comprising: detecting, by the active area, a plurality of first events indicative of changes in light intensity; detecting, by the optically black region, a plurality of second events that do not indicate a change in light intensity; generating a plurality of event signals based on the plurality of first events; generating a plurality of noise signals based on the plurality of second events; determining a real event among the plurality of first events by performing a subtraction operation based on the plurality of event signals and the plurality of noise signals; as well as The plurality of event signals are selectively output based on a result of the determination and a threshold value.

17. The method according to claim 16, wherein: Determination includes: determining a number of first events included in the plurality of first events based on the plurality of event signals; determining a number of second events included in the plurality of second events based on the plurality of noise signals; performing a subtraction operation on the number of the first events and the number of the second events to generate a value representing a real event; and The plurality of event signals are selectively output based on the value representing a real event and the threshold value.

18. The method according to claim 17, further comprising: classifying the plurality of event signals with reference to a plurality of first addresses of the active area where the first event occurs; as well as The noise signal is classified with reference to a plurality of second addresses of the optical black area where the second event occurs.

19. The method according to claim 17, wherein: Optional output includes: outputting the plurality of event signals in response to the value representing a real event being greater than the threshold; and In response to the value representing a real event being less than the threshold, outputting the plurality of event signals is blocked.

20. The method according to claim 17, wherein: The threshold value is a value of the number of the second events.

Images