Vision sensor and operating method thereof

Abstract

A method by which a power supply device controls power by using wireless communication, according to an embodiment of the present invention, may include the steps in which: a first processor of the power supply device recognizes a near field communication (NFC) terminal supporting NFC; the first processor obtains permission information from the NFC terminal; the first processor transmits a permission signal for power supply to a second processor included in the power supply device according to the obtained permission information; and the power supply is activated in response to the obtaining of the permission signal.

Description

Vision sensor and its operating method The present invention relates to a vision sensor and an operating method thereof. Specifically, the present invention relates to a vision sensor to which a filtering method for noise reduction is applied, and an operating method thereof. A vision sensor, for example, generates information about the event, that is, an event signal, when an event (e.g., a change in light intensity) occurs, and transmits the event signal to a processor. Meanwhile, events can be triggered by valid signals representing changes in light intensity, but they can also be triggered by noise. Research is needed on vision sensors that can accurately sense events triggered by valid signals, determining whether an event is triggered by noise or a valid signal. The purpose of the present invention is to provide a vision sensor that reduces noise through multiple filters and an operating method thereof. The technical tasks to be achieved by this embodiment are not limited to the technical tasks described above, and other technical tasks can be inferred from the following embodiments. A vision sensor according to one embodiment of the present invention comprises a pixel array including a plurality of pixels arranged in an M x N (M, N are natural numbers) matrix form and generating an event signal; and A filter unit that performs a filtering operation on the generated event signal by utilizing at least one buffer and at least one filter, wherein the filter unit can perform the filtering operation based on the size of the kernel, the number of filters, and the number of buffers by applying a kernel of a preset size. In one embodiment, when the size of the kernel is '3x3' and the number of the filters and the number of the buffers are '1', the filter unit includes a buffer that stores event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; and a filter that performs a filtering operation from the first row to the Mth row for the first column, the second column, and the third column by applying the kernel; wherein the first column, the second column, and the third column may be configured as consecutive columns. In one embodiment, when the size of the kernel is '3x3' and the number of the filters and the number of the buffers are '2', the filter unit includes: a first buffer storing event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; a first filter performing a filtering operation from a first row to an M-th row on the first column, the second column, and the third column by applying the kernel; a second buffer storing a result produced by the filtering operation of the first filter unit and event information of each of a plurality of pixels included in the third column and event information of each of a plurality of pixels included in a fourth column; and a second filter performing a filtering operation from a first row to an M-th row on the second column, the third column, and the fourth column by applying the kernel; wherein the first column, the second column, the third column, and the fourth column may be configured as consecutive columns. In one embodiment, when the size of the kernel is '3x3' and the number of the filters and the number of the buffers are '3', the filter unit includes: a first buffer storing event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; a first filter performing a filtering operation from a first row to an M-th row on the first column, the second column, and the third column by applying the kernel; a second buffer storing a result produced by the filtering operation of the first filter unit and event information of each of a plurality of pixels included in the third column and event information of each of a plurality of pixels included in a fourth column; and a second filter performing a filtering operation from a first row to an M-th row on the second column, the third column, and the fourth column by applying the kernel. A third buffer storing the result produced by the filtering operation of the second filter unit and the event information of each of the plurality of pixels included in the fourth column and the event information of each of the plurality of pixels included in the fifth column; and a third filter performing a filtering operation from the first row to the Mth row on the third column, the fourth column, and the fifth column by applying the kernel; wherein the first column, the second column, the third column, the fourth column, and the fifth column may be configured as consecutive columns. In one embodiment, when the size of the kernel is '3x3' and the number of the filters and the number of the buffers are '4', the filter unit includes: a first buffer storing event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; a first filter unit performing a filtering operation from a first row to an M-th row on the first column, the second column, and the third column by applying the kernel; a second buffer storing a result produced by the filtering operation of the first filter unit and event information of each of a plurality of pixels included in the third column and event information of each of a plurality of pixels included in a fourth column; and a second filter performing a filtering operation from a first row to an M-th row on the second column, the third column, and the fourth column by applying the kernel. A third buffer storing the result produced by the filtering operation of the second filter unit and the event information of each of the plurality of pixels included in the fourth column and the event information of each of the plurality of pixels included in the fifth column; A third filter performing a filtering operation from the first row to the Mth row on the third column, the fourth column and the fifth column by applying the kernel; A fourth buffer storing the result produced by the filtering operation of the third filter and the event information of each of the plurality of pixels included in the fifth column and the event information of each of the plurality of pixels included in the sixth column; And a fourth filter unit performing a filtering operation from the first row to the Mth row on the fourth column, the fifth column and the sixth column by applying the kernel; wherein the first column, the second column, the third column, the fourth column, the fifth column and the sixth column may be configured as consecutive columns. An operating method of a vision sensor according to one embodiment of the present invention may include a step of detecting an event signal from each of the plurality of pixels included in the pixel array; and a step of applying a kernel of a preset size and performing the filtering operation based on the size of the kernel, the number of buffers, and the number of filters. In one embodiment, when the size of the kernel is '3x3' and the number of the filters and the number of the buffers are '1', the step of performing the filtering operation includes the step of storing event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; and the step of performing a filtering operation from the first row to the Mth row for the first column, the second column, and the third column by applying the kernel, wherein the first column, the second column, and the third column may be configured as consecutive columns. In one embodiment, when the size of the kernel is '3x3' and the number of the filters and the number of the buffers are '2', the step of performing the filtering operation includes: storing event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; performing a filtering operation from a first row to an M-th row on the first column, the second column, and the third column by applying the kernel; storing a result produced by the filtering operation of the first filter unit and event information of each of a plurality of pixels included in the third column and event information of each of a plurality of pixels included in a fourth column; and performing a filtering operation from a first row to an M-th row on the second column, the third column, and the fourth column by applying the kernel, wherein the first column, the second column, and the third column may be configured as consecutive columns. In one embodiment, when the size of the kernel is '3x3' and the number of the filters and the number of the buffers are '3', the step of performing the filtering operation includes: storing event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; performing a filtering operation from a first row to an M-th row on the first column, the second column, and the third column by applying the kernel; storing a result produced by the filtering operation of the first filter unit and event information of each of a plurality of pixels included in the third column and event information of each of a plurality of pixels included in a fourth column; performing a filtering operation from a first row to an M-th row on the second column, the third column, and the fourth column by applying the kernel; storing a result produced by the filtering operation of the second filter unit and event information of each of a plurality of pixels included in the fourth column and event information of each of a plurality of pixels included in a fifth column; And a step of performing a filtering operation from the first row to the Mth row for the third column, the fourth column and the fifth column by applying the kernel, wherein the first column, the second column and the third column may be configured as consecutive columns. In one embodiment, when the size of the kernel is '3x3' and the number of the filters and the number of the buffers are '4', the step of performing the filtering operation includes: storing event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; performing a filtering operation from a first row to an M-th row on the first column, the second column, and the third column by applying the kernel; storing a result produced by the filtering operation of the first filter unit and event information of each of a plurality of pixels included in the third column and event information of each of a plurality of pixels included in a fourth column; performing a filtering operation from a first row to an M-th row on the second column, the third column, and the fourth column by applying the kernel; storing a result produced by the filtering operation of the second filter unit and event information of each of a plurality of pixels included in the fourth column and event information of each of a plurality of pixels included in a fifth column; A method for performing a filtering operation from the first row to the Mth row on the third column, the fourth column, and the fifth column by applying the kernel; a step of storing a result produced by the filtering operation of the third filter and event information of each of a plurality of pixels included in the fifth column and event information of each of a plurality of pixels included in the sixth column; and a step of performing a filtering operation from the first row to the Mth row on the fourth column, the fifth column, and the sixth column by applying the kernel; wherein the first column, the second column, and the third column may be configured as consecutive columns. According to the vision sensor and its operating method according to the present invention, events caused by noise can be reduced compared to the prior art. In addition, according to the vision sensor and its operating method according to the present invention, a plurality of consecutive filters are utilized, but the plurality of filters operate in parallel, so that the delay occurring in filtering can be reduced even when utilizing the plurality of filters. FIG. 1 is a block diagram illustrating an image processing system according to an embodiment of the present invention. FIG. 2 is a block diagram showing the configuration of a vision sensor according to one embodiment of the present invention. FIG. 3 is a conceptual diagram showing a schematic configuration of a pixel array according to one embodiment. Figure 4 is a flowchart showing the operation of a vision sensor according to one embodiment of the present invention. FIG. 5a is a block diagram showing a schematic configuration of a filter unit included in a vision sensor according to one embodiment of the present invention. FIG. 5b is a conceptual diagram illustrating the operation of a filter unit included in a vision sensor according to one embodiment of the present invention. FIG. 6a is a block diagram showing a schematic configuration of a filter unit of a vision sensor according to one embodiment of the present invention. Figure 6b is a conceptual diagram illustrating the operation of a filter unit included in a vision sensor according to one embodiment of the present invention. FIG. 7a is a block diagram showing a schematic configuration of a filter unit included in a vision sensor according to one embodiment of the present invention. Figure 7b is a conceptual diagram illustrating the operation of a filter unit included in a vision sensor according to one embodiment of the present invention. FIG. 8A is a block diagram showing a schematic configuration of a filter unit included in a vision sensor according to one embodiment of the present invention. Figure 8b is a conceptual diagram illustrating the operation of a filter unit included in a vision sensor according to one embodiment of the present invention. Hereinafter, various embodiments of the present invention will be described with reference to the accompanying drawings. It should be understood that the present invention is not limited to specific embodiments, but rather encompasses various modifications, equivalents, and / or alternatives of the embodiments of the present invention. In connection with the description of the drawings, similar reference numerals may be used for similar components. In this document, the expressions "have", "may have", "include", or "may include" indicate the presence of a feature (e.g., a number, function, operation, or component such as a part), and do not exclude the presence of additional features. In this document, the expressions "A or B," "at least one of A and / or B," or "one or more of A and / or B" can include all possible combinations of the listed items. For example, "A or B," "at least one of A and B," or "at least one of A or B" can all refer to cases where (1) at least one A is included, (2) at least one B is included, or (3) at least one A and at least one B are included. The terms "first," "second," "first," or "second" used herein may describe various components, regardless of order and / or importance, and are used only to distinguish one component from another, without limiting the components. For example, without departing from the scope of the rights set forth in this document, the first component may be renamed the second component, and similarly, the second component may be renamed the first component. The expression "configured to" as used herein can be used interchangeably with, for example, "suitable for", "having the capacity to", "designed to", "adapted to", "made to", or "capable of". The term "configured to" does not necessarily mean "specifically designed to". In this document, the words "command", "instruction", "control information", "message", "information", "data", "packet", "data packet", "intent" and / or "signal" transmitted and received between the first electronic device(s) and the second electronic device(s) may include or refer to human-perceivable ideas or specific electrical expressions (e.g., digital codes / analog physical quantities) regardless of their expressions. It will be apparent to those skilled in the art to which the invention disclosed in this document pertains that the exemplary expressions listed above may be interpreted in various ways depending on the context in which they are used. In this document, "is greater than B" not only simply means "is greater than B" but also includes the meaning of "is equal to or greater than B." The terms used in this document are used only to describe specific embodiments and may not be intended to limit the scope of other embodiments. The singular expression may include the plural expression unless the context clearly indicates otherwise. Terms used herein, including technical or scientific terms, may have the same meaning as commonly understood by those of ordinary skill in the art described in this document. Terms defined in general dictionaries among the terms used in this document may be interpreted as having the same or similar meaning in the context of the relevant technology, and shall not be interpreted in an idealized or overly formal sense unless explicitly defined in this document. In some cases, even if a term is defined in this document, it cannot be interpreted to exclude the embodiments of this document. Fig. 1 is a block diagram for explaining an image processing system (1) according to an embodiment of the present invention. An image processing system (1) may include a light source (10) and an electronic device (20). According to an embodiment of the present invention, the light source (10) may include at least one light-emitting element. For example, the light source (10) may include an LED, a laser diode, and a VCSEL (Vertical Cavity Surface Emitting Laser). According to an embodiment of the present invention, the electronic device (20) may include a vision sensor (21) and a processor (25). The vision sensor (21) can collect light incident from the light source (10). The vision sensor (21) can detect a change in the intensity of the incident light and output an event signal. For example, when an event occurs in which the intensity of the light increases, the vision sensor (21) can output a corresponding positive event. Conversely, when an event occurs in which the intensity of the light decreases, the vision sensor (21) can output a negative event. According to an embodiment of the present invention, the vision sensor (21) may be a dynamic vision sensor. For example, the vision sensor (21) may access a pixel where a change in light intensity is detected and output an event signal. For example, the change in light intensity may be generated from the movement of a subject being photographed by the vision sensor (21) or from the movement of the vision sensor (21) itself. The event signal may be mainly generated from the outline of the subject. Since the vision sensor (21) outputs only values ​​corresponding to light whose intensity changes, the amount of processing data can be reduced compared to other sensors (e.g., CMOS image sensors, etc.). The vision sensor (21) can detect changes in the intensity of the collected light and provide an event signal to the processor (25). The processor (25) can receive an event signal from the vision sensor (21). The processor (25) can respond appropriately based on the event signal. For example, the processor (25) can execute or change the operation of the electronic device (20) based on the event signal. Figure 2 is a block diagram showing the configuration of a vision sensor (21) according to one embodiment of the present invention. According to one embodiment of the present invention, the vision sensor (21) may include a pixel array (210), a column AER (Address Event Representation) circuit (220), a row AER circuit (230), a filter unit (240), and an output buffer (250). The pixel array (210) may include at least one pixel that detects an event based on a change in incident light brightness. The pixel array (210) may include at least one pixel in which at least one row and at least one column are arranged in a matrix form. A signal notifying that an event in which the light intensity increases or decreases in a pixel included in the pixel array (210) has occurred can be provided from the pixel to the column AER circuit (220). Specifically, the pixel array (210) can detect a change in light from each pixel, output an output voltage, and compare the output voltage with a preset threshold value to determine whether an event has occurred. When an event occurs, the pixel array (210) can provide an event signal to the column AER circuit (220). The column AER circuit (220) can provide a response signal (ACK) to the pixel in response to a signal received from a pixel that has detected an event. The pixel that has received the response signal (ACK) can provide polarity information of the occurred event to the row AER circuit (230). The column AER circuit (220) can generate a column address of a pixel that detected an event based on a signal received from the pixel that detected the event. The low AER circuit (230) can receive polarity information from a pixel that detects an event. Based on the polarity information, the low AER circuit (230) can generate a timestamp containing information about the time at which the event occurred. The low AER circuit (230) can provide a reset signal to the pixel where the event occurred in response to polarity information. The reset signal can reset the pixel where the event occurred. Furthermore, the low AER circuit (230) can generate a row address of the pixel where the event occurred. The filter unit (240) can filter the event signal according to a preset method. According to an embodiment of the present invention, the filter unit (240) can move a kernel having a preset size on a pixel where an event signal is detected, perform an operation on an area overlapping the kernel, and output the calculated result as an event value on the corresponding pixel. According to an embodiment of the present invention, the kernel may have a size of 3x3. According to an embodiment of the present invention, the filter unit (250) can filter the event signal by utilizing at least one or more techniques among erosion, dilation, opening, and closing. The output buffer (250) can generate a packet based on column address, row address, and polarity information. The output buffer (250) can add a header indicating the start of the packet to the front of the packet and a tail indicating the end of the packet to the back of the packet. FIG. 3 is a conceptual diagram showing a schematic configuration of a pixel array (210) according to one embodiment. A plurality of unit pixels (PX) arranged in an MxN array form can be connected to a plurality of row lines (ROW 0 to ROW M-1) and a plurality of column lines (COL 0 to COL N-1). A plurality of unit pixels (PX) included in the pixel array (210) can be scanned in units of rows (ROW) or columns (COL). In one embodiment, while sequentially scanning a plurality of columns (COL), pixel signals can be detected from the unit pixels (PX) and an event signal can be generated using the pixel signals. Figure 4 is a flowchart showing the operation of a vision sensor (21) according to one embodiment of the present invention. In step S401, the vision sensor (21) can detect an event signal. In step S403, the vision sensor (21) can filter the detected event signal using a preset method. According to an embodiment of the present invention, the vision sensor (21) can perform a filtering operation using a preset method on events corresponding to pixels included in a column to be filtered (hereinafter, “target column”) based on the size of the kernel, information about the event detected in step S401, etc. According to an embodiment of the present invention, the vision sensor (21) can generate a kernel of size 'axb' (a and b are natural numbers) to be utilized for filtering operations. Preferably, the vision sensor (21) can generate a kernel of size 3x3 to be utilized for filtering operations. For convenience of explanation, the kernel size is assumed to be 3x3. According to an embodiment of the present invention, the vision sensor (21) Dog's filter( is a natural number) by using '3(the value corresponding to column (b) of the kernel size) + 'Located in the middle of a series of columns Filtering operations can be performed on the target column of the dog. for example, If is '1', the vision sensor (21) can perform a filtering operation on the target column located in the middle of three consecutive columns. Another example is, When is '2', the vision sensor (21) can perform a filtering operation on two target columns located in the middle of four consecutive columns. Another example is, When is '3', the vision sensor (21) can perform a filtering operation on three target columns located in the middle of five consecutive columns. Another example is, When is '4', the vision sensor (21) can perform a filtering operation on the four target columns located in the middle of six consecutive columns. The specific filtering operation process of the vision sensor (21) is described in FIGS. 5a to 8b. In step S405, the vision sensor (21) can perform coding on the filtered event signal. In step S407, the vision sensor (21) can output a coded event. Fig. 5a is a block diagram showing a schematic configuration of a filter unit (500) according to one embodiment of the present invention. Fig. 5b is a conceptual diagram showing the operation of a filter unit (500) according to one embodiment of the present invention. In particular, in Figs. 5a and 5b When is '1', it shows the configuration and operation of the filter unit (500). The filter unit (500) illustrated in Fig. 5a may correspond to the filter unit (240) illustrated in Fig. 2. In Figs. 5a and 5b, the size of the kernel is assumed to be '3x3' for convenience of explanation. According to one embodiment of the present invention, the filter unit (500) may include a first buffer (510) and a first filter (530). The filter unit (500) can store the received event in the first buffer (510). Information about the event (row address, column address, polarity information, etc.) can be stored in the first buffer (510). Based on the stored event, the first filter (520) can perform filtering using a preset method. According to an embodiment of the present invention, the filter unit (500) may include a core buffer, although not illustrated in the drawing. The core buffer may store event information for each of a plurality of pixels included in the pixel array (210). The filter unit (500) may receive event information for each of the pixels included in three consecutive columns to be filtered for the filtering operation of the first filter (530) from the core buffer and store the information in the first buffer (510). According to an embodiment of the present invention, the first filter (520) can generate a kernel of a preset size. For example, the first filter (520) can generate a kernel of a 3x3 size. The first filter (520) can apply the 3x3 kernel to perform a filtering operation on three consecutive columns from the first row to the Mth row. According to an embodiment of the present invention, the first filter (520) can perform a filtering operation based on an event of each of the pixels included in a target column located in the middle of three consecutive columns and an event of each of the pixels adjacent to each of the pixels included in the target column. According to an embodiment of the present invention, the first filter (520) can perform filtering on stored events by utilizing an erosion or dilation technique. Referring to FIG. 5b, the first filter (530) can generate a 3x3 kernel (511) to perform a filtering operation on the first column (COL[1]) to the third column (COL[3]). According to an embodiment of the present invention, the first filter (530) sets the second column (COL[2]) located between the first column (COL[1]) and the third column (COL[3]) as a target column, and selects the first row (ROW[1]) to the eleventh row (ROW[1]) included in the second column (COL[2]) based on the second column (COL[2]).

[0011] ) can sequentially perform filtering operations on each pixel. Referring to FIG. 5b, the first filter (530) is included in the event and kernel (511) of the target pixel (515) to perform a filtering operation on the target pixel (515), and can perform the filtering operation by considering the events of eight pixels adjacent to the target pixel (515). The filter unit (500) can output an event produced by a filtering operation performed by the first filter (530). Although not shown in the drawing, the first filter (530) can perform a filtering operation on the fourth column (COL[4]) to the sixth column (COL[6]) after performing a filtering operation on the first column (COL[1]) to the third column (COL[3]). That is, the first filter (530) can continue to sequentially perform a filtering operation on every three consecutive columns. According to various embodiments of the present invention, the first filter (530) can perform filtering operations on the fourth column (COL[4]) to the sixth column (COL[6]) in parallel with the filtering operations on the first column (COL[1]) to the third column (COL[3]). That is, the first filter (530) can perform filtering operations on every three consecutive columns in parallel. Fig. 6a is a block diagram showing a schematic configuration of a filter unit (600) according to one embodiment of the present invention. Fig. 6b is a conceptual diagram showing the operation of a filter unit (600) according to one embodiment of the present invention. In particular, in Figs. 6a and 6b When is '2', it shows the configuration and operation of the filter unit (500). The filter unit (600) illustrated in Fig. 6a may correspond to the filter unit (240) illustrated in Fig. 2. In Figs. 6a and 6b, the size of the kernel is assumed to be '3x3' for convenience of explanation. According to one embodiment of the present invention, the filter unit (600) may include a first buffer (610), a first filter (620), a second buffer (630), and a second filter (640). The filter unit (600) can store the received event in the first buffer (610). Information about the event (row address, column address, polarity information, etc.) can be stored in the first buffer (610). Based on the stored event, the first filter (620) can perform filtering using a preset method. According to an embodiment of the present invention, the filter unit (600) may include a core buffer, although not illustrated in the drawing. The core buffer may store event information for each of a plurality of pixels included in the pixel array (210). The filter unit (600) may receive event information for each of the pixels included in three consecutive columns to be filtered for the filtering operation of the first filter (620) from the core buffer and store the information in the first buffer (610). According to an embodiment of the present invention, the first filter (620) can generate a kernel of a preset size. For example, the first filter (620) can generate a kernel of a 3x3 size. The first filter (620) can perform a filtering operation on three consecutive columns from the first row to the Mth row by utilizing the 3x3 kernel. According to an embodiment of the present invention, the first filter (620) may perform a filtering operation based on an event of each pixel included in a target column located in the middle of three consecutive columns and an event of each pixel adjacent to each pixel included in the target column. According to an embodiment of the present invention, the result of the filtering operation performed by the first filter (620) may be a changed value (or maintained value) for event information of each pixel included in the target column. The filter unit (600) can store the result produced by the filtering operation of the first filter (620) in the second buffer (630). The second buffer (630) can store information about the event produced by the filtering operation of the first filter (620). The second filter (640) can perform filtering in a preset manner based on the event stored in the second buffer (630). In addition, the filter unit (600) can receive event information of each pixel included in three consecutive columns including the target column to be filtered for the filtering operation of the second filter (640) from the core buffer and store it in the second buffer (630). Specifically, the second filter (640) can generate a kernel of a preset size. For example, the second filter (640) can generate a kernel of a 3x3 size. The second filter (640) can perform a filtering operation on three consecutive columns using the 3x3 kernel. According to an embodiment of the present invention, the second filter (640) can perform a filtering operation based on an event of each pixel included in a target column located in the middle of three consecutive columns and an event of each pixel adjacent to each pixel included in the target column. According to an embodiment of the present invention, the second filter (640) can perform a filtering operation on three consecutive columns including at least one of the columns utilized in the filtering operation of the first filter (620). Preferably, the second filter (640) can perform a filtering operation by including two of the three columns utilized in the filtering operation of the first filter (620). According to an embodiment of the present invention, the filter unit (600) can perform a filtering operation on four consecutive columns by utilizing the first filter (620) and the second filter (640). For the convenience of explanation, the operation order of the first filter (620) and the second filter (640) is described as being fixed, but this is only one embodiment, and the first filter (620) and the second filter (640) can operate simultaneously in parallel. The vision sensor (21) according to the embodiment of the present invention can reduce the delay caused by filtering by having multiple filters operate in parallel. Referring to FIG. 6b, the first filter (620) can generate a first kernel (611) of 3x3 size and perform a filtering operation on the first column (COL[1]) to the third column (COL[3]) by applying the first kernel (611). According to an embodiment of the present invention, the first filter (620) sets the second column (COL[2]) located between the first column (COL[1]) and the third column (COL[3]) as the first target column, and sequentially performs a filtering operation on each of the pixels located from the first row (ROW[1]) to the eleventh row (ROW

[0011] ) included in the second column (COL[2]) based on the second column (COL[2]). Referring to FIG. 6b, the first filter (620) is included in the original event of the first target pixel (615) and the first kernel (611) to perform a filtering operation on the first target pixel (615), and can perform the filtering operation by considering the events of pixels adjacent to the first target pixel (615). The filter unit (600) can store the event produced by the filtering operation performed by the first filter (620) in the second buffer (630). The second filter (640) can generate a second kernel (621) of 3x3 size and perform a filtering operation on the second column (COL[2]) to the fourth column (COL[4]) by applying the second kernel (621). The second filter (640) can perform a filtering operation including the second column (COL[2]) and the third column (COL[3]) considered in the filtering operation of the first filter (620). According to an embodiment of the present invention, the second filter (640) sets the third column (COL[3]) located between the second column (COL[2]) and the fourth column (COL[4]) as the second target column, and sequentially performs a filtering operation on each of the pixels located from the first row (ROW[1]) to the eleventh row (ROW

[0011] ) included in the third column (COL[3]) based on the third column (COL[3]). According to an embodiment of the present invention, the second filter (640) is included in the original event of the second target pixel (625) and the second kernel (621) to perform a filtering operation on the second target pixel (625), and can perform the filtering operation by considering the events of eight pixels adjacent to the second target pixel (625). The filter unit (600) can output an event produced by a filtering operation performed by the second filter (640). Fig. 7a is a block diagram showing a schematic configuration of a filter unit (700) according to one embodiment of the present invention. Fig. 7b is a conceptual diagram showing the operation of a filter unit (700) according to one embodiment of the present invention. In particular, in Figs. 7a and 7b When is '3', it shows the configuration and operation of the filter unit (500). The filter unit (700) illustrated in Fig. 7a may correspond to the filter unit (240) illustrated in Fig. 2. In Figs. 7a and 7b, the size of the kernel is assumed to be '3x3' for convenience of explanation. According to one embodiment of the present invention, the filter unit (700) may include a first buffer (710), a first filter (720), a second buffer (730), a second filter (740), a third buffer (750), and a third filter (760). The filter unit (700) can store the received event in the first buffer (710). Information about the event (row address, column address, polarity information, etc.) can be stored in the first buffer (710). Based on the stored event, the first filter (720) can perform filtering using a preset method. According to an embodiment of the present invention, the filter unit (700) may include a core buffer, although not illustrated in the drawing. The core buffer may store event information for each of a plurality of pixels included in the pixel array (210). The filter unit (700) may receive event information for each of the pixels included in three consecutive columns to be filtered for the filtering operation of the first filter (720) from the core buffer and store the information in the first buffer (710). According to an embodiment of the present invention, the first filter (720) can generate a kernel of a preset size. For example, the first filter (720) can generate a kernel of a 3x3 size. The first filter (720) can perform a filtering operation on three consecutive columns from the first row to the Mth row by utilizing the 3x3 kernel. According to an embodiment of the present invention, the first filter (720) may perform a filtering operation based on an event of each pixel included in a target column located in the middle of three consecutive columns and an event of each pixel adjacent to each pixel included in the target column. According to an embodiment of the present invention, the result of the filtering operation performed by the first filter (720) may be a changed value (or maintained value) for event information of each pixel included in the target column. The filter unit (700) can store the result produced by the filtering operation of the first filter (720) in the second buffer (730). The second buffer (730) can store information about the event produced by the filtering operation of the first filter (720). The second filter (740) can perform filtering in a preset manner based on the event stored in the second buffer (730). In addition, the filter unit (700) can receive event information of each pixel included in three consecutive columns including the first target column to be filtered for the filtering operation of the second filter (740) from the core buffer and store it in the second buffer (730). According to an embodiment of the present invention, the second filter (740) can generate a kernel of a preset size. For example, the second filter (740) can generate a kernel of a 3x3 size. The second filter (740) can perform a filtering operation on three consecutive columns by utilizing the 3x3 kernel. According to an embodiment of the present invention, the second filter (740) may perform a filtering operation based on the events of each of the pixels included in the target column located in the middle of three consecutive columns, and the events of each of the pixels adjacent to each of the pixels included in the target column. The result of the filtering operation performed by the second filter (740) may be a changed value (or maintained value) for the event information of each of the pixels included in the target column. According to an embodiment of the present invention, the second filter (740) can perform a filtering operation on three consecutive columns, including at least one column utilized in the filtering operation of the first filter (720). Preferably, the second filter (740) can perform a filtering operation by including two columns among the three columns utilized in the filtering operation of the first filter (720). The filter unit (700) can store the result produced by the filtering operation of the second filter (740) in the third buffer (750). The third buffer (750) can store information about the event produced by the filtering operation of the second filter (740). The third filter (760) can perform filtering in a preset manner based on the event stored in the third buffer (750). In addition, the filter unit (700) can receive event information of each pixel included in three consecutive columns including the second target column to be filtered for the filtering operation of the third filter (760) from the core buffer and store it in the third buffer (750). According to an embodiment of the present invention, the third filter (760) can generate a kernel of a preset size. For example, the third filter (760) can generate a kernel of a 3x3 size. The third filter (760) can perform a filtering operation on three consecutive columns by utilizing the 3x3 kernel. According to an embodiment of the present invention, the third filter (760) can perform a filtering operation based on an event of each of the pixels included in a target column located in the middle of three consecutive columns and an event of each of the pixels adjacent to each of the pixels included in the target column. According to an embodiment of the present invention, the third filter (760) may perform a filtering operation on at least one column that overlaps with a column utilized in the filtering operation of the second filter (740). Preferably, the third filter (760) may perform a filtering operation by including two columns among the three columns utilized in the filtering operation of the second filter (740). According to an embodiment of the present invention, the filter unit (700) can perform a filtering operation on five consecutive columns by utilizing the first filter (720), the second filter (740), and the third filter (760). For the convenience of explanation, the operation order of the first filter (720), the second filter (740), and the third filter (760) is described as being fixed, but this is only one embodiment, and the first filter (720), the second filter (740), and the third filter (760) may operate simultaneously in parallel. The vision sensor (21) according to the embodiment of the present invention may reduce the delay caused by filtering by having multiple filters operate in parallel. Referring to FIG. 7b, the first filter (720) can generate a first kernel (711) of 3x3 size and perform a filtering operation on the first column (COL[1]) to the third column (COL[3]) by applying the first kernel (711). According to an embodiment of the present invention, the first filter (720) sets the second column (COL[2]) located between the first column (COL[1]) and the third column (COL[3]) as the first target column, and sequentially performs a filtering operation on each of the pixels located from the first row (ROW[1]) to the eleventh row (ROW

[0011] ) included in the second column (COL[2]) based on the second column (COL[2]). Referring to FIG. 7b, the first filter (720) is included in the original event of the first target pixel (715) and the first kernel (711) to perform a filtering operation on the first target pixel (715), and can perform the filtering operation by considering the events of pixels adjacent to the first target pixel (715). The filter unit (700) can store the event produced by the filtering operation performed by the first filter (720) in the second buffer (730). The second filter (740) can generate a second kernel (721) of 3x3 size and perform a filtering operation on the second column (COL[2]) to the fourth column (COL[4]) by applying the second kernel (721). The second filter (740) can perform a filtering operation including the second column (COL[2]) and the third column (COL[3]) considered in the filtering operation of the first filter (720). According to an embodiment of the present invention, the second filter (740) sets the third column (COL[3]) located in the middle of the second column (COL[2]) and the fourth column (COL[4]) as the second target column, and sequentially performs a filtering operation on each of the pixels located in the first row (ROW[1]) to the eleventh row (ROW

[0011] ) included in the third column (COL[3]) based on the third column (COL[3]). According to an embodiment of the present invention, the second filter (740) is included in the original event of the second target pixel (725) and the second kernel (721) to perform a filtering operation on the second target pixel (725), and can perform the filtering operation by considering the events of eight pixels adjacent to the second target pixel (725). The filter unit (700) can store the event produced by the filtering operation performed by the second filter (740) in the third buffer (750). The third filter (760) can generate a third kernel (731) of 3x3 size and perform a filtering operation on the third column (COL[2]) to the fifth column (COL[4]) by applying the third kernel (731). The third filter (760) can perform a filtering operation including the third column (COL[3]) and the fourth column (COL[4]) considered in the filtering operation of the second filter (740). According to an embodiment of the present invention, the third filter (760) sets the fourth column (COL[4]) located between the third column (COL[3]) and the fifth column (COL[5]) as the third target column, and sequentially performs a filtering operation on each of the pixels included in the fourth column (COL[4]) and located in the first row (ROW[1]) to the eleventh row (ROW

[0011] ) based on the fourth column (COL[4]). According to an embodiment of the present invention, the third filter (760) is included in the original event of the third target pixel (735) and the third kernel (731) to perform a filtering operation on the third target pixel (735), and can perform the filtering operation by considering the events of eight pixels adjacent to the third target pixel (735). The filter unit (700) can output an event produced by a filtering operation performed by the third filter (760). Fig. 8a is a block diagram showing a schematic configuration of a filter unit (800) according to one embodiment of the present invention. Fig. 8b is a conceptual diagram showing the operation of a filter unit (800) according to one embodiment of the present invention. In particular, in Figs. 8a and 8b When '4' is shown, the configuration and operation of the filter unit (500) are shown. The filter unit (800) shown in Fig. 8a may correspond to the filter unit (240) shown in Fig. 2. In Figs. 8a and 8b, the size of the kernel is assumed to be '3x3' for convenience of explanation. According to one embodiment of the present invention, the filter unit (800) may include a first buffer (810), a first filter (820), a second buffer (830), a second filter (840), a third buffer (850), a third filter (860), a fourth buffer (870), and a fourth filter (880). The filter unit (800) can store the received event in the first buffer (810). Information about the event (row address, column address, polarity information, etc.) can be stored in the first buffer (810). Based on the stored event, the first filter (820) can perform filtering using a preset method. According to an embodiment of the present invention, the filter unit (800) may include a core buffer, although not illustrated in the drawing. The core buffer may store event information for each of a plurality of pixels included in the pixel array (210). The filter unit (800) may receive event information for each of the pixels included in three consecutive columns to be filtered for the filtering operation of the first filter (820) from the core buffer and store the information in the first buffer (810). According to an embodiment of the present invention, the first filter (820) can generate a kernel of a preset size. For example, the first filter (820) can generate a kernel of a 3x3 size. The first filter (820) can perform a filtering operation on three consecutive columns from the first row to the Mth row by utilizing the 3x3 kernel. According to an embodiment of the present invention, the first filter (820) may perform a filtering operation based on an event of each of the pixels included in a target column located in the middle of three consecutive columns, and an event of each of the pixels adjacent to each of the pixels included in the target column. According to an embodiment of the present invention, the result of the filtering operation performed by the first filter (720) may be a changed value (or maintained value) for event information of each of the pixels included in the first target column. The filter unit (800) can store the result produced by the filtering operation of the first filter (820) in the second buffer (830). The second buffer (830) can store information about the event produced by the filtering operation of the first filter (820). The second filter (840) can perform filtering in a preset manner based on the event stored in the second buffer (830). In addition, the filter unit (800) can receive event information of each pixel included in three consecutive columns including the first target column to be filtered for the filtering operation of the second filter (840) from the core buffer and store it in the second buffer (830). According to an embodiment of the present invention, the second filter (840) can generate a kernel of a preset size. For example, the second filter (840) can generate a kernel of a 3x3 size. The second filter (840) can perform a filtering operation on three consecutive columns using the 3x3 kernel. The result of the filtering operation performed by the second filter (840) can be a changed value (or maintained value) for the event information of each pixel included in the second target column. According to an embodiment of the present invention, the second filter (840) can perform a filtering operation based on an event of each pixel included in a target column located in the middle of three consecutive columns and an event of each pixel adjacent to each pixel included in the target column. According to an embodiment of the present invention, the second filter (840) can perform a filtering operation on three consecutive columns, including at least one column utilized in the filtering operation of the first filter (820). Preferably, the second filter (840) can perform a filtering operation by including two columns among the three columns utilized in the filtering operation of the first filter (820). The filter unit (800) can store the result produced by the filtering operation of the second filter (840) in the third buffer (850). The third buffer (850) can store information about the event produced by the filtering operation of the second filter (840). The third filter (860) can perform filtering in a preset manner based on the event stored in the third buffer (850). In addition, the filter unit (800) can receive event information of each pixel included in three consecutive columns including the third target column to be filtered for the filtering operation of the third filter (860) from the core buffer and store it in the third buffer (850). According to an embodiment of the present invention, the third filter (860) can generate a kernel of a preset size. For example, the third filter (860) can generate a kernel of a 3x3 size. The third filter (860) can perform a filtering operation on three consecutive columns by utilizing the 3x3 kernel. According to an embodiment of the present invention, the third filter (860) can perform a filtering operation based on an event of each pixel included in a target column located in the middle of three consecutive columns and an event of each pixel adjacent to each pixel included in the target column. According to an embodiment of the present invention, the third filter (860) may perform a filtering operation on at least one overlapping column with a column used in the filtering operation of the second filter (840). Preferably, the third filter (860) may perform the filtering operation by including two columns among the three columns used in the filtering operation of the second filter (840). The result of the filtering operation performed by the third filter (860) may be a changed value (or maintained value) for the event information of each of the pixels included in the third target column. The filter unit (800) can store the result produced by the filtering operation of the third filter (860) in the fourth buffer (870). The fourth buffer (870) can store information about the event produced by the filtering operation of the third filter (860). The fourth filter (880) can perform filtering in a preset manner based on the event stored in the fourth buffer (870). In addition, the filter unit (800) can receive event information of each pixel included in three consecutive columns including the third target column to be filtered for the filtering operation of the fourth filter (880) from the core buffer and store it in the fourth buffer (870). According to an embodiment of the present invention, the fourth filter (870) can generate a kernel of a preset size. For example, the fourth filter (870) can generate a kernel of a 3x3 size. The fourth filter (870) can perform a filtering operation on three consecutive columns by utilizing the 3x3 kernel. According to an embodiment of the present invention, the fourth filter (870) can perform a filtering operation based on an event of each of the pixels included in a target column located in the middle of three consecutive columns and an event of each of the pixels adjacent to each of the pixels included in the target column. According to an embodiment of the present invention, the fourth filter (870) may perform a filtering operation on at least one column that overlaps with the column utilized in the filtering operation of the third filter (860). Preferably, the fourth filter (880) may perform a filtering operation by including two columns among the three columns utilized in the filtering operation of the third filter (860). The filter unit (800) can output the result of the filtering operation performed by the fourth filter (880) to the outside. According to an embodiment of the present invention, the filter unit (800) can perform a filtering operation on six consecutive columns by utilizing the first filter (820), the second filter (840), the third filter (860), and the fourth filter (880). For the convenience of explanation, the operation order of the first filter (820), the second filter (840), the third filter (860), and the fourth filter (880) is described as being fixed, but this is only one embodiment, and the first filter (820), the second filter (840), the third filter (860), and the fourth filter (880) can operate simultaneously in parallel. The vision sensor (21) according to the embodiment of the present invention can reduce the delay caused by filtering by operating multiple filters in parallel. Referring to FIG. 8b, the first filter (820) can generate a first kernel (811) of 3x3 size and perform a filtering operation on the first column (COL[1]) to the third column (COL[3]) by applying the first kernel (811). According to an embodiment of the present invention, the first filter (820) sets the second column (COL[2]) located between the first column (COL[1]) and the third column (COL[3]) as the first target column, and sequentially performs a filtering operation on each of the pixels located from the first row (ROW[1]) to the eleventh row (ROW

[0011] ) included in the second column (COL[2]) based on the second column (COL[2]). Referring to FIG. 8b, the first filter (820) is included in the original event of the first target pixel (815) and the first kernel (811) to perform a filtering operation on the first target pixel (815), and can perform the filtering operation by considering the events of pixels adjacent to the first target pixel (815). The filter unit (800) can store the event produced by the filtering operation performed by the first filter (820) in the second buffer (830). The second filter (840) can generate a second kernel (821) of 3x3 size and perform a filtering operation on the second column (COL[2]) to the fourth column (COL[4]) by applying the second kernel (821). The second filter (840) can perform a filtering operation including the second column (COL[2]) and the third column (COL[3]) considered in the filtering operation of the first filter (820). According to an embodiment of the present invention, the second filter (840) sets the third column (COL[3]) located between the second column (COL[2]) and the fourth column (COL[4]) as the second target column, and sequentially performs a filtering operation on each of the pixels located from the first row (ROW[1]) to the eleventh row (ROW

[0011] ) included in the third column (COL[3]) based on the third column (COL[3]). According to an embodiment of the present invention, the second filter unit (840) is included in the original event of the second target pixel (825) and the second kernel (821) to perform a filtering operation on the second target pixel (825), and can perform the filtering operation by considering the events of eight pixels adjacent to the second target pixel (825). The filter unit (800) can store the event produced by the filtering operation performed by the second filter (840) in the third buffer (850). The third filter (860) can generate a third kernel (831) of 3x3 size and perform a filtering operation on the third column (COL[3]) to the fifth column (COL[5]) by applying the third kernel (831). The third filter (860) can perform a filtering operation including the third column (COL[3]) and the fourth column (COL[4]) considered in the filtering operation of the second filter (840). According to an embodiment of the present invention, the third filter (860) sets the fourth column (COL[4]) located between the third column (COL[3]) and the fifth column (COL[5]) as the third target column, and sequentially performs a filtering operation on each of the pixels included in the fourth column (COL[4]) and located in the first row (ROW[1]) to the eleventh row (ROW

[0011] ) based on the fourth column (COL[4]). According to an embodiment of the present invention, the third filter (860) is included in the original event of the third target pixel (835) and the third kernel (831) to perform a filtering operation on the third target pixel (835), and can perform the filtering operation by considering the events of pixels adjacent to the third target pixel (835). The filter unit (800) can store the event produced by the filtering operation performed by the third filter (860) in the fourth buffer (870). The fourth filter (880) can generate a fourth kernel (841) of 3x3 size and perform a filtering operation on the fourth column (COL[4]) to the sixth column (COL[6]) by applying the fourth kernel (841). The fourth filter (880) can perform a filtering operation including the fourth column (COL[4]) and the fifth column (COL[5]) considered in the filtering operation of the third filter (860). According to an embodiment of the present invention, the fourth filter (880) sets the fifth column (COL[5]) located between the fourth column (COL[4]) and the sixth column (COL[6]) as the fourth target column, and sequentially performs a filtering operation on each of the pixels included in the fifth column (COL[5]) and located from the first row (ROW[1]) to the eleventh row (ROW

[0011] ) based on the fifth column (COL[5]). According to an embodiment of the present invention, the fourth filter (880) is included in the original event of the fourth target pixel (845) and the fourth kernel (841) to perform a filtering operation on the fourth target pixel (845), and can perform the filtering operation by considering the events of pixels adjacent to the fourth target pixel (845). The filter unit (800) can output the result produced by the filtering operation performed by the fourth filter (880) to the outside. Although all components constituting the embodiments of the present invention have been described as being combined or operating in combination as one, the present invention is not necessarily limited to such embodiments. That is, within the scope of the present invention, all of the components may be selectively combined and operated one or more times. Meanwhile, the various embodiments described herein may be implemented by hardware, middleware, microcode, software, and / or a combination thereof. For example, the various embodiments may be implemented in one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, other electronic units designed to perform the functions presented herein, or a combination thereof. Additionally, for example, various embodiments may be embodied or encoded in a computer-readable medium containing instructions. Instructions embodied or encoded in the computer-readable medium may cause a programmable processor or other processor to perform a method when the instructions are executed, for example. The computer-readable medium includes a computer storage medium, which may be any available medium that can be accessed by a computer. For example, such a computer-readable medium may include a RAM, a ROM, an EEPROM, a CD-ROM or other optical disk storage medium, a magnetic disk storage medium, or other magnetic storage devices. Such hardware, software, firmware, etc. may be implemented within the same device or within separate devices to support the various operations and functions described herein. Additionally, components, units, modules, components, etc. described as “units” in the present invention may be implemented together or individually as separate but interoperable logic devices. The depiction of different features for modules, units, etc. is intended to highlight different functional embodiments and does not necessarily imply that they must be realized by separate hardware or software components. Rather, the functionality associated with one or more modules or units may be performed by separate hardware or software components, or integrated into common or separate hardware or software components. Although operations are depicted in the drawings in a particular order, this should not be construed as requiring that these operations be performed in the particular order depicted, or in any sequential order, or that all depicted operations be performed to achieve the desired results. In certain circumstances, multitasking and parallel processing may be advantageous. Furthermore, the distinction between various components in the embodiments described above should not be construed as requiring such distinction in all embodiments, and it should be understood that the components depicted may generally be integrated together into a single software product or packaged into multiple software products. The electronic device, server, or external device according to the various embodiments of the present document described above may include, for example, at least one of a smartphone, a tablet PC, a mobile phone, a video phone, a desktop PC, a laptop PC, a personal digital assistant (PDA), a portable multimedia player (PMP), an MP3 player, a mobile medical device, a camera, or a wearable device. According to various embodiments, the wearable device may include at least one of an accessory type (e.g., a watch, a ring, a bracelet, an anklet, a necklace, glasses, contact lenses, or a head-mounted device (HMD)), a fabric or clothing-integrated type (e.g., an electronic garment), a body-attached type (e.g., a skin pad or tattoo), or a bio-implant type (e.g., an implantable circuit). In some embodiments, the electronic device or external device may be a home appliance. The home appliance may include, for example, at least one of a television, a digital video disk player (DVD player), an audio device, a refrigerator, an air conditioner, a vacuum cleaner, an oven, a microwave oven, a washing machine, an air purifier, a set-top box, a home automation control panel, a security control panel, a TV box, a game console, an electronic dictionary, an electronic key, a camcorder, or an electronic picture frame. In another embodiment, the electronic device, external device, or wearable device may include at least one of various medical devices (e.g., various portable medical measuring devices (such as a blood glucose meter, a heart rate meter, a blood pressure meter, or a body temperature meter), magnetic resonance angiography (MRA), magnetic resonance imaging (MRI), computed tomography (CT), a camera, or an ultrasound machine), a navigation device, a satellite navigation system (Global Navigation Satellite System (GNSS)), an event data recorder (EDR), a flight data recorder (FDR), an automobile infotainment device, a home robot, or an internet of things device (e.g., a light bulb, various sensors, an electric or gas meter, a sprinkler device, a fire alarm, a thermostat, a streetlight, an exercise machine, a hot water tank, a heater, a boiler, or the like). As described above, the best practice embodiments have been disclosed in the drawings and specifications. While specific terminology has been used herein, it is solely for the purpose of describing the present invention and is not intended to limit the scope of the invention as defined in the claims. Therefore, those skilled in the art will understand that various modifications and equivalent embodiments are possible. Therefore, the true technical protection scope of the present invention should be determined by the technical spirit of the appended claims.

Claims

1. A pixel array including a plurality of pixels arranged in a matrix form of M x N (M, N are natural numbers) and generating an event signal; and A filter unit that performs a filtering operation on the generated event signal by utilizing at least one buffer and at least one filter. Including, The above filter part, Applying a kernel of a preset size, performing the filtering operation based on the size of the kernel, the number of filters, and the number of buffers. Vision sensor.

2. In claim 1, If the size of the above kernel is '3x3' and the number of filters and the number of buffers are '1', The above filter part A buffer storing event information of each of a plurality of pixels included in the first column, event information of each of a plurality of pixels included in the second column, and event information of each of a plurality of pixels included in the third column; and A filter that performs a filtering operation from the first row to the Mth row for the first column, the second column, and the third column by applying the kernel; Including, The first column, the second column and the third column are composed of consecutive columns. Vision sensor.

3. In claim 1, If the size of the above kernel is '3x3' and the number of filters and the number of buffers are '2', The above filter part A first buffer that stores event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; A first filter that performs a filtering operation from the first row to the Mth row for the first column, the second column, and the third column by applying the kernel; A second buffer that stores the result produced by the filtering operation of the first filter unit, event information of each of the plurality of pixels included in the third column, and event information of each of the plurality of pixels included in the fourth column; and A second filter that performs a filtering operation from the first row to the Mth row for the second column, the third column, and the fourth column by applying the kernel; Including, The first column, the second column, the third column and the fourth column are composed of consecutive columns. Vision sensor.

4. In claim 1, If the size of the above kernel is '3x3' and the number of filters and the number of buffers are '3', The above filter part A first buffer that stores event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; A first filter that performs a filtering operation from the first row to the Mth row for the first column, the second column, and the third column by applying the kernel; A second buffer that stores the result produced by the filtering operation of the first filter unit, event information of each of the plurality of pixels included in the third column, and event information of each of the plurality of pixels included in the fourth column; and A second filter that performs a filtering operation from the first row to the Mth row for the second column, the third column, and the fourth column by applying the kernel; A third buffer that stores the result produced by the filtering operation of the second filter unit, event information of each of the plurality of pixels included in the fourth column, and event information of each of the plurality of pixels included in the fifth column; and A third filter that performs a filtering operation from the first row to the Mth row for the third column, the fourth column, and the fifth column by applying the kernel; Including, The first column, the second column, the third column, the fourth column and the fifth column are composed of consecutive columns. Vision sensor.

5. In claim 1, If the size of the above kernel is '3x3' and the number of filters and the number of buffers are '4', The above filter part A first buffer that stores event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; A first filter unit that performs a filtering operation from the first row to the Mth row for the first column, the second column, and the third column by applying the kernel; A second buffer that stores the result produced by the filtering operation of the first filter unit, event information of each of the plurality of pixels included in the third column, and event information of each of the plurality of pixels included in the fourth column; and A second filter that performs a filtering operation from the first row to the Mth row for the second column, the third column, and the fourth column by applying the kernel; A third buffer that stores the result produced by the filtering operation of the second filter unit, event information of each of the plurality of pixels included in the fourth column, and event information of each of the plurality of pixels included in the fifth column; A third filter that performs a filtering operation from the first row to the Mth row for the third column, the fourth column, and the fifth column by applying the kernel; A fourth buffer storing the result produced by the filtering operation of the third filter and the event information of each of the plurality of pixels included in the fifth column and the event information of each of the plurality of pixels included in the sixth column; and A fourth filter unit that performs a filtering operation from the first row to the Mth row for the fourth column, the fifth column, and the sixth column by applying the kernel; Including, The first column, the second column, the third column, the fourth column, the fifth column and the sixth column are composed of consecutive columns. Vision sensor.

6. A method of operating a vision system, comprising a pixel array that includes a plurality of pixels arranged in a matrix form of M x N (M, N are natural numbers) and generates an event signal, and a filter unit that performs a filtering operation on the generated event signal by utilizing at least one buffer and at least one filter, A step of detecting an event signal from each of the plurality of pixels included in the pixel array; and A step of applying a kernel of a preset size and performing the filtering operation based on the size of the kernel, the number of buffers, and the number of filters. Including How the vision sensor works.

7. In claim 6, If the size of the above kernel is '3x3' and the number of filters and the number of buffers are '1', The steps for performing the above filtering operation are A step of storing event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; and A step of performing a filtering operation from the first row to the Mth row for the first column, the second column, and the third column by applying the kernel. Including, The first column, the second column and the third column are composed of consecutive columns. How the vision sensor works.

8. In claim 6, If the size of the above kernel is '3x3' and the number of filters and the number of buffers are '2', The steps for performing the above filtering operation are A step of storing event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; A step of performing a filtering operation from the first row to the Mth row for the first column, the second column, and the third column by applying the kernel; A step of storing the result produced by the filtering operation of the first filter unit and the event information of each of the plurality of pixels included in the third column and the event information of each of the plurality of pixels included in the fourth column; and A step of performing a filtering operation from the first row to the Mth row for the second column, the third column, and the fourth column by applying the kernel. Including, The first column, the second column and the third column are composed of consecutive columns. How the vision sensor works.

9. In claim 6, If the size of the above kernel is '3x3' and the number of filters and the number of buffers are '3', The steps for performing the above filtering operation are A step of storing event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; A step of performing a filtering operation from the first row to the Mth row for the first column, the second column, and the third column by applying the kernel; A step of storing the result produced by the filtering operation of the first filter unit and the event information of each of the plurality of pixels included in the third column and the event information of each of the plurality of pixels included in the fourth column; A step of performing a filtering operation from the first row to the Mth row for the second column, the third column, and the fourth column by applying the kernel; A step of storing the result produced by the filtering operation of the second filter unit and the event information of each of the plurality of pixels included in the fourth column and the event information of each of the plurality of pixels included in the fifth column; and A step of performing a filtering operation from the first row to the Mth row for the third column, the fourth column, and the fifth column by applying the kernel. Including, The first column, the second column and the third column are composed of consecutive columns. How the vision sensor works.

10. In claim 6, If the size of the above kernel is '3x3' and the number of filters and the number of buffers are '4', The steps for performing the above filtering operation are A step of storing event information of each of a plurality of pixels included in a first column, event information of each of a plurality of pixels included in a second column, and event information of each of a plurality of pixels included in a third column; A step of performing a filtering operation from the first row to the Mth row for the first column, the second column, and the third column by applying the kernel; A step of storing the result produced by the filtering operation of the first filter unit and the event information of each of the plurality of pixels included in the third column and the event information of each of the plurality of pixels included in the fourth column; A step of performing a filtering operation from the first row to the Mth row for the second column, the third column, and the fourth column by applying the kernel; A step of storing the result produced by the filtering operation of the second filter unit and the event information of each of the plurality of pixels included in the fourth column and the event information of each of the plurality of pixels included in the fifth column; A step of performing a filtering operation from the first row to the Mth row for the third column, the fourth column, and the fifth column by applying the kernel; A step of storing the result produced by the filtering operation of the third filter and the event information of each of the plurality of pixels included in the fifth column and the event information of each of the plurality of pixels included in the sixth column; and A step of performing a filtering operation from the first row to the Mth row for the fourth column, the fifth column, and the sixth column by applying the kernel; Including, The first column, the second column and the third column are composed of consecutive columns. How the vision sensor works.

Images