A noise point identification method, device, equipment and storage medium

Abstract

The application provides a noise point identification method, device and equipment and a storage medium. The method comprises the following steps: for a target pixel point generating an event at a current time in an EVS pixel array, planning a target array region with the target pixel point as the center; calculating a time difference between a time stamp of the last event generated by other pixel points in the target array region except the target pixel point and the current time; and judging whether the target pixel point is a noise point according to the number of pixel points with a time difference less than or equal to a time length threshold. Through the implementation of the application, an array region composed of neighborhood pixel points is planned with the current event point as the center, then the number of pixel points with a time interval less than a time length threshold between the current event point and the event generated by all pixel points in the region is counted, and whether the current event point is a noise point is judged according to the number, which can realize accurate noise removal, has a good noise removal effect on small-range gathered noise and scattered noise, and reduces the communication bandwidth and computing resources of event data processing requirements.

Description

Technical Field

[0001] This application relates to the field of electronic technology, and in particular to a noise identification method, apparatus, device, and storage medium. Background Technology

[0002] With the continuous development of science and technology, image sensors are widely used in products such as smartphones, digital cameras, and security cameras. EVS (Event-based Vision Sensor) is a new type of image sensor that simulates the human retina and responds to pixel pulses caused by changes in brightness due to motion. Therefore, it can capture changes in scene brightness (i.e., changes in light intensity) at an extremely high frame rate, record events at specific times and specific locations in the image, forming an event stream instead of a frame stream. This can solve the problems of information redundancy, large data storage requirements, and large real-time processing requirements of traditional cameras.

[0003] In practical applications, EVS is extremely sensitive to background activity (BA) noise, which is mainly related to thermal noise and leakage current. BA noise reduces the quality of event data generated by the sensor and leads to unnecessary consumption of communication bandwidth and computing resources. Summary of the Invention

[0004] This application provides a noise identification method, apparatus, device, and storage medium, which can at least solve the problem in related technologies where event monitoring visual sensors are affected by noise, resulting in unnecessary consumption of communication bandwidth and computing resources.

[0005] The first aspect of this application provides a noise identification method, comprising: for a target pixel in an EVS pixel array that generates an event at a current time, planning a target array region centered on the target pixel in the EVS pixel array; calculating a first time difference between the current time and the timestamp of the most recent event generated by other pixels in the target array region excluding the target pixel; and determining whether the target pixel is noise based on the number of first pixels whose first time difference is less than or equal to a preset duration threshold.

[0006] A second aspect of this application provides a noise identification device, comprising: a planning module, configured to plan a target array region centered on a target pixel in an EVS pixel array for a target pixel that generates an event at a current time; a calculation module, configured to calculate a first time difference between the current time and the timestamp of the most recent event generated by other pixels in the target array region excluding the target pixel; and a judgment module, configured to determine whether the target pixel is noise based on a first number of pixels whose first time difference is less than or equal to a preset duration threshold.

[0007] A third aspect of this application provides an electronic device, including a memory and a processor, wherein the processor is used to execute a computer program stored in the memory, and when the processor executes the computer program, it implements the steps of the noise recognition method provided in the first aspect of this application.

[0008] The fourth aspect of this application provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the steps of the noise recognition method provided in the first aspect of this application.

[0009] As can be seen from the above, according to the noise identification method, apparatus, device, and storage medium provided in this application, for the target pixel in the EVS pixel array that generates an event at the current time, a target array region centered on the target pixel is planned in the EVS pixel array; the first time difference between the current time and the timestamp of the most recent event generated by other pixels in the target array region excluding the target pixel is calculated; and the target pixel is determined to be a noise point based on the number of pixels whose first time difference is less than or equal to a preset duration threshold. Through the implementation of this application, an array region composed of neighboring pixels is planned centered on the current event point, and then the number of pixels in the region whose time interval between the event generated by the current event point and the event generated by the current event point is less than the duration threshold is counted. This number is then used to determine whether the current event point is a noise point, achieving accurate noise reduction. It has a good noise reduction effect on both small-scale clustered noise and scattered noise, effectively reducing the communication bandwidth and computing resources required for event data processing. Attached Figure Description

[0010] Figure 1 A schematic diagram of the basic process of a noise recognition method provided in an embodiment of this application;

[0011] Figure 2 A schematic diagram of the planning of a target array region is provided in one embodiment of this application;

[0012] Figure 3 This application provides a schematic diagram of the planning of an expanded array region according to an embodiment of the present application;

[0013] Figure 4 A detailed flowchart illustrating a noise recognition method provided in an embodiment of this application;

[0014] Figure 5 A schematic diagram of the program modules of a noise recognition device provided in an embodiment of this application;

[0015] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0016] To make the inventive objectives, features, and advantages of this application more apparent and understandable, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0017] In the description of the embodiments of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0018] The following will describe in detail, with reference to the accompanying drawings, a noise identification method, apparatus, device, and storage medium according to embodiments of this application.

[0019] To address the problem that noise in event monitoring visual sensors in related technologies leads to unnecessary consumption of communication bandwidth and computing resources, one embodiment of this application provides a noise detection method, such as... Figure 1 This is a basic flowchart of a noise recognition method provided in an embodiment of this application. The noise recognition method specifically includes the following steps:

[0020] Step 101: For the target pixel in the EVS pixel array that generates the event at the current moment, plan a target array region centered on the target pixel in the EVS pixel array.

[0021] Specifically, in this embodiment, the overall pixel array of the event monitoring visual sensor includes multiple EVS pixels. Each EVS pixel is also referred to as the pixel point in this embodiment. During the operation of the event monitoring visual sensor, the pixel point generates an event in response to changes in light intensity. For the target pixel point that generates the event at the current moment, this embodiment plans a target array area centered on the target pixel point on the overall pixel array.

[0022] In practical applications, a rectangular array region can be planned centered on the target pixel, such as a rectangular array region of size a*b or a square array region of size S*S. However, this array planning method has poor robustness when facing application scenarios with large differences in event sparsity. The window size, as the only variable, becomes difficult to control. In addition, it is difficult to identify object edges and a small amount of noise clustered in a small area, because they will mutually confirm each other as correct event points.

[0023] Based on this, in this embodiment, a rectangular array centered on the target pixel is planned in the EVS pixel array; then, a target array region with the target pixel as the center of rotational symmetry is planned in the rectangular array; wherein, from the middle row to the edge row of the rectangular array, the number of pixels belonging to the target array region in each row gradually decreases.

[0024] like Figure 2 The diagram shown illustrates the planning of a target array region according to this embodiment. First, this embodiment plans an S*S rectangular array region centered on the target pixel P(x,y) where the event occurs at the current moment. Then, based on this array region, a rotationally symmetric target array region is further planned, with a rotation angle of 90°. Figure 2 The pixels defined by the solid lines in the rectangle are the pixels that make up the target array region, which can be understood as a region of interest within a rectangular array. In this embodiment, when planning the region of interest, as many pixels as possible from the middle row and middle column of the rectangular array containing the target pixel can be planned into the target array region. Preferably, all pixels from the middle row and middle column can be planned into the target array region. Then, planning continues outward from the middle row or middle column of the rectangular array. Each time a row or column is planned outward, fewer pixels are planned into the target array region compared to the previous row or column. Figure 2For example, the five pixels in the middle row of the rectangular array are all allocated to the target array region. Then, for the two adjacent rows of the middle row, three pixels are selected from the five pixels in each row and allocated to the target array region. Next, for the two outermost edge rows, one pixel is selected from the five pixels in each row and allocated to the target array region. This fully considers the distance between pixels and the universality of object motion, giving more attention to pixels in the row and column direction where the current event point is located, making the selection of the region of interest more reasonable.

[0025] It is worth noting that this embodiment can cache the time of the most recent event for each pixel using the Surface of Active Events (SAE). Therefore, when executing the Background Activity Filter (BAF) denoising algorithm, array region planning can be performed on the SAE. In practical applications, when a pixel generates a new event, the timestamp of the pixel's coordinates in the SAE is updated. This embodiment is described above. Figure 2 This can be understood as an SAE (Self-Enhancing Image). The values ​​in the image record the timestamp (in milliseconds) of the most recent event that occurred at the corresponding pixel. For example, the timestamp of the event that occurred at pixel p is the current time t.

[0026] Step 102: Calculate the first time difference between the current time and the timestamp of the most recent event generated by other pixels in the target array region excluding the target pixel.

[0027] Specifically, in this embodiment, for each pixel in the target array region, the difference between the timestamp of the most recent event generated by the target pixel at the center of the region and that of each pixel is calculated to obtain the first time difference, that is, the current time t corresponding to the target pixel is used as the minuend, and the timestamp t of the most recent event generated by other pixels is used as the minuend. SAE As the subtrahend, perform subtraction on the two. Figure 2 Taking a pixel with a gray fill mark as an example, the calculation process for the first time difference of this pixel can be expressed as: t-(t-10)=10ms.

[0028] Step 103: Determine whether the target pixel is noise based on the number of first pixels whose first time difference is less than or equal to the preset duration threshold.

[0029] Specifically, in this embodiment, a corresponding first time difference can be calculated for all pixels in the target pixel array. Then, all first time differences are compared with a preset duration threshold τ. The number of pixels whose first time difference is less than or equal to the duration threshold is counted, which is the number of first pixels. Finally, the number of first pixels is used as a reference to determine whether the target pixel is noise. Continuing with the above... Figure 2 Taking the SAE example shown, if τ = 10ms, then the number of the first pixel is 1, that is... Figure 2 Chinese SAE = t-10 pixels.

[0030] In one optional embodiment of this example, the step of determining whether a target pixel is noise based on the number of first pixels whose first time difference is less than or equal to a preset duration threshold includes: comparing the number of first pixels whose first time difference is less than or equal to the preset duration threshold with a preset first quantity threshold and a preset second quantity threshold; determining that the target pixel is a correct event point when the number of first pixels is greater than or equal to the first quantity threshold; and determining that the target pixel is noise when the number of first pixels is less than the second quantity threshold.

[0031] Specifically, unlike related technologies that typically determine whether a target pixel is noise by comparing only a single threshold, this embodiment sets multiple levels of quantity thresholds. The first quantity threshold n is a positive integer greater than or equal to 2, and the second quantity threshold m is a positive integer less than the first quantity threshold. More preferably, the value of the second quantity threshold can be... That is, n / 2 rounded up. In practical applications, if there are actually n or more pixels in the target array region that satisfy tt SAE If ≤τ, then the target pixel p is considered a correct event point, if the target array region satisfies tt SAE If the number of pixels ≤τ is less than m, then the target pixel p is considered noise and needs to be denoised.

[0032] Further, in an optional embodiment of this example, after comparing the number of first pixels whose first time difference is less than or equal to a preset duration threshold with a preset first quantity threshold and a preset second quantity threshold, the method further includes: when the number of first pixels is greater than or equal to the second quantity threshold and less than the first quantity threshold, selecting pixels to be expanded that satisfy the condition that the first time difference is less than or equal to the duration threshold; planning an expansion array region centered on the pixel to be expanded in the EVS pixel array; calculating the second time difference between the timestamp of the most recent event generated by the pixel to be expanded and the timestamp of the most recent event generated by each pixel in the expansion array region; and determining whether the target pixel is noise based on the number of second pixels whose second time difference is less than or equal to the duration threshold.

[0033] Specifically, in this embodiment, when the target array region satisfies tt SAEThe number of pixels ≤τ is greater than or equal to m and less than n. In this embodiment, the region of interest (ROI) is expanded for pixels that meet the condition. If there are multiple pixels that meet the condition, in practical applications, one of the multiple pixels can be selected for ROI expansion, or ROI expansion can be performed on all pixels. In a preferred embodiment, this embodiment can plan an expansion array region centered on the pixel to be expanded in the EVS pixel array according to the planning form of the target array region. In the aforementioned... Figure 2 Based on the example, this embodiment Figure 3 A schematic diagram of the planning of an expanded array region is provided, showing the t SAE =t-10 is the pixel to be expanded as the array region expansion principle. This pixel and all pixels with gray background fill form the expansion array region.

[0034] Next, for all pixels in the expanded array region, the timestamp of each pixel is used as the subtrahend, and the timestamp of the pixel to be expanded is used as the minuend. The subtraction operation is then performed to obtain the second time difference. Furthermore, the number of pixels in the expanded array region that satisfy the condition that the second time difference is less than or equal to the duration threshold is counted. That is, for the above... Figure 3 For example, within the statistical grayscale region, 0 ≤ (t-10) - t SAE The number of pixels ≤τ is used to determine whether the target pixel is noise. It should be understood that this embodiment may only calculate the second time difference between the timestamp of the most recent event of the pixel to be expanded and the timestamps of the most recent events of all pixels in the expanded array region excluding the target pixel, i.e. Figure 3 The middle pixel p(x,y) is not included in the statistics, that is, in Figure 3 In the expanded array region, the pixels that meet the conditions are the two pixels to the left and right of the pixel to be expanded, so the number of the second pixel is 2.

[0035] In one optional embodiment of this example, the step of determining whether a target pixel is noise based on the number of second pixels whose second time difference is less than or equal to a duration threshold includes: comparing the number of second pixels whose second time difference is less than or equal to a duration threshold with a preset third quantity threshold; wherein the third quantity threshold is a positive integer greater than or equal to the second quantity threshold; when the number of second pixels is greater than or equal to the third quantity threshold, the target pixel is determined to be a correct event point; when the number of second pixels is less than the third quantity threshold, the target pixel is determined to be noise.

[0036] Furthermore, in an optional embodiment of this example, the step of determining the target pixel as the correct event point when the number of second pixels is greater than or equal to the third quantity threshold includes: when the number of second pixels is greater than or equal to the third quantity threshold, obtaining the relative distance between the pixel and the target pixel when the second time difference between the pixel and the timestamp of the most recent event generated by the event point to be expanded is less than or equal to the duration threshold; when the relative distance is less than or equal to a preset distance threshold, determining the target pixel as the correct event point.

[0037] Specifically, in practical applications, if the determination of whether a target pixel is noise is based solely on the time difference, the robustness is poor in scenarios with high event sparsity. Therefore, in this embodiment, when the number of pixels in the expanded array region that satisfy the second time difference being less than or equal to the duration threshold is greater than the number threshold, the relative distance between the pixels in the expanded array region that satisfy the second time difference being less than or equal to the duration threshold and the target pixel is further determined. When the relative distance is small, it indicates that the neighborhood of the target pixel also has a high probability of event generation. At this time, the determination that the target pixel is the correct event point has high accuracy.

[0038] Based on the technical solution of the above embodiments of this application, for the target pixel in the EVS pixel array that generates an event at the current time, a target array region centered on the target pixel is planned in the EVS pixel array; a first time difference is calculated between the current time and the timestamp of the most recent event generated by other pixels in the target array region excluding the target pixel; and the target pixel is determined to be a noise point based on the number of pixels whose first time difference is less than or equal to a preset duration threshold. By implementing the solution of this application, an array region composed of neighboring pixels is planned centered on the current event point, and then the number of pixels in the region whose time interval between the event generated by the current event point and the event generated by the current event point is less than the duration threshold is counted. This number is then used to determine whether the current event point is a noise point, achieving accurate noise reduction. It has a good noise reduction effect on both small-scale clustered noise and scattered noise, effectively reducing the communication bandwidth and computing resources required for event data processing.

[0039] Next, this embodiment further provides a refined noise recognition method, such as... Figure 4 The diagram shown is a flowchart illustrating a refined noise identification method according to an embodiment of this application. The noise identification method specifically includes the following steps:

[0040] Step 401: For the target pixel in the EVS pixel array that generates an event at the current moment, plan a rectangular array centered on the target pixel in the EVS pixel array, and plan a target array region in the rectangular array with the target pixel as the rotational symmetry center.

[0041] Step 402: Calculate the first time difference between the current time and the timestamp of the most recent event generated by other pixels in the target array region excluding the target pixel;

[0042] Step 403: Compare the number of first pixels whose first time difference is less than or equal to a preset duration threshold with a preset first quantity threshold and a preset second quantity threshold;

[0043] Step 404: When the number of first pixels is greater than or equal to the first quantity threshold, determine the target pixel as the correct event point;

[0044] Step 405: When the number of first pixels is less than the second threshold, the target pixel is determined to be noise.

[0045] Step 406: When the number of first pixels is greater than or equal to the second number threshold and less than the first number threshold, select the pixels to be expanded that satisfy the first time difference being less than or equal to the duration threshold.

[0046] Step 407: According to the planning form of the target array area, plan the expansion array area centered on the pixel to be expanded in the EVS pixel array;

[0047] Step 408: Calculate the second time difference between the timestamp of the most recent event generated by the pixel to be expanded and the timestamp of the most recent event generated by each pixel in the expanded array region;

[0048] Step 409: Compare the number of second pixels whose second time difference is less than or equal to the duration threshold with the preset third quantity threshold.

[0049] Step 410: When the number of second pixels is greater than or equal to the third threshold, the target pixel is determined to be the correct event point;

[0050] Step 411: When the number of second pixels is less than the third threshold, the target pixel is determined to be noise.

[0051] It should be understood that the sequence number of each step in this embodiment does not imply the order in which the steps are executed. The execution order of each step should be determined by its function and internal logic, and should not constitute a unique limitation on the implementation process of this application embodiment.

[0052] Based on the above technical solutions of this application, by turning SAE event points into variable parameters and further adjusting the SAE window, denoising robustness under different scenarios can be ensured. The region of interest is designed as a rhombus-shaped region, giving more attention to event points in the row and column directions of the target pixel. Finally, through temporal and spatial expansion, the SAE neighborhood characteristics of the required event points in earlier time intervals are further considered, achieving more accurate denoising. Since temporal and spatial expansion has a good denoising effect on small-scale clustered noise points, it can also remove scattered noise near object edges without affecting the correct event points at the object edges, resulting in good edge preservation.

[0053] Figure 5 A noise recognition device is provided in one embodiment of this application. This noise recognition device can be used to implement the noise recognition method in the foregoing embodiments. The noise recognition device mainly includes:

[0054] The planning module 501 is used to plan a target array region centered on the target pixel in the EVS pixel array, for the target pixel that generates an event at the current moment.

[0055] The calculation module 502 is used to calculate the first time difference between the current time and the timestamp of the most recent event generated by other pixels in the target array region excluding the target pixel;

[0056] The judgment module 503 is used to determine whether a target pixel is noise based on the number of first pixels whose first time difference is less than or equal to a preset duration threshold.

[0057] In some embodiments of this example, the planning module is specifically used to: plan a rectangular array centered on the target pixel in the EVS pixel array; plan a target array region in the rectangular array with the target pixel as the center of rotational symmetry; wherein, from the middle row to the edge row of the rectangular array, the number of pixels belonging to the target array region in each row gradually decreases.

[0058] In some embodiments of this example, the determination module is specifically used to: compare the number of first pixels whose first time difference is less than or equal to a preset duration threshold with a preset first quantity threshold and a preset second quantity threshold; wherein, the first quantity threshold is a positive integer greater than or equal to 2, and the second quantity threshold is a positive integer less than the first quantity threshold; when the number of first pixels is greater than or equal to the first quantity threshold, the target pixel is determined to be a correct event point; when the number of first pixels is less than the second quantity threshold, the target pixel is determined to be noise.

[0059] In some embodiments of this example, the noise identification device further includes: a selection module, configured to select pixels to be expanded that satisfy a first time difference less than or equal to a duration threshold when the number of first pixels is greater than or equal to a second quantity threshold and less than the first quantity threshold. Correspondingly, the planning module is further configured to: plan an expansion array region centered on the pixel to be expanded in the EVS pixel array; the calculation module is further configured to: calculate a second time difference between the timestamp of the most recent event generated by the pixel to be expanded and the timestamp of the most recent event generated by each pixel in the expansion array region; and the judgment module is further configured to: determine whether the target pixel is noise based on the number of second pixels whose second time difference is less than or equal to the duration threshold.

[0060] In some embodiments of this example, when the determination module performs the function of determining whether a target pixel is noise based on the number of second pixels whose second time difference is less than or equal to the duration threshold, it is specifically used to: compare the number of second pixels whose second time difference is less than or equal to the duration threshold with a preset third quantity threshold; wherein, the third quantity threshold is a positive integer greater than or equal to the second quantity threshold; when the number of second pixels is greater than or equal to the third quantity threshold, the target pixel is determined to be a correct event point; when the number of second pixels is less than the third quantity threshold, the target pixel is determined to be noise.

[0061] In some embodiments of this example, when the determination module performs the function of determining the target pixel as the correct event point when the number of second pixels is greater than or equal to the third quantity threshold, it is specifically used to: when the number of second pixels is greater than or equal to the third quantity threshold, obtain the relative distance between the target pixel and the pixel whose second time difference between the second pixel and the timestamp of the most recent event generated by the pixel to be expanded is less than or equal to the duration threshold; when the relative distance is less than or equal to the preset distance threshold, determine the target pixel as the correct event point.

[0062] It should be noted that the noise recognition methods in the foregoing embodiments can all be implemented based on the noise recognition device provided in this embodiment. Those skilled in the art can clearly understand that, for the sake of convenience and brevity, the specific working process of the noise recognition device described in this embodiment can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0063] Figure 6 An electronic device is provided as an embodiment of this application. This electronic device can be used to implement the noise recognition method in the foregoing embodiments, and mainly includes:

[0064] The system includes a memory 601, a processor 602, and a computer program 603 stored on the memory 601 and executable on the processor 602. The memory 601 and the processor 602 are communicatively connected. When the processor 602 executes the computer program 603, it implements the method described in the foregoing embodiments. The number of processors can be one or more.

[0065] The memory 601 can be a high-speed random access memory (RAM) or a non-volatile memory, such as a disk storage device. The memory 601 is used to store executable program code, and the processor 602 is coupled to the memory 601.

[0066] Furthermore, embodiments of this application also provide a computer-readable storage medium, which may be disposed in the electronic device described in the above embodiments, and the computer-readable storage medium may be as described above. Figure 6 The memory in the illustrated embodiment.

[0067] The computer-readable storage medium stores a computer program that, when executed by a processor, implements the noise recognition method described in the foregoing embodiments. Furthermore, the computer-readable storage medium can also be a USB flash drive, a portable hard drive, a read-only memory (ROM), RAM, a magnetic disk, or an optical disk, or any other medium capable of storing program code.

[0068] It should also be noted that the apparatuses and methods disclosed in the several embodiments provided in this application can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0069] The modules described as separate components may or may not be physically separate. Similarly, the components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0070] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0071] If the integrated module is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned readable storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0072] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0073] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0074] The above is a description of the noise recognition method, apparatus, device and storage medium provided in this application. For those skilled in the art, based on the ideas of the embodiments of this application, there will be changes in the specific implementation and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A noise detection method, characterized in that, include: For the target pixel in the EVS pixel array that generates an event at the current moment, a rectangular array centered on the target pixel is planned in the EVS pixel array; a target array region is planned in the rectangular array with the target pixel as the rotational symmetry center; wherein, from the middle row to the edge row of the rectangular array, the number of pixels belonging to the target array region in each row gradually decreases, and the target array region is used to analyze the spatial distribution characteristics of the surrounding area of ​​the target pixel to identify background activity noise; Calculate the first time difference between the current time and the timestamp of the most recent event generated by other pixels in the target array region excluding the target pixel; Based on the number of first pixels whose first time difference is less than or equal to a preset duration threshold, it is determined whether the target pixel is noise.

2. The noise identification method according to claim 1, characterized in that, The step of determining whether the target pixel is noise based on the number of first pixels whose first time difference is less than or equal to a preset duration threshold includes: The number of first pixels whose first time difference is less than or equal to a preset duration threshold is compared with a preset first quantity threshold and a preset second quantity threshold; wherein, the first quantity threshold is a positive integer greater than or equal to 2, and the second quantity threshold is a positive integer less than the first quantity threshold; When the number of the first pixel points is greater than or equal to the first number threshold, the target pixel point is determined to be a correct event point; When the number of the first pixel points is less than the second number threshold, the target pixel point is determined to be noise.

3. The noise identification method according to claim 2, characterized in that, After the step of comparing the number of first pixels whose first time difference is less than or equal to a preset duration threshold with a preset first quantity threshold and a preset second quantity threshold, the method further includes: When the number of the first pixel is greater than or equal to the second number threshold and less than the first number threshold, select the pixel to be expanded that satisfies the condition that the first time difference is less than or equal to the duration threshold. In the EVS pixel array, plan an expansion array region centered on the pixel to be expanded; Calculate the second time difference between the timestamp of the most recent event generated by the pixel to be expanded and the timestamp of the most recent event generated by each pixel in the expanded array region; Based on the number of second pixels whose second time difference is less than or equal to the duration threshold, it is determined whether the target pixel is noise.

4. The noise identification method according to claim 3, characterized in that, The step of planning the expansion array region centered on the pixel to be expanded in the EVS pixel array includes: According to the planning form of the target array region, an expansion array region centered on the pixel to be expanded is planned in the EVS pixel array.

5. The noise identification method according to claim 3, characterized in that, The step of determining whether the target pixel is noise based on the number of second pixels whose second time difference is less than or equal to the duration threshold includes: The number of second pixels whose second time difference is less than or equal to the duration threshold is compared with a preset third quantity threshold; wherein the third quantity threshold is a positive integer greater than or equal to the second quantity threshold; When the number of the second pixel is greater than or equal to the third quantity threshold, the target pixel is determined to be a correct event point; When the number of the second pixel is less than the third threshold, the target pixel is determined to be noise.

6. The noise identification method according to claim 5, characterized in that, The step of determining the target pixel as a correct event point when the number of the second pixel points is greater than or equal to the third quantity threshold includes: When the number of the second pixel is greater than or equal to the third number threshold, obtain the relative distance between the pixel whose second time difference with the timestamp of the most recent event generated by the pixel to be expanded is less than or equal to the duration threshold and the target pixel. When the relative distance is less than or equal to a preset distance threshold, the target pixel is determined to be the correct event point.

7. A noise detection device, characterized in that, include: The planning module is used to plan a rectangular array centered on the target pixel in the EVS pixel array for the target pixel that generates an event at the current time; and to plan a target array region in the rectangular array with the target pixel as the rotational symmetry center; wherein, from the middle row to the edge row of the rectangular array, the number of pixels belonging to the target array region in each row gradually decreases, and the target array region is used to analyze the spatial distribution characteristics of the surrounding area of ​​the target pixel to identify background activity noise; The calculation module is used to calculate the first time difference between the current time and the timestamp of the most recent event generated by other pixels in the target array region excluding the target pixel. The judgment module is used to determine whether the target pixel is noise based on the number of first pixels whose first time difference is less than or equal to a preset duration threshold.

8. An electronic device, characterized in that, Includes memory and processor, of which: The processor is used to execute computer programs stored in the memory; When the processor executes the computer program, it implements the steps in the noise recognition method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps in the noise recognition method according to any one of claims 1 to 6.

Images