A method and device for collecting event camera motion blur data set

Abstract

The invention discloses a method and device for collecting a motion blur data set of an event camera, which is applied to a collection system of a motion blur data set of an event camera, the system comprising an event camera capable of simultaneously outputting frame images and event information, a mobile computer PC, three linear guides, two servo motors, two servo motor drivers, a motion control card and a target object; the method comprises: controlling the servo motor to drive the linear guide by the PC to drive the target object to perform different modes of motion, during which the motion blur image and the corresponding event information are collected by the event camera; moving the target object to an initial position, displaying the event information, and setting adjustable parameters according to different motion modes to quickly align to a clear image position corresponding to a previously collected motion blur image, thereby obtaining a data set including the motion blur image, the clear image and the corresponding event information.

Description

Technical Field

[0001] The present invention belongs to the technical field of data acquisition and mechanical control, and in particular relates to a method and device for acquiring an event camera motion blur data set. Background Art

[0002] As a new type of bionic visual sensor, the event camera benefits from its asynchronous output of the brightness change information of each pixel, with the advantages of high temporal resolution, high dynamic range and low power consumption. It has broad application prospects in many fields such as image restoration. In particular, the high temporal resolution of the event camera can well assist the grayscale camera with lower temporal resolution to remove motion blur.

[0003] However, most of the current motion blur datasets of event cameras are generated through simulation, while most of the motion blur datasets of event cameras in real scenes are achieved by synthesizing blurred images from multiple images or by building a dual-optical path design which is more difficult to build. There may often be problems such as unrealistic blurred images, no true clear image truth, complex system construction, and misalignment between event information and grayscale images.

[0004] In summary, although event cameras have great potential in the field of motion blur removal, it is difficult to collect event camera motion blur datasets in real scenes. Therefore, an efficient event camera motion blur dataset collection method is needed, which can not only collect real motion blur images and event information, but also obtain the corresponding real clear images as the true value, and the system construction can be relatively simple. Summary of the invention

[0005] The purpose of the embodiments of the present application is to address the above-mentioned problems and provide a method and device for collecting event camera motion blur data sets, which can collect real motion blur images, event information and corresponding real clear image truth values, and the system construction is relatively simple.

[0006] According to a first aspect of an embodiment of the present application, a method for collecting a motion blur data set of an event camera is provided, which is applied to a system for collecting a motion blur data set of an event camera, the system comprising an event camera capable of simultaneously outputting frame images and event information, a mobile computer PC, three linear guides, two servo motors, two servo motor drivers, a motion control card and a target object, wherein the event camera is connected to the PC and is responsible for collecting frame images and event information; the transmission shafts of the two linear guides are connected in series, and a servo motor cooperates with the servo motor driver to drive the motion in the y direction; a linear guide is fixed on the sliders of the two linear guides, and a servo motor cooperates with the servo motor driver to drive the motion in the x direction; the two servo motor drivers are connected to the motion control card; the PC is connected to the motion control card via a network cable to control the servo motor and the linear guide; the method comprises:

[0007] The servo motor drives the linear guide rail through the PC to drive the target object to move in different modes. During this period, the event camera collects motion blur images and corresponding event information.

[0008] The target object is moved to the initial position, the event information is displayed, and the adjustable parameters are set according to different motion modes to quickly align to the clear image position corresponding to the previously collected motion blurred image, thereby obtaining a data set including motion blurred images, clear images and corresponding event information.

[0009] Furthermore, the requirements that the linear guide needs to meet are that for an event camera with a resolution of (h, w), an exposure time of t, an imaging distance of L, and a field of view FOV of θ, the movement speed V of the linear guide is greater than 2Nltan(θ / 2) / (tw), where N represents the degree of motion blur, and the repeatability of the linear guide needs to reach 0.1*(2Ltan(θ / 2)) / w.

[0010] Furthermore, the different modes of motion include uniform linear motion, uniformly accelerated linear motion, variable accelerated linear motion, uniform circular motion, and variable speed circular motion.

[0011] Furthermore, adjustable parameters are set according to different sports modes, specifically:

[0012] The middle moment of the frame exposure time is t, and all the event information to be displayed at this time is ∑ T∈[t-125us,t+125us] (x T ,y T ,T,P T ), where (x T ,y T ) represents the spatial location of the event information, T represents the timestamp of the event information, and P T Represents the polarity of event information.

[0013] Furthermore, the PC controls the servo motor to drive the linear guide rail to drive the target object to move in different modes. During the movement, the event camera collects motion blur images and corresponding event information, including:

[0014] (1) For uniform linear motion, the differential equation of motion is dX=VT, dY=0, where V is the uniform linear motion speed and T is the sampling time interval of the event camera frame image. The adjustable parameter step is set to be close to Vdt, and the step length of the target object's x-direction motion is controlled so that the target object can move quickly from the previous aligned position to the next aligned position. Then, the alignment event information is fine-tuned to collect clear images.

[0015] (2) For uniformly accelerated linear motion, the differential equation of motion can be approximately expressed as dX = V b T+0.5aT 2 ,dY=0,where V b is the motion speed at the beginning of frame image sampling, T represents the event camera frame image sampling

[0016] The sample time interval is a, a is the motion acceleration, and the adjustable parameters step and A are set to approach V b and 0.5aT 2 , control the target object to move step+A in the x direction, and the step parameter is automatically updated to step+2A, so that the target object can quickly move from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear pictures;

[0017] (3) For variable acceleration linear motion, the differential equation of motion is approximately expressed as dX = ωTR (R 2 -(XC x ) 2 ) 0.5 / R,dY=0, where ω represents the angular velocity, T represents the sampling time interval of the event camera frame image, R represents the cosine motion amplitude length, X represents the current x-direction coordinate of the target object, and C x represents the center coordinate of the linear guide in the x direction, and sets the adjustable parameter step to make it close to ωTR, controlling the x-direction motion step*(R 2 -(XC x ) 2 ) 0.5 / R length, so that the target object can quickly move from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear pictures;

[0018] (4) For circular motion, the differential equation of motion is approximately expressed as dX = ωTR (R 2 -(XCx ) 2 ) 0.5 / R,dY=ωTR(YC Y ) / R, where ω represents the angular velocity, t represents the current moment, R represents the radius of circular motion, X represents the current x-coordinate of the target object, Y represents the current y-coordinate of the target object, and C x Indicates the center coordinate of the linear guide in the x direction, C Y represents the center coordinate of the linear guide in the y direction, and sets the adjustable parameter step to make it close to ωTR, controlling the x-direction movement of the target object step*(R 2 -(XC x ) 2 ) 0.5 / R length, controls the target object's y-direction movement step*(YC Y ) / R length, so that the target object can quickly move from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear pictures;

[0019] (5) For uniformly accelerated circular motion, the differential equation of motion is approximately expressed as dX = (ωTR + 0.5ωaT 2 )(R 2 -(XC x ) 2 ) 0.5 / R,dY=(ωTR+0.5ωaT 2 )(YC Y ) / R, where ω represents the angular velocity, t represents the current moment, R represents the radius of circular motion, X represents the current x-coordinate of the target object, Y represents the current y-coordinate of the target object, and C x Indicates the center coordinate of the linear guide in the x direction, C Y Indicates the center coordinate of the linear guide in the y direction. Set the adjustable parameters step and A to approach ωTR and 0.5ωaT respectively. 2 , control the target object to move in the x direction (step+A)*(R 2 -(XC x ) 2 ) 0.5 / R length, controls the target object's y-direction movement (step+A)*(YC Y ) / R length, and the step parameter is automatically updated to step+2A, so that the target object can quickly move from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear pictures.

[0020] According to a second aspect of an embodiment of the present application, a device for collecting motion blur data sets of an event camera is provided, which is applied to a system for collecting motion blur data sets of an event camera, the system comprising an event camera capable of simultaneously outputting frame images and event information, a mobile computer PC, three linear guides, two servo motors, two servo motor drivers, a motion control card and a target object, wherein the event camera is connected to the PC and is responsible for collecting frame images and event information; the transmission shafts of the two linear guides are connected in series, and a servo motor cooperates with the servo motor driver to drive the movement in the y direction; a linear guide is fixed on the sliders of the two linear guides, and a servo motor cooperates with the servo motor driver to drive the movement in the x direction; the two servo motor drivers are connected to the motion control card; the PC is connected to the motion control card via a network cable to realize the control of the servo motor and the linear guide; the device comprises:

[0021] The acquisition module is used to control the servo motor to drive the linear guide through the PC to drive the target object to move in different modes. During this period, the event camera is used to collect motion blur images and corresponding event information;

[0022] The setting module is used to move the target object to the initial position, display the event information, and set adjustable parameters according to different motion modes to quickly align to the clear image position corresponding to the previously collected motion blurred image, thereby obtaining a data set including motion blurred images, clear images and corresponding event information.

[0023] According to a third aspect of an embodiment of the present application, there is provided an electronic device, including:

[0024] one or more processors;

[0025] A memory for storing one or more programs;

[0026] When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in the first aspect.

[0027] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, on which computer instructions are stored, and when the instructions are executed by a processor, the steps of the method described in the first aspect are implemented.

[0028] The technical solution provided by the embodiments of the present application may have the following beneficial effects:

[0029] It can be seen from the above embodiments that the present application uses an event camera with both frame image output and event information output, avoiding the alignment problem of frame images and event information, and using high-speed motion rails to achieve real motion blurred image acquisition, and by visualizing event information and setting adjustable parameters according to different motion modes, the rapid acquisition of real clear true value images is achieved. The system construction complexity is relatively low, and the acquisition process is relatively fast. The data set collected by the method of the present invention is basically consistent with the motion blur in the real scene, and can be well applied to various image restoration task studies based on event cameras.

[0030] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0032] Figure 1 The present invention is a flowchart of a method for collecting an event camera motion blur data set according to an exemplary embodiment.

[0033] Figure 2 It is a schematic diagram of constructing a linear guide rail according to an exemplary embodiment.

[0034] Figure 3 is a collected motion blur image according to an exemplary embodiment.

[0035] Figure 4 The figure is a collected clear picture according to an exemplary embodiment.

[0036] Figure 5 The invention is a block diagram of a device for collecting event camera motion blur dataset according to an exemplary embodiment. DETAILED DESCRIPTION

[0037] Here, exemplary embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application.

[0038] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0039] It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0040] Figure 1 is a flow chart of a method for collecting event camera motion blur data set according to an exemplary embodiment. Figure 1 As shown, the method is applied to an event camera motion blur data set acquisition system, the system includes an event camera capable of simultaneously outputting frame images and event information, a mobile computer PC, three linear guides, two servo motors, two servo motor drivers, a motion control card and a target object, wherein the event camera is connected to the PC and is responsible for the acquisition of frame images and event information; the transmission shafts of the two linear guides are connected in series, and a servo motor cooperates with the servo motor driver to drive the motion in the y direction; a linear guide is fixed on the sliders of the two linear guides, and a servo motor cooperates with the servo motor driver to drive the motion in the x direction; the two servo motor drivers are connected to the motion control card; the PC is connected to the motion control card via a network cable to realize the control of the servo motor and the linear guide; the method may include the following steps:

[0041] Step S11: Controlling the servo motor to drive the linear guide through the PC to drive the target object to move in different modes, during which the motion blur image and corresponding event information are collected through the event camera;

[0042] Step S12: Move the target object to the initial position, display the event information, and set adjustable parameters according to different motion modes to quickly align to the clear image position corresponding to the previously collected motion blurred image. The final data set includes three parts: motion blurred image, clear image and corresponding event information.

[0043] It can be seen from the above embodiments that the present application uses an event camera with both frame image output and event information output, avoiding the alignment problem of frame images and event information, and using high-speed motion rails to achieve real motion blurred image acquisition, and by visualizing event information and setting adjustable parameters according to different motion modes, the rapid acquisition of real clear true value images is achieved. The system construction complexity is relatively low, and the acquisition process is relatively fast. The data set collected by the method of the present invention is basically consistent with the motion blur in the real scene, and can be well applied to various image restoration task studies based on event cameras.

[0044] In one embodiment, an event camera capable of simultaneously outputting frame images and event information is selected as a visual sensor. In this example, the SEEM1 model DAVIS camera of Insightness is selected. The camera resolution is 320*262, the camera field of view FOV is 130°, the frame image output frequency is up to 30Hz, and the imaging distance is about 0.5m.

[0045] In the specific implementation of building this system, according to the imaging distance and resolution of the selected event camera, three linear guide rails that meet the positioning accuracy requirements and movement speed requirements are selected to build a dual-axis motion system. Two of the guide rails are used for y-direction motion control, and their transmission shafts are connected in series. The other guide rail is placed on the two guide rails and is responsible for x-direction motion control. The actual construction is shown in the figure below. Figure 2 As shown. Use two servo motors to drive the two linear guide rails for y-direction motion control and the x-direction motion control rail connected in series; use a dual-axis motion control card to connect the two servo motors, and connect the motion control card to the PC; aim the event camera at the rail and fix it, set the grayscale camera sampling frequency to 8Hz to 16Hz to make motion blur images easier to collect, and connect it to the PC; fix the target object on the slider of the x-direction motion control linear guide rail, and choose the target object with iconic and rich texture patterns such as checkerboard, school emblem, and text printed on A4 paper.

[0046] In the specific implementation of step S11, the PC controls the servo motor to drive the linear guide rail to drive the target object to perform different modes of movement, and the event camera collects motion blur images and corresponding event information during the movement;

[0047] Specifically, the requirements that the linear guide needs to meet are as follows: for an event camera with a resolution of (h, w), an exposure time of t, an imaging distance of L, and a field of view FOV of θ, in order to produce obvious motion blur, the movement speed V of the linear guide needs to be greater than 2Nltan(θ / 2) / (tw), where N represents the degree of motion blur, which generally needs to be above 5, and the repeatability of the linear guide needs to reach 0.1*(2Ltan(θ / 2)) / w.

[0048] Substituting the parameters of the event camera selected in this example, it can be obtained that the running speed V of the linear guide needs to satisfy V>0.48m / s, and the positioning accuracy needs to reach 0.5mm. The linear guide selected in this example has a repeatability positioning accuracy of 0.05mm and a running speed of 1.5m / s, which meets the data set collection requirements.

[0049] By calling the motion control card API on the PC side, the motion control trajectory is edited, including uniform linear motion, uniformly accelerated linear motion, variable accelerated linear motion, uniform circular motion, and variable circular motion, to simulate motion blur in different situations and control the target object to move accordingly; the event camera is used to collect frame images and event information with motion blur, and the frame images collected include motion blur, such as Figure 3 shown.

[0050] In the specific implementation of step S12, the target object is moved to the initial position, the event information is displayed, and the adjustable parameters are set according to different motion modes to quickly align to the clear picture position corresponding to the previously collected motion blurred picture, thereby obtaining a data set including motion blurred pictures, clear pictures and corresponding event information.

[0051] Specifically, the target object is moved to the initial position, and the event information in the middle time interval [-125us, +125us] corresponding to the motion blurred frame image is displayed on the event camera image acquisition display screen on the PC. Selecting this time interval can make the edge of the event information sharp, which is conducive to assisting alignment and finding the clear image position; according to different motion modes, adjustable parameters are designed to control the motion control card so that the target object can be controlled to move to the approximate position each time to achieve fast alignment; the edge of the target object is aligned according to the event information, and the corresponding clear image is collected. The collected clear image is as follows: Figure 4 As shown; repeated event information is displayed on the display screen, the moving target object is aligned, and a clear picture is collected; thereby obtaining motion blurred pictures, clear pictures and corresponding event information, and making corresponding sequences according to different motion modes and target objects to complete the collection of the entire data set.

[0052] Specifically, the specific implementation of displaying event information is that the middle moment of the frame image exposure time is t, and all event information to be displayed at this time is ∑ T∈[t-125us,t+125us] (x T ,y T ,T,P T ), where (x T ,y T ) represents the spatial location of the event information, T represents the timestamp of the event information, and P TRepresents the polarity of event information, and draws the event information on the event camera image acquisition interface according to the spatial position.

[0053] Specifically, according to different motion modes, adjustable parameters are designed to control the motion control card, which is specifically achieved through the following settings:

[0054] (1) For uniform linear motion, the differential equation of motion is dX=VT, dY=0, where V is the uniform linear motion speed and T is the sampling time interval of the event camera frame image. The adjustable parameter step is set to be close to Vdt, and the step length of the target object's x-direction motion is controlled, so that the target object can move quickly from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear images;

[0055] (2) For uniformly accelerated linear motion, the differential equation of motion can be approximately expressed as dX = V b T+0.5aT 2 ,dY=0,where V b is the motion speed at the beginning of frame image sampling, T represents the event camera frame image sampling time interval, a is the motion acceleration, and the adjustable parameters step and A are set to approach V respectively. b and 0.5aT 2 , control the target object's x-direction movement step+A length, and the step parameter is automatically updated to step+2A, which allows the target object to quickly move from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear images;

[0056] (3) For variable acceleration linear motion, in order to facilitate acquisition, the motion change is adjusted to cosine form, that is, the motion differential equation can be approximately expressed as dX = ωTR (R 2 -(XC x ) 2 ) 0.5 / R,dY=0, where ω represents the angular velocity, T represents the sampling time interval of the event camera frame image, R represents the cosine motion amplitude length, X represents the current x-direction coordinate of the target object, and C x represents the center coordinate of the linear guide in the x direction, and sets the adjustable parameter step to make it close to ωTR, controlling the x-direction motion step*(R 2 -(XC x ) 2 ) 0.5 / R length can make the target object move quickly from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear pictures;

[0057] (4) For circular motion, the differential equation of motion can be approximately expressed as dX = ωTR (R 2-(XC x ) 2 ) 0.5 / R,dY=ωTR(YC Y ) / R, where ω represents the angular velocity, t represents the current moment, R represents the radius of circular motion, X represents the current x-coordinate of the target object, Y represents the current y-coordinate of the target object, and C x Indicates the center coordinate of the linear guide in the x direction, C Y represents the center coordinate of the linear guide in the y direction, and sets the adjustable parameter step to make it close to ωTR, controlling the x-direction movement of the target object step*(R 2 -(XC x ) 2 ) 0.5 / R length, controls the target object's y-direction movement step*(YC Y ) / R length, the target object can be quickly moved from the previous alignment position to the next alignment position, and then the alignment event information is fine-tuned to collect clear pictures;

[0058] (5) For uniformly accelerated circular motion, the differential equation of motion can be approximately expressed as dX = (ωTR + 0.5ωaT 2 )(R 2 -(XC x ) 2 ) 0.5 / R,dY=(ωTR+0.5ωaT 2 )(YC Y ) / R, where ω represents the angular velocity, t represents the current moment, R represents the radius of circular motion, X represents the current x-coordinate of the target object, Y represents the current y-coordinate of the target object, and C x Indicates the center coordinate of the linear guide in the x direction, C Y Indicates the center coordinate of the linear guide in the y direction. Set the adjustable parameters step and A to approach ωTR and 0.5ωaT respectively. 2 , control the target object to move in the x direction (step+A)*(R 2 -(XC x ) 2 ) 0.5 / R length, controls the target object's y-direction movement (step+A)*(YC Y ) / R length, and the step parameter is automatically updated to step+2A, which allows the target object to quickly move from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear pictures.

[0059] Corresponding to the aforementioned embodiment of the method for collecting event camera motion blur dataset, the present application also provides an embodiment of a device for collecting event camera motion blur dataset.

[0060] Figure 5 is a block diagram of a device for collecting event camera motion blur data sets according to an exemplary embodiment. Figure 5 The device is applied to an event camera motion blur data set acquisition system, which includes an event camera that can simultaneously output frame images and event information, a mobile computer PC, three linear guides, two servo motors, two servo motor drivers, a motion control card and a target object, wherein the event camera is connected to the PC and is responsible for the acquisition of frame images and event information; the transmission shafts of the two linear guides are connected in series, and a servo motor cooperates with the servo motor driver to drive the movement in the y direction; a linear guide is fixed on the sliders of the two linear guides, and a servo motor cooperates with the servo motor driver to drive the movement in the x direction; the two servo motor drivers are connected to the motion control card; the PC is connected to the motion control card via a network port cable to realize the control of the servo motor and the linear guide; the device may include:

[0061] The acquisition module 21 is used to control the servo motor to drive the linear guide rail through the PC to drive the target object to move in different modes, during which the motion blur image and corresponding event information are collected through the event camera;

[0062] The setting module 22 is used to move the target object to the initial position, display the event information, and set adjustable parameters according to different motion modes to quickly align to the clear picture position corresponding to the previously collected motion blurred picture, thereby obtaining a data set including the motion blurred picture, the clear picture and the corresponding event information.

[0063] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0064] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application scheme. A person of ordinary skill in the art can understand and implement it without paying any creative work.

[0065] Correspondingly, the present application also provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the method for collecting event camera motion blur data sets as described above.

[0066] Accordingly, the present application also provides a computer-readable storage medium on which computer instructions are stored, and when the instructions are executed by a processor, the method for collecting an event camera motion blur data set as described above is implemented.

[0067] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the contents disclosed herein. The present application is intended to cover any variations, uses or adaptations of the present application, which follow the general principles of the present application and include common knowledge or customary technical means in the art that are not disclosed in the present application.

[0068] It should be understood that the present application is not limited to the exact construction that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof.

Claims

1. A method for collecting event camera motion blur datasets, characterized in that: The invention relates to a collection system for event camera motion blur data set, which comprises an event camera capable of simultaneously outputting frame images and event information, a mobile computer PC, three linear guides, two servo motors, two servo motor drivers, a motion control card and a target object, wherein the event camera is connected to the PC and is responsible for collecting frame images and event information; the transmission shafts of the two linear guides are connected in series, and a servo motor cooperates with the servo motor driver to drive the motion in the y direction; a linear guide is fixed on the sliders of the two linear guides, and a servo motor cooperates with the servo motor driver to drive the motion in the x direction; the two servo motor drivers are connected to the motion control card; the PC is connected to the motion control card via an Ethernet cable to realize the control of the servo motor and the linear guide; the method comprises: The servo motor drives the linear guide rail through the PC to drive the target object to move in different modes. During this period, the event camera collects motion blur images and corresponding event information. The target object is moved to the initial position, the event information is displayed, and adjustable parameters are set according to different motion modes to quickly align to the clear image position corresponding to the previously acquired motion blurred image, thereby obtaining a data set including the motion blurred image, the clear image and the corresponding event information; The different modes of motion include uniform linear motion, uniformly accelerated linear motion, variable accelerated linear motion, uniform circular motion, and variable circular motion; Adjustable parameters can be set according to different sports modes, specifically: The middle moment of the frame exposure time is t , all the event information to be displayed at this time is, among which ( x T ,y T ) represents the spatial location of event information, T The timestamp representing the event information, P T Represents the polarity of event information; The PC controls the servo motor to drive the linear guide to drive the target object to move in different modes. During the movement, the event camera collects motion blur images and corresponding event information, including: (1) For uniform linear motion, the differential equation of motion is dX=VT, dY=0, where V is the uniform linear motion speed and T is the sampling time interval of the event camera frame image. The adjustable parameter step is set to be close to Vdt, and the step length of the target object's x-direction motion is controlled so that the target object can move quickly from the previous aligned position to the next aligned position. Then, the alignment event information is fine-tuned to collect clear images. (2) For uniformly accelerated linear motion, the differential equation of motion can be approximately expressed as dX = V b T+0.5aT 2 ,dY=0, where V b is the motion speed at the beginning of frame image sampling, T represents the event camera frame image sampling The sample time interval is a, a is the motion acceleration, and the adjustable parameters step and A are set to approach V b and 0.5aT 2 , control the target object to move step+A in the x direction, and the step parameter is automatically updated to step+2A, so that the target object can quickly move from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear pictures; (3) For variable acceleration linear motion, the differential equation of motion is approximately expressed as dX = ωTR (R 2 -(XC x ) 2 ) 0.5 / R, dY=0, where ω represents the angular velocity, T represents the sampling time interval of the event camera frame image, R represents the cosine motion amplitude length, X represents the current x-direction coordinate of the target object, and C x represents the center coordinate of the linear guide in the x direction, and sets the adjustable parameter step to make it close to ωTR, controlling the x-direction motion step*(R 2 -(XC x ) 2 ) 0.5 / R length, so that the target object can quickly move from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear pictures; (4) For circular motion, the differential equation of motion is approximately expressed as dX = ωTR (R 2 -(XC x ) 2 ) 0.5 / R, dY=ωTR(YC Y ) / R, where ω represents the angular velocity, T represents the timestamp of the event information, R represents the radius of the circular motion, X represents the current x-coordinate of the target object, Y represents the current y-coordinate of the target object, and C x Indicates the center coordinate of the linear guide in the x direction, C Y represents the center coordinate of the linear guide in the y direction, and sets the adjustable parameter step to make it close to ωTR, controlling the x-direction movement of the target object step*(R 2 -(XC x ) 2 ) 0.5 / R length, controls the target object's y-direction movement step*(YC Y ) / R length, so that the target object can quickly move from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear pictures; (5) For uniformly accelerated circular motion, the differential equation of motion is approximately expressed as dX = (ωTR + 0.5ωaT 2 )(R 2 -(XC x ) 2 ) 0.5 / R, dY=(ωTR+0.5ωaT 2 ) (YC Y ) / R, where ω represents the angular velocity, T represents the timestamp of the event information, R represents the radius of the circular motion, X represents the current x-coordinate of the target object, Y represents the current y-coordinate of the target object, and C x Indicates the center coordinate of the linear guide in the x direction, C Y Indicates the center coordinate of the linear guide in the y direction. Set the adjustable parameters step and A to approach ωTR and 0.5ωaT respectively. 2 , control the target object to move in the x direction (step+A)*(R 2 -(XC x ) 2 ) 0.5 / R length, controls the target object's y-direction movement (step+A)*(YC Y ) / R length, and the step parameter is automatically updated to step+2A, so that the target object can quickly move from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear pictures.

2. The method according to claim 1, characterized in that The requirements that the linear guide needs to meet are that for an event camera with a resolution of (h, w), an exposure time of t, an imaging distance of L, and a field of view FOV of θ, the movement speed V of the linear guide is greater than 2NLtan(θ / 2) / (tw), where N represents the degree of motion blur, and the repeatability of the linear guide needs to reach 0.1*(2Ltan(θ / 2)) / w.

3. A device for collecting event camera motion blur datasets, characterized in that: A collection system of event camera motion blur data set is applied, the system includes an event camera capable of simultaneously outputting frame images and event information, a mobile computer PC, three linear guides, two servo motors, two servo motor drivers, a motion control card and a target object, wherein the event camera is connected to the PC and is responsible for collecting frame images and event information; the transmission shafts of the two linear guides are connected in series, and a servo motor cooperates with the servo motor driver to drive the motion in the y direction; a linear guide is fixed on the sliders of the two linear guides, and a servo motor cooperates with the servo motor driver to drive the motion in the x direction; the two servo motor drivers are connected to the motion control card; the PC is connected to the motion control card via an Ethernet cable to realize the control of the servo motor and the linear guide; the device includes: The acquisition module is used to control the servo motor to drive the linear guide through the PC to drive the target object to move in different modes. During this period, the event camera is used to collect motion blur images and corresponding event information; A setting module is used to move the target object to the initial position, display the event information, and set adjustable parameters according to different motion modes to quickly align to the clear image position corresponding to the previously collected motion blurred image, thereby obtaining a data set including the motion blurred image, the clear image and the corresponding event information; The different modes of motion include uniform linear motion, uniformly accelerated linear motion, variable accelerated linear motion, uniform circular motion, and variable circular motion; Adjustable parameters can be set according to different sports modes, specifically: The middle moment of the frame exposure time is t , all the event information to be displayed at this time is, among which ( x T ,y T ) represents the spatial location of event information, T The timestamp representing the event information, P T Represents the polarity of event information; The PC controls the servo motor to drive the linear guide to drive the target object to move in different modes. During the movement, the event camera collects motion blur images and corresponding event information, including: (1) For uniform linear motion, the differential equation of motion is dX=VT, dY=0, where V is the uniform linear motion speed and T is the sampling time interval of the event camera frame image. The adjustable parameter step is set to be close to Vdt, and the step length of the target object's x-direction motion is controlled so that the target object can move quickly from the previous aligned position to the next aligned position. Then, the alignment event information is fine-tuned to collect clear images. (2) For uniformly accelerated linear motion, the differential equation of motion can be approximately expressed as dX = V b T+0.5aT 2 ,dY=0, where V b is the motion speed at the beginning of frame image sampling, T represents the event camera frame image sampling The sample time interval is a, a is the motion acceleration, and the adjustable parameters step and A are set to approach V b and 0.5aT 2 , control the target object to move step+A in the x direction, and the step parameter is automatically updated to step+2A, so that the target object can quickly move from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear pictures; (3) For variable acceleration linear motion, the differential equation of motion is approximately expressed as dX = ωTR (R 2 -(XC x ) 2 ) 0.5 / R, dY=0, where ω represents the angular velocity, T represents the sampling time interval of the event camera frame image, R represents the cosine motion amplitude length, X represents the current x-direction coordinate of the target object, and C x represents the center coordinate of the linear guide in the x direction, and sets the adjustable parameter step to make it close to ωTR, controlling the x-direction motion step*(R 2 -(XC x ) 2 ) 0.5 / R length, so that the target object can quickly move from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear pictures; (4) For circular motion, the differential equation of motion is approximately expressed as dX = ωTR (R 2 -(XC x ) 2 ) 0.5 / R, dY=ωTR(YC Y ) / R, where ω represents the angular velocity, T represents the timestamp of the event information, R represents the radius of the circular motion, X represents the current x-coordinate of the target object, Y represents the current y-coordinate of the target object, and C x Indicates the center coordinate of the linear guide in the x direction, C Y represents the center coordinate of the linear guide in the y direction, and sets the adjustable parameter step to make it close to ωTR, controlling the x-direction movement of the target object step*(R 2 -(XC x ) 2 ) 0.5 / R length, controls the target object's y-direction movement step*(YC Y ) / R length, so that the target object can quickly move from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear pictures; (5) For uniformly accelerated circular motion, the differential equation of motion is approximately expressed as dX = (ωTR + 0.5ωaT 2 )(R 2 -(XC x ) 2 ) 0.5 / R, dY=(ωTR+0.5ωaT 2 ) (YC Y ) / R, where ω represents the angular velocity, T represents the timestamp of the event information, R represents the radius of the circular motion, X represents the current x-coordinate of the target object, Y represents the current y-coordinate of the target object, and C x Indicates the center coordinate of the linear guide in the x direction, C Y Indicates the center coordinate of the linear guide in the y direction. Set the adjustable parameters step and A to approach ωTR and 0.5ωaT respectively. 2 , control the target object to move in the x direction (step+A)*(R 2 -(XC x ) 2 ) 0.5 / R length, controls the target object's y-direction movement (step+A)*(YC Y ) / R length, and the step parameter is automatically updated to step+2A, so that the target object can quickly move from the previous alignment position to the next alignment position, and then fine-tune the alignment event information to collect clear pictures.

4. An electronic device, characterized in that: include: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 2.

5. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the instruction is executed by a processor, the steps of the method according to any one of claims 1 to 2 are implemented.

Images