Event signal acquisition method for event-based vision cameras

Abstract

The present invention provides an event signal acquisition method for an event-based vision camera, the method acquiring an event signal from the event-based vision camera that captures a motionless subject and an unchanging surrounding environment as a photography area.　A first event generating means 2A comprising a slit body 21 and a slit body drive mechanism 22 is placed between an event-based vision camera 1 and a photography area PA. When the slit body 21 is reciprocated in a crossing direction CD by the slit body drive mechanism 22, positions at which the photography area PA is visible and positions at which the photography area PA is invisible change over time to generate a continuous change in brightness and hence an event signal ES can be acquired by the event-based vision camera 1 that captures a motionless subject 3 and an unchanging surrounding environment as the photography area PA.

Description

[Technical Field]

[0001] The present invention relates to a method for acquiring an event signal that an event-based vision camera outputs when it detects a change in brightness. [Background technology]

[0002] Machine vision, which captures and processes images to ensure proper equipment operation, is used in a variety of industrial applications, including security applications such as surveillance, as well as automatic inspection and process control. Accurately capturing movement information is particularly important in machine vision, but conventional frame-based sensors must handle the information from all pixels in a captured video frame, resulting in large amounts of data. This creates processing delays due to data transfer, making real-time processing difficult and placing a heavy processing burden on the image processing device. In contrast, event-based vision sensors asynchronously output only pixel information for which there is a change in brightness within the image, at the timing of the change. This reduces the amount of data and enables highly efficient capture of only the movement information required for machine vision.

[0003] Event-based vision sensors are widely known, which asynchronously output event signals (e.g., data signals including pixel brightness change information, pixel position coordinates, and the time the brightness change was detected) from pixels that detect brightness changes caused by the movement of a subject in a shooting area or changes in the surrounding environment. Event-based vision sensors can only acquire instantaneous information for each pixel that captures a brightness change, and various processing processes are performed on the event signals to recognize moving objects in the shooting area and capture their movement. For example, an information processing method has been proposed in which events identified by event signals from a device equipped with an event-based vision sensor (an event-based vision camera) are grouped according to a connectivity criterion, thereby extracting features of objects, etc. with high accuracy (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Special Publication No. 2022-523505 Summary of the Invention [Problem to be solved by the invention]

[0005] However, although the event-based vision camera described in Patent Document 1 is configured to receive incident signals from the shooting area through an acquisition optical system (such as a telephoto lens) equipped with one or more lenses, no consideration is given to focus adjustment of the acquisition optical system. In other words, like a normal camera, an event-based vision sensor also requires focus adjustment of the optical system according to the shooting distance from the subject, but if the subject is stationary, it is difficult to adjust the focus of the event-based vision camera before shooting.

[0006] In the first place, an event-based vision camera asynchronously outputs an event signal from pixels that detect changes in brightness caused by the movement of a subject in the shooting area or changes in the surrounding environment, and if the subject is stationary and there are no changes in the surrounding environment, no event signal is output from the event-based vision sensor. Naturally, if the event-based vision camera cannot obtain an event signal, it cannot check the image of the shooting area and therefore cannot adjust the focus.

[0007] Furthermore, if there is a constantly moving subject or a constantly changing surrounding environment within the shooting area, an event signal will be output due to changes in brightness caused by the subject's movement or changes in the surrounding environment, allowing the image of the shooting area to be confirmed, making it possible to adjust the focus of the optical system according to the shooting distance. Therefore, one possible method is to forcibly cause a brightness change (to generate an event) by forcing the subject to move by vibrating it, or by displaying an image with brightness changes on a display installed near the target. In this way, if a vibrating subject or a flickering display is photographed with an event-based vision camera, the image of the subject or display can be confirmed, making it possible to adjust the focus of the optical system according to the shooting distance.

[0008] However, the method of triggering an event by vibrating the object to be photographed carries the risk of the object shifting position and falling outside the photographing area, or of the object being physically damaged, making it undesirable. Furthermore, the method of triggering an event using a flickering display requires the effort and cost of procuring the display, as well as the additional time and cost required for running power and communication lines, making it undesirable. For starters, if the photographing area is so far away that it requires a telephoto lens from the event-based vision camera, or if the photographing area is so high up that it is difficult to work in, it is difficult to vibrate the object to be photographed or to set up a display.

[0009] Therefore, an object of the present invention is to provide an event signal acquisition method for an event-based vision camera that acquires an event signal from an event-based vision camera whose shooting area is a stationary subject or an unchanging surrounding environment. [Means for solving the problem]

[0010] In order to solve the above problems, the present invention provides an event signal acquisition method for an event-based vision camera equipped with an event-based vision sensor that outputs an event signal asynchronously from pixels that detect brightness changes caused by the movement of a subject in a shooting area or changes in the surrounding environment, and acquires the event signal in response to incident light from a motionless subject in the shooting area or from the unchanged surrounding environment, the method comprising: a first step of placing an event generation means between the event-based vision camera and the shooting area where the event-based vision camera receives incident light; and a second step of using the event generation means to change over time the position of a shielding area that blocks incident light from the shooting area and the position of a transparent area that transmits incident light from the shooting area, thereby generating a brightness change due to the change in position of the shielding area and / or the change in position of the transparent area, and acquiring the event signal.

[0011] Furthermore, in the above configuration, the event generation means may include a slit body in which the transparent area is formed as a slit portion between each of the shielding materials by arranging a plurality of shielding materials that function as the shielding area at intervals, and may perform the first step of arranging the slit body so that the shielding material and the slit portion are simultaneously captured by the event-based vision camera, and may perform the second step of changing the position of the shielding area and the position of the transparent area over time by moving the slit body in an intersecting direction that intersects with the shooting direction of the event-based vision camera.

[0012] In the above configuration, the event generating means may include a slit body driving mechanism that reciprocates the slit body in the intersecting direction.

[0013] In the above configuration, the event generating means may support the slit body via a swing support that can swing the slit body in the intersecting direction.

[0014] Furthermore, in the above configuration, the event generation means may include a rotary slit body having a plurality of shielding sections formed radially from a center of rotation and functioning as the shielding areas, and slit sections formed between the plurality of shielding sections and functioning as the transparent areas, and the first step may be performed in which the rotary slit body is positioned so that the shielding sections and the slit sections are simultaneously captured by the event-based vision camera, and the second step may be performed in which the rotary slit body is rotated to change the positions of the shielding areas and the transparent areas over time.

[0015] In the above configuration, the event generating means may include a rotary slit body rotating mechanism that rotates the rotary slit body at least in a fixed direction at a fixed speed.

[0016] In addition, in the above configuration, the event generation means may be provided with a liquid crystal shutter that can be converted between a transparent state that transmits light and a blocking state that blocks light by adjusting the voltage applied to a liquid crystal cell, and may perform the first step of positioning the shutter surface of the liquid crystal shutter so that it blocks the shooting direction of the event-based vision camera, and the second step of changing the position of the blocking area and the transparent area over time by alternately creating a state where the shutter surface is in the blocking state so that the blocking area is positioned so that it blocks the shooting direction of the event-based vision camera, and a state where the shutter surface is in the transparent state so that the transparent area is positioned so that it blocks the shooting direction of the event-based vision camera.

[0017] In the above configuration, the event generating means may include an applied voltage control mechanism that controls the applied voltage so that the liquid crystal shutter alternates between the transparent state and the blocked state at regular intervals. [Effects of the Invention]

[0018] According to the event signal acquisition method for an event-based vision camera of the present invention, by performing the first and second steps, a change in brightness occurs due to a change in the position of the occluded area and / or a change in the position of the transparent area, making it possible to acquire an event signal from an event-based vision camera whose shooting area is a stationary subject or an unchanging surrounding environment. [Brief explanation of the drawings]

[0019] [Figure 1] This is a configuration diagram in which a first event generating means consisting of a slit body and a slit body driving mechanism is placed between an event-based vision camera and a shooting area. [Figure 2] FIG. 4 is an explanatory diagram of a slit body driving mechanism. [Figure 3] FIG. 10 is an explanatory diagram of the transition state of the imaging area visualized based on the event signal acquired from the event-based vision camera by the operation of the first event generating means. [Figure 4] This is a configuration diagram in which a second event generating means consisting of a slit body and a swinging support is placed between the event-based vision camera and the shooting area. [Figure 5] FIG. 10 is an explanatory diagram of the transition state of the imaging area visualized based on the event signal acquired from the event-based vision camera by the operation of the second event generating means. [Figure 6] This is a configuration diagram in which a manually operable slit body is placed between an event-based vision camera and a shooting area. [Figure 7] FIG. 10 is an explanatory diagram of the transition state of the imaging area visualized based on the event signal acquired from the event-based vision camera by manual operation of the slit body. [Figure 8] This is a configuration diagram in which a third event generating means consisting of a rotating slit body and a rotating slit body rotation mechanism is placed between the event-based vision camera and the shooting area. [Figure 9] FIG. [Figure 10]FIG. 10 is an explanatory diagram of the transition state of the imaging area visualized based on the event signal acquired from the event-based vision camera by the operation of the third event generating means. [Figure 11] This is a configuration diagram in which a fourth event generating means consisting of a liquid crystal shutter and an applied voltage control mechanism is placed between the event-based vision camera and the shooting area. [Figure 12] FIG. 10 is an explanatory diagram of the transition state of the imaging area visualized based on the event signal acquired from the event-based vision camera by the operation of the fourth event generating means. DETAILED DESCRIPTION OF THE INVENTION

[0020] An embodiment of the present invention will be described in detail below with reference to the accompanying drawings. Fig. 1 is a configuration diagram in which a first event generation means 2A consisting of a slit body 21 and a slit body drive mechanism 22 is placed between an event-based vision camera 1 and a photographing area PA. A subject 3, which is the subject of photographing by the event-based vision camera 1, is stationary and is placed, for example, in a photographing box 4 with a low-light interior so that it fits within the photographing area PA. An event signal ES output from the event-based vision camera 1 is input via a signal cable 5 to an information processing device 6 (e.g., configured by a personal computer executing a processing program). This information processing device 6 performs appropriate data processing to visualize the situation in the photographing area PA (e.g., moving objects or moving parts).

[0021] The event-based vision camera 1 is equipped with an event-based vision sensor 1a, which asynchronously outputs an event signal ES from pixels that detect brightness changes caused by the movement of a subject 3 in the imaging area PA or changes in the surrounding environment. The event signal ES is a data signal that includes information about the pixel's brightness change, its position coordinates, and the time the brightness change was detected. It is output only from pixels that capture a change in the imaging area PA; no event signal ES is output from pixels that do not experience a change. The event-based vision camera 1 is also equipped with an optical system 1b, such as a telephoto lens, which requires focus adjustment according to the distance to the subject 3. However, since the subject 3 is stationary and there is no movement within the imaging box 4, which is the environment surrounding the subject 3, even if the event-based vision camera 1 captures the imaging area PA, the subject 3 and the surrounding environment cannot be confirmed, and therefore focus adjustment by the optical system 1b is not possible.

[0022] Therefore, a method for acquiring an event signal by the event-based vision camera 1 that receives incident light from a motionless subject 3 in the imaging area PA or from the unchanging surrounding environment and acquires the event signal ES will be described.

[0023] First, a first step is performed in which a first event generation means 2A is placed between the event-based vision camera 1 and the imaging area PA where the event-based vision camera 1 receives incident light. As shown in FIG. 2, the slit body 21 constituting the first event generation means 2A is formed by providing shielding materials 212 at equal intervals vertically to block the inner space of a frame body 211 surrounding the four sides. By arranging multiple shielding materials 212 that function as shielding areas at intervals, transparent areas are formed as slit portions 213 between each shielding material 212. In the first step, the slit body 21 is positioned so that the shielding materials 212 and slit portions 213 of the slit body 21 are simultaneously captured by the event-based vision camera 1. For example, if the slit body 21 is positioned parallel to the imaging direction of the event-based vision camera 1, only the side of the frame body 211 of the slit body 21 will be captured, and the shielding materials 212 and slit portions 213 of the slit body 21 will not be simultaneously captured by the event-based vision camera 1. That is, when the slit body 21 is placed in the cross direction CD that intersects with the shooting direction of the event-based vision camera 1 so as to block the shooting direction of the event-based vision camera 1, the shielding material 212 and the slit portion 213 of the slit body 21 are positioned so that they are simultaneously captured by the event-based vision camera 1.

[0024] Next, a second step is performed in which the position of the shielding material 212, which is the shielding region, and the position of the slit portion 213, which is the transparent region, are changed over time by using the slit body drive mechanism 22 of the first event generation means 2A to move the slit body 21 in the intersecting direction CD that intersects with the shooting direction of the event-based vision camera 1 so as to block the shooting direction of the event-based vision camera 1. That is, in this second step, the position of the shielding region that blocks light incident from the shooting area PA to the event-based vision camera 1 and the position of the transparent region that transmits light incident from the shooting area PA to the event-based vision camera 1 are changed over time. Note that the width of the slit body 21 cannot be made extremely long, so by performing a reciprocating motion in which the slit body 21 is moved a fixed distance in the reference direction, then returned a fixed distance in the opposite direction, and then moved again a fixed distance in the reference direction, it is possible to continuously generate changes over time in the position of the shielding region and the position of the transparent region.

[0025] The configuration of the slit body drive mechanism 22 is not particularly limited, but for example, an example configuration in which the rotational power of the motor 221, which is the drive source, is converted into the reciprocating motion of the slit body 21 is shown in Figure 2. A first interlocking belt 223 is wound around a drive pulley 221b attached to a motor shaft 221a of the motor 221 and a large diameter pulley portion 222a of a first pulley 222, so that when the motor 221 rotates, for example, clockwise, the first pulley 222 also rotates clockwise. A second interlocking belt 225 is wound around a small diameter pulley portion 222b of the first pulley 222 and a second pulley 224, so that when the first pulley 222 rotates clockwise, the second pulley 224 also rotates clockwise. One end of the connecting piece 226 is loosely fitted into a driving protrusion 224a protruding from one side surface of the second pulley 224, and the other end of the connecting piece 226 is loosely fitted into a driven protrusion 227a of the slide base 227. A first slide portion 227b1 on one side of the slide base 227 and a second slide portion 227b2 on the other side are guided by a guide rail 228 arranged parallel to the cross direction CD, and the movement direction of the slide base 227 is restricted to the cross direction CD. Therefore, the pushing and pulling action of the connecting piece 226 accompanying the rotation of the second pulley 224 causes the slide base 227 to reciprocate in the cross direction CD, and the slit body 21 held by the slit body fixing portion 227c of this slide base 227 also reciprocates in the cross direction CD.

[0026] As described above, performing the first and second steps causes a change in brightness due to a change in the position of the blocked area and the position of the transparent area, so that an event signal ES can be acquired by the event-based vision camera 1, whose imaging area PA is an unmoving subject 3 and an unchanging surrounding environment. Therefore, the status of the imaging area PA, which is an unmoving subject 3 and an unchanging surrounding environment, can be confirmed, and the focus of the optical system 1b of the event-based vision camera 1 can also be adjusted.

[0027] FIG. 3 shows the transition state of the imaging area PA visualized based on the event signal ES acquired from the event-based vision camera 1 by the operation of the first event generation means 2A. Without the second step performed by the first event generation means 2A, nothing would be visible in the imaging area PA. However, when the second step is performed by the first event generation means 2A, the object 3 becomes visible through the slit portion 213 along with the shielding material 212 of the slit body 21. Once the state of the object 3 can be confirmed, the focus of the optical system 1b can be adjusted to match the focal length to the object 3. Note that if the brightness of the object 3 and the surrounding environment is the same as the brightness of the slit body 21, the event-based vision sensor 1a cannot detect a clear change in brightness. Therefore, it is desirable to make the brightness of the object 3 and the surrounding environment different from the brightness of the slit body 21. Because the interior of the imaging box 4 described above is relatively dark, if the slit body 21 is a light color (e.g., white), the event-based vision sensor 1a can detect a clear change in brightness. The width, number, sliding period and vibration period of the shielding material 212 and slit portion 213 in the slit body 21 have optimum values ​​depending on the distance to the subject, the size of the subject, the surrounding lighting conditions and the like.

[0028] While the first event generation means 2A described above uses a slit body drive mechanism 22 to perform the second step, it is also possible to generate an event (forcefully cause a change in brightness) with a simpler structure. Figure 4 shows second event generation means 2B, which is composed of a slit body 21 and a swinging support 23 that can swing the slit body 21 in the cross direction CD. The swinging support 23 has an elastic body 23a, such as a coil spring, leaf spring, or rubber material, whose lower part is fixed to a fixed plate 23b, and the upper part of the elastic body 23a supports the bottom of the slit body 21. In other words, the slit body 21, supported by the elastic body 23a, swings in front of the event-based vision camera 1 due to elastic deformation according to the magnitude and direction of the force applied to the elastic body 23a.

[0029] Even when using this second event generation means 2B to acquire an event signal ES, the first step involves placing the second event generation means 2B between the event-based vision camera 1 and the imaging area PA, and the second step involves changing the positions of the shielded and transparent areas in the slit body 21 over time using the oscillating support 23. The oscillation of the slit body 21 causes changes in brightness due to the changes in the positions of the shielded and transparent areas, so the event signal ES can be acquired by the event-based vision camera 1, whose imaging area PA is an unmoving subject 3 or an unchanging surrounding environment. Therefore, the status of the imaging area PA can be confirmed, including the unmoving subject 3 and the unchanging surrounding environment, and it is also possible to adjust the focus of the optical system 1b of the event-based vision camera 1.

[0030] Figure 5 shows the transition state of the imaging area PA visualized based on the event signal ES acquired from the event-based vision camera 1 by the operation of the second event generation means 2B. The second event generation means 2B detects a greater amount of displacement in areas farther from the elastic body 23a of the oscillating support 23, so the frequency of detecting brightness changes increases in areas where the upper part of the slit body 21, farther from the elastic body 23a, is captured. The more frequently the event-based vision sensor 1a detects brightness changes, the more detailed the event signal ES can be acquired, contributing to the visualization of a clearer image. Therefore, acquiring the event signal ES using the second event generation means 2B is effective when the subject 3 is located relatively high in the imaging area PA.

[0031] The first and second event generation means 2A and 2B described above use a slit body drive mechanism 22 or a swinging support 23 to perform the second step, but it is also possible to generate an event more simply (to forcibly cause a change in brightness). Figure 6 shows a configuration in which a slit body 21 is placed between the event-based vision camera 1 and the photographing area PA with a hand 7, and the slit body 21 is moved in the cross direction CD with the hand 7. Figure 7 shows the transition state of the photographing area PA visualized based on the event signal ES acquired from the event-based vision camera 1 as the slit body 21 is moved with the hand 7. Even when the slit body 21 is moved with the hand 7, the subject 3 in the photographing area PA is accurately captured.

[0032] Furthermore, while the first and second event initiation means 2A and 2B reciprocate the slit body 21 in the cross direction CD, this is not limitative. The third event initiation means 2C is comprised of a circular plate-shaped rotary slit body 24 and a rotary slit body rotation mechanism 25 that rotates the rotary slit body 24 at least in a fixed direction at a fixed speed, and the rotary slit body 24 performs the same function as the slit body 21.

[0033] An example of the rotary slit body 24 is shown in Figure 9. The rotary slit body 24 has a rotary base 241 having a rotary shaft mounting hole 241a serving as the center of rotation, and an outer peripheral ring 242 concentrically spaced from the rotary base 241, which are connected by a plurality of shielding portions 243, and slit portions 244 are formed between the shielding portions 243. That is, the shielding portions 243 function as a plurality of shielding regions formed radially, and the slit portions 244 function as transmission regions formed between the plurality of shielding portions 243. The structure of the rotary slit body rotation mechanism 25 is not particularly limited, and can be easily realized, for example, by transmitting the rotational force of a motor serving as a drive source to the rotary slit body 24 via a pulley and a connecting belt.

[0034] To acquire an event signal ES using this third event generation means 2C, a first step involves placing the rotary slit body 24 of the third event generation means 2C between the event-based vision camera 1 and the imaging area PA, and a second step involves changing the positions of the shielding area and the transmitting area of ​​the rotary slit body 24 of the third event generation means 2C over time. In the first step, the rotary slit body 24 is positioned so that the shielding portion 243 and the slit portion 244 are simultaneously captured by the event-based vision camera 1. In the second step, the rotary slit body rotation mechanism 25 rotates the rotary slit body 24 in a fixed direction (e.g., clockwise CW in FIG. 8 ), changing the positions of the shielding area and the transmitting area over time. The fixed-direction rotation of the rotary slit body 24 causes a change in brightness due to the change in the positions of the shielding area and the transmitting area, allowing the event signal ES to be acquired by the event-based vision camera 1, whose imaging area PA is a stationary subject 3 and an unchanging surrounding environment. Therefore, the situation of the photographing area PA, such as the stationary subject 3 and the unchanging surrounding environment, can be confirmed, and the focus of the optical system 1b of the event-based vision camera 1 can also be adjusted.

[0035] FIG. 10 shows the transition state of the imaging area PA visualized based on the event signal ES acquired from the event-based vision camera 1 when the rotary slit body 24 is operated by the rotary slit body rotation mechanism 25. Even when the rotary slit body 24 is operated by the rotary slit body rotation mechanism 25, the subject 3 in the imaging area PA is accurately captured. Note that the event signal ES cannot be acquired from pixels capturing areas where the positions of the shielding portion 243 (a shielded area) and the slit portion 244 (a transparent area) do not change over time. In other words, the positions of the rotating base 241 and the areas outside the outer edge ring 242 only function as shielded areas and do not change to transparent areas. Therefore, events cannot be generated and these areas cannot be visualized. Therefore, it is desirable to adjust the size of the rotary slit body 24 itself and its placement distance from the front of the event-based vision camera 1. Note that, because the interior of the imaging box 4 is relatively dark, if the rotary slit body 24 is a light color (e.g., white), the event-based vision sensor 1a can detect a clear change in brightness. In addition, the width of the shielding section 243 and the width of the slit section 244 in the rotating slit body 24, the number of shielding sections 243 and slit sections 244 arranged, and the rotation speed have optimal values ​​depending on the distance to the subject to be photographed, the size of the subject to be photographed, the surrounding lighting conditions, etc.

[0036] The first to third event generation means 2A to 2C described above are methods of generating events by changing the positions of the shielded area and the transparent area due to movement, but this is not limited to this. Figure 11 shows fourth event generation means 2D made up of liquid crystal shutter 26 and applied voltage control mechanism 27 that controls the voltage applied to this liquid crystal shutter 26.

[0037] The liquid crystal shutter 26 in the fourth event generation means 2D has a bezel portion 261 that holds the liquid crystal panel 262, and can be switched between a transmission state that transmits light and a blocking state that blocks light by adjusting the voltage applied to each liquid crystal cell that makes up the liquid crystal panel 262. That is, when the liquid crystal panel 262 is in the blocking state, the shutter surface of the liquid crystal panel 262 is reflected by the event-based vision camera 1, and when the liquid crystal panel 262 is in the transmission state, the shooting area PA is reflected through the liquid crystal panel 262 by the event-based vision camera 1. In addition, the applied voltage control mechanism 27 is connected to the liquid crystal shutter 26 via a cable 271, and controls the applied voltage so that the transmission state and the blocking state of the liquid crystal shutter 26 are repeated at regular intervals by, for example, turning on an operation switch 272.

[0038] When this fourth event generating means 2D is used to acquire an event signal ES, the first step is to place liquid crystal shutter 26 of fourth event generating means 2D between event-based vision camera 1 and shooting area PA, and the second step is to change the positions of the shielded area and the transparent area over time by controlling applied voltage control mechanism 27 for liquid crystal shutter 26 in fourth event generating means 2D. In the first step, liquid crystal shutter 26 is positioned so that the shutter surface of liquid crystal panel 262 blocks the shooting direction of event-based vision camera 1. In the second step, the shutter surface of liquid crystal panel 262 is set to a blocking state, thereby alternately creating a state in which the shielded area is positioned to block the shooting direction of event-based vision camera 1, and a state in which the shutter surface of liquid crystal panel 262 is set to a transparent state, thereby changing the positions of the shielded area and the transparent area over time. The periodic state change of the liquid crystal shutter 26 causes a change in brightness in accordance with the change in the position of the blocked area and the position of the transparent area, so that an event signal ES can be acquired by the event-based vision camera 1, which has an unmoving subject 3 and an unchanging surrounding environment as the shooting area PA. Therefore, it is possible to confirm the status of the unmoving subject 3 and the unchanging surrounding environment in the shooting area PA, and it is also possible to adjust the focus of the optical system 1b of the event-based vision camera 1.

[0039] Figure 12 shows the transitional states of the imaging area PA visualized based on the event signal ES acquired from the event-based vision camera 1 when the liquid crystal shutter 26 is periodically changed in state under the control of the applied voltage control mechanism 27. Even when the liquid crystal shutter 26 is periodically changed in state under the control of the applied voltage control mechanism 27, the subject 3 in the imaging area PA is accurately captured. Note that the liquid crystal panel 262 of the liquid crystal shutter 26 is positioned to completely block the field of view of the event-based vision camera 1. Therefore, when the liquid crystal shutter 26 is changed from the blocking state to the transparent state, an event occurs across the entire surface, and event signals ES can be acquired from all pixels of the event-based vision sensor 1a. This has the advantage of making it easier to obtain a clear image of the entire imaging area PA, including the subject 3 and its surrounding environment. Note that the optimum value for the cycle between the transparent and blocked states depends on factors such as the distance to the subject, the size of the subject, and the ambient lighting conditions.

[0040] In addition to the above-mentioned first to fourth event generating means 2A to 2D, there are other methods for acquiring the event signal ES from the event-based vision camera 1, such as the following.

[0041] A device that generates minute vibrations is attached to the event-based vision camera 1. For example, a small vibration device like those used in smartphones is attached to the body of the event-based vision camera 1 in an appropriate position, causing the body of the event-based vision camera 1 to vibrate. The vibration of the body of the event-based vision camera 1 simulates changes in the subject 3 and surrounding environment within the shooting area PA, making it possible to acquire an event signal ES. Note that there is an optimal value for the vibration period of the vibration device depending on factors such as the distance to the subject and the size of the subject.

[0042] A flickering light is projected toward the lens of the event-based vision camera 1. For example, if flickering light is projected obliquely onto the lens of the event-based vision camera 1, a change in brightness occurs between when the subject 3 and the surrounding environment in the shooting area PA are visible and when the light is projected, making it possible to acquire an event signal ES. The optimum value for the flickering cycle depends on factors such as the distance to the subject, the size of the subject, and the surrounding lighting conditions.

[0043] The subject is illuminated with a flickering light. For example, if the subject is illuminated with flickering light from an angle, a change in brightness occurs between when no light is shining on the subject 3 or the surrounding environment in the photographing area PA and when light is shining on the subject 3 or the surrounding environment in the photographing area PA, and therefore an event signal ES can be acquired. The optimum value for the flickering cycle depends on factors such as the distance to the subject, the size of the subject, and the surrounding lighting conditions.

[0044] The above describes embodiments of the signal acquisition method for an event-based vision camera according to the present invention based on the accompanying drawings. However, the present invention is not limited to these embodiments, and may be implemented by adapting publicly known, existing equivalent technical means as long as the configuration described in the claims is not changed. [Explanation of symbols]

[0045] 1. Event-based vision camera 1a Event-based vision sensor 1b Optical system 2A First Event Generation Method 21 Slit body 22 Slit body drive mechanism 3. Subject 4. Photo booth 5 Signal Cable 6. Information processing equipment PA shooting area CD cross direction ES Event Signal

Claims

[Claim 1] An event signal acquisition method for an event-based vision camera equipped with an event-based vision sensor that asynchronously outputs an event signal from a pixel that detects a change in brightness caused by movement of a subject in a shooting area or a change in the surrounding environment, the method receiving incident light from the subject that is not moving in the shooting area or the surrounding environment that is not changing and acquiring the event signal, comprising: a first step of disposing an event generating means between the event-based vision camera and the photographing area where the event-based vision camera receives incident light; a second step of changing, over time, a position of a shielding area that blocks incident light from the photographing area and a position of a transmission area that transmits incident light from the photographing area by the event generating means; to cause a change in luminance in accordance with a change in the position of the shielding region and / or a change in the position of the transmitting region, and to acquire the event signal; the event generating means includes a slit body in which the transmission area is formed as a slit portion between each of the shielding materials by arranging a plurality of shielding materials at intervals to function as the shielding area, and performing the first step of arranging the slit body so that the shielding material and the slit portion are simultaneously captured by the event-based vision camera; the second step of changing the positions of the blocking area and the transmitting area over time by moving the slit body in an intersecting direction that intersects with the shooting direction of the event-based vision camera so as to block the shooting direction of the event-based vision camera; The event signal acquisition method for an event-based vision camera, wherein the event generating means supports the slit body via a swingable support that can swing the slit body in the intersecting direction.

Images