KR20220108920A - Illumination device for lighting based on possibility of pedestrian collision - Google Patents

Abstract

The present invention relates to a lighting device based on pedestrian collision possibility recognition, which comprises: a light emitting unit which includes a plurality of light sources and emits light; at least one lens which is disposed in front of the light source; at least one camera which is installed in the light emitting unit and photographs the front of the lighting device; and a control unit which receives an image from the camera, recognizes an object, and controls the light emitting unit based on the same. The control unit selects an object of which the collision probability value is greater than or equal to a predetermined threshold point among a plurality of objects in the front, and controls the light emitting unit to emit light to the selected object.

Description

Pedestrian collision potential recognition based lighting device {ILLUMINATION DEVICE FOR LIGHTING BASED ON POSSIBILITY OF PEDESTRIAN COLLISION}

The present invention relates to a lighting apparatus based on recognition of the possibility of pedestrian collision, and more particularly, by photographing the front of the lighting apparatus through a camera, and recognizing an object through the photographed image, the movement direction of the object and movement within the image Technology that collects distance and analyzes the moving direction of the object and the moving distance in the image through the controller to determine the possibility of collision with the object, then selects the object with high probability of collision and scans the light to prevent collision with the object is about

Personal means of transportation are suitable for the aspect of multi-modal transportation that utilizes various means of transportation according to the purpose of transportation, and transportation that can be used to the final destination after using public transportation while linking with public transportation means It can be used as a means (last-mile mobility) and is attracting attention in that it is suitable for car-sharing from the point of view of use rather than ownership.

Personal means of transportation, which have recently been widely used, are mainly used for commuting and short-distance transportation.

In order to safely drive bicycles and electric scooters, which are representative personal means of transportation at night, a headlight is essential to secure the rider's view. The demand for high-brightness and expensive lighting is gradually increasing.

In the case of such a headlamp, it is a situation that makes a lot of accidents because it has a disadvantage of causing glare to an oncoming target for the purpose of only high luminous intensity.

The fatality rate of personal mobility accidents at night is three times higher than that of daytime accidents. Failure to do so is the most common cause.

In addition, since such nighttime accidents may occur in unmanned vehicles operated unmanned and vehicles equipped with autonomous driving in recent years, efforts to prevent the root cause of night accidents are urgently needed.

The present invention is to solve the problems of the prior art, and after determining the collision possibility by determining the distance to the object at night or in a dark situation, by injecting light to an object with a high probability of collision to reduce the probability of an accident. It aims to provide a lighting device based on possibility recognition.

However, the technical task to be achieved by the present embodiment is not limited to the above-described technical task, and other technical tasks may exist.

In the lighting device based on the possibility of pedestrian collision, a light emitting unit that emits light including a plurality of light sources, at least one lens disposed in front of the light source, is installed in the light emitting unit and installed in the light emitting unit to view the front of the lighting device at least one camera for photographing; and a controller for receiving the image from the camera, recognizing the object, and controlling the light emitting unit based on this, wherein the controller selects and selects an object whose collision probability value is equal to or greater than a preset threshold among a plurality of objects in front Pedestrian collision possibility recognition-based lighting device comprising controlling a light emitting unit to inject light with respect to the object.

In addition, the camera may be configured as a fixed type or a rotation type, and in the case of a stationary type, it is fixed without a change in position, and in the case of a rotation type, the lighting device further includes an electric motor to rotate the camera 360 degrees. have.

In addition, the type of light source constituting the light emitting unit may include a straight beam, a high beam, and a low beam.

In addition, a plurality of light sources constituting the straight beam, the high beam and the low beam may be arranged horizontally with respect to the ground.

In addition, as the substrate supporting the light source of the light emitting unit is formed to be curved, the angle at which the plurality of light sources scan may have an obtuse angle.

In addition, a plurality of lenses may be installed to correspond to each light source constituting the straight beam, the high beam, and the low beam.

In addition, the camera is disposed in an area facing the front of the lighting device, and continuously shoots the front of the lighting device at an obtuse angle and transmits it to the control unit, but a plurality of light sources are disposed at positions corresponding to the angle of the camera It may be implemented to inject light into objects in different directions.

Also, the control unit may calculate the collision probability value through the collision prediction model.

In addition, the collision prediction model extracts the object from the accident image, and tracks the change in the position of the object over time through a Kalman filter-based SORT (SIMPLE ONLINE AND REALTIME TRACKING) algorithm and detects a collision between the objects. And, the collision probability value and the collision expected time can be calculated by collecting the time until the collision occurs, the moving direction, and the moving distance in the image for each object.

In addition, the control unit extracts and analyzes the object in the image through the image transmitted from the camera, calculates the movement direction of the object and the movement distance in the image, and calculates the movement direction of the object and By inputting a movement distance within the image to the collision prediction model, the collision probability value and the collision expected time are output, and the selected object is selected by selecting an object whose collision probability value is equal to or greater than a preset threshold point. It may include controlling the light emitting unit to scan the light with respect to the.

In addition, the controller may include calculating a collision probability value for each object when there are a plurality of objects extracted and analyzed from the image received from the camera.

In addition, when the control unit controls the light emitting unit, the control unit detects an object having the collision probability value exceeding a value corresponding to a preset threshold among the collision probability values for each object, and then directs the object to the light emitting unit The light emitting unit can be controlled by providing a value and a blinking value.

In addition, when the collision probability value with the object is greater than or equal to a numerical value corresponding to a preset threshold, the control unit provides a direction value and a blinking value of the object to the light emitting unit to control the injection of light to the object through the high beam, and collision with the object When the likelihood value falls below a value corresponding to a preset threshold, stopping the control of the high beam and operating the low beam.

According to an embodiment of the present invention, in a lighting device based on recognition of the possibility of pedestrian collision, an image of the front of the lighting device is received from a camera, the object is recognized and analyzed through the image, and the movement direction of each object and the inside of the image are received. After calculating the movement distance in the , inputting it into the collision prediction model to determine the possibility of collision with the object, the outputted collision probability value is selected for the object corresponding to the preset critical point or more, and light is injected to the selected object. As the object and the rider recognize each other, the incidence of nighttime accidents can be reduced. In addition, the present invention can be applied to various transportation means such as delivery robots, unmanned vehicles, and automobile headlights.

1 is a view showing the outline of a lighting device based on the recognition of the possibility of pedestrian collision, according to an embodiment of the present invention.
2 is a diagram showing the configuration of a lighting device based on recognition of the possibility of pedestrian collision according to an embodiment of the present invention.
3A is a diagram illustrating a configuration of a light emitting unit in a lighting device according to an embodiment of the present invention.
3B is an exemplary diagram illustrating a structure of a light emitting unit in a lighting device according to an embodiment of the present invention.
3C is an exemplary view illustrating a direction of a light source scanned by a light emitting unit in a lighting device and a scanning area according to an embodiment of the present invention.
4 is a diagram illustrating the configuration of a control unit of a lighting apparatus based on recognition of the possibility of pedestrian collision according to an embodiment of the present invention.
5 is an overall flowchart of an operation of a lighting device based on recognition of the possibility of pedestrian collision according to an embodiment of the present invention.
6 is a flowchart for explaining the construction of a collision prediction model of a lighting apparatus based on recognition of the possibility of pedestrian collision according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention. However, the present invention may be embodied in several different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

Throughout the specification, when a part is "connected" with another part, this includes not only the case of being "directly connected" but also the case of being "electrically connected" with another element interposed therebetween. . Also, when a part "includes" a certain component, it means that other components may be further included, rather than excluding other components, unless otherwise stated.

In this specification, a "part" includes a unit realized by hardware, a unit realized by software, and a unit realized using both. In addition, one unit may be implemented using two or more hardware, and two or more units may be implemented by one hardware. Meanwhile, '~ unit' is not limited to software or hardware, and '~ unit' may be configured to be in an addressable storage medium or may be configured to reproduce one or more processors. Thus, as an example, '~' denotes components such as software components, object-oriented software components, class components, and task components, and processes, functions, properties, and procedures. , subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays and variables. The functions provided in the components and '~ units' may be combined into a smaller number of components and '~ units' or further separated into additional components and '~ units'. In addition, components and '~ units' may be implemented to play one or more CPUs in a device or secure multimedia card.

The "device" referred to below may be implemented as a computer or a portable terminal capable of accessing a server or other terminal through a network. Here, the computer is, for example, a laptop, desktop, laptop, VR HMD (eg, HTC VIVE, Oculus Rift, GearVR, DayDream, PSVR, etc.) equipped with a web browser (WEB Browser), etc. may include Here, VR HMD is for PC (e.g., HTC VIVE, Oculus Rift, FOVE, Deepon, etc.), mobile (e.g., GearVR, DayDream, Storm Horse, Google Cardboard, etc.) Independently implemented Stand Alone models (eg Deepon, PICO, etc.) are included. A portable terminal is, for example, a wireless communication device that guarantees portability and mobility, and includes not only a smart phone, a tablet PC, a wearable device, but also Bluetooth (BLE, Bluetooth Low Energy), NFC, RFID, and ultrasound (Ultrasonic). , infrared, Wi-Fi, Li-Fi, etc. may include various devices equipped with a communication module. In addition, "network" refers to a connection structure capable of exchanging information between each node, such as terminals and servers, and includes a local area network (LAN), a wide area network (WAN), and the Internet. (WWW: World Wide Web), wired and wireless data networks, telephone networks, wired and wireless television networks, and the like. Examples of wireless data communication networks include 3G, 4G, 5G, 3rd Generation Partnership Project (3GPP), Long Term Evolution (LTE), World Interoperability for Microwave Access (WIMAX), Wi-Fi, Bluetooth communication, infrared communication, ultrasound communication, Visible Light Communication (VLC), LiFi, and the like, but are not limited thereto.

Hereinafter, 'mobility' means a means of transport that can transport people, such as bicycles, electric boards, and kickboards, and may include delivery robots, unmanned vehicles, and vehicles equipped with autonomous driving functions. In addition, 'rider' hereinafter means a person driving mobility.

Hereinafter, with reference to FIG. 1 , the lighting device 30 based on the recognition of the possibility of pedestrian collision according to an embodiment of the present invention will be described in general as follows. The lighting device 30 according to an embodiment of the present invention is to brighten the field of vision in front of the mobility, and may be installed in front of the mobility.

The lighting device 30 may recognize each object located in the front and measure a distance value for each object. Based on the measured distance value, the lighting device 30 may calculate the probability of collision with each object.

The lighting device 30 may scan the light source toward the calculated high probability of collision so that the object and the rider can recognize each other, thereby reducing the probability of collision.

In this case, when a plurality of objects are highly likely to collide, the lighting device 30 may scan a light source toward each object.

Hereinafter, with reference to FIGS. 2 and 3A , the configuration of the lighting device 30 based on pedestrian collision possibility recognition according to an embodiment of the present invention will be described in detail.

According to an embodiment of the present invention, the lighting device 30 includes a light emitting unit 100 , a camera 300 , and a control unit 400 , and is photographed by photographing the front through the camera 300 disposed therein. The image is transmitted to the control unit 400, and the control unit 400 recognizes the object, calculates the moving direction of the object and the moving distance in the image, and calculates the moving direction of the object and the moving distance in the image as a collision prediction model. A signal is provided to the light emitting unit 100 so as to output a collision probability value with each object by input to (not shown), but to inject light to an object having a high probability of collision.

The signal provided from the control unit 400 to the light emitting unit 100 may include a blinking value and a direction value, and the blinking value represents the blinking cycle of each light source, the number of blinks, and the time to turn on when blinking once. , and the direction value may be to control which light source among the light sources constituting the light emitting unit 100 is operated or in which direction the light source is operated.

The light source of the light emitting unit 100 is composed of a straight lamp 110 , a high beam 120 , and a low beam 130 , and according to a signal received from the control unit 400 , the light emitting unit 100 is a straight lamp 110 , a high beam. (120), the low beam 130 is controlled, it is possible to inject light to the object so that the rider and the object can recognize each other.

In the lighting device 30 , the lens 200 may be installed in front of the light source in order to focus and scan the light on the object. A plurality of lenses 200 may be installed to correspond to each light source constituting the straight lamp 110 , the high beam 120 , and the low beam 130 .

A plurality of light sources constituting the straight lamp 110 , the high beam 120 , and the low beam 130 may be arranged horizontally with respect to the ground, and the angle of the light source scanned through the lens 200 may be adjusted.

The camera 300 continuously shoots the front of the lighting device and provides the captured image to the control unit 400 , and the control unit 400 continuously calculates a collision probability value with the object and transmits the information to the light emitting unit 100 . However, when the collision probability value with the object is greater than or equal to the preset threshold, the high beam 120 is operated toward the object, and when the collision probability value with the object falls below the preset threshold, the operation of the high beam 120 is stopped and A blinking value and a direction value may be provided to control the operation of the low beam 130 .

In addition, when a plurality of objects are detected, the controller 400 collects and calculates information for each object, and inputs the calculated movement direction and movement distance in the image to a collision prediction model (not shown). , the collision prediction model (not shown) calculates each collision probability value based on the movement direction of each object received and the movement distance in the image, It can provide the blinking value and direction value of . The light emitting unit 100 may operate a plurality of high beams 120 or a plurality of low beams 130 according to the blinking value and direction value provided from the control unit 400 to inject light to each object.

At least one camera 300 is disposed on the lighting device 30 , the front of the lighting device 30 is photographed at an obtuse angle, and a plurality of light sources corresponding to the angle of the camera 300 are respectively disposed, and in different directions. It can be implemented to detect objects of

In addition, the camera 300 is disposed in an area facing the front of the lighting device, and can photograph the front of the lighting device 30 at an obtuse angle.

According to the shape of the lighting device 30 , the camera 300 is disposed toward the front when the lighting device 30 is fixed, and when the lighting device 30 is rotary, it is implemented to rotate 360 degrees through an electric motor. Accordingly, the camera 300 may be rotated by 360 degrees through an electric motor.

As an additional embodiment, instead of the camera 300, an object detection sensor may be installed, and the object detection sensor may be implemented as an ultrasonic sensor or a time of flight (TOF) sensor, and further, the object detection sensor is provided from the front. The luminance value can be determined by measuring the amount of reflected light.

In a further embodiment, the controller 400 may control the light emitting unit 100 based on the amount of light reflected from the front detected by the object detection sensor to adjust the amount of light of the light source scanned in the front. When the amount of reflected light is small, the amount of light of the light source can be increased, and when the amount of reflected light is large, the amount of light of the light source can be decreased.

The control unit 400 captures the front of the lighting device 30 from the camera 300, recognizes the object through the captured image, calculates the moving direction of the object and the moving distance in the image, and determines the possibility of collision based on this. After calculating and determining, the light emitting unit 100 can be controlled. In addition, when there are a plurality of objects in front, the light emitting unit 100 may be controlled to select an object corresponding to a predetermined threshold or higher of collision possibility among the plurality of objects and inject light to the selected object.

In this case, when the mobility is in operation, the type of light source constituting the light emitting unit 100 may be selectively controlled according to a preset speed and direction by sensing the movement speed of the mobility.

For example, if the distance from the object recognized in front is not far and the mobility is fast, the control unit 400 controls the high beam 120 instead of the low beam 130 because the collision probability is high to control the rider and the object. can make everyone aware of the possibility of an accident.

Referring to Figure 3a, the straight lamp 110 is number 13, the high beam 120 is 7, 8, 9, 10, 11, 12, and the low beam 130 is 1,2,3,4,5,6. consists of numbers. The number of lights is arbitrarily assigned, and the number of lights can be changed depending on the number or location.

The straight lamp 110 , the high beam 120 , and the low beam 130 are composed of a plurality of light sources, and each light source is periodically turned on and off based on PWM (Pulse Width Modulation) to operate.

Referring to FIG. 3B , in the structure of the light emitting unit 100 , the straight lamp 110 is disposed at the uppermost portion of the central portion. In the drawing, the straight light 110 is marked as one, but a plurality of them may be disposed. In addition, the straight light may be disposed at a position other than the upper side.

The plurality of high beams 120 are disposed below the straight lamp 110 , and the plurality of low beams 130 are disposed at the bottom. In addition, the camera 300 is disposed between the high beam 120 and the low beam 130 . The arrangement position of the camera 300 is only an example, and may be arranged in another area.

In the straight lamp 110 , the high beam 120 and the low beam 130 , the lens 200 is disposed in front of each light source so that the forward-scanned light can be focused on the object.

The numbers indicated in FIG. 3B may indicate areas and directions scanned by each light source of FIG. 3C, which will be described later.

When the light emitting unit 100 of the lighting device 30 based on pedestrian collision possibility recognition according to an embodiment of the present invention will be described with reference to FIG. 3C , the angle and angle at which a plurality of light sources of the light emitting unit 100 scan You can check the direction the light source scans. Specifically, the area of light emitted by the light source No. 3 in FIG. 3B means the area indicated by No. 3 in FIG. 3C. In this way, the area and angle illuminated by each light source are different as shown in FIG. 3C . In this case, the angle of the total light scanned by the plurality of light sources may be in the range of an obtuse angle. On the other hand, the substrate (not shown) supporting the light source of the light emitting part may be configured to be flat, but if it is configured to be slightly curved (convexly curved with respect to the direction in which the mobility is directed) in a convex shape, the angle at which the light source illuminates will become wider. can

In addition, the light source of the light emitting unit 100 may operate differently according to the speed and direction of mobility. The control unit 400 detects the movement speed of the mobility, and when running at a speed less than a preset speed, blinks the low beam 130, and when driving at a speed higher than a preset speed, sends a signal to the light emitting unit 100 so that the high beam 120 blinks. ) is provided for This can also be applied when the rider manipulates the steering wheel, and when steering to the left, the light source located on the left blinks, and when steering to the right, the light source located on the right side may blink.

When the light source of the light emitting unit 100 operates according to the speed, it is possible to illuminate a wide radius when driving at a slow speed, and to illuminate a narrow radius further when driving at a high speed.

For example, when the mobility running at a preset speed or less rotates to the right, the number 3, 4, 5, 6, etc. may blink. In addition, when the mobility running at a preset speed or more goes straight ahead, the number 9, 10, and 13 flickers so that the light source can illuminate farther away from the front, and illuminate with a narrower radius.

In addition, each light source of the light emitting unit 100 may be operated through a user's manipulation.

In a further embodiment, an object or other driver may be notified via a light source to recognize the rider. The light emitting unit 100 may operate to recognize the rider by blinking or blinking the high beam 120 through the light source in a specific section according to the position of the object or other driver, and after blinking, the low beam 130 in a specific section can be operated to illuminate.

Hereinafter, with reference to FIG. 4 , the controller 400 of the lighting device 30 based on the recognition of the possibility of pedestrian collision according to an embodiment of the present invention will be described as follows.

In detail, the communication module 110 provides a communication interface necessary to provide a transmission/reception signal between the control unit 400 and the camera 300 and the light emitting unit 100 in the form of packet data in connection with a communication network. Furthermore, the communication module 410 may serve to receive a data request from each device and transmit data in response thereto.

Here, the communication module 410 may be a device including hardware and software necessary for transmitting and receiving signals such as control signals or data signals through wired/wireless connection with other network devices.

In the memory 420, a collision prediction model (not shown) for performing the operation of the lighting device 30 based on recognition of the possibility of pedestrian collision is recorded. In addition, the processor 130 performs a function of temporarily or permanently storing the processed data. A collision prediction model (not shown) recorded in the memory 420 will be described later with reference to FIG. 5 . Here, the memory 420 may include magnetic storage media or flash storage media, but the scope of the present invention is not limited thereto.

The processor 430 is a kind of central processing unit and controls the operation of the lighting device 30 based on recognition of the possibility of pedestrian collision. Each step performed by the processor 430 will be described later with reference to FIG. 6 .

Here, the processor 430 may include all kinds of devices capable of processing data, such as a processor. Here, the 'processor' may refer to, for example, a data processing device embedded in hardware having a physically structured circuit to perform a function expressed as a code or an instruction included in a program. As an example of the data processing device embedded in the hardware as described above, a microprocessor, a central processing unit (CPU), a processor core, a multiprocessor, an application-specific integrated (ASIC) circuit) and a processing device such as a field programmable gate array (FPGA), but the scope of the present invention is not limited thereto.

Hereinafter, with reference to FIG. 5 , the function of the lighting device 30 based on recognition of the possibility of pedestrian collision according to an embodiment of the present invention will be described in detail.

According to an embodiment of the present invention, in the lighting device 30 based on recognition of the possibility of pedestrian collision, the camera 300 disposed in the lighting device 30 continuously shoots the front of the lighting device, and transmits the captured image to the controller. can be transmitted (S510).

The controller 400 may extract an object from the image transmitted from the camera 300 and calculate a movement direction for each object and a movement distance within the image. (S520)

The controller 400 recognizes and analyzes the object through the image transmitted from the camera 300, extracts the object, calculates the movement direction for each object and the movement distance in the image, but if there are a plurality of recognized objects The movement direction for each object and the movement distance in the image can be calculated.

The controller 400 may input the calculated moving direction and the moving distance in the image to the collision prediction model (not shown). (S530)

Through the collision prediction model (not shown), the collision probability and collision prediction time for each object can be calculated, and the collision probability value can be output. (S540)

The control unit 400 inputs the calculated moving direction and the moving distance in the image to a collision prediction model (not shown), and the collision prediction model (not shown) calculates a collision probability value and a collision prediction time through the calculated values. can In this case, the collision prediction model may continuously receive values from the control unit 400 , and accordingly calculate and output a collision probability value and a collision prediction time continuously.

A collision prediction model (not shown) may be pre-built through machine learning. The collision prediction model automatically collects accident images through the Internet, extracts objects from the collected images, recognizes information about collisions in the images, and analyzes them by matching the moving direction, speed, and time of the object. . And, it can be constructed by repeatedly performing machine learning with values such as images, accident occurrence, movement direction and speed, and time of an object as input values, and numerical values for the possibility of collision as output values. This will be described later in detail with reference to FIG. 6 .

The control unit 400 may select an object whose collision probability value output by the collision prediction model (not shown) is equal to or greater than a preset critical point, and may inject light onto the selected object. (S550)

When there is an object whose calculated collision probability is greater than or equal to a preset threshold, the controller 400 may control the object by transmitting a signal to the light emitting unit 100 to select the object and inject light to the selected object.

The light emitting unit 100 may receive the blinking value and direction value of the selected object from the control unit 400 and may scan the light toward the object in front.

At this time, depending on the blinking value, the blinking period of each light source, the blinking frequency, and the time to turn on at one blink are controlled. It may be determined whether to operate or which direction of the light sources constituting the high beam is to be controlled.

In a further embodiment, the control unit may be divided into a plurality of stages such as recognition, recognition, attention, warning, and the like, and transmit the flashing value to the light emitting unit at each stage, and the light emitting unit may control the light source according to the flashing value for each stage.

As an additional embodiment, when the collision probability value with the object in front is greater than or equal to a preset threshold, both the rider and the object may be aware of the high probability of collision through a warning sound.

In addition, when the collision probability value with an object in front is greater than or equal to a preset critical point, it is possible to make the rider aware of the high probability of collision with the object through the vibration of the handle.

At this time, the vibration of the handle may operate the vibration on the handle opposite to the object to steer the handle toward the vibration sounding side. For example, when the probability of collision with an object located on the left side with respect to the front is greater than or equal to a preset critical point, vibration may be generated in the rider's right handle to induce the rider to steer to the right.

Hereinafter, with reference to FIG. 6, the above-described collision prediction model (not shown) will be described in more detail.

The collision prediction model (not shown) is built through machine learning before being applied to the lighting device, and after the collision prediction model is built, it is applied to the lighting device to predict the collision in real time. can be implemented

Referring to FIG. 6 , each step may be added or simplified as necessary.

In the step of constructing a collision prediction model (not shown), first, an image may be automatically collected through the Internet, and an object may be extracted from the collected image. (S610)

An accident image or an image including an image with a high probability of an accident can be automatically collected through the Internet, and objects can be extracted from the collected images.

After the object is extracted, it is possible to track a change in the position of the object over time and detect a collision between the objects through a Kalman filter-based SORT (Simple ONLINE AND REALTIME TRACKING) algorithm. (S620)

It is possible to construct a collision prediction model (not shown) that calculates the collision probability value and the expected collision time by collecting the time until the collision for each object, the moving direction, and the moving distance in the image. (S630)

Building a collision prediction model (not shown) that calculates the collision probability value and expected collision time through repeated learning by collecting the time, movement direction, and moving distance within the image for each object through the image. , the control unit 400 analyzes the image received from the camera in real time to recognize the object, collects the movement direction of each object and the movement distance within the image, and inputs it to a collision prediction model (not shown), thereby providing a collision probability value and an expected collision time.

Information necessary for calculating the possibility of collision with an object may be a movement speed of the mobility, a movement direction of the mobility, a movement speed of the object, and a movement direction of the object.

In a further embodiment, the process of calculating the collision probability value may be performed by a machine learning algorithm. For example, the controller 400 matches and collects the calculated collision probability value and data on whether a collision actually occurs. The collision probability value may be calculated and stored by the controller 400 through information continuously collected and transmitted from the camera 300 . Data on the actual occurrence of collisions can be entered by recording whether collisions occur by actual collision possibility. And, after setting the accident image, the moving speed of the mobility, the moving direction of the mobility, the moving speed of the object and the moving direction of the object as input values, and setting the collision probability value and the expected collision time as output values, a preset machine learning model A collision prediction model can be built by repeating the learning process applied to

After the collision prediction model learning is completed, the device collects the images captured in real time and inputs them into the collision prediction model, so that the collision probability value and the collision expected time can be calculated immediately, and this can be directly applied to the driving of mobility. can

An embodiment of the present invention may also be implemented in the form of a recording medium including instructions executable by a computer, such as a program module executed by a computer. Computer-readable media can be any available media that can be accessed by a computer and includes both volatile and non-volatile media, removable and non-removable media. Also, computer-readable media may include all computer storage media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data.

Although the methods and systems of the present invention have been described with reference to specific embodiments, some or all of their components or operations may be implemented using a computer system having a general purpose hardware architecture.

The foregoing description of the present invention is for illustration, and those of ordinary skill in the art to which the present invention pertains can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. For example, each component described as a single type may be implemented in a distributed manner, and likewise components described as distributed may also be implemented in a combined form.

The scope of the present invention is indicated by the following claims rather than the above detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention. do.

30: lighting device
100: light emitting part
110: straight light
120: high beam
130: low beam
200: lens
300: camera
400: control unit
410: communication module
420: memory
430: processor

Claims (13)

In the lighting device based on the recognition of the possibility of pedestrian collision,
a light emitting unit including a plurality of light sources to emit light;
at least one lens disposed in front of the light source;
at least one camera installed in the light emitting unit to photograph the front of the lighting device; and
A control unit for receiving the image from the camera, recognizing the object, determining the possibility of collision with each object, and controlling the light emitting unit based on this;
The control unit selects an object whose collision probability value is greater than or equal to a preset value among a plurality of objects in front, and controls the light emitting unit to inject light to the selected object. . 2. The method of claim 1
The camera may be of a fixed or rotary configuration,
In the case of the fixed type, it is fixed without a change in position,
In the case of the rotation, the lighting device further includes an electric motor, including being implemented to rotate the camera 360 degrees
A lighting device based on the recognition of the possibility of pedestrian collision. 2. The method of claim 1
The type of the light source constituting the light emitting unit will include a straight lamp, a high beam, and a low beam,
A lighting device based on the recognition of the possibility of pedestrian collision. 4. The method of claim 3
A plurality of light sources constituting the straight lamp, the high beam and the low beam are arranged horizontally with respect to the ground,
A lighting device based on recognition of the possibility of pedestrian collision. 4. The method of claim 3
As the substrate supporting the light source of the light emitting unit is formed to be curved, the angle at which the plurality of light sources scan has an obtuse angle,
A lighting device based on the recognition of the possibility of pedestrian collision. The method of claim 1,
The lens is provided in plurality to correspond to each light source constituting the straight beam, high beam, and low beam.
A lighting device based on the recognition of the possibility of pedestrian collision. 2. The method of claim 1
The camera is disposed in an area facing the front of the lighting device, and continuously captures the front of the lighting device at an obtuse angle and transmits it to the control unit, but a light source is disposed at a position corresponding to the angle of the camera and is different from each other which is implemented to inject light to an object in a direction,
A lighting device based on the recognition of the possibility of pedestrian collision. The method of claim 1,
The control unit is to calculate the collision probability value through the collision prediction model,
A lighting device based on the recognition of the possibility of pedestrian collision. 9. The method of claim 8,
The collision prediction model automatically collects accident images through the Internet, extracts the objects from the collected images, but follows the passage of time through a Kalman filter-based SORT (SIMPLE ONLINE AND REALTIME TRACKING) algorithm. To calculate the collision probability value and the collision expected time by tracking changes and detecting the collision between the objects, collecting the time until the collision for each object, the moving direction, and the moving distance in the image,
A lighting device based on the recognition of the possibility of pedestrian collision. The method of claim 1,
The control unit extracts and analyzes the object in the image through the image transmitted from the camera, calculates the moving direction of the object and the moving distance in the image, and calculates the moving direction of the object and the image Input the movement distance within the collision prediction model to output the collision probability value and the collision expected time, and select an object whose output collision probability value is equal to or greater than a preset value, and the selected object Including controlling the light emitting unit to scan the light,
A lighting device based on the recognition of the possibility of pedestrian collision. 11. The method of claim 10
The control unit extracts an object in the image through the image transmitted from the camera and, when there are a plurality of analyzed objects, calculating a movement direction for each object and a movement distance in the image sign,
A lighting device based on the recognition of the possibility of pedestrian collision. 2. The method of claim 1
That the control unit controls the light emitting unit,
The control unit detects an object having the collision probability value exceeding a value corresponding to a preset threshold among the collision probability values for each object, and then provides a direction value and a blinking value of the object to the light emitting unit to control the light emitting unit that is
A lighting device based on recognition of the possibility of pedestrian collision. 13. The method of claim 12
the control unit
When the collision probability value with the object is greater than or equal to a numerical value corresponding to the preset critical point, the light emitting unit provides a direction value and a blinking value of the object to control to inject light to the object through a high beam,
When the collision probability value with the object falls below a value corresponding to the preset critical point, stopping the control of the high beam and operating the low beam,
A lighting device based on the recognition of the possibility of pedestrian collision.

Images