KR20220032792A - Edge computing-based headlight inspection device and method - Google Patents

Abstract

The present invention relates to an edge computing-based headlight inspection device and method. The device comprises: a light distribution image acquisition unit that acquires a light distribution image of headlights emitted toward a screen in front of a vehicle; a binarized image generation unit that generates a binarized image for headlight inspection based on the light distribution image; an elbow point detection unit that detects an elbow point by sequentially applying scan lines defined in horizontal and vertical directions to the binarized image; a cut-off line determination unit that determines a cut-off line based on the elbow point; and a headlight inspection unit that obtains a reference headlight model based on identification information of the vehicle, and compares and analyzes the reference headlight model with the elbow point and the cut-off line to determine whether or not the headlight is abnormal.

Description

EDGE COMPUTING-BASED HEADLIGHT INSPECTION DEVICE AND METHOD

The present invention relates to a headlight inspection technology, and more particularly, to an edge computing-based headlight inspection apparatus and method capable of quickly and accurately performing a vehicle headlight inspection in conjunction with a cloud server.

The headlights (or headlights) installed in the vehicle need to meet the safety standards stipulated by laws and regulations for the safe operation of the vehicle and the protection of the life of the vehicle driver. Therefore, headlight inspection is included as an essential item among vehicle inspection items performed immediately before the vehicle is shipped out or at a vehicle inspection station.

In general, the headlight inspection of a vehicle includes the inspection of high and low beams, and an automatic headlight inspection device using a CCD (Charge Coupled Device) sensor camera is used to precisely inspect the optical axis of the headlight. Since the device is still expensive, there is a problem in that the cost burden is large to be individually provided at a regular automobile inspection station or an on-site inspection station.

In addition, since the headlamp inspection device is implemented as a single device, there is a problem in that it is difficult to accurately inspect all the headlamps that are recently released in various ways. That is, the headlight inspection device has been developed technology to inspect headlamps having various specifications, but there are difficulties in solving the problem smoothly.

Korean Patent Registration No. 10-1076315 (2011.10.18)

An embodiment of the present invention is to provide an edge computing-based headlight inspection apparatus and method capable of quickly and accurately performing a vehicle headlight inspection in conjunction with a cloud server.

An embodiment of the present invention is to provide an edge computing-based headlight inspection apparatus and method capable of smoothly inspecting various types of headlights based on a reference headlight model provided from a cloud server.

An embodiment of the present invention is to provide an edge computing-based headlight inspection apparatus and method capable of accumulating inspection data for various vehicles and headlights and performing a quick and accurate headlight inspection through big data analysis .

In embodiments, the edge computing-based headlight inspection apparatus includes: a light distribution image acquisition unit for obtaining a light distribution image of a headlight irradiated toward a screen in front of the vehicle; a binarized image generator for generating a binarized image for headlight inspection based on the light distribution image; an elbow point detector configured to sequentially apply scan lines defined in horizontal and vertical directions to the binarized image, respectively, to detect an elbow point; a cut-off line determining unit for determining a cut-off line based on the elbow point; and a headlight inspection unit that acquires a reference headlight model based on the identification information of the vehicle, and compares and analyzes the reference headlight model with the elbow point and the cut-off line to determine whether the headlight is abnormal.

The light distribution image may include light distribution images related to low beam and high beam of the headlight.

The elbow point detector detects horizontal and vertical boundary points by sequentially applying the scan lines in the case of the binarized image for the high beam, and the cut-off line determiner determines a light center based on the horizontal and vertical boundary points, The headlight inspector may compare and analyze the reference headlight model, the horizontal and vertical boundary points, and the optical axis to determine whether the headlight is abnormal.

The binarized image generator may perform a preprocessing operation by applying at least one image filter to the binarized image.

The elbow point detection unit defines a horizontal H-scan line and a vertical V-scan line, moves each scan line in a vertical direction and a horizontal direction, respectively, at least one candidate whose rate of change in brightness exceeds a threshold value It is possible to detect the elbow point by acquiring points.

The headlight inspection unit includes an input module for receiving the identification information; a reference receiving module for receiving a reference headlight model corresponding to the identification information from the cloud server; a comparative analysis module for obtaining respective deviation information by comparing the elbow point and the cut-off line with the reference headlight model; and an inspection module for determining whether the headlight is abnormal based on the deviation information.

The headlight inspection unit transmits the binarized image to the cloud server to request a detailed inspection on the headlight, and the cloud server learns the black and white gradation image as a standard light source image to build a detailed inspection model, and the binarization A gradation image converted corresponding to the image may be provided as an input to the detailed inspection model to receive an output related to abnormality as an inspection result, and the inspection result may be transmitted to the headlight inspection unit in response to the request.

In embodiments, the edge computing-based headlight inspection method includes: acquiring a light distribution image of a headlight irradiated toward a screen in front of the vehicle; generating a binarized image for headlight inspection based on the light distribution image; detecting an elbow point by sequentially applying scan lines respectively defined in horizontal and vertical directions to the binarized image; determining a cut-off line based on the elbow point; and obtaining a reference headlight model based on the identification information of the vehicle, and comparing and analyzing the reference headlight model, the elbow point, and the cut-off line to determine whether the headlight is abnormal.

The disclosed technology may have the following effects. However, this does not mean that a specific embodiment should include all of the following effects or only the following effects, so the scope of the disclosed technology should not be understood as being limited thereby.

Edge computing-based headlight inspection apparatus and method according to an embodiment of the present invention can smoothly perform inspection of various types of headlights based on a reference headlight model provided from a cloud server.

The edge computing-based headlight inspection apparatus and method according to an embodiment of the present invention can accumulate inspection data for various vehicles and headlights and perform a fast and accurate headlight inspection through big data analysis.

1 is a view for explaining a headlight inspection system according to the present invention.
FIG. 2 is a view for explaining the system configuration of the headlight inspection apparatus of FIG. 1 .
FIG. 3 is a view for explaining a functional configuration of the headlight inspection apparatus of FIG. 1 .
4 is a flowchart illustrating an edge computing-based headlight inspection process according to an embodiment of the present invention.
5 is an exemplary view for explaining a vehicle headlight according to the position of the headlight.
6 is a view for explaining the physical configuration of the headlight inspection apparatus according to the present invention.
7 is an exemplary view for explaining a light distribution image related to the low beam of the headlight.
8 is an exemplary view for explaining a headlight inspection process according to the present invention.

Since the description of the present invention is merely an embodiment for structural or functional description, the scope of the present invention should not be construed as being limited by the embodiment described in the text. That is, since the embodiment may have various changes and may have various forms, it should be understood that the scope of the present invention includes equivalents capable of realizing the technical idea. In addition, since the object or effect presented in the present invention does not mean that a specific embodiment should include all of them or only such effects, it should not be understood that the scope of the present invention is limited thereby.

On the other hand, the meaning of the terms described in the present application should be understood as follows.

Terms such as “first” and “second” are for distinguishing one component from another, and the scope of rights should not be limited by these terms. For example, a first component may be termed a second component, and similarly, a second component may also be termed a first component.

When a component is referred to as being “connected to” another component, it may be directly connected to the other component, but it should be understood that other components may exist in between. On the other hand, when it is mentioned that a certain element is "directly connected" to another element, it should be understood that the other element does not exist in the middle. Meanwhile, other expressions describing the relationship between elements, that is, "between" and "between" or "neighboring to" and "directly adjacent to", etc., should be interpreted similarly.

The singular expression is to be understood as including the plural expression unless the context clearly dictates otherwise, and terms such as "comprises" or "have" refer to the embodied feature, number, step, action, component, part or these It is intended to indicate that a combination exists, and it should be understood that it does not preclude the possibility of the existence or addition of one or more other features or numbers, steps, operations, components, parts, or combinations thereof.

In each step, identification numbers (eg, a, b, c, etc.) are used for convenience of description, and identification numbers do not describe the order of each step, and each step clearly indicates a specific order in context. Unless otherwise specified, it may occur in a different order from the specified order. That is, each step may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

The present invention can be embodied as computer-readable codes on a computer-readable recording medium, and the computer-readable recording medium includes all types of recording devices in which data readable by a computer system is stored. . Examples of the computer-readable recording medium include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, and the like. In addition, the computer-readable recording medium is distributed in a computer system connected to a network, so that the computer-readable code can be stored and executed in a distributed manner.

All terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs, unless otherwise defined. Terms defined in general used in the dictionary should be interpreted as having the meaning consistent with the context of the related art, and cannot be interpreted as having an ideal or excessively formal meaning unless explicitly defined in the present application.

1 is a view for explaining a headlight inspection system according to the present invention.

Referring to FIG. 1 , the headlight inspection system 100 may include a vehicle 110 , a headlight inspection device 130 , a database 150 , and a cloud server 170 .

The vehicle 110 is a transportation means for transporting passengers or cargo using power produced by an engine, and may correspond to a vehicle as a representative example. The vehicle 110 may include not only automobiles, but also ships, airplanes, and the like, and is not necessarily limited thereto, and may include various transportation means capable of moving using power.

The headlight inspection apparatus 130 may correspond to a computing device capable of inspecting the presence or absence of abnormality regarding the headlight of the vehicle 110 and generating and providing an inspection result. Here, the headlight may be disposed in the forward direction of the vehicle 110 , and may be symmetrically coupled to both sides to operate. The headlight inspection apparatus 130 may be connected to at least one cloud server 170 through a network, and may exchange various information necessary for a headlight inspection.

In addition, the headlight inspection apparatus 130 may be implemented by including various sensors necessary for the headlight inspection of the vehicle 110 . For example, the headlight inspection apparatus 130 may be implemented including a distance sensor, a camera sensor, a radar (RADAR) sensor, a LIDAR sensor, and the like.

In one embodiment, the headlight inspection apparatus 130 may be implemented in a form that is fixedly installed on the ground, or may be implemented in a form that is movable so that the position can be freely changed during the inspection process. On the other hand, the headlight inspection apparatus 130 may be included and operated as a component of an automatic vehicle inspection system. That is, when the vehicle 110 is positioned in a fixed place, the automatic vehicle inspection system may sequentially or simultaneously perform inspections on various items according to a series of inspection scenarios, and integrate and provide the results.

In addition, the headlight inspection apparatus 130 may be implemented including the database 150 , and may be implemented independently of the database 150 . When implemented independently of the database 150 , the headlight inspection device 130 may be connected to the database 150 by wire or wirelessly to exchange data.

The database 150 may store various pieces of information necessary for the inspection of the headlight of the vehicle 110 . The database 150 may store information about the vehicle 110 and the headlight, and may store information about inspection rules and conditions for the headlight inspection, but is not necessarily limited thereto, and the headlight of the vehicle 110 . In the process of performing an inspection, information collected or processed in various forms may be stored.

The cloud server 170 may correspond to a service providing server capable of providing various functions required for headlight inspection while operating in conjunction with the headlight inspection device 130 . The cloud server 170 may be connected to the headlight inspection device 130 through a network, and if necessary, it is connected to a user terminal owned by the user of the vehicle 110 to obtain various information related to the order and method of inspection and inspection results. can provide

FIG. 2 is a view for explaining the system configuration of the headlight inspection apparatus of FIG. 1 .

Referring to FIG. 2 , the headlight inspection apparatus 130 may be implemented to include a processor 210 , a memory 230 , a user input/output unit 250 , and a network input/output unit 270 .

The processor 210 may execute a procedure for processing each step in the process in which the headlight inspection device 130 operates, and manage the memory 230 that is read or written throughout the process, and the memory ( 230) may schedule a synchronization time between the volatile memory and the non-volatile memory. The processor 210 may control the overall operation of the headlight inspection device 130 , and is electrically connected to the memory 230 , the user input/output unit 250 , and the network input/output unit 270 to control data flow therebetween. can do. The processor 210 may be implemented as a central processing unit (CPU) of the headlight inspection apparatus 130 .

The memory 230 is implemented as a non-volatile memory, such as a solid state drive (SSD) or a hard disk drive (HDD), and may include an auxiliary storage device used to store overall data required for the headlight inspection device 130 and , and may include a main memory implemented as a volatile memory such as random access memory (RAM).

The user input/output unit 250 may include an environment for receiving a user input and an environment for outputting specific information to the user. For example, the user input/output unit 250 may include an input device including an adapter such as a touch pad, a touch screen, an on-screen keyboard, or a pointing device, and an output device including an adapter such as a monitor or a touch screen. In an embodiment, the user input/output unit 250 may correspond to a computing device connected through a remote connection, and in such a case, the headlight inspection device 130 may be performed as a server.

The network input/output unit 270 includes an environment for connecting with an external device or system through a network, for example, a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), and a VAN (Wide Area Network) (VAN). It may include an adapter for communication such as Value Added Network).

FIG. 3 is a view for explaining a functional configuration of the headlight inspection apparatus of FIG. 1 .

Referring to FIG. 3 , the headlight inspection device 130 includes a light distribution image acquisition unit 310 , a binarized image generation unit 320 , an elbow point detection unit 330 , a cut-off line determination unit 340 , and a headlight inspection unit 350 . ) and a control unit 360 .

The light distribution image acquisition unit 310 may obtain a light distribution image of the headlight irradiated toward the screen in front of the vehicle 110 . In order to inspect the headlight, the vehicle 110 should be placed in front of the headlight inspection device 130 , and in a state in which the illumination of the headlight is irradiated toward the headlight inspection device 130 disposed in the front of the vehicle 110 . Inspection may be performed. In this case, the screen on which the light distribution image is reflected may be implemented as being included in the headlight inspection apparatus 130 , or may be implemented independently outside the headlight inspection apparatus 130 according to the usage environment. The light distribution image acquisition unit 310 may obtain a light distribution image formed on the screen through a camera sensor.

In an embodiment, the light distribution image may include light distribution images regarding low beam and high beam of a headlight. The headlight of the vehicle 110 may be divided into high beam and low beam according to a function, and the light distribution image acquisition unit 310 may independently acquire a light distribution image regarding the low beam and high beam of the headlight.

The binarized image generator 320 may generate a binarized image for headlight inspection based on the light distribution image. The light distribution image may be expressed in various colors, which may act as a cause of making a precise inspection difficult. Accordingly, the binarized image generator 320 may support smooth progress of the headlight inspection thereafter by converting the light distribution image into a black and white binarized image.

Meanwhile, the binarization image generator 320 may use various binarization (thresholding) algorithms to generate a corresponding binarized image from the light distribution image. The binarization algorithm can convert a light distribution image into a binarized image by determining an arbitrary brightness value as a reference value and changing it to white if it is greater than the reference value and black if it is smaller than the reference value for each pixel of the light distribution image.

In an embodiment, the binarized image generator 320 may perform a preprocessing operation by applying at least one image filter to the binarized image. For example, the binarized image generator 320 may apply a Kalman filter, a bilateral filter, a Gaussian filter, etc. to the binarized image. Meanwhile, the binarized image generator 320 may generate a binarized image by applying various image filters to the light distribution image itself and then applying a binarization algorithm.

The elbow point detector 330 may detect an elbow point by sequentially applying scan lines defined in horizontal and vertical directions to the binarized image, respectively. Here, the scan line may correspond to a search line for searching a binarized image for detecting an elbow point. That is, the scan line may act as a reference for partial motion for detecting the elbow point while moving in the horizontal or vertical direction of the binarized image. The elbow point detection unit 330 may detect the elbow point by applying the scan line in the horizontal direction and then the scan line in the vertical direction, and may operate in the reverse order. On the other hand, the elbow point detection unit 330 may perform the elbow point detection operation by dynamically setting the movement interval of the scan line according to the inspection accuracy, and basically the scan line may be set to move at intervals of 1 pixel. there is.

In an embodiment, the elbow point detection unit 330 may detect horizontal and vertical boundary points by sequentially applying scan lines in the case of a binarized image of a high beam. That is, the elbow point detection unit 330 may perform different operations for the high beam and low beam of the headlight, respectively. If the binarized image corresponds to the high beam, the elbow point detector 330 may apply vertical and horizontal scan lines to detect a boundary point at which the light distribution area corresponding to the high beam and the scan line intersect first or last, respectively. When it intersects a horizontal scan line, it may correspond to a horizontal boundary point, and when it intersects a vertical scan line, it may correspond to a vertical boundary point.

In an embodiment, the elbow point detection unit 330 defines a horizontal H-scan line and a vertical V-scan line, and moves each scan line in the vertical and horizontal directions, respectively, so that the rate of change of the brightness value is a threshold value. The elbow point may be detected by acquiring at least one candidate point exceeding . More specifically, the elbow point detection unit 330 may search the binarized image while moving the H-scan line in the vertical direction, and may detect a point at which the brightness value rapidly changes. If the rate of change of the brightness value exceeds a preset threshold value, it may correspond to a point at which the brightness value rapidly changes. The elbow point detection unit 330 may determine a point at which the brightness value rapidly changes in each of the vertical and horizontal directions, and may determine the elbow point based on the corresponding points.

For example, the intersection point between a straight line defined in the horizontal direction at a first point detected through the H-scan line and a straight line defined by a second point and a third point detected through the V-scan line is in the corresponding binarized image. It may correspond to the corresponding elbow point. On the other hand, when the first to third points are determined through the scan line search, the elbow point detection unit 330 obtains the location of the elbow point by applying the equation for calculating the elbow point based on the coordinate values of each of the points. can do.

The cut-off line determiner 340 may determine a cut-off line based on the elbow point. When the elbow point is determined by the elbow point detection unit 330, the cut-off line determining unit 340 includes a horizontal cut-off line defined horizontally in the left direction based on the position of the corresponding elbow point and the right direction based on the position of the corresponding elbow point. It is possible to determine the slope cutoff lines defined to be sloped, respectively. In this case, the inclination for determining the inclination cut-off line may be predefined by the headlight inspection apparatus 130 and may be different depending on the model of the vehicle 110 .

In an embodiment, the cutoff line determiner 340 may determine a light center based on horizontal and vertical boundary points. Here, the optical axis (light center) may correspond to the center of the high beam. That is, in the case of the binarized image of the high beam, the cutoff line determiner 340 may determine the coordinates of the optical axis by using two horizontal boundary points detected in the horizontal direction and two vertical boundary points detected in the vertical direction. For example, the cut-off line determiner 340 may define horizontal straight lines passing through two horizontal boundary points, respectively, and vertical straight lines passing through the two vertical boundary points, where each straight line intersects. Thus, the center of the rectangular region formed by the method may be determined as the optical axis.

The headlight inspection unit 350 obtains a reference headlight model based on the identification information of the vehicle 110 , and compares and analyzes the reference headlight model with an elbow point and a cut-off line to determine whether the headlight is abnormal. Here, the reference headlight model may correspond to reference information according to the model of the vehicle 110 . That is, the vehicle 110 may have different specifications of the headlight depending on the model, so the headlight inspection unit 350 may perform the headlight inspection by applying the inspection standard suitable for the vehicle 110 . In this case, the reference headlight model may be stored and stored in the database 150 , or may be provided from the cloud server 170 through a network.

In an embodiment, the headlight inspection unit 350 may determine whether the headlight is abnormal by comparing and analyzing the reference headlight model with horizontal and vertical boundary points and optical axes. That is, the headlight inspection unit 350 derives a deviation between the reference information of the reference headlight model and the actually measured values, and can finally determine whether the headlight is abnormal depending on whether the deviations are within the allowed range. there is.

In one embodiment, the headlight inspection unit 350 receives the input module for receiving identification information, the reference receiving module for receiving the reference headlight model corresponding to the identification information from the cloud server 170, the elbow point and the cut-off line to the reference head Comparison with the light model may include a comparative analysis module for obtaining each deviation information, and an inspection module for determining the presence or absence of abnormality of the headlight based on the deviation information.

More specifically, the input module may receive identification information of the vehicle 110 from the user. To this end, the input module may provide an interface for user input. The reference receiving module may query the cloud server 170 for vehicle identification information and receive a reference headlight model corresponding to the corresponding vehicle 110 as a response thereto. The comparative analysis module may compare and analyze the received information to calculate the deviation, and the inspection module may finally determine whether the headlight is abnormal based on the calculated deviation.

In an embodiment, the headlight inspection unit 350 may transmit a binarized image to the cloud server 170 to request a detailed inspection of the headlight. At this time, the cloud server 170 learns the black and white gradation image as a standard light source image to build a detailed inspection model, and provides the converted gradation image corresponding to the binarized image as an input to the detailed inspection model to determine whether there is an abnormality The output may be received as an inspection result, and the inspection result may be transmitted to the headlight inspection unit 350 in response to the request. That is, the headlight inspection apparatus 130 may perform the headlight inspection by itself, but may also request a separate inspection from the cloud server 170 and receive a response therefor as an inspection result if necessary.

More specifically, the cloud server 170 may perform a headlight inspection by utilizing the results of artificial intelligence learning of various inspection data. On the other hand, the cloud server 170 performs big data analysis and artificial intelligence learning, which are difficult to perform smoothly in the headlight inspection device 130 in that it can utilize sufficient resources, and can utilize the results of the existing head It is possible to effectively provide rapid and accurate inspection of various models that the light inspection system could not provide and the modification of existing inspection methods according to changes in inspection regulations.

In addition, the cloud server 170 may use the binarized image as it is, but may perform an operation of converting the binarized image into a gradation image based on the binarized image to improve inspection accuracy. In this case, while the binarized image includes only black and white color values, the gradation image may include 256 color values in which both ends are black and white, respectively. As a result of the headlight distribution image being expressed as a gradation image, the cloud server 170 may provide higher accuracy than the inspection performed by the headlight inspection apparatus 130 .

That is, the cloud server 170 may learn a black and white gradation image as a standard light source image to build a detailed inspection model, and then use it to conduct a headlight inspection. That is, the cloud server 170 can provide inspection accuracy above a certain level for headlights implemented in various forms in that it analyzes and utilizes big data of various inspection cases that have been previously performed, and the inspection results through the detailed inspection model can be provided quickly.

In an embodiment, when the cloud server 170 receives the binarized image of the vehicle 110 , it searches for a pre-stored standard light source image based on the identification information of the vehicle 110 , and there is no existing standard light source image. If not, a standard light source image for the corresponding vehicle 110 may be generated and stored based on the detailed inspection result, and may then be used for headlight inspection on the same vehicle 110 .

The control unit 360 controls the overall operation of the headlight inspection apparatus 130, the light distribution image acquisition unit 310, the binarized image generation unit 320, the elbow point detection unit 330, the cut-off line determiner 340 and A control flow or data flow between the headlight inspection units 350 may be managed.

4 is a flowchart illustrating a headlight inspection process based on edge computing according to an embodiment of the present invention.

Referring to FIG. 4 , the headlight inspection apparatus 130 may acquire a light distribution image of the headlight irradiated toward the screen in front of the vehicle 110 through the light distribution image acquisition unit 310 (step S410 ). The headlight inspection apparatus 130 may generate a binarized image for headlight inspection based on the light distribution image through the binarized image generator 320 (step S420).

In addition, the headlight inspection apparatus 130 may detect an elbow point by sequentially applying the scan lines respectively defined in the horizontal and vertical directions to the binarized image through the elbow point detection unit 330 (step) S430). The headlight inspection apparatus 130 may determine a cut-off line based on the elbow point through the cut-off line determiner 340 (step S440). The headlight inspection device 130 obtains a reference headlight model based on the identification information of the vehicle 110 through the headlight inspection unit 350 and compares and analyzes the reference headlight model with the elbow point and cut-off line to obtain a headlight It is possible to determine whether there is an abnormality (step S450).

5 is an exemplary view for explaining a vehicle headlight according to the position of the headlight.

Referring to FIG. 5 , the headlight inspection apparatus 130 may perform vehicle alignment according to the height of the vehicle. Here, the vehicle head may correspond to the operation of adjusting the height of the headlight inspection apparatus 130 according to the height of the vehicle, that is, the height of the headlight. Even for the vehicle 110 of the same model, the height of the actual headlight may vary depending on the use state or inspection environment, and in particular, when the height of the vehicle is increased by 30 mm (lower), it is formed on the screen for the headlight inspection. The elbow point of the light distribution image may also be increased (lower) by the same distance. If the headlight inspection apparatus 130 fails to track the actual vehicle height h, misaiming may occur as much as the vehicle height. On the other hand, the headlight inspection apparatus 130 can measure the distance to the vehicle 110 and adjust the magnification of the light distribution image according to the distance to perform a precise headlight inspection even if the vehicle 110 is not placed in a fixed position. It may be possible.

6 is a view for explaining the physical configuration of the headlight inspection apparatus according to the present invention.

Referring to FIG. 6 , the headlight inspection apparatus 130 may be disposed in front of the headlight 610 of the vehicle 110 to perform the headlight inspection. The headlight inspection device 130 may operate by being combined in a form in which the front lens 630, the reflector 650, and the camera 670 are arranged side by side. Here, the front lens 630 may be implemented as a planar convex lens, and may serve to collect the light source of the headlight 610 by being located at a point 1 m or more in front of the vehicle 110 . In addition, the front lens 630 exists in the form of a rectangle and provides a function of collecting the original light source in a reduced form, for example, it can be reduced by 1/4 or more.

The reflector 650 may be implemented as a translucent reflector, and may correspond to a screen on which an image of a light source passing through the front lens 630 is formed. The reflector 650 may be implemented with a plastic or glass material, and may include an image reference line for determining the optical axis. The camera 670 may correspond to a device for photographing a light distribution image formed on the reflector 650 , and may include a color camera of 5 MP or more.

Meanwhile, the headlight inspection apparatus 130 may be implemented to include an optical analysis system, and the optical analysis system may perform photometric analysis, optical axis analysis, refraction point analysis, and refraction angle analysis of the light source. In addition, the headlight inspection apparatus 130 may transmit measurement data to an external system by utilizing a network, and may operate in conjunction with the cloud server 170 . Also, the headlight inspection apparatus 130 may provide various UIs and MMIs for the headlight inspection using a touch screen.

7 is an exemplary view for explaining a light distribution image related to the low beam of the headlight.

Referring to FIG. 7 , the headlight inspection apparatus 130 may inspect the low beam among the headlights of the vehicle 110 . The light distribution image for the low beam of the headlight may be divided into four zones based on a horizontal line (H) and a vertical line (V). The upper left region may correspond to a left front direction based on the traveling direction of the vehicle 110 , and the upper left region may correspond to a right front shoulder direction of the vehicle 110 . The low beams of the headlights need to be dimmed to avoid dazzle towards vehicles traveling in the opposite lane to the left forward direction.

In addition, in the light distribution image, the elbow point 720 may be located in the downward direction with respect to the headlight center 710 where the horizontal line (H) and the vertical line (V) intersect, and the elbow point 720 may be located in the left direction based on the elbow point 720 . A horizontal cut-off line 741 corresponding to a horizontal line of , and an inclined cut-off line 743 corresponding to an inclined line in an upward right direction may be formed. In addition, a photometric point 730 may be located in the lower right region of the light distribution image. The position of the elbow point 720 may correspond to the downward beam irradiation direction, that is, the optical axis. Brightness such as mark confirmation may be required for the upward direction based on the slope cutoff line 743 , that is, the right front and top direction of the vehicle 110 , and the downward direction of the slope cutoff line 743 , that is, the vehicle 110 . Sufficient brightness may be required to allow people, cars, and the like to see the right front shoulder of the vehicle.

8 is an exemplary view for explaining a headlight inspection process according to the present invention.

Referring to FIG. 8 , the headlight inspection apparatus 130 may inspect whether there is an abnormality with respect to the low beam and the high beam based on the light distribution image of the headlight. In the case of FIGS. 8A and 8B , the headlight inspection apparatus 130 may acquire a binarized image 810 for a light distribution image related to a low beam, and a horizontal H-scan line 830 for the binarized image 810 . and V-scan lines 850 in the vertical direction are defined, respectively, and then, while moving each scan line in the vertical and horizontal directions, candidate points are obtained according to the rate of change of the brightness value to detect the elbow point.

More specifically, the headlight inspection apparatus 130 may determine a point at which the brightness value rapidly changes as the first candidate point 831 while moving the H-scan line 830 from the bottom to the top at 1-pixel intervals. The headlight inspection apparatus 130 may determine a first virtual straight line defined in a horizontal direction while passing through the first candidate point 831 .

In addition, the headlight inspection apparatus 130 moves the V-scan line 850 from left to right at intervals of 1 pixel, and sets the point at which the brightness value rapidly changes as second and third candidate points 851 and 852 . each can be decided. The headlight inspection apparatus 130 may determine a second virtual straight line passing through both the second and third candidate points 851 and 852 . The headlight inspection apparatus 130 may detect an intersection point of the first and second virtual straight lines, and may determine the intersection point as an elbow point in the corresponding light distribution image. On the other hand, it goes without saying that the search direction of each scan line for detecting the elbow point may be reversed if necessary.

In the case of FIGS. 8C and 8D , the headlight inspection apparatus 130 may acquire a binarized image 810 with respect to the high beam-related light distribution image, and a horizontal H-scan line 830 with respect to the binarized image 810 . and V-scan lines 850 in the vertical direction are defined, respectively, and then boundary points are obtained according to the rate of change of the brightness value while moving each scan line in the vertical and horizontal directions to detect the optical axis.

More specifically, the headlight inspection apparatus 130 moves the H-scan line 830 from the bottom to the top at 1-pixel intervals, and sets the point at which the brightness value rapidly changes as the first and second boundary points 832 and 833. can be determined individually. In addition, the headlight inspection apparatus 130 moves the V-scan line 850 from left to right at intervals of 1 pixel as the third and fourth boundary points 853 and 854, respectively. can decide The headlight inspection apparatus 130 may determine an optical axis in the corresponding light distribution image by using the first to fourth boundary points.

For example, the headlight inspection apparatus 130 may determine a virtual circle passing at least three or more of the first to fourth boundary points, and then determine the center of the virtual circle as the optical axis. As another method, the headlight inspection apparatus 130 may determine, as an optical axis, the center of a quadrangle formed by virtual straight lines that pass through each of the first to fourth boundary points and are defined in horizontal and vertical directions.

Although the above has been described with reference to preferred embodiments of the present invention, those skilled in the art can variously modify and change the present invention within the scope without departing from the spirit and scope of the present invention as set forth in the claims below. You will understand that it can be done.

100: headlight inspection system
610: headlight 630: front lens
650: reflector 670: camera
710: headlight center 720: elbow point
730: photometric point 741: horizontal cutoff line
743: bevel cutoffline
810: binarized image 830: H-scan line
831: first candidate point 832: first boundary point
833: second boundary point 850: V-scan line
851: second candidate point 852: third candidate point
853: third boundary point 854: fourth boundary point

Claims (8)

A light distribution image acquisition unit for obtaining a light distribution image of the headlight irradiated toward the screen in front of the vehicle;
a binarized image generator for generating a binarized image for headlight inspection based on the light distribution image;
an elbow point detector configured to sequentially apply scan lines defined in horizontal and vertical directions to the binarized image, respectively, to detect an elbow point;
a cut-off line determining unit for determining a cut-off line based on the elbow point;
Edge computing base including a headlight inspection unit that acquires a reference headlight model based on the identification information of the vehicle, and compares and analyzes the reference headlight model, the elbow point, and the cut-off line to determine whether the headlight is abnormal headlight inspection device.
According to claim 1, wherein the light distribution image
Edge computing-based headlight inspection device, characterized in that it includes a light distribution image for the low beam and high beam of the headlight.
3. The method of claim 2,
The elbow point detector detects horizontal and vertical boundary points by sequentially applying the scan lines in the case of the binarized image of the high beam,
The cut-off line determining unit determines a light center based on the horizontal and vertical boundary points,
The headlight inspection unit compares and analyzes the reference headlight model with the horizontal and vertical boundary points and the optical axis to determine whether the headlight is abnormal.
According to claim 1, wherein the binarized image generating unit
Edge computing-based headlight inspection apparatus, characterized in that the pre-processing operation is performed by applying at least one image filter to the binarized image.
According to claim 1, wherein the elbow point detection unit
By defining a horizontal H-scanline and a vertical V-scanline, moving each scanline in the vertical and horizontal directions, respectively, obtaining at least one candidate point whose rate of change in brightness exceeds a threshold value, Edge computing-based headlight inspection device, characterized in that it detects the elbow point.
According to claim 1, wherein the headlight inspection unit
an input module for receiving the identification information;
a reference receiving module for receiving a reference headlight model corresponding to the identification information from the cloud server;
a comparative analysis module for obtaining respective deviation information by comparing the elbow point and the cut-off line with the reference headlight model; and
Edge computing-based headlight inspection apparatus comprising an inspection module for determining whether the headlight is abnormal based on the deviation information.
The method of claim 1,
The headlight inspection unit transmits the binarized image to the cloud server to request a detailed inspection of the headlight,
The cloud server learns a black and white gradation image as a standard light source image to build a detailed inspection model, and provides a gradation image converted to correspond to the binarized image as an input to the detailed inspection model to check the output regarding the presence or absence of abnormality Edge computing-based headlight inspection apparatus, characterized in that receiving as a result, and transmitting the inspection result to the headlight inspection unit in response to the request.
obtaining a light distribution image of a headlight irradiated toward a screen in front of the vehicle;
generating a binarized image for headlight inspection based on the light distribution image;
detecting an elbow point by sequentially applying scan lines respectively defined in horizontal and vertical directions to the binarized image;
determining a cut-off line based on the elbow point; and
Edge computing-based head comprising the step of obtaining a reference headlight model based on the identification information of the vehicle and comparing and analyzing the reference headlight model, the elbow point, and the cut-off line to determine whether the headlight is abnormal Light inspection method.

Images