Vehicle and method of controlling the same

Abstract

The present invention relates to a vehicle including a camera that acquires a surrounding image of a periphery of the vehicle, and a controller configured to derive spatial recognition data by learning the surrounding image of the vehicle as an input value of the controller, derive object recognition data including wheel region data of a surrounding vehicle in the periphery of the vehicle by learning the surrounding image of the vehicle as the input value of the controller, determine a ground clearance between a vehicle body bottom surface of the surrounding vehicle and a ground surface by using the spatial recognition data and the wheel region data, and control the vehicle to park the vehicle in accordance with the ground clearance.

Description

[0001] Cross-reference to related applications

[0002] This application claims priority to Korean Patent Application No. 10-2020-0106850, filed on August 25, 2020, with the Korean Intellectual Property Office, the entire contents of which are incorporated herein by reference for all purposes. Technical Field

[0003] This invention relates to a vehicle that performs automatic parking and a control method thereof. Background Technology

[0004] Autonomous driving technology is a technology that allows a vehicle to understand road conditions and drive itself even when the driver does not control the brakes, steering wheel, or accelerator pedal.

[0005] Autonomous driving technology is a core technology for realizing intelligent vehicles. For autonomous driving, technologies can include Highway Driving Assist (HDA, a technology that automatically maintains a safe following distance), Blind Spot Detection (BSD, a technology that detects surrounding vehicles and issues warnings while reversing), Automatic Emergency Braking (AEB, a technology that activates the braking system when the vehicle cannot recognize a vehicle in front), Lane Departure Warning System (LDWS), Lane Keeping Assist System (LKAS, a technology that compensates for lane departures without a turn signal), Advanced Smart Cruise Control (ASCC, a technology that maintains a constant following distance and drives at a set speed), Traffic Jam Assist (TJA), Parking Collision Avoidance Assist (PCA), and Remote Smart Parking Assist (RSPA), among others.

[0006] In technologies that identify surrounding objects and parking spaces for automatic parking control of vehicles, ultrasonic signals are used to perform parking.

[0007] On the other hand, in recent years, research has been actively conducted on automatic parking systems that utilize cameras to perform parking.

[0008] By using the spatial recognition results of existing semantic segmentation to label vehicles in a visually perceptible form, ground clearance errors occur. Therefore, when determining the actual distance to the vehicle, the distance is sometimes set to a value greater than the actual distance.

[0009] The information disclosed in the background section of this invention is only intended to enhance the understanding of the general background of this invention, and should not be construed as an admission or implication in any way that such information constitutes prior art known to those skilled in the art. Summary of the Invention

[0010] Various aspects of the present invention aim to accurately determine the distance to surrounding vehicles by fusing spatial recognition and object recognition to determine and correct the vehicle’s ground clearance.

[0011] According to one aspect of the invention, a vehicle includes: a camera for acquiring surrounding images of the vehicle's surroundings; and a controller configured to derive spatial recognition data by learning the surrounding images of the vehicle as input values ​​to the controller, to derive object recognition data including wheel area data of surrounding vehicles by learning the surrounding images of the vehicle as input values ​​to the controller, to determine the ground clearance between the underside of the vehicle body and the ground using the spatial recognition data and the wheel area data, and to control the vehicle to park based on the ground clearance.

[0012] The controller can be configured to reflect the ground clearance of the surrounding vehicle area, excluding the wheel area.

[0013] The controller can be configured to correct the distance between the vehicle and surrounding vehicles by reflecting a determined ground clearance.

[0014] Object recognition data may include at least one of the following: surrounding vehicles, bumper area data of surrounding vehicles, and pedestrian data around the vehicles.

[0015] The controller can be configured to, when identifying the first and second wheels on the same side of surrounding vehicles, export object recognition data including first wheel area data and second wheel area data, and determine the ground clearance between surrounding vehicles and the ground by using the first wheel area data and the second wheel area data.

[0016] The controller can be configured to identify the midpoint of a line segment passing through the center of the ground in contact with the first wheel and the center of the ground in contact with the second wheel by using first wheel area data and second wheel area data, and to determine the ground clearance by using the length of a vertical line at the midpoint of the line segment from the bottom surface of the surrounding vehicle body.

[0017] The controller can be configured to, when it identifies the first and second wheels on the same side of a surrounding vehicle, as well as any of the front and rear bumpers of the surrounding vehicle, export object recognition data including first wheel area data, second wheel area data, and bumper area data, and determine ground clearance by using the bumper area data, first wheel area data, and second wheel area data.

[0018] The controller can be configured to identify a straight line passing through the center of the ground in contact with the first wheel and the center of the ground in contact with the second wheel, and to determine the ground clearance by using the length of the straight line and the vertical line between the identified bumper.

[0019] The controller can be configured to export object recognition data, including data of the first wheel region, when the controller identifies the first wheel of a surrounding vehicle, and to determine the ground clearance by using the identified first wheel region data.

[0020] The controller can be configured to convert the midpoint of the top of the first wheel and the midpoint of the bottom of the first wheel into world coordinates, respectively, and determine the ground clearance by using the converted world coordinates and the height of the camera above the ground to determine the distance between the midpoint of the top of the first wheel and the ground.

[0021] According to another aspect of the present invention, a vehicle control method includes: deriving spatial recognition data by learning an image of the vehicle's surroundings as an input value to a controller; deriving object recognition data including wheel area data of surrounding vehicles around the vehicle by learning the image of the vehicle's surroundings as an input value to the controller; determining the ground clearance between the underside of the vehicle body of the surrounding vehicles and the ground by using the spatial recognition data and the wheel area data; and controlling the vehicle to park the vehicle based on the ground clearance.

[0022] The control method may further include: reflecting the ground clearance of the remaining areas of the surrounding vehicle, excluding the wheel area.

[0023] The control method may further include: correcting the distance between the vehicle and surrounding vehicles by reflecting a determined ground clearance.

[0024] Object recognition data may include at least one of the following: surrounding vehicles, bumper area data of surrounding vehicles, and pedestrian data around the vehicles.

[0025] The control method may further include: when identifying the first wheel and the second wheel on the same side of the surrounding vehicles, deriving object recognition data including the first wheel area data and the second wheel area data; and determining the ground clearance between the surrounding vehicles and the ground by using the first wheel area data and the second wheel area data.

[0026] The control method may further include: identifying the midpoint of a line segment passing through the center of the ground in contact with the first wheel and the center of the ground in contact with the second wheel by using first wheel area data and second wheel area data; and determining the ground clearance by using the length of a vertical line at the midpoint of the line segment from the bottom surface of the surrounding vehicle body.

[0027] The control method may further include: when identifying the first and second wheels on the same side of the surrounding vehicles and any one of the front and rear bumpers of the surrounding vehicles, deriving object recognition data including first wheel area data, second wheel area data and bumper area data; and determining ground clearance by using the bumper area data, first wheel area data and second wheel area data.

[0028] The control method may further include: identifying a straight line passing through the center of the ground in contact with the first wheel and the center of the ground in contact with the second wheel; and determining the ground clearance by using the length of the straight line and the vertical line between the identified bumper.

[0029] The control method may further include: when the first wheel of a surrounding vehicle is identified, deriving object recognition data including data of the first wheel region; and determining the ground clearance by using the identified first wheel region data.

[0030] The control method may further include: converting the midpoint of the top of the first wheel and the midpoint of the bottom of the first wheel into world coordinates respectively; and determining the ground clearance by using the converted world coordinates and the height of the camera above the ground to determine the distance between the midpoint of the top of the first wheel and the ground.

[0031] The methods and apparatus of the present invention have other features and advantages, which will be apparent from or set forth in more detail from the accompanying drawings incorporated herein and the following detailed description, which together serve to explain certain principles of the invention. Attached Figure Description

[0032] Figure 1A , Figure 1B and Figure 1C This is a diagram illustrating the prior art according to the present invention.

[0033] Figure 2 This is a control block diagram of a vehicle according to various exemplary embodiments of the present invention.

[0034] Figure 3A , Figure 3B and Figure 3C This is an illustration showing images of spatial recognition and object recognition of a vehicle according to various exemplary embodiments of the present invention.

[0035] Figure 4 This is a perspective view illustrating, exemplarily, a method for determining the ground clearance of a vehicle according to various exemplary embodiments of the present invention.

[0036] Figure 5This is a perspective view illustrating, exemplarily, a method for determining the ground clearance of a vehicle according to various exemplary embodiments of the present invention.

[0037] Figure 6A and Figure 6B This is a perspective view illustrating a method for determining the ground clearance of a vehicle according to various exemplary embodiments of the present invention.

[0038] Figure 7A , Figure 7B and Figure 7C This is a diagram illustrating a method for determining the ground clearance of a vehicle according to various exemplary embodiments of the present invention.

[0039] Figure 8A and Figure 8B These are images illustrating the ground clearance of a vehicle according to various exemplary embodiments of the present invention.

[0040] Figure 9 These are images reflecting the ground clearance of a vehicle according to various exemplary embodiments of the present invention.

[0041] Figure 10 This is a flowchart of a vehicle control method according to various exemplary embodiments of the present invention.

[0042] It is understood that the accompanying drawings are not necessarily drawn to scale and are slightly simplified representations of various features illustrating the basic principles of the invention. Specific design features of the invention, including, for example, particular dimensions, orientations, positions, and shapes, as included herein, will be determined in part by the specific intended application and environment of use.

[0043] In the accompanying drawings, reference numerals refer to the same or equivalent parts of the invention in several figures throughout the drawings. Detailed Implementation

[0044] Referring now to various embodiments of the invention, examples of which are shown in the accompanying drawings and described below. While the invention will be described in conjunction with exemplary embodiments, it will be understood that this specification is not intended to limit the invention to those exemplary embodiments. On the other hand, the invention is intended to cover not only the exemplary embodiments of the invention, but also various alternatives, modifications, equivalents, and other embodiments that may be included within the spirit and scope of the invention as defined by the appended claims.

[0045] Throughout this specification, similar reference numerals refer to similar elements. Not all elements of embodiments of the invention are described, and descriptions of content well-known in the art or overlapping with each other in exemplary embodiments are omitted. Terms such as “~part,” “~module,” “~component,” and “~block” as used throughout this specification may be implemented in software or hardware, and multiple “~parts,” “~modules,” “~components,” or “~blocks” may be implemented as a single element, or a single “~part,” “~module,” “~component,” or “~block” may include multiple elements.

[0046] It will be further understood that the term "connection" and its derivatives refer to both direct and indirect connections, with indirect connections including connections via wireless communication networks.

[0047] It will be further understood that, unless the context clearly indicates otherwise, when used in this specification, the terms “comprising” and / or “including” identify the presence of the stated features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0048] In this specification, it will be understood that when a component is referred to as being "on / under" another component, the component may be directly on / under the other component, or there may be one or more intermediate components.

[0049] Terms including ordinal numbers such as "first" and "second" can be used to explain various components, but components are not limited by terms. Terms are only used to distinguish one component from another.

[0050] When used in this document, the singular forms “a,” “one,” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise.

[0051] The reference numerals used in the accompanying drawings for the purpose of explanation are used only for convenience of interpretation and do not restrict the order of the steps. Therefore, unless the context clearly specifies otherwise, the written order may be otherwise.

[0052] The operating principles and embodiments of the present invention will be described below with reference to the accompanying drawings.

[0053] Figures 1A to 1C These are illustrations of related technologies according to the present invention.

[0054] Reference Figure 1A The image shows images of surrounding vehicles labeled as is, using spatial recognition with semantic segmentation performed by a panoramic imaging system (SVM) camera in the vehicle.

[0055] Reference Figure 1BThe image shows the area actually occupied by surrounding vehicles. More specifically, it can be seen that the space between the underside of the surrounding vehicles and the ground is also an area occupied by the vehicles.

[0056] However, as Figure 1A As shown, when performing spatial recognition using semantic segmentation, not all areas actually occupied by surrounding vehicles are displayed. Therefore, the following problem exists: the distance between a vehicle and its surrounding vehicles is determined to be greater than the actual distance between the vehicle and its surrounding vehicles.

[0057] Reference Figure 1C The image shows an image of spatial recognition using semantic segmentation performed by a panoramic imaging system (SVM) camera in the vehicle, marking surrounding vehicles after error correction based on the distance between the underside of the surrounding vehicles and the ground. More specifically, when correcting and displaying the ground clearance of all surrounding vehicles, a problem arises: because the wheel areas of surrounding vehicles are in contact with the ground, the distance between the vehicle and surrounding vehicles is determined to be closer than the actual distance between the vehicle and surrounding vehicles.

[0058] Figure 2 This is a control block diagram of a vehicle according to various exemplary embodiments of the present invention.

[0059] Reference Figure 2 Vehicle 1 may include: a camera 110 for acquiring images of the vehicle's surroundings; a controller 100 for deriving spatial recognition data by learning the images of the vehicle's surroundings as input values ​​to the controller, deriving object recognition data including wheel area data of surrounding vehicles by learning the images of the vehicle's surroundings as input values ​​to the controller, determining the ground clearance between the underside of the vehicle body and the ground by using the spatial recognition data and the wheel area data, and controlling the vehicle to park by reflecting the ground clearance; and a drive unit 120.

[0060] Camera 110 can have multiple channels and can acquire images of the vehicle's surroundings.

[0061] Camera 110 can be installed in vehicle 1 to perform the operation of a panoramic imaging system (SVM).

[0062] Camera 110 may include a CCD (charge-coupled device) camera or a CMOS color image sensor. Here, both CCD and CMOS can refer to sensors that convert and store light input through the lens of camera 110 as electrical signals.

[0063] Camera 110 can acquire images of the surroundings of vehicle 1, including parking lines for parking vehicle 1.

[0064] The drive unit 120 can be configured as a device for driving the vehicle 1.

[0065] The drive unit 120 may include an engine and may include various components configured to drive the engine. More specifically, the drive unit 120 may include brakes and steering mechanisms, and there are no particular limitations on the configuration of the devices, as long as the configuration is one that enables the drive of the vehicle 1.

[0066] The controller 100 can derive spatial recognition data by learning images of the vehicle's surroundings acquired from the camera 110 as input values. For example, the controller 100 can identify the space surrounding the vehicle 1, surrounding vehicles, and obstacles by receiving 4-channel images from the SVM system of the camera 110 and performing a pre-learning algorithm.

[0067] The controller 100 can derive object recognition data by learning images of the vehicle's surroundings acquired from the camera 110 as input values. For example, the controller 100 receives 4-channel images from the camera 110's SVM system and performs a pre-learning algorithm, and can identify surrounding vehicles, the wheels and bumpers of surrounding vehicles, and pedestrians around vehicle 1 as objects.

[0068] Here, the controller 100 can derive the wheel areas of surrounding vehicles as object recognition data from the surrounding images. Furthermore, the controller 100 can derive at least one of the following as object recognition data: surrounding vehicles, the bumper areas of surrounding vehicles, and surrounding pedestrians.

[0069] The controller 100 can form a distance mapping by mapping object recognition data to spatial recognition data.

[0070] The controller 100 can determine the ground clearance between the underside of surrounding vehicles and the ground by using wheel area data and bumper area data from object recognition data.

[0071] More specifically, the controller 100 can identify the first and second wheels on the same side of surrounding vehicles via the camera 110. The controller 100 can export object recognition data including first wheel area data and second wheel area data of surrounding vehicles.

[0072] In the current situation, the controller 100 can determine the midpoint of the ground contacting the first wheel by using first wheel area data and second wheel area data. The midpoint of the line segment passing through the center of the ground contacting the first wheel and the center of the ground contacting the second wheel can be determined. Ground clearance can be determined by using the length of the vertical line drawn from the underside of surrounding vehicles at the midpoint of the line segment.

[0073] More specifically, the controller 100 can identify the first and second wheels on the same side of surrounding vehicles via the camera 110. Furthermore, the controller 100 can identify the front or rear bumpers of surrounding vehicles via the camera 110. The controller 100 can export object recognition data including first wheel area data, second wheel area data, and bumper area data.

[0074] In the current situation, the controller 100 can determine the midpoint of the ground contacting the first wheel by using first wheel area data and second wheel area data. The midpoint of the line segment passing through the center of the ground contacting the first wheel and the center of the ground contacting the second wheel can be determined. Ground clearance can be determined by using the length of the vertical line drawn from the underside of surrounding vehicles at the midpoint of the line segment.

[0075] Additionally, the controller 100 can identify a straight line passing through the center of the ground in contact with the first wheel and the center of the ground in contact with the second wheel, and the controller 100 can determine the ground clearance by using the length of the vertical line between the identified straight line and the bumper.

[0076] More specifically, the controller 100 can identify the first wheel on the same side of surrounding vehicles via the camera 110. The controller 100 can export object recognition data including data of the first wheel region.

[0077] In the current situation, the controller 100 can use the first wheel region data to convert the midpoint of the top of the first wheel and the midpoint of the bottom of the first wheel into world coordinate values, respectively. Here, the controller 100 can determine the distance between the midpoint of the top of the first wheel and the ground by using the converted world coordinate values ​​and the height of the camera 110 above the ground. Furthermore, the controller 100 can determine the ground clearance by using the distance between the midpoint of the top of the first wheel and the ground.

[0078] The controller 100 can convert the wheel area into a background area and map object recognition data to spatial recognition data.

[0079] The controller 100 can reflect the ground clearance of the remaining area of ​​the surrounding vehicles, excluding the wheel area.

[0080] The controller 100 can correct the distance between vehicle 1 and surrounding vehicles by reflecting a determined ground clearance.

[0081] The controller 100 can apply the calibrated distance to the PCA system that assists in preventing collisions when parking the vehicle 1 and the RSPA system that searches for parking spaces and automatically parks the vehicle 1.

[0082] The controller 100 can control the drive unit 120 to park the vehicle 1 according to the control of the PCA system and the RSPA system.

[0083] Figure 3A , Figure 3B and Figure 3C This is an illustration showing images of spatial recognition and object recognition of a vehicle according to various exemplary embodiments of the present invention.

[0084] Reference Figure 3A The image shows the surrounding vehicles marked as is by performing spatial recognition through the camera 110 of vehicle 1.

[0085] Reference Figure 3B Vehicle 1 can divide the spatially recognized image into cell regions based on the coordinates of camera 110. For example, vehicle 1 can divide the image into 30 cells by dividing it into cells every 6° from the front or rear, and into 60 cells by dividing it into cells every 3° from the side, but is not limited to these.

[0086] Reference Figure 3C The image shows vehicle 1 performing object recognition and classifying surrounding vehicles 2 as objects. For example, vehicle 1 can classify each of the surrounding vehicles (vehicle 0.6) and the wheels (wheel 0.7) of the surrounding vehicles (vehicle 0.6) as objects.

[0087] Vehicle 1 can convert the coordinates of surrounding vehicles (vehicle 0.6) and the wheels (wheel 0.7) of the surrounding vehicles (vehicle 0.6) in the image where object recognition has already been performed into the coordinates of the original image. More specifically, since object recognition is performed by converting the original image into a distortion-corrected image, in order to convert back to the original image after performing the above object recognition, Vehicle 1 can convert the coordinates of the object-recognized image back into the coordinates of the original image by applying the coordinates of the object-recognized image to the four points that constitute the bounding box.

[0088] As mentioned above, converting the coordinates of the object-recognized image back to the coordinates of the original image can be done by converting the object-recognized image back to the coordinates of the original image and mapping it to the spatially recognized original image.

[0089] Figure 4 This is a perspective view illustrating, exemplarily, a method for determining the ground clearance of a vehicle according to various exemplary embodiments of the present invention.

[0090] Reference Figure 4 This illustrates a method for determining ground clearance when the first wheel 210 and the second wheel 220 on the same side of a surrounding vehicle 2 are identified by the camera 110 of vehicle 1.

[0091] Vehicle 1 can export object recognition data, which includes data of a first wheel region 210-1 corresponding to a first wheel 210 of the surrounding vehicle 2 and data of a second wheel region 220-1 corresponding to a second wheel 220.

[0092] Vehicle 1 can determine the midpoint of the ground in contact with the first wheel 210 by using data from the first wheel region 210-1 and the second wheel region 220-1. A line segment 2a passing through the center of the ground in contact with the first wheel 210 and the center of the ground in contact with the second wheel 220 can be determined.

[0093] Vehicle 1 can identify the midpoint of line segment 2a. Vehicle 1 can identify the vertical line 2b of the bottom surface of the surrounding vehicle 2 at the midpoint of line segment 2a. Vehicle 1 can use the length of the vertical line 2b to determine the ground clearance between the bottom surface of the surrounding vehicle 2 and the ground.

[0094] Figure 5 This is a perspective view illustrating, exemplarily, a method for determining the ground clearance of a vehicle according to various exemplary embodiments of the present invention.

[0095] Reference Figure 5 This illustrates a method for determining ground clearance when vehicle 1 identifies the first wheel 210 and second wheel 220 on the same side of surrounding vehicles via camera 110 and identifies the front or rear bumper 230 of surrounding vehicles.

[0096] Vehicle 1 can export object recognition data, which includes data of a first wheel region 210-1 corresponding to a first wheel 210 of the surrounding vehicle 2 and data of a second wheel region 220-1 corresponding to a second wheel 220.

[0097] Vehicle 1 can determine the midpoint of the ground in contact with the first wheel 210 by using data from the first wheel region 210-1 and the second wheel region 220-1. A line segment 2a passing through the center of the ground in contact with the first wheel 210 and the center of the ground in contact with the second wheel 220 can be determined. Vehicle 1 can determine the midpoint of line segment 2a. Vehicle 1 can determine the perpendicular line 2b of the underside of the surrounding vehicle 2 at the midpoint of line segment 2a.

[0098] Vehicle 1 can identify a line segment that passes through the center of the ground in contact with the first wheel 210 and the center of the ground in contact with the second wheel 220. By extending line segment 2a in the direction of bumper 230, the extended line segment 2a and the vertical line 2c between the bumper 230 can be identified.

[0099] Vehicle 1 can determine the ground clearance between the bottom surface of the vehicle body of the surrounding vehicle 2 and the ground between the first wheel 210 and the second wheel 220, as well as the ground clearance between the bottom surface of the vehicle body of the surrounding vehicle and the ground between the wheels 210 and 220 and the bumper 230.

[0100] Figure 6A and Figure 6B This is a perspective view illustrating a method for determining the ground clearance of a vehicle according to various exemplary embodiments of the present invention.

[0101] Reference Figure 6A This illustrates a method for determining ground clearance when vehicle 1 uses camera 110 to identify only the first wheel 210 on the same side of surrounding vehicles 2.

[0102] Vehicle 1 can export object recognition data, which includes data of the first wheel region 210-1 corresponding to the first wheel 210 of the surrounding vehicle 2.

[0103] Vehicle 1 can convert the midpoint of the top of the first wheel 210 and the midpoint of the bottom of the first wheel 210 into world coordinates using data from the first wheel region 210-1. For more details, see [link to documentation]. Figure 6B Vehicle 1 can determine the distance y2 between the midpoint of the top of the first wheel 210 and the ground by using the transformed world coordinates and the distance between the camera 110 and the ground. For example, the distance y2 between the midpoint of the top of the first wheel 210 and the ground can be determined by using the proportional equation x1:y1=x2:y2.

[0104] Vehicle 1 can determine the ground clearance between the underside of the vehicle body of the surrounding vehicle 2 and the ground by using the distance between the ground and the midpoint of the top of the first wheel 210.

[0105] Figure 7A , Figure 7B and Figure 7C This is a diagram illustrating a method for determining the ground clearance of a vehicle according to various exemplary embodiments of the present invention.

[0106] Reference Figure 7A , Figure 7B and Figure 7C This paper illustrates a method for determining the ground clearance between the underside of a vehicle and the ground by using wheel area data of surrounding vehicles from spatial recognition data and object recognition data.

[0107] Reference Figure 7A , Figure 7B and Figure 7CFor each cell, vehicle 1 extracts the point closest to vehicle 1 and determines the ground clearance when the extracted point is a surrounding vehicle 2. More specifically, when the cell near vehicle 1 is an empty space or corresponds to a stop line, vehicle 1 can confirm the next cell in the direction away from vehicle 1. As described above, it can be confirmed whether a cell corresponds to a surrounding vehicle by confirming cells in the direction away from vehicle 1, starting from the area near vehicle 1.

[0108] Reference Figure 7A , Figure 7B and Figure 7C When the confirmed cell corresponds to the surrounding vehicle 2, vehicle 1 can determine the ground clearance between the bottom surface of the vehicle body and the ground by using the data of the second wheel area 220-1 corresponding to the second wheel 220 of the surrounding vehicle.

[0109] Vehicle 1 can reflect the ground clearance between the underside of the vehicle body and the ground in the remaining areas (220-3) excluding the second wheel area 220-2.

[0110] Figure 8A and Figure 8B These are images illustrating the ground clearance of a vehicle according to various exemplary embodiments of the present invention.

[0111] Figure 8A This is an image of a vehicle parked at Z=0. Figure 8B The image is based on a vehicle parked at Z=300mm. Figure 8A and Figure 8B Each cell in the text can have a length and width of 25cm.

[0112] like Figure 8A As shown, a1 can represent the point on the ground projected from a2. At this time, on the ground with Z=0, the actual location of the surrounding vehicle 2 is a1, but the distance of the surrounding vehicle 2 when it is located at a2 can be determined by the ground clearance.

[0113] By using ground clearance, vehicle 1 can determine the distance between vehicle 1 and surrounding vehicles 2 as approximately 70cm-100cm greater than the actual distance between vehicle 1 and surrounding vehicles 2.

[0114] like Figure 8B As shown, when the surrounding vehicle 2 is located at b2, vehicle 1 can determine the distance between vehicle 1 and the surrounding vehicle 2 located at b2. In the current situation, the distance between vehicle 1 and the surrounding vehicle 2 located at b2 can be similar to the actual distance between vehicle 1 and the surrounding vehicle 2 located at a1.

[0115] As described above, by reflecting the ground clearance between the underside of the vehicle body of surrounding vehicles and the ground, it can be corrected to be similar to the actual distance between vehicle 1 and surrounding vehicles.

[0116] Figure 9 These are images reflecting the ground clearance of a vehicle according to various exemplary embodiments of the present invention.

[0117] Figure 9 ① is an image showing the actual area occupied by the surrounding vehicles.

[0118] Figure 9 ② is an image showing the area occupied by surrounding vehicles after spatial recognition is performed. More specifically, referring to the post-processing output of ②, there is a problem that the distance between vehicle 1 and surrounding vehicles is determined to be farther than the actual distance because the area between the bottom surface of the surrounding vehicles and the ground is displayed as empty space.

[0119] Figure 9 Image ③ shows the area occupied by surrounding vehicles after ground clearance correction for all surrounding vehicles following spatial recognition. More specifically, referring to the post-processing output of image ③, the following problem arises because the displayed occupied area is larger than the actual occupied area of ​​the surrounding vehicles: the distance between vehicle 1 and the surrounding vehicles is determined to be closer than the actual distance.

[0120] Figure 9 Image ④ shows the area occupied by surrounding vehicles, which is corrected for ground clearance for the remaining area of ​​surrounding vehicles except for the wheels, by performing object recognition after spatial recognition to determine the ground clearance between the bottom surface of the vehicle body and the ground. More specifically, referring to the post-processing output of image ④, since the area shown resembles the actual occupied area of ​​surrounding vehicles, it can be determined that the distance between vehicle 1 and surrounding vehicles is similar to the actual distance.

[0121] Figure 10 This is a flowchart of a vehicle control method according to various exemplary embodiments of the present invention.

[0122] Vehicle 1 can receive raw images (310) of the surroundings of vehicle 1 via camera 110.

[0123] Vehicle 1 can use the original image to perform spatial recognition and object recognition (320).

[0124] In the current situation, the original image could be an image of the area surrounding vehicle 1.

[0125] More specifically, vehicle 1 can derive spatial recognition data by learning images of the vehicle's surroundings as input values.

[0126] For example, vehicle 1 can identify its surrounding space, surrounding vehicles, and obstacles by receiving 4-channel images from the SVM system of camera 110 and performing a pre-learning algorithm. Vehicle 1 can derive object recognition data by learning the surrounding images of the vehicle acquired from camera 110 as input values.

[0127] For example, vehicle 1 receives 4-channel images from the SVM system of camera 110 and performs a pre-learning algorithm, thereby identifying surrounding vehicles, the wheels and bumpers of surrounding vehicles, and pedestrians around vehicle 1 as objects. In the current situation, vehicle 1 can derive at least one of the surrounding vehicles, the wheel areas of surrounding vehicles, the bumper areas of surrounding vehicles, and surrounding pedestrians as object recognition data from the surrounding images.

[0128] Vehicle 1 can set cell regions in spatial recognition data, convert the original image into a distortion-corrected image, and classify the wheel region and bumper region in object recognition data as objects.

[0129] In the current situation, vehicle 1 can classify the wheel area and bumper area as objects and reconvert them back to the original image coordinates (330).

[0130] Vehicle 1 can use at least one of the wheel area data and bumper area data from the object recognition data to determine the ground clearance (340) between the bottom surface of the vehicle body of surrounding vehicles and the ground.

[0131] Vehicle 1 can separate the wheel region from the vehicle 1 region in the spatial recognition data (350). More specifically, vehicle 1 can convert the wheel region into the background region.

[0132] Vehicle 1 can correct the distance (360) between itself and surrounding vehicles by reflecting the ground clearance between the underside of the vehicle body of surrounding vehicles and the ground in the areas other than the wheel area.

[0133] It can output the ground clearance correction result (370).

[0134] In various exemplary embodiments of the present invention, the types of surrounding vehicles are not classified, but the types of surrounding vehicles, such as SUVs and SEDANs, can be classified to apply different ground clearances.

[0135] Various exemplary embodiments of the present invention enable accurate determination of the distance to surrounding vehicles when parking a vehicle.

[0136] Furthermore, by accurately determining the distance to surrounding vehicles, parking routes can be optimized during parking and collisions with surrounding vehicles can be prevented.

[0137] When ground clearance correction is performed for vehicle category without excluding the wheel area, it may be incorrectly judged that the distance to the vehicle is closer than it actually is. However, in vehicles according to various exemplary embodiments of the present invention, by correcting the ground clearance excluding the wheel area, incorrect vehicle judgments can be reduced.

[0138] Furthermore, the disclosed exemplary embodiments can be implemented in the form of a storage medium storing computer-executable instructions. The instructions can be stored as program code, and when run by a processor, program modules can be generated to perform the operations of the disclosed exemplary embodiments. The recording medium can be implemented as a computer-readable recording medium.

[0139] Computer-readable recording media include various recording media that store instructions that can be decoded by a computer, such as read-only memory (ROM), random access memory (RAM), magnetic tape, magnetic disk, flash memory, optical data storage devices, etc.

[0140] For ease of interpretation and precise definition in the appended claims, the terms “above,” “below,” “inside,” “outside,” “up,” “down,” “upward,” “downward,” “front,” “back,” “rear,” “inside,” “outside,” “inward,” “outward,” “inner,” “outer,” “inner,” “outer,” “forward,” and “backward” are used to describe features in reference to the locations of features in the exemplary embodiments shown in the accompanying drawings. It will be further understood that the term “connection” or its derivatives refer to both direct and indirect connections.

[0141] For the purposes of illustration and description, the foregoing description has provided specific exemplary embodiments of the invention. The foregoing description is not intended to be exhaustive or to limit the invention to the precise forms disclosed; obviously, many modifications and variations are possible in light of the above teachings. Exemplary embodiments were chosen and described to explain certain principles of the invention and its practical application, so that those skilled in the art can implement and utilize the various exemplary embodiments of the invention, as well as various alternatives and modifications thereof. The scope of the invention is intended to be defined by the appended claims and their equivalents.

Claims

1. A vehicle comprising: The camera acquires images of the surrounding area around the vehicle. as well as A controller, connected to the camera, determines spatial recognition data by learning images of the vehicle's surroundings as input values ​​to the controller. The controller determines object recognition data, including wheel area data of surrounding vehicles, by learning images of the vehicle's surroundings as input values. It then determines the ground clearance between the underside of the surrounding vehicles and the ground using the spatial recognition data and the wheel area data, and controls the vehicle to park based on the determined ground clearance. The wheel region data includes first wheel region data and second wheel region data, and The controller is configured to determine the ground clearance based on the first wheel area data and the second wheel area data when it determines that the controller has identified the first wheel and the second wheel on the same side of the surrounding vehicles.

2. The vehicle according to claim 1, wherein, The controller reflects the ground clearance of the surrounding vehicle's remaining areas, excluding the wheel area.

3. The vehicle according to claim 1, wherein, The controller corrects the distance between the vehicle and surrounding vehicles by reflecting the determined ground clearance.

4. The vehicle according to claim 1, wherein, The object recognition data further includes at least one of the surrounding vehicles, the bumper area data of the surrounding vehicles, and the pedestrian data around the vehicles.

5. The vehicle according to claim 1, wherein, The controller is configured to determine object identification data, including first wheel region data and second wheel region data, when it determines that the controller has identified a first wheel and a second wheel on the same side of the surrounding vehicle, and to determine the ground clearance between the surrounding vehicle and the ground by using the first wheel region data and the second wheel region data.

6. The vehicle according to claim 5, wherein, The controller uses the first wheel area data and the second wheel area data to identify the midpoint of a line segment passing through the center of the ground contacting the first wheel and the center of the ground contacting the second wheel, and determines the ground clearance by using the length of a vertical line from the bottom surface of the surrounding vehicles at the midpoint of the line segment.

7. The vehicle according to claim 1, wherein, The wheel region data in the object recognition data includes first wheel region data and second wheel region data, and The controller is configured to determine object identification data, including first wheel area data, second wheel area data, and bumper area data, when it determines that the controller identifies one of the first wheel and second wheel on the same side of the surrounding vehicle and one of the front and rear bumpers of the surrounding vehicle, and to determine the ground clearance by using the bumper area data, the first wheel area data, and the second wheel area data.

8. The vehicle according to claim 7, wherein, The controller is configured to identify a straight line passing through the center of the ground in contact with the first wheel and the center of the ground in contact with the second wheel, and to determine the ground clearance by using the length of the straight line and the vertical line between the identified bumper.

9. The vehicle according to claim 1, wherein, The wheel region data in the object recognition data includes first wheel region data, and The controller is configured to determine object recognition data including data of the first wheel region when it is determined that the controller has identified the first wheel of the surrounding vehicle, and to determine the ground clearance by using the identified first wheel region data.

10. The vehicle according to claim 9, wherein, The controller is configured to convert the midpoint of the top of the first wheel and the midpoint of the bottom of the first wheel into world coordinates, respectively, and to determine the ground clearance by using the converted world coordinates and the height of the camera above the ground to determine the distance between the midpoint of the top of the first wheel and the ground.

11. A method for controlling a vehicle, the method comprising: The controller determines spatial recognition data by learning images of the vehicle's surroundings, which serve as input values ​​to the controller. The controller determines object recognition data, including wheel area data of surrounding vehicles, by learning images of the vehicle's surroundings as input values ​​to the controller. The controller determines the ground clearance between the underside of the surrounding vehicles and the ground by using the spatial recognition data and the wheel area data; and The controller controls the vehicle to park based on the determined ground clearance. The wheel region data includes first wheel region data and second wheel region data, and The control method further includes: when the controller identifies a first wheel and a second wheel on the same side of a surrounding vehicle, the controller determines the ground clearance based on the first wheel area data and the second wheel area data.

12. The control method according to claim 11, further comprising: The controller reflects the ground clearance of the surrounding vehicle's remaining areas, excluding the wheel area.

13. The control method according to claim 11, further comprising: The controller corrects the distance between the vehicle and surrounding vehicles by reflecting the determined ground clearance.

14. The control method according to claim 11, wherein, The object recognition data further includes at least one of the surrounding vehicles, the bumper area data of the surrounding vehicles, and the pedestrian data around the vehicles.

15. The control method according to claim 11, wherein, Determining the ground clearance based on the first wheel area data and the second wheel area data includes: When the controller determines that it identifies the first and second wheels on the same side of the surrounding vehicles, the controller determines the object identification data including the first wheel region data and the second wheel region data; and The controller determines the ground clearance between the surrounding vehicles and the ground by using the first wheel area data and the second wheel area data.

16. The control method according to claim 15, further comprising: The controller uses the first wheel region data and the second wheel region data to determine the midpoint of a line segment passing through the center of the ground contacting the first wheel and the center of the ground contacting the second wheel; and The controller determines the ground clearance by using the length of a vertical line drawn from the underside of the surrounding vehicles at the midpoint of the line segment.

17. The control method according to claim 11, wherein, The wheel region data in the object recognition data includes first wheel region data and second wheel region data, and The control method further includes: When the controller determines that it identifies one of the first and second wheels on the same side of the surrounding vehicles, and one of the front and rear bumpers of the surrounding vehicles, the controller determines the object identification data including the first wheel area data, the second wheel area data, and the bumper area data; and The controller determines the ground clearance by using the bumper area data, the first wheel area data, and the second wheel area data.

18. The control method according to claim 17, further comprising: The controller confirms a straight line passing through the center of the ground in contact with the first wheel and the center of the ground in contact with the second wheel; as well as The controller determines the ground clearance by using the length of the straight line and the vertical line between the identified bumper.

19. The control method according to claim 11, wherein, The wheel region data in the object recognition data includes first wheel region data, and The control method further includes: When the controller determines that the first wheel of the surrounding vehicle has been identified, it determines the object identification data including data of the first wheel region; and The controller determines the ground clearance by using the identified data from the first wheel area.

20. The control method according to claim 19, further comprising: The controller converts the midpoint of the top of the first wheel and the midpoint of the bottom of the first wheel into world coordinates, respectively. as well as The controller determines the ground clearance by using the converted world coordinates and the camera's height above the ground to determine the distance between the midpoint of the top of the first wheel and the ground.

Images