Method, device, and system for providing a solution for processing real-image printing linked to painting and wrapping using AI trained on vehicle exterior and actual measurement information

The AI-driven system addresses uniformity and efficiency issues in vehicle painting and wrapping by optimizing 3D scanning, real-time control, and material management, enhancing quality and reducing rework and equipment downtime.

KR103046488B1Active Publication Date: 2026-11-11SAMGUKJI CORP CO LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
KR1020260059398
Authority / Receiving Office
KR · KR
Patent Type
Patents
Current Assignee / Owner
Filing Date
2026-04-01
Publication Date
2026-11-11
Estimated Expiration
2046-04-01

AI Technical Summary

Technical Problem

Conventional vehicle painting and wrapping processes face challenges in maintaining uniformity and consistency due to manual operations, complex surface handling, and lack of real-time control, leading to defects and inefficiencies in custom vehicle applications.

Method used

A system utilizing AI-based processing that includes a main frame with driving devices, a camera, detachable painting/wrapping units, and a control device, which uses 3D scanning and AI learning to optimize paint application and film placement, providing real-time control and simulation through augmented reality, and managing materials and equipment maintenance.

Benefits of technology

Ensures uniform paint film thickness on complex surfaces, reduces operator dependence, minimizes rework, optimizes material usage, and maximizes equipment efficiency by predicting failures and scheduling.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 112026039943785-PAT00001_ABST
    Figure 112026039943785-PAT00001_ABST
Patent Text Reader

Abstract

The present invention relates to a method, apparatus, and system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information. The system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information comprises: a main frame including an auxiliary frame that connects vertical frames, which are spaced apart from each other and stand upright from the ground so that the vehicle can move towards the center, in the longitudinal and transverse directions; a plurality of driving devices provided on the vertical frame or the auxiliary frame and moving in the longitudinal direction; a camera device provided on any one of the driving devices to capture an image of the vehicle exterior while the driving device is moving; a detachment device in which any one of a painting unit that sprays paint provided from the outside and a wrapping unit (60) to which a cut wrapping film is fixed is selectively fixed to the driving device; a control device that controls the driving device and the detachment device; and an analysis of images, actual dimensions, 3D curvature data, and surface condition data from the image obtained from the camera device, and based on the 3D curvature data and surface condition data of the vehicle exterior By including a main server that sets the painting path of the painting unit or sets film cutting dimensions corresponding to the shape of the vehicle exterior according to the generated normal vector map, it is possible to optimize the movement speed and paint spraying amount for painting by extracting 3D curvature data and a normal vector map from the vehicle exterior image in real time, thereby ensuring uniformity of the paint film thickness not only on flat parts but also on complex 3D curved surfaces such as bumper corners, wheel arches, and pillars, which significantly improves painting quality compared to manual methods. Furthermore, since it is possible to continuously learn and infer through AI using accumulated process data and automatically correct process parameters by reflecting quality inspection feedback, it is possible to provide painting and wrapping services of consistent quality that do not rely on the skill level of the operator.
Need to check novelty before this filing date? Find Prior Art

Description

Technology Field

[0001] The following embodiments relate to a method, apparatus, and system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information. Background Technology

[0003] The painting and wrapping processes of a vehicle are core processes responsible for the aesthetic appeal and protective functions of the vehicle's exterior. Conventional vehicle painting processes have relied on manual methods in which a worker manually operates a spray gun to spray paint; however, this method has resulted in problems where the film thickness and color uniformity vary significantly depending on the worker's skill level. In particular, since the vehicle's exterior consists of various three-dimensional curved surfaces such as door panels, hoods, trunks, bumpers, and fenders, it is difficult to maintain a consistent nozzle approach angle and paint spray volume, and defects such as paint bleeding or running frequently occur at the boundaries.

[0004] In the conventional vehicle wrapping process, workers typically measured the dimensions of each panel manually using a tape measure, cut the wrapping film (such as PVC or thermoformed film) based on these measurements, and then applied it to the vehicle's exterior. This method was prone to film wrinkles or air bubbles on complex curved surfaces such as bumper corners, A, B, and C pillars, and wheel arches, and resulted in reduced finishing quality due to the difficulty of precisely processing the seams between panels.

[0005] Recently, robotic arm-based automated painting devices have been developed and applied to automotive manufacturing production lines; however, these are optimized for environments where new cars of the same specifications are repeatedly produced. Consequently, they have limitations in that they are difficult to apply to custom painting and wrapping processes for private vehicles, where the vehicle model, damage condition, and customer design requirements vary. Furthermore, existing automated equipment lacks the capability to scan vehicles in 3D to generate normal vector maps and control the spray head's approach angle and paint output in real time based on these maps, making it difficult to guarantee paint film uniformity on complex curved surfaces.

[0006] Meanwhile, composite images created using 2D image editing software have been used as a means to visually confirm the results of painting and wrapping to customers in advance. However, these images often differ from the actual construction results because they fail to accurately reflect the 3D shape and lighting response of the actual vehicle, leading to frequent rework issues due to the discrepancy between customer expectations and the actual results. Although some vehicle exterior visualization services utilizing augmented reality (AR) technology are also provided, they are limited to overlaying virtual colors onto a smartphone screen and have not evolved into an integrated solution that links with actual painting and wrapping equipment to initiate an automated process immediately upon customer approval.

[0007] Furthermore, in painting and wrapping workshops, inventory management of consumable materials such as paints, wrapping films, and cleaning agents, as well as preventive maintenance of equipment parts, has been carried out manually. This has led to recurring problems, including process stoppages due to material shortages and unpredictable production losses caused by equipment failures.

[0008] Therefore, there is a continuously growing technical need for an integrated solution that generates a digital twin based on 3D scan data of the vehicle's exterior, infers an optimized paint and wrapping design through an AI deep learning model, provides a high-quality preview to the customer using physically based rendering (PBR), and then performs an automated process after obtaining customer approval. Prior art literature

[0010] 1. Korean Patent Application No. 10-2016-0072502 (June 10, 2016) 2. Korean Patent Application No. 10-2024-0037000 (March 18, 2024) The problem to be solved

[0011] The present invention provides a method, apparatus, and system for providing a painting and wrapping linked real-image output processing solution utilizing AI learned based on vehicle exterior and actual measurement information, which improves the uniformity of the paint film by precisely scanning the vehicle exterior to optimize the approach angle, movement speed, and paint spray amount of the spray head in real time for each part of the vehicle, and simulates the painting and wrapping design using physically based rendering (PBR) through an AI deep learning model learned from the vehicle exterior scan data to provide a realistic preview mode to customers through an augmented reality (AR) or mixed reality (XR) interface. means of solving the problem

[0013] The present invention relates to a method, apparatus, and system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information. The system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information comprises: a main frame including auxiliary frames that connect vertical frames, which are spaced apart from each other and stand upright from the ground to allow the vehicle to move towards the center, in the longitudinal and transverse directions; a plurality of driving devices provided on the vertical frames or auxiliary frames and moving in the longitudinal direction; a camera device provided on any one of the driving devices to capture an image of the vehicle exterior during the movement of the driving device; a detachment device in which either a painting unit that sprays paint provided from the outside or a wrapping unit to which a cut wrapping film is fixed is selectively fixed to the driving device; a control device that controls the driving device and the detachment device; and a normal vector generated based on the 3D curvature data and surface condition data of the vehicle exterior, which analyzes images, actual dimensions, 3D curvature data, and surface condition data from the images acquired from the camera device. It includes a main server that sets the painting path of the painting unit or sets film cutting dimensions corresponding to the shape of the vehicle exterior according to the map.

[0014] In addition, the main server comprises: a vehicle exterior analysis module that analyzes images received from the camera device to extract data on the vehicle's shape, 3D curvature, panel boundaries, and surface damage, and generates a surface normal vector map for each panel; a digital twin generation module that generates twin data modeled 1:1 with the vehicle based on the shape and curvature data extracted by the vehicle exterior analysis module, and corrects the twin data by receiving a database on vehicle types corresponding to the vehicle exterior analysis information through a connection with an external server; an AI learning inference module that receives input data on the vehicle type, surface damage status, and paint or wrapping design requested by the customer, infers paint patterns, color combinations, and wrapping layouts through a trained deep learning model, and continuously learns using accumulated process data; a process simulation module that simulates the results of painting and wrapping in advance on the twin data generated by the digital twin generation module before process execution, predicts paint usage, film cutting volume, and work time, and generates the simulation results as real-world images to provide to the customer; and the spray head approach angle of the painting unit according to the normal vector map generated by the vehicle exterior analysis module, It may include a painting process control module that controls the movement speed and paint spraying amount of the above-mentioned drive device in real time.

[0015] In addition, the main server may include a wrapping process control module that generates a wrapping film cutting pattern according to preset default values, a wrapping unfolding diagram generated according to the twin data of the digital twin generation model, film tension according to vehicle curvature, and an attachment path; a quality inspection module that detects defects such as uneven coating thickness, color deviation, bubbles, and boundary bleeding by re-photographing the result through the camera device after painting is completed; a material management module that calculates the estimated time of depletion by tracking the current inventory of paint, wrapping film, and cleaning agent and the actual usage amount by process in real time; a micro-boundary processing module that generates a precision boundary processing path in units of several μm at the masking boundary of the painting and wrapping process, trim lines between panels, and curvature inflection points; and an optimization module that predicts the time of failure by learning the component degradation patterns of the painting unit and the wrapping unit through analysis of accumulated process data, and corrects painting and wrapping process parameters by receiving feedback from the quality inspection module.

[0016] Additionally, the driving device comprises a body portion that encloses a portion of the outer side of the auxiliary frame, a seating portion containing a motor that extends from the body portion and engages with a gear to move the body portion along the longitudinal direction of the auxiliary frame, and a connector portion that is bent and extended from the body portion toward the lower side of the auxiliary frame opposite to the seating portion. The connector portion is supported on the inner side of the auxiliary frame by a roller provided at the end of a connecting bar extended into the interior of the auxiliary frame, and includes a cylinder that is extendable in multiple stages toward the opposite side of the auxiliary frame, and a bracket at the end of the cylinder to which the painting unit and the wrapping unit are detachably attached. The painting unit may include a head equipped with a nozzle for spraying paint, and a painting unit body to which the head is fixed, which is bent toward the opposite direction of the paint sprayed through the nozzle and fixed to the connector portion.

[0017] Additionally, the wrapping unit may include a film fixing body that secures the film by tightly fixing it with an elastic member, comprising: a fixing part having an incision formed in the longitudinal direction so that a portion of the end side of the film is inserted therein for fixing the film; a wrapping unit body connected to the fixing part and folded in the opposite direction to where the film is fixed and fixed to the connector part; an elastic member provided on the inner upper surface of the incision; a guide bar provided parallel to each other and passing through toward the incision from the outer side of the lower side of the fixing part; a flange connected from the outer side of the guide bar; an adsorption part connecting the guide bars within the incision part and on which the film is seated; and a fastening member that passes through the flange, presses one side of the lower side of the fixing part, and presses the flange. Effects of the invention

[0019] According to the method, device, and system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information of the present invention, the following effects are achieved.

[0020] First, according to the method, apparatus, and system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information according to the present invention, by extracting 3D curvature data and a normal vector map from a vehicle exterior image, the movement speed and paint spraying amount for painting can be controlled and optimized in real time, thereby ensuring uniformity of the paint film thickness not only on flat parts but also on complex 3D curved surfaces such as bumper corners, wheel arches, and pillars, and thus the painting quality is significantly improved compared to manual methods.

[0021] Secondly, according to the method, apparatus, and system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information according to the present invention, continuous AI learning through accumulated process data enables inference, and since process parameters can be automatically corrected by reflecting quality inspection feedback, it has the effect of providing painting and wrapping services of consistent quality that do not depend on the skill level of the operator.

[0022] Thirdly, according to the method, device, and system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information according to the present invention, painting and wrapping results can be simulated in advance using physically based rendering (PBR) and provided to customers in real time in a manner similar to the actual object by applying an augmented reality (AR) or mixed reality (XR) interface. Since customers can verify and approve the results prior to actual construction at a level close to real-image, this resolves the discrepancy between customer expectations and actual results, significantly reduces requests for rework, increases customer satisfaction, and improves the cost efficiency of the company.

[0023] Fourth, according to the method, apparatus, and system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information according to the present invention, the actual usage amount of paint, film, and cleaning agent is tracked in real time to calculate the expected time of depletion in advance, thereby preventing process interruption due to material shortage, and the time of equipment failure is predicted through learning component degradation patterns, thereby enabling preventive maintenance and maximizing the equipment utilization rate.

[0024] Fifth, according to the method, apparatus, and system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information according to the present invention, by efficiently dynamically distributing work for multiple vehicles, the idle time of the painting unit and the wrapping unit is minimized and the work throughput is maximized. Brief explanation of the drawing

[0026] FIG. 1 is a block diagram illustrating the overall configuration of a system providing a solution for painting and wrapping linked to a real-image output processing solution using AI learned based on vehicle exterior and actual measurement information according to one embodiment. FIG. 2 is a block diagram showing the internal module configuration of a main server according to one embodiment. Figure 3 is a diagram schematically showing the configuration of Figure 1 from the front. Figure 4 is a drawing showing Figure 3 from the left. Figure 5 is a diagram showing the operation of Figure 3. Figure 6 is a drawing showing the state in which the connecting frame is provided in Figure 3. FIGS. 7 and FIGS. 8 are drawings showing embodiments in which the system of the present invention is used in other vehicles. Figure 9 is a drawing showing the painting unit of Figure 3. FIG. 10 is a drawing showing the wrapping unit of FIG. 3. Specific details for implementing the invention

[0027] Hereinafter, embodiments are described in detail with reference to the attached drawings. However, various modifications may be made to the embodiments, and thus the scope of the patent application is not limited or restricted by these embodiments. It should be understood that all modifications, equivalents, and substitutions to the embodiments are included within the scope of the rights.

[0028] Specific structural or functional descriptions of the embodiments are disclosed for illustrative purposes only and may be modified and implemented in various forms. Accordingly, the embodiments are not limited to the specific disclosed forms, and the scope of this specification includes modifications, equivalents, or substitutions that fall within the technical concept.

[0029] Terms such as "first" or "second" may be used to describe various components, but these terms should be interpreted solely for the purpose of distinguishing one component from another. For example, the first component may be named the second component, and similarly, the second component may be named the first component.

[0030] When it is stated that a component is "connected" to another component, it should be understood that it may be directly connected to or coupled with that other component, or that there may be other components in between.

[0031] The terms used in the embodiments are for illustrative purposes only and should not be interpreted as intended to be limiting. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this specification, terms such as "comprising" or "having" are intended to indicate the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0032] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the embodiments pertain. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this application.

[0033] In addition, when describing with reference to the attached drawings, identical components are assigned the same reference numeral regardless of drawing symbols, and redundant descriptions thereof are omitted. In describing the embodiments, if it is determined that a detailed description of related prior art could unnecessarily obscure the essence of the embodiments, such detailed description is omitted.

[0034] The advantages and features of the present invention and the methods for achieving them will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below but may be implemented in various different forms. These embodiments are provided merely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention, and the present invention is defined only by the scope of the claims.

[0035] In the embodiments of the present invention, unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the present invention pertains. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in the embodiments of the present invention.

[0036] The shapes, sizes, ratios, angles, numbers, etc. disclosed in the drawings for explaining embodiments of the present invention are exemplary, and therefore the present invention is not limited to the depicted details. Furthermore, in describing the present invention, if it is determined that a detailed description of related known technology may unnecessarily obscure the essence of the present invention, such detailed description is omitted. Where terms such as "includes," "has," or "is made up" are used in this specification, other parts may be added unless "only" is used. Where a component is expressed in the singular, it includes cases where it includes the plural unless specifically stated otherwise.

[0037] In interpreting the components, they are interpreted to include a margin of error even in the absence of a separate explicit statement.

[0038] In the case of describing a positional relationship, for example, when the positional relationship between two parts is described using expressions such as 'on,' 'upper,' 'lower,' or 'next to,' one or more other parts may be located between the two parts unless 'immediately' or 'directly' is used.

[0039] When elements or layers are referred to as "on" another element or layer, this includes cases where another layer or element is placed directly on top of or in between. Throughout the specification, the same reference numerals refer to the same components.

[0040] The size and thickness of each component shown in the drawings are illustrated for convenience of explanation, and the present invention is not necessarily limited to the size and thickness of the illustrated components.

[0041] The features of each of the various embodiments of the present invention may be combined or combined with one another, either partially or wholly, and as will be fully understood by those skilled in the art, various technical interlocking and operation are possible, and each embodiment may be implemented independently of one another or together in an interlocking relationship.

[0042] Hereinafter, with reference to the attached drawings, an embodiment of a method, apparatus, and system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information according to the present invention will be described.

[0043] FIG. 1 is a block diagram illustrating the overall configuration of a system providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information according to one embodiment; FIG. 2 is a block diagram showing the internal module configuration of a main server according to one embodiment; FIG. 3 is a diagram schematically showing the configuration of FIG. 1 from the front; FIG. 4 is a diagram showing FIG. 3 from the left; FIG. 5 is a diagram showing the operation of FIG. 3; FIG. 6 is a diagram showing the state in which a connecting frame is provided in FIG. 3; FIG. 7 and FIG. 8 are diagrams showing an embodiment in which the system of the present invention is used in another vehicle; FIG. 9 is a diagram showing the painting unit of FIG. 3; and FIG. 10 is a diagram showing the wrapping unit of FIG. 3.

[0044] Referring to FIGS. 1 to 10, the method, apparatus, and system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information according to the present invention are described as follows: the system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle (V) exterior and actual measurement information includes a main frame (10), a driving device (20), a camera device (30), a detachable device (50, 60), a control device (C), and a main server (S).

[0045] The main frame (10) includes vertical frames (11, 12) arranged upright from the ground so that the vehicle (V) can move between centers spaced apart from each other, and auxiliary frames (111, 121) connecting the vertical frames (11, 12) to each other in the longitudinal and transverse directions.

[0046] The vertical frame (11, 12) includes a pair of first vertical frame sections (11) spaced apart from each other in the direction in which a vehicle enters, and a second vertical frame section (12) including a third vertical frame and a fourth vertical frame facing the first vertical frame section.

[0047] These first vertical frame section (11) and second vertical frame section (12) may include a driving motor (not shown) that moves the auxiliary frame (111, 121) vertically along the cut section (112, 122) cut along the longitudinal direction.

[0048] The auxiliary frames (111, 121) may include a first auxiliary frame (111) connecting the first vertical frame and the second vertical frame, and a second auxiliary frame (121) connecting the third vertical frame and the fourth vertical frame.

[0049] Additionally, the main frame (10) includes a connecting frame (14) that connects the first auxiliary frame (111) and the second auxiliary frame (121) to each other. For example, as shown in FIGS. 6 and 7, the connecting frame (14) forms a gantry structure that connects the upper portions of the first vertical frame section (11) and the second vertical frame section (12), and a driving device (20) is movably mounted on the connecting frame (14) so ​​that painting of the upper surface of the vehicle roof and the hood area is possible, and various vehicles are possible.

[0050] Each of these first auxiliary frames (111) is provided with a gear (1111) along the longitudinal direction on the upper side, and the driving device (20) can move along the longitudinal direction of the first auxiliary frame (111) by engaging with the gear (1111). Similarly, the second auxiliary frame (121) can also be provided with a gear like the first auxiliary frame (111) so that the driving device (20) engages with it.

[0051] The driving device (20) may be provided on a vertical frame (11, 12) or an auxiliary frame (111, 121) and may be provided in multiple units that move in the longitudinal direction.

[0052] The driving device (20) includes a body portion (211) that encloses a portion of the outer side of the auxiliary frame (111, 121), a seating portion (212) that extends from the body portion (211) and has a motor embedded therein that engages with a gear to move the body portion (211) along the longitudinal direction of the auxiliary frame (111, 121), and a connector portion (213) that is bent and extended from the body portion (211) toward the lower side of the auxiliary frame (111, 121) in the opposite direction of the seating portion (212).

[0053] The connector part (213) is supported on the inner side of the auxiliary frame (111, 121) by a roller provided at the end side of a connecting bar (2131) that extends into the interior of the auxiliary frame (111, 121), and includes a cylinder (214) that can be extended in multiple stages to the opposite side of the auxiliary frame (111, 121) and a bracket (215) on the end side of the cylinder (214) to which a painting unit (50) and a wrapping unit (60) are detachably attached.

[0054] As shown in FIGS. 9 and 10, the bracket (215) may include an insertion groove (2151) into which a painting unit (50) or a wrapping unit (60) is inserted.

[0055] The cylinder (214) can precisely adjust the distance between the vehicle exterior and the painting unit (50) or wrapping unit (60) by means of a multi-stage extendable structure, thereby maintaining a constant spray distance and film adhesion even in parts of the vehicle with different curvatures.

[0056] A camera device (30) is provided in one of the driving devices (20) and captures an image of the exterior of the vehicle while the driving device (20) is moving.

[0057] The camera device (30) may include a depth sensor using a structured light method or a Time-of-Flight (ToF) method along with an RGB camera, and the captured image data is transmitted to the main server (S) in real time.

[0058] The camera device (30) can obtain a multi-view image covering the entire exterior of the vehicle by continuously shooting the front, side, rear, and roof of the vehicle while the driving device (20) moves along the auxiliary frame (111, 121).

[0059] The detachable device (50, 60) is selectively fixed to the driving device (20) either the painting unit (50) or the wrapping unit (60).

[0060] The painting unit (50) sprays paint supplied from the outside. The painting unit (50) can be fixed by a painting unit body (52) to which a head (51) equipped with a nozzle for spraying paint is fixed, which is bent in the opposite direction of the paint sprayed through the nozzle and inserted into an insertion groove (2151) by a press fit.

[0061] The head (51) may include a nozzle drive motor (not shown) that rotates the nozzle inside so that the nozzle protrudes outward in a spherical shape and can rotate vertically around the ground. When the nozzle drive motor is driven according to a control signal from the painting process control module (S5), the head (51) can change its orientation in real time so that it becomes perpendicular to the normal vector direction of the vehicle surface.

[0062] For example, in the flat door panel area, the head (51) is directed horizontally toward the outer surface of the vehicle, and in the area of ​​sharp curvature at the bottom of the bumper, the head (51) rotates within a maximum range of ±90° by a control signal from the main server (S) so that paint can always be sprayed in the direction of the surface normal. The nozzle can be controlled by the main server (S) by receiving paint from the outside.

[0063] The wrapping unit (60) has a cut wrapping film (F) fixed thereto. The wrapping unit (60) can be fixed by the fixing part (61, 63) to which the film (F) is fixed and the wrapping unit body (62) connected to the fixing part (61, 63) being folded in the opposite direction to where the film (F) is fixed and inserted into the insertion groove (2151) by a press fit.

[0064] The wrapping unit (60) includes a fixed part (61), a wrapping unit body (62), a film fixing body (63), and an elastic member (64).

[0065] The fixing part (61) may be formed with an incision (611) that is cut in the longitudinal direction so that a portion of the end side of the film is inserted to fix the film.

[0066] The wrapping unit body (62) is connected to the fixing part (61) and can be fixed to the connector by being folded in the opposite direction to where the film is fixed.

[0067] The elastic member (64) may be provided on the inner upper surface of the cut portion.

[0068] The film fixing member (63) can secure the film by including a guide bar (632) that passes through the lower side of the fixing member (61) in a parallel manner toward the cut portion (611), a flange (631) that connects the guide bar (632) on the outer side, an adsorption member (633) that connects the guide bar (632) within the cut portion (611) and on which the film is seated, and a fastening member (634) that passes through the flange (631) and presses the lower side of the fixing member (61) and presses the flange (631).

[0069] Accordingly, a film (F) that has been pre-cut by an unillustrated film cutter according to a wrapping cutting pattern calculated by an image output module (S14) can be precisely attached to the exterior of a vehicle by the movement of the driving device (20) and the extension / extension movement of the cylinder (214) while fixed by the film fixing body (63).

[0070] The control device (C) supplies power to each unit and device and provides signals provided by the main server (S) to each unit and device. The control device (C) transmits process commands (painting path, movement speed, spray amount, head angle, film attachment path, etc.) transmitted from the main server (S) to the driving device (20) and each detaching device (50, 60) in real time, and feeds back the operating status and sensor data of each device to the main server (S).

[0071] The main server (S) analyzes the image, actual dimensions, three-dimensional curvature data, and surface condition data from the image obtained from the camera device, and sets the painting path of the painting unit or the film cutting dimensions corresponding to the shape of the vehicle exterior according to the normal vector map generated based on the three-dimensional curvature data and surface condition data of the vehicle exterior.

[0072] The main server (S) includes a vehicle exterior analysis module (S1), a digital twin generation module (S2), an AI learning inference module (S3), a process simulation module (S4), a painting process control module (S5), a wrapping process control module (S6), a quality inspection module (S7), a material management module (S8), a fine boundary processing module (S9), an optimization module (S10), a multi-vehicle scheduling module (S11), a rendering module (S12), a coating module (S13), and an image output module (S14).

[0073] The vehicle exterior analysis module (S1) analyzes the image received from the camera device to extract data on the vehicle's shape, 3D curvature, panel boundaries, and surface damage, and generates a surface normal vector map for each panel.

[0074] For example, the vehicle exterior analysis module (S1) generates a 3D point cloud by applying a Structure from Motion (SfM) algorithm or a stereo vision algorithm to a multi-view image obtained by a camera device (30) continuously capturing the vehicle exterior along the movement path of a driving device (20). After constructing a 3D mesh model of the vehicle exterior using the Poisson Surface Reconstruction method from the generated point cloud, the boundaries of each panel, such as the door panel, hood, trunk, bumper, fender, and roof, are identified and separated by applying a RANSAC (Random Sample Consensus)-based planar segmentation algorithm and an edge detection algorithm. For each panel, a normal vector is calculated through PCA (Principal Component Analysis) analysis of the covariance matrix between each point of the point cloud and adjacent points, and a surface normal vector map is generated by aggregating this by panel. For example, in the case of a vehicle door panel, the normal vector maintains a constant direction in the central part that is close to a plane, whereas the character line part on the outer edge of the door can identify a curvature inflection point due to a sudden change in the direction of the normal vector.

[0075] In addition, surface damage data extraction can be performed by applying a CNN (Convolutional Neural Network)-based defect detection model (e.g., a binary classifier or semantic segmentation model of the ResNet family) to the acquired RGB image to extract defect types and location coordinates such as scratches (width 50 μm or more), dents (displacement 2 mm or more), paint peeling, and oxidation, and the extracted defect information can be transmitted to a digital twin generation module (S2) and an AI learning inference module (S3) to be reflected in the painting and wrapping design.

[0076] The digital twin generation module (S2) generates twin data modeled 1:1 with the vehicle based on shape and curvature data extracted from the vehicle exterior analysis module, and corrects the twin data by receiving a database of vehicle types corresponding to the vehicle exterior analysis information through a connection with an external server.

[0077] The digital twin generation module (S2) constructs twin data with a 1:1 scale with the actual vehicle based on the point cloud and mesh model generated by the vehicle exterior analysis module (S1). At this time, vehicle type recognition identifies the manufacturer, vehicle type, and generation from the captured image using a deep learning-based vehicle type classification model (e.g., EfficientNet series classifier), obtains standard panel dimension and shape data of the vehicle type by linking with an external server (a vehicle manufacturer's CAD database or a public standard vehicle type dimension database), and generates precise twin data that minimizes measurement error while reflecting actual damaged areas such as scratches and dents by comparing the obtained standard data with the scan data and correcting the scan error to within ±0.5mm using an ICP (Iterative Closest Point) alignment algorithm. For example, if the scanned vehicle is identified as a specific sedan model, the twin model is corrected by receiving data such as the standard door panel width (e.g., 1,030mm) and roof curvature radius (e.g., R=2,500mm) of the vehicle type.

[0078] The AI ​​learning inference module (S3) receives data on vehicle type, surface damage condition, and paint or wrapping design requested by the customer, infers paint patterns, color combinations, and wrapping layouts through a learned deep learning model, and can continuously learn using accumulated process data.

[0079] The AI ​​learning inference module (S3) uses a generative model based on a Generative Adversarial Network (GAN) or a Diffusion Model to infer and output a paint pattern, a color combination based on RAL code, and a wrapping layout (arrangement of film color and texture per panel) using the vehicle type, surface damage condition, and customer color and design requests (e.g., solid color, two-tone, metallic pearl, matte black, carbon look wrapping, front wrapping, or partial wrapping, etc.) as input conditions. For example, if a customer requests 'white matte front wrapping,' the AI ​​learning inference module (S3) infers a film splitting method and tension conditions suitable for each panel area and curvature of the vehicle twin data and transmits them to the process simulation module (S4) and the wrapping process control module (S6). For continuous learning using accumulated process data (input parameters of previous work, final quality inspection score, customer feedback), the Elastic Weight Consolidation (EWC) technique is applied so that new process data can be continuously reflected while preserving previously learned knowledge.

[0080] The process simulation module (S4) simulates the results of painting and wrapping in advance on the digital twin module before the process is performed, predicts the amount of paint used, the amount of film to be cut, and the time required for the work, and generates the simulation results as a real image and provides it to the customer.

[0081] The process simulation module (S4) can render using the PBR (Physically Based Rendering) method by applying the painting and wrapping design inferred by the AI ​​learning inference module (S3) onto the twin data generated by the digital twin generation module (S2). Here, PBR rendering generates a painting and wrapping result image that is close to actual conditions by combining physical property parameters such as paint reflectance (Roughness), metallicity (Metalness), and transparency (Alpha) with a lighting environment (HDR image-based environmental lighting). The paint usage is calculated by multiplying the target film thickness (e.g., base paint 30~60μm, urethane clear coat 50~80μm) by the area of ​​each panel calculated from the normal vector map of the vehicle exterior analysis module (S1), and applying the paint solid content ratio and spray efficiency (e.g., HVLP method 70~80%) to calculate the total paint consumption (e.g., 200ml per door panel). The amount of wrapping film to be cut is calculated by adding a construction allowance (e.g., 3 cm added to the outer edge of the panel) to the film area for each panel calculated by the unfolding algorithm of the wrapping process control module (S6), and the time required for the work can be predicted by adding the unit replacement time and drying waiting time to the value obtained by dividing the total length of each painting path by the movement speed of the driving device.

[0082] The painting process control module (S5) controls the spray head approach angle of the painting unit, the movement speed of the driving device, and the amount of paint sprayed in real time according to the normal vector map generated by the vehicle exterior analysis module (S1).

[0083] The painting process control module (S5) reads the surface normal vector direction at each point of the normal vector map and transmits a rotation command to the nozzle drive motor of the head (51), thereby controlling the nozzle to always spray paint in a direction perpendicular to the vehicle surface (normal direction). The movement speed of the drive device (20) is differentially controlled according to the curvature of the vehicle surface. For example, the movement speed of the drive device is set to 300 mm / s in flat parts such as door panels, and the movement speed is reduced to 80-100 mm / s in the lower corner of the bumper (a sharp curved part with a radius of curvature R=50 mm) to maintain a uniform paint film thickness. The amount of paint sprayed is adjusted in real time in inverse proportion to the movement speed, and the amount of sprayed is reduced in low-speed sections to prevent overloading of paint. Additionally, when the camera device (30) provides real-time images during the painting operation, the painting process control module (S5) monitors the paint layering state through optical flow analysis and can immediately correct the movement speed and spray amount if a deviation from the target value of the paint film thickness occurs.

[0084] The wrapping process control module (S6) generates a wrapping film cutting pattern according to preset default values ​​for the wrapping unfolding diagram generated according to the twin data of the digital twin generation model, film tension according to the vehicle curvature, and attachment path.

[0085] The wrapping process control module (S6) generates an unwrapped layout by unfolding each panel of the twin data into a two-dimensional plane using a UV unwrapping algorithm. In this process, double-curved areas with large Gaussian curvature (e.g., bumper corners, door character line joints) are divided into multiple pieces so that the film cutting pattern can be evenly distributed within the film stretching tolerance (typically within 3% for PVC cast film and within 5% for polyurethane PPF film). For example, the double-curved area at the bottom of the bumper is divided and cut into 5 to 7 pieces so that each piece is attached within the allowable tension. The film attachment path is set in the order from flat parts to curved parts to minimize the occurrence of wrinkles and bubbles during operation.

[0086] The quality inspection module (S7) re-photographs the result through the camera device after the painting is completed to detect defects such as uneven film thickness, color deviation, bubbles, and boundary bleeding.

[0087] The quality inspection module (S7) compares and analyzes the image obtained by re-scanning the vehicle exterior with the camera device (30) after the completion of the painting and wrapping process with the scanned image before the process. For the detection of non-uniformity in film thickness, the film thickness is estimated using laser reflection intensity analysis or a structured light interferometer, and areas exceeding ±10μm relative to the target film thickness (e.g., 60μm) are determined as defects. For color deviation, the captured image is converted to the CIE Lab color space to calculate the color difference from the target color, and color deviation defects can be classified. Additionally, bubbles are detected by detecting local expansion patterns in the RGB image, and boundary blurring is determined as a defect when the penetration width of an adjacent color at the panel boundary area exceeds 0.5mm.

[0088] The material management module (S8) tracks the current inventory of paint, wrapping film, and cleaning agent and the actual usage amount by process in real time to calculate the estimated time of depletion.

[0089] The material management module (S8) measures the amount of paint sprayed during the process in real time from a flow meter built into the painting unit (50) and calculates the length of film used from an encoder provided in the film supply unit of the wrapping unit (60), thereby accumulating and recording actual usage data for each process. The estimated time of depletion is calculated by dividing the current inventory amount by the daily average consumption amount over the past 7 days (e.g., paint inventory 10kg ÷ daily average consumption 1.5kg / day = estimated depletion after 6.7 days), and an order notification is sent to the manager on the estimated date of depletion. In addition, by comparing the material usage amount of the next process predicted by the process simulation module (S4) with the current inventory amount, if the inventory is insufficient, the corresponding process schedule can be adjusted in conjunction with the multi-vehicle scheduling module (S11).

[0090] The micro-boundary processing module (S9) generates a precision boundary processing path in the order of several μm at the masking boundary of the painting and wrapping process, trim lines between panels, and curvature inflection points.

[0091] The micro-boundary processing module (S9) detects a normal vector discontinuity point (panel boundary trim line) between adjacent panels in the normal vector map of the vehicle exterior analysis module (S1), and interpolates the boundary using a cubic B-spline curve to generate a boundary processing path with precision in units of several μm (e.g., ±5 μm). At the masking boundary (e.g., the boundary between the window and the door panel), the nozzle of the painting unit (50) approaches within exactly 0.5 mm of the boundary line to set a path so that the paint does not encroach on the masking area. At the curvature inflection point (the point where the second derivative becomes zero), the paint spraying temporary suspension and resumption point is automatically set to prevent the phenomenon where the film thickness before and after the inflection point becomes uneven. In addition, during the wrapping process, the movement path of the wrapping unit (60) is precisely controlled so that the film seam is aligned with an error of within ±0.5 mm at the trim line between panels.

[0092] The optimization module (S10) learns the component degradation patterns of the painting unit and wrapping unit through the analysis of accumulated process data to predict the time of failure, and corrects the painting and wrapping process parameters by receiving feedback from the quality inspection module.

[0093] The component degradation prediction of the optimization module (S10) can calculate the expected time of failure (e.g., remaining nozzle life of 72 hours) by inputting time-series data, such as the cumulative number of nozzle sprays of the painting unit (50), operating temperature, pressure history, and the number of times the flange (631) of the film fixing part operates, into a Long Short-Term Memory (LSTM)-based life prediction model, such as nozzle clogging, suction pad wear, and drive motor bearing degradation. The calculated prediction result is linked with the multi-vehicle scheduling module (S11) to automatically rearrange the work schedule of the corresponding unit before the time of preventive maintenance. Process parameter correction can be performed by a reinforcement learning agent that uses the inspection results of the quality inspection module (S7) as a compensation signal to correct parameters such as the movement speed, spray volume, and head approach angle of the next identical or similar process.

[0094] In addition, the main server (S) generates a 3D digital twin model of the vehicle based on vehicle information obtained from a camera device, generates image output data by simulating the painting and wrapping processes on the generated 3D digital twin model using physically based rendering (PBR), provides this data to the customer through an augmented reality (AR) or mixed reality (XR) interface, and can start the actual process using the painting unit and the wrapping unit only when the customer's approval signal is received.

[0095] The multi-vehicle scheduling module (S11) simultaneously schedules painting and wrapping operations for multiple vehicles and performs dynamic work distribution by identifying the real-time availability status and movement paths of the painting unit and the wrapping unit.

[0096] The multi-vehicle scheduling module (S11) optimizes the order of painting and wrapping operations and unit assignments for multiple vehicles (e.g., 5 vehicles that can be processed simultaneously) through a Critical Path Method (CPM) based schedule optimization algorithm or a Genetic Algorithm. For example, it can determine in real-time the predicted work time for each vehicle (calculated by the process simulation module (S4)), the current location of each drive unit (20), the type of mounted unit (painting unit / wrapping unit), and availability, and calculate the shortest travel path that minimizes the idle travel distance of each drive unit through the Traveling Salesman Problem (TSP) algorithm. For example, when the painting process of vehicle A is completed and the drive unit (21) becomes idle, the vehicle among the other waiting vehicles B to E that has the minimum travel distance from the current location of the drive unit (21) can be assigned as the next work target.

[0097] The rendering module (S12) transmits to the customer in real time a result overlaid to be painted and wrapped on an actual vehicle through an augmented reality (AR) or mixed reality (XR) interface based on data simulated in the process simulation module (S4).

[0098] The rendering module (S12) receives actual vehicle images captured through the customer's smartphone camera or HMD (Head Mounted Display) device, synchronizes the PBR rendering result image generated by the process simulation module (S4) with the digital twin coordinate system of the vehicle to register it on the actual vehicle images, and then transmits the AR overlay image to the customer terminal in real time (at least 30fps). At this time, in XR mode, the customer can wear the HMD and move around the vehicle 360° to experience the appearance of the vehicle that has been virtually painted and wrapped.

[0099] The coating module (S13) selects a special coating material according to customer requirements or vehicle applications, and sets paint unit control parameters optimized for the physical properties (viscosity, curing temperature, spray pressure) of the material.

[0100] The coating module (S13) searches for a coating material selected by the customer (e.g., inorganic ceramic coating, organic nanoceramic, PPF (Paint Protection Film), photochromic paint, military camouflage paint, solvent ink paint for vehicle advertising wrapping, etc.) in the material property database and automatically sets the optimal spraying conditions for the corresponding material in the painting unit (50). For example, if an inorganic ceramic coating material is selected, control parameters are automatically set to a viscosity of 40 to 60 cP, a curing temperature of 60°C, and a spraying pressure of 2.0 to 2.5 bar, and if a high-gloss urethane clear coat is selected, a viscosity of 15 to 25 cP, a spraying pressure of 1.5 to 2.0 bar, and a command to enter a drying chamber at 60°C after spraying can be automatically generated. In addition, special functional materials (e.g., UV reflective coating, water-repellent / oil-repellent coating, electromagnetic shielding coating) may be selected depending on the vehicle application, and parameters optimized for the physical properties of each material are transmitted to the painting process control module (S5) by the coating module (S13).

[0101] The image output module (S14) converts the image output data generated as a result of the process simulation into a high-resolution realistic rendering image, a 2D layout unfolded diagram, and a 3D interactive viewer format and outputs it, and automatically generates a print output file based on the image output data and transmits it to the painting unit (50) and the wrapping unit (60), thereby enabling automated processing from the image output data into actual process commands.

[0102] According to the method, apparatus, and system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information according to the present invention, by extracting 3D curvature data and normal vector maps from vehicle exterior images, the movement speed and paint spraying amount for painting can be controlled and optimized in real time, thereby ensuring uniformity of paint film thickness not only on flat surfaces but also on complex 3D curved surfaces such as bumper corners, wheel arches, and pillars, resulting in significantly improved painting quality compared to manual methods. Furthermore, by enabling continuous AI learning and inference through accumulated process data, process parameters can be automatically corrected by reflecting quality inspection feedback, thereby providing painting and wrapping services of consistent quality that do not rely on the skill level of the operator.

[0103] In addition, according to the method, apparatus, and system for providing a solution for processing real-image output linked to painting and wrapping using AI learned based on vehicle exterior and actual measurement information according to the present invention, painting and wrapping results can be simulated in advance using physically based rendering (PBR) and provided to customers in real time in a manner similar to the actual product by applying an augmented reality (AR) or mixed reality (XR) interface. This allows customers to verify and approve the results before actual construction at a level close to real-image, thereby resolving the discrepancy between customer expectations and actual results and significantly reducing requests for rework. Furthermore, it can increase customer satisfaction and improve the cost efficiency of the company. Additionally, by tracking the actual usage of paint, film, and cleaning agent in real time and calculating the expected time of depletion in advance, it prevents process interruptions due to material shortages. It also maximizes equipment utilization by enabling preventive maintenance through the prediction of equipment failure times by learning component degradation patterns. Moreover, by efficiently dynamically distributing work for multiple vehicles, it minimizes idle time for painting units and wrapping units and maximizes work throughput.

[0104] Although embodiments of the present invention have been described in more detail with reference to the attached drawings, the present invention is not necessarily limited to these embodiments and may be modified in various ways within the scope of the technical spirit of the present invention. Accordingly, the embodiments disclosed in the present invention are intended to explain, not limit, the technical spirit of the present invention, and the scope of the technical spirit of the present invention is not limited by these embodiments. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. The scope of protection of the present invention shall be interpreted by the claims below, and all technical spirits within an equivalent scope shall be interpreted as being included within the scope of rights of the present invention.

[0105] Therefore, other implementations, other embodiments, and equivalents to the claims also fall within the scope of the claims set forth below. Explanation of the symbols

[0107] 10: Mainframe 20: Driving device 30: Camera device C: Control unit S: Main Server F: Film

Claims

Claim 1 A main frame comprising an auxiliary frame connecting vertical frames that are spaced apart from each other and stand upright from the ground to allow the vehicle to move towards the center in longitudinal and transverse directions; a plurality of driving devices provided on the vertical frames or auxiliary frames and moving in the longitudinal direction; a camera device provided on any one of the driving devices to capture images of the vehicle's exterior while the driving device is moving; a detachment device in which either a painting unit that sprays paint provided from the outside or a wrapping unit to which a cut wrapping film is fixed is selectively fixed to the driving device; a control device that controls the driving device and the detachment device; and a main server that analyzes images, actual dimensions, 3D curvature data, and surface condition data from images acquired from the camera device, and sets a painting path for the painting unit or sets film cutting dimensions corresponding to the shape of the vehicle's exterior according to a normal vector map generated based on the 3D curvature data and surface condition data of the vehicle's exterior.The system includes, wherein the main server comprises: a vehicle exterior analysis module that analyzes images received from the camera device to extract data on the vehicle's shape, 3D curvature, panel boundaries, and surface damage, and generates a surface normal vector map for each panel; a digital twin generation module that generates twin data modeled 1:1 with the vehicle based on the shape and curvature data extracted by the vehicle exterior analysis module, and corrects the twin data by receiving a database for vehicle types corresponding to the vehicle exterior analysis information through a connection with an external server; an AI learning inference module that receives input data on the vehicle type, surface damage status, and paint or wrapping design requested by the customer, infers paint patterns, color combinations, and wrapping layouts through a trained deep learning model, and continuously learns using accumulated process data; a process simulation module that simulates the results of painting and wrapping in advance on the twin data generated by the digital twin generation module before process execution, predicts paint usage, film cutting volume, and work time, and generates the simulation results as real-world images to provide to the customer; and the spray head of the painting unit according to the normal vector map generated by the vehicle exterior analysis module A system for providing a painting and wrapping linked real-image output processing solution utilizing AI learned based on vehicle exterior and actual measurement information, characterized by comprising: a painting process control module that controls the approach angle, the movement speed of the driving device, and the amount of paint sprayed in real time; a wrapping process control module that generates a wrapping film cutting pattern according to preset default values ​​for the wrapping unfolding diagram generated according to the twin data of the digital twin generation module, the film tension according to the vehicle curvature, and the attachment path; and a quality inspection module that detects defects such as uneven coating thickness, color deviation, bubbles, and boundary blurring by re-photographing the result through the camera device after painting is completed. Claim 2 delete Claim 3 In claim 1, the main server comprises: a material management module that tracks the current inventory of paint, wrapping film, and cleaning agent and the actual usage amount by process in real time to calculate the expected time of depletion; a micro-boundary processing module that generates a precision boundary processing path in units of several μm at the masking boundary of the painting and wrapping process, trim lines between panels, and curvature inflection points; and an optimization module that learns the component degradation patterns of the painting unit and the wrapping unit through analysis of accumulated process data to predict the time of failure occurrence and receives feedback from the quality inspection module to correct the painting and wrapping process parameters; characterized in that the system provides a painting and wrapping linked real-image output processing solution utilizing AI learned based on vehicle exterior and actual measurement information.

Citation Information

Patent Citations

  • Evaluation system for painting simulation

    JP2020086373A

  • Systems, methods, and interfaces for defect tracking

    KR1020240065124A

  • Electronic device for providing painting information, and operation method of the same

    KR1020250098647A

  • Electronic device for providing painting information, and operation method of the same

    KR1020240095958A

  • A system for controlling the painting of a road surface performed by a painting device based on augmented reality

    KR102719837B1