Pre-owned Product Management System

Abstract

The present invention is characterized in that a used product management system comprises: a display unit for displaying a screen of a used user terminal purchase and shipment management program; and a computing device provided with a user input unit for user input.

Description

Technology Field

[0001] The present invention relates to a used product management system that manages the processes of purchasing, evaluating, and reselling used user terminals. More specifically, it relates to a used product management system that improves the efficiency and reliability of used user terminal transactions through a system that integrates the evaluation of the condition, pricing, and sale of used user terminals. Background Technology

[0002] Generally, the sale of used smartphones takes the form of individual sellers finding buyers directly through online marketplaces or social media, or selling to stores specializing in buying used goods.

[0003] However, this form of sales involves direct transactions between the seller and the buyer, which can lead to reliability issues. Furthermore, if the seller fails to accurately describe the condition of the product or intentionally conceals defects, the buyer may find it difficult to detect them and suffer damages.

[0004] Selling through second-hand stores generally results in sellers failing to receive fair value because purchase prices are set lower than market rates. Additionally, there is a risk that lost or stolen devices may be circulated during the transaction process, and the lack of a systematic verification system leaves room for potential problems.

[0005] Due to these issues, there is a growing need for a more efficient and reliable buying and selling system in the used smartphone trading market. The problem to be solved

[0006] The present invention aims to solve the aforementioned problems by including a function that automatically receives purchase requests through user input and can rapidly process product information, thereby significantly improving the efficiency of the purchasing process.

[0007] By verifying the legality of the device from the purchase stage through the loss / theft lookup function, the distribution of illegal devices can be effectively blocked; furthermore, by automatically processing and displaying physical inspection information and results, the condition of the purchased device can be accurately evaluated, and an appropriate price can be calculated based on this.

[0008] The goal is to provide a used product management system that contributes to increasing the speed and efficiency of distribution in the used terminal market by rapidly switching used terminals from a standby status to be sold and effectively resupplying them to the market. means of solving the problem

[0009] A used product management system according to one embodiment of the present invention is characterized by comprising: a display unit for displaying a screen of a used user terminal purchase and shipment management program; and a computing device provided with a user input unit for user input.

[0010] The above-mentioned computing device includes a server connected to a wired or wireless network, wherein the server includes a storage unit; a communication unit; and a control unit; wherein the control unit, upon receiving a purchase request for a first user terminal subject to purchase from the user input unit, requests first product information including a customer's expected price and first image information for the first user terminal; wherein the control unit, upon receiving the first product information from the user input unit, identifies the first image information to identify first defect information regarding the display screen and appearance of the first user terminal, generates inspection peculiarity information, and displays it on the display unit; wherein the control unit, upon receiving first physical inspection information and a first physical inspection price including an inspection result of at least one of vibration, speaker, face recognition, fingerprint sensor, communication connection, camera, touch display sensitivity, and button for the first user terminal from the user input unit, generates a first inspection result list for the first user terminal and displays it on the display unit.

[0011] The control unit switches the first user terminal, for which the first inspection result list has been generated, to a sales waiting state and registers the information as a used user terminal available for sale in the program; when the control unit receives user input regarding the manufacturer-based first terminal source information stored in the first user terminal from the user input unit, it inputs the first terminal source information through a pre-configured loss / theft solution lookup API for blocking the distribution of illegal terminals and performs a loss / theft lookup for the first user terminal; when the control unit identifies the first user terminal as being in a normal release state as a result of the loss / theft lookup, it displays a popup window on the display unit for deleting existing data and connecting a function diagnostic program for the first user terminal; when the control unit receives user input for a purchase request for the first user terminal from the user input unit, it identifies the first physical sales price, which is the amount of a pre-configured ratio added to the first physical inspection price, and displays the first physical sales price on the display unit; and the control unit The external state of the first user terminal is 3D modeled to provide visualization of the external state and functions of the first user terminal in a virtual reality environment, and the control unit generates a 3D model of the first user terminal by combining image capture technology and point cloud data with multiple images of the first user terminal placed on an automated rotating imaging tool and captured from multiple angles.

[0012] The control unit receives a plurality of images captured while stationary at each rotation interval and rotation angle information corresponding to each image for the first user terminal placed on an automated rotating imaging tool; extracts reference feature points by identifying the outline, corner line, display boundary line, camera placement area boundary line, and button placement area boundary line of the first user terminal in each of the plurality of images; identifies the connection relationship between images by matching reference feature points representing the same location among images corresponding to adjacent rotation angles; unifies the coordinate reference of each image based on the identified connection relationship; calculates the depth of the surface point using the parallax between feature points identified as the same surface location in the plurality of images with unified coordinate references; generates a point cloud representing the surface shape of the front, side, and rear of the first user terminal by continuously connecting the calculated surface points; compares the color value change, brightness value change, and boundary line position change by angle for the area identified as the same surface location in the plurality of images with unified coordinate references; excludes areas where only the brightness value changes with the angle change and there is no change in the boundary line position as noise areas caused by reflection. Areas where linear color discontinuities are continuous regardless of angle change are identified as scratches, areas where local color degradation forms in the shape of a closed curve are identified as dents, and areas where the boundary line position deviates beyond a standard and remains in a continuous section are identified as warping; an appearance inspection model is generated by superimposing the locations of the identified scratches, dents, and warping onto a generated point cloud to display the location, length, area, and type of defects together; the appearance inspection model is displayed on the display unit so that rotation, zooming, and selection by area are possible, and control is provided so that defects are displayed separately by the display area, side frame area, rear case area, and camera placement area.When a purchase request is entered, the above-mentioned appearance inspection model is superimposed on the real-time captured video, and when a user selects a specific area, the defect type, defect location, rotation angle information at the time of inspection, and the original image of that area are displayed together, characterized by control.

[0013] The control unit performs a virtual cross-section reconstruction procedure to which the principle of tomographic reconstruction of medical images is substituted, wherein a plurality of virtual cutting planes are set based on the length, width, and thickness directions of the first user terminal, and surface points intersecting each virtual cutting plane are rearranged according to rotation angles to reconstruct the outlines for each cutting plane; for each reconstructed outline of the cutting plane, the amount of curvature deviating from a reference straight line, the gap deviation between two opposing outlines, the amount of reduction in corner angle, and the deviation in separation between the display boundary line and the frame boundary line are calculated; if the calculated amount of curvature accumulates in the same direction across three or more consecutive virtual cutting planes, it is identified as frame warping; if the deviation in separation is repeated in a section adjacent to the display boundary line, it is identified as panel lifting; and if the reduction in corner angle is concentrated in a corner section, it is identified as corner deformation; and a cross-section distortion map is generated by arranging the occurrence sections of the identified frame warping, panel lifting, and corner deformation according to the length direction order of the first user terminal, and is displayed on the display unit together with the appearance inspection model, and the occurrence of frame warping in the cross-section distortion map It is characterized by lowering the appearance grade and identifying it as a target for re-inspection if the section is continuous up to the section adjacent to the display boundary line, if panel lifting is repeated in three or more virtual cutting planes continuous along one side of the first user terminal, or if corner deformation is simultaneously identified in two or more of the four corner sections.

[0014] The control unit performs an identity verification procedure to which the feature point correspondence principle used in fingerprint identification is substituted, wherein at the time of purchase, the time of registration for sale, and the time immediately before shipment, it secures enlarged exterior images including the perimeter of the camera placement area, the arrangement of speaker holes, the perimeter of the charging terminal, the button boundary line, and the distribution of fine scratches on the side frame, respectively, and extracts the start and end points of linear scratches, the inflection points of curved scratches, the center points of the hole arrangement, the corner bending points, and the discontinuity points of the paint boundary line as feature points from each enlarged exterior image, generates a unique exterior pattern for each terminal by combining the relative distance between the extracted feature points, the pinch angle with adjacent feature points, the arrangement order of the feature points, and the location of repeated appearance within the same area, identifies whether the exterior housing has been replaced or parts have been replaced by comparing the unique exterior pattern at the time of purchase with the unique exterior pattern at the time of registration for sale, identifies whether new damage has occurred during the storage or packaging process by re-comparing the unique exterior pattern at the time of registration for sale with the unique exterior pattern at the time immediately before shipment, and wherein pairs of corresponding feature points that maintain the same arrangement order of the feature points are camera The invention is characterized by controlling the process such that if the same terminal is identified simultaneously in at least two of the placement area, the charging terminal perimeter, and the side frame, it is determined to be the same terminal, and if non-corresponding feature points are concentrated in a specific area, it is identified as a replacement of parts or the occurrence of new damage in that area, and if it is not determined to be the same terminal or if a replacement of parts or the occurrence of new damage is identified, sales are suspended and a re-photograph and re-inspection are requested, and only if it is determined to be the same terminal and the occurrence of new damage is not identified, final shipment approval is performed. Effects of the invention

[0015] By including features that automatically receive purchase requests via user input and rapidly process product information, it significantly improves the efficiency of the purchasing process,

[0016] By verifying the legality of the device from the purchase stage through the loss / theft lookup function, the distribution of illegal devices can be effectively blocked; furthermore, by automatically processing and displaying physical inspection information and results, the condition of the purchased device can be accurately evaluated, and an appropriate price can be calculated based on this.

[0017] We can provide a used product management system that contributes to increasing the speed and efficiency of distribution in the used terminal market by rapidly switching the sales waiting status of used terminals and effectively resupplying them to the market. Brief explanation of the drawing

[0018] FIG. 1 is a diagram illustrating the schematic configuration of a used product management system according to one embodiment of the present invention. FIG. 2 is a drawing illustrating the specific configuration of a used product management system according to one embodiment of the present invention. FIG. 3 is a diagram illustrating an algorithm in which a used product management system according to one embodiment of the present invention operates. FIG. 4 is a drawing illustrating an example of a screen of a program in which a used product management system according to one embodiment of the present invention operates. Specific details for implementing the invention

[0019] Embodiments of the present invention will be described in detail below with reference to the accompanying drawings. In the drawings, identical reference numbers or symbols refer to components that perform substantially the same function, and the size of each component in the drawings may be exaggerated for clarity and convenience of explanation. However, the technical concept of the present invention and its core components and operations are not limited only to the components or operations described in the following embodiments. In describing the present invention, if it is determined that a detailed description of known technologies or components related to the present invention may unnecessarily obscure the essence of the present invention, such detailed description will be omitted.

[0020] In embodiments of the present invention, terms including ordinal numbers, such as first, second, etc., are used solely for the purpose of distinguishing one component from another, and singular expressions include plural expressions unless the context clearly indicates otherwise. Furthermore, in embodiments of the present invention, terms such as 'composed of,' 'include,' 'have,' etc., should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof. Additionally, in embodiments of the present invention, 'module' or 'part' performs at least one function or operation and may be implemented in hardware or software, or as a combination of hardware and software, or may be integrated into at least one module and implemented as at least one processor. Furthermore, in embodiments of the present invention, 'at least one' among a plurality of elements refers not only to all of the plurality of elements but also to each individual element excluding the remainder or all combinations thereof. Additionally, "configured to" may be used interchangeably with, depending on the context, for example, "suitable for," "having the capacity to," "designed to," "adapted to," "made to," or "capable of." "Configured to" does not necessarily mean that it is "specifically designed to" in hardware. Instead, in some situations, the expression "device configured to" may mean that the device is "capable of" doing so in conjunction with other devices or components.For example, the phrase “a processor configured (or set) to perform A, B, and C” may mean a dedicated processor for performing said operations (e.g., an embedded processor), or a generic-purpose processor (e.g., a CPU or an application processor) capable of performing said operations by executing one or more software programs stored in a memory device.

[0021] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings. This description is intended to be detailed enough for a person skilled in the art to easily practice the invention, and it should be noted that the technical scope and concept of the present invention are not limited thereby.

[0022] FIG. 1 is a diagram illustrating the schematic configuration of a used product management system according to one embodiment of the present invention, FIG. 2 is a diagram illustrating the specific configuration of a used product management system according to one embodiment of the present invention, FIG. 3 is a diagram illustrating an algorithm in which a used product management system according to one embodiment of the present invention operates, and FIG. 4 is a diagram illustrating an example of a screen of a program in which a used product management system according to one embodiment of the present invention operates.

[0023] Referring to FIGS. 1 to 4, a used product management system according to one embodiment of the present invention includes a computing device having a display unit for displaying a screen of a used user terminal purchase and shipment management program and a user input unit for user input, and a server (200) connected to the computing device (100) via a wired or wireless network.

[0024] A computing device (100) according to one embodiment of the present invention includes, for example, a personal computer, a server computer, a handheld or laptop device, a mobile device (mobile phone, PDA, media player, etc.), a multiprocessor system, a consumer electronic device, a mini computer, a mainframe computer, any of the aforementioned systems or devices, a distributed computing environment in which data is processed at the edge where data is generated rather than at a central server, and the configuration is not limited to only those described.

[0025] The computing device (100) may include at least one processor and memory. Here, the processor may include, for example, a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor, an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), and may have multiple cores.

[0026] The memory may be volatile memory (e.g., RAM, etc.), non-volatile memory (e.g., ROM, flash memory, etc.), or a combination thereof. Additionally, the computing device (100) may include additional storage. The storage includes, but is not limited to, magnetic storage, optical storage, etc. The storage may store computer-readable instructions for implementing one or more embodiments disclosed herein, and may also store other computer-readable instructions for implementing an operating system, an application program, etc. Computer-readable instructions stored in the storage may be loaded into memory to be executed by a processor.

[0027] Additionally, the computing device (100) may include a user input unit (110) and an output device. The user input unit (110) may include, for example, a keyboard, a mouse, a pen, a voice input device, a touch input device, an infrared camera, a video input device, or any other input device. Additionally, the output device may include, for example, one or more displays, speakers, printers, or any other output devices. Additionally, the computing device may use an input device or an output device provided in another computing device as the user input unit (110) or the output device. Additionally, the computing device may include a communication module that enables the computing device to communicate with another device. Here, the communication module may include a modem, a network interface card (NIC), an integrated network interface, a radio frequency transmitter / receiver, an infrared port, a USB connection, or other interfaces for connecting the computing device to another computing device. The communication module may include a wired connection or a wireless connection.

[0028] Each component of the computing device (100) may be connected by various interconnections such as a bus (e.g., peripheral component interconnection (PCI), USB, firmware (IEEE 1394), optical bus structure, etc.) or may be interconnected by a network. Terms such as “component,” “system,” etc. used in this specification generally refer to computer-related entities that are hardware, a combination of hardware and software, software, or running software.

[0029] A computing device (100) according to one embodiment of the present invention may include a display unit (120), and the method of implementing the display of the display unit (120) is not limited and may be implemented in various display methods such as liquid crystal, plasma, light-emitting diode, organic light-emitting diode, surface-conduction electron-emitter, carbon nanotube, nanocrystal, etc. In the case of a liquid crystal method, the display unit (120) includes a liquid crystal display panel, a backlight unit that supplies light to the liquid crystal display panel, and a panel driving unit that drives the liquid crystal display panel. Meanwhile, the display unit (120) may be implemented as an OLED panel, which is a self-emissive element, without a backlight unit.

[0030] A used user terminal according to one embodiment of the present invention may include electronic devices such as smartphones, tablets, and laptops that have been previously used. The used user terminal according to the present invention has the advantage of being highly accessible due to its lower price compared to new products, and also contributes to resource recycling in terms of environmental protection. An example includes a smartphone purchased after being used by a consumer for a certain period; this device is prepared for resale after undergoing a factory reset to delete all existing data and, if necessary, software updates or hardware repairs. Furthermore, the purchased used terminal is graded according to various criteria, such as the condition of the exterior, battery performance, and screen condition, and this grade is used as an important criterion for calculating the final selling price.

[0031] Purchase and shipment management according to one embodiment of the present invention refers to systematically managing the entire process from the purchase of used user terminals to shipment to end users or other business operators through a system, and includes initial inspection of purchased terminals, price calculation, data deletion and restoration, function checks, and appearance evaluation. Through this, detailed information about each terminal is recorded in a database, and inventory management and optimal shipment strategies are established based on this data.

[0032] The types of purchasing and outbound management include individual purchasing and bulk purchasing, and different processing procedures may be applied depending on each type. For example, individual purchasing primarily applies to cases where terminals are purchased directly from consumers, while bulk purchasing refers to cases where multiple devices previously used by companies or institutions are purchased at once.

[0033] An example includes a process in which a purchased smartphone undergoes an initial inspection. Here, the smartphone's functionality, battery life, and the presence of screen damage are checked. Subsequently, the purchase price is calculated based on this information. Once the purchase price is determined, the smartphone is registered in a database, and after performing necessary repairs or data deletion, it undergoes a preparation process for shipment. During this process, information regarding the device's condition and price is provided to consumers or other businesses to ensure transparent transactions.

[0034] A program according to one embodiment of the present invention refers to software that is executed on a computer or mobile device to automate the management of purchasing and shipping used user terminals. The program according to the present invention provides a user interface to input, manage, and track the status of the terminal, stores information about the purchased terminal in a database, and can be implemented to automate tasks such as inspection, price calculation, and preparation for shipping based on this information.

[0035] Meanwhile, the program according to the present invention may include a purchase management program, a shipment management program, and an inventory management program, and may be implemented to manage the inspection, data deletion, and function checks of terminals required during the purchase process. Additionally, the shipment management program may be implemented to coordinate the timing and process of shipment of terminals that are ready for sale, and the inventory management program may be implemented to perform the function of updating and managing the inventory status of all purchased and shipped terminals in real time.

[0036] According to one embodiment of the present invention, a software program used for purchasing a used smartphone receives a purchase request, receives information such as the model, manufacturer, and condition of the smartphone from the user, and stores it in a database. Subsequently, it analyzes the condition of the smartphone through an automated algorithm, checks the degree of external damage and the presence of functional abnormalities, and finally calculates the purchase price. All inspection and price calculation processes are automatically handled within the program without additional user input, and the result can be implemented to be displayed immediately on the user's display screen.

[0037] A server (200) according to one embodiment of the present invention may include a communication unit (210), a storage unit (220), and a control unit (230).

[0038] A communication unit (210) according to one embodiment of the present invention can communicate with a computing device (100) or other external electronic devices, etc., via wired or wireless communication methods. Therefore, in addition to a connection unit including a connector or terminal for wired connection, it can be implemented in various other communication methods. For example, it can be configured to perform one or more of Wi-Fi, Bluetooth, Zigbee, infrared communication, Radio Control, Ultra-Wide Band (UWM), Wireless USB, and Near Field Communication (NFC). The communication unit (210) may include communication modules such as Bluetooth Low Energy (BLE), Serial Port Profile (SPP), Wi-Fi Direct, infrared communication, Zigbee, and Near Field Communication (NFC). Additionally, the communication unit (210) can be implemented in the form of a device, a software module, a circuit, a chip, etc.

[0039] A communication unit (210) according to one embodiment of the present invention may include various communication modules as described above, and may include an IoT communication module having an IoT network for each carrier. An IoT communication module may refer to any IoT communication network in which a plurality of objects having separate communication units (210) are connected through a network to enable services based on various platforms. By using such an IoT communication module, a smoother communication network can be provided within a set area.

[0040] A storage unit (220) according to one embodiment of the present invention may receive information from a computing device (100) or from an external search platform through a communication unit (210) and store it, and may receive text information and image information included in a plurality of web pages received by a control unit (230) and store them. The storage unit (220) may store various data according to the processing and control of the control unit (230) described later. The storage unit (220) may be accessed by the control unit (230) to perform reading, recording, modification, deletion, updating, etc. of data. The storage unit (220) may include non-volatile memory such as flash memory, a hard disk drive, and an SSD (solid-state drive) so as to preserve data regardless of whether system power is provided to the server (200). Additionally, the storage unit (220) may include volatile memory, such as a buffer or RAM, for temporarily loading data processed by the control unit (230).

[0041] A control unit (230) according to one embodiment of the present invention can perform control for the operation of various components of a server (200). The control unit (230) may include a control program (or instruction) that enables such control operation, an inactive memory in which the control program is installed, a volatile memory in which at least a portion of the installed control program is loaded, and at least one processor or CPU (Central Processing Unit) that executes the loaded control program. In addition, such a control program may be stored in an external electronic device other than the server (200).

[0042] The control program may include program(s) implemented in at least one form among BIOS, device driver, operating system, firmware, platform, and application. In one embodiment, the application may be pre-installed or stored in the server (200) at the time of manufacturing the server (200), or may be installed in the server (200) based on the received data by receiving data of the application from an external source when it is used later. The data of the application may be downloaded to the server (200) from an external server, such as an application market, for example, a platform according to the present invention, but is not limited thereto. Meanwhile, the control unit (230) may be implemented in the form of a device, S / W module, circuit, chip, etc., or a combination thereof.

[0043] When a control unit (230) according to one embodiment of the present invention receives a purchase request for a first user terminal to be purchased from a user input unit, it may request first product information including the customer's expected price and first image information for the first user terminal.

[0044] When a control unit (230) according to one embodiment of the present invention receives a request to purchase a used smartphone from a user input unit, it requests product information including the customer's estimated price and video information regarding the smartphone. In this process, the control unit collects the estimated price information provided by the customer and photo or video data showing the current state of the smartphone. The information collected in this way can be used as an important criterion for evaluating the value of the smartphone in the purchase process and making an actual purchase decision.

[0045] A control unit (230) according to one embodiment of the present invention, upon receiving first product information from a user input unit, can identify first image information, identify first defect information regarding the display screen and appearance of the first user terminal, generate inspection specific information, and display it on the display unit.

[0046] A control unit (230) according to one embodiment of the present invention receives first product information regarding a used smartphone from a user input unit, and then analyzes the submitted first image information to inspect the display screen and the exterior of the smartphone. In this process, the control unit identifies defect information such as cracks, pixel damage, touch sensitivity issues that may exist on the display screen, and scratches, dents, and discoloration on the exterior. These various defects are organized into inspection specific information and displayed on the display unit of the system. This accurately reflects the condition of the used smartphone and is provided to the end user or the next buyer, thereby enabling transparent and reliable transactions.

[0047] A control unit (230) according to one embodiment of the present invention can generate a first inspection result list for the first user terminal and display it on a display unit when it receives first physical inspection information and a first physical inspection price including at least one inspection result of vibration, speaker, face recognition, fingerprint sensor, communication connection, camera, touch display sensitivity, and button for the first user terminal from a user input unit.

[0048] A control unit (230) according to one embodiment of the present invention receives physical inspection information and inspection price of a used smartphone from a user input unit, identifies and organizes them, and the data received from the user includes information regarding the operating status of the smartphone, such as vibration, speaker, face recognition, fingerprint sensor, communication connection, camera, touch display sensitivity, and button function. Based on this, the control unit systematically organizes the operating status of each item and generates an inspection result list that encompasses them. The generated inspection result list is displayed on a display unit in a format designed to be easily understood by the user, and in this process, the status of each item and the assigned physical inspection price are clearly presented, thereby transparently evaluating the condition of the used smartphone and providing information necessary for a decision to purchase or sell.

[0049] A control unit (230) according to one embodiment of the present invention can convert the first user terminal, for which the first inspection result list has been generated, into a sales waiting state and register information as a used user terminal available for sale in the program.

[0050] A control unit (230) according to one embodiment of the present invention converts a used smartphone into a sales-ready state when the physical inspection of the used smartphone is completed and the information is organized into an inspection result list. In this process, the control unit registers the smartphone as a sellable used user terminal by including detailed information, status, and inspection results of the smartphone within the program. The registered information is stored in a database accessible for future trading and is provided to potential buyers through a sales platform, enabling buyers to clearly understand the status and features of the smartphone and make a purchase decision. This contributes to efficiently managing the smartphone sales process and enhancing the transparency and reliability of the transaction.

[0051] A control unit (230) according to one embodiment of the present invention, upon receiving user input regarding manufacturer-based first terminal source information stored in the first user terminal from a user input unit, can input the first terminal source information through a pre-configured lost / theft solution lookup API for blocking the distribution of illegal terminals and perform a lost / theft lookup for the first user terminal.

[0052] A control unit (230) according to one embodiment of the present invention receives first terminal source information based on the manufacturer of the smartphone from a user when purchasing a used smartphone. This information may generally include the IMEI number or other serial number of the smartphone. The control unit inputs this information into a pre-configured lost / theft solution lookup API to look up the lost or stolen status of the first user terminal. At this time, the API compares the source information of the terminal with domestic and international databases to identify whether there is a record of the terminal being lost or stolen. If the lookup result indicates that the terminal is registered as lost or stolen, this is recorded in the system, further purchases are stopped, and the relevant information is notified to the administrator. Through this process, the distribution of illegal terminals in the used smartphone trading market can be effectively blocked.

[0053] According to one embodiment of the present invention, when the control unit (230) identifies the first user terminal as being in a normal release state as a result of performing a loss / theft inquiry, it may display a pop-up window on the display unit for deleting existing data and connecting a function diagnosis program for the first user terminal.

[0054] If a loss / theft check on a used smartphone confirms that the device has been properly deactivated, a pop-up window is displayed on the screen requesting action from the user to prepare for the next step. This pop-up window instructs the user to delete the smartphone's existing data and begin a functional diagnostic. When the user selects an option presented in the pop-up, the control unit automatically initiates the process of executing the data deletion and functional diagnostic programs. This ensures privacy protection by formatting the smartphone's memory and verifies that various hardware and software functions are operating normally to finally evaluate the condition of the used smartphone; this step serves to guarantee that the used smartphone is in a safe and reliable condition before being sold to the next user.

[0055] A control unit (230) according to one embodiment of the present invention, upon receiving a user input of a purchase request for a first user terminal from a user input unit, can identify a first physical product sales price in which an amount of a pre-set ratio is added to a first physical product inspection price and display the first physical product sales price on a display unit.

[0056] A control unit (230) according to one embodiment of the present invention, upon receiving a purchase request for a used smartphone from a user input unit, calculates a final selling price by applying a preset ratio to the physical inspection price of the smartphone. For example, after the physical inspection price is determined, a specific ratio (e.g., a 10% profit) is added to the price to calculate a first physical selling price. The first physical selling price calculated in this way is clearly displayed on the display unit of the system by the control unit, allowing the user to easily check the final price. This process plays an important role in processing the sale of a used smartphone quickly and efficiently and ensuring transparency between the buyer and the seller.

[0057] A control unit (230) according to one embodiment of the present invention, upon receiving a request to purchase a used terminal from a user input unit, specifically records detailed information such as the model name, year of manufacture, and usage status of the terminal in a database. This information is stored together with a unique identification number and is utilized for tracking and managing the terminal during the purchase process. The control unit also calls an online price comparison API to evaluate the real-time market value of the terminal and derives an estimated purchase price in the current market through this.

[0058] A control unit (230) according to one embodiment of the present invention 3D models the external condition of a purchased terminal in real time and visualizes the appearance and function of the terminal in a virtual reality environment. This 3D model allows inspection personnel to examine all aspects of the terminal in a virtual space without directly operating the terminal during the purchase process, and can be used to identify defects hidden in the internal structure of the terminal or damage that is difficult to detect with general inspection methods.

[0059] For devices ready for sale, the control unit provides an augmented reality (AR)-based interactive interface, allowing buyers to virtually place the device in their environment. This feature enables buyers to directly verify the device's actual size and appearance within their usage environment, helping them make more informed decisions when selecting a product.

[0060] In one embodiment of the present invention, the process of the control unit (230) 3D modeling the state of the purchased terminal in real time serves to more accurately evaluate and visualize the state of the used terminal, and this 3D modeling function is implemented by combining high-resolution image capture technology and point cloud data. Each part of the terminal is photographed from various angles by a high-precision camera, and these images are analyzed through a computer vision algorithm to generate a 3D point cloud of the terminal's exterior.

[0061] More specifically, the used terminal is placed on an automated rotating imaging tool, and high-resolution images are captured from multiple angles, which is designed to capture all external features of the terminal.

[0062] The captured images are processed by computer vision software to extract detailed information such as surface curvature, color, and reflection from each image, and this data is overlapped and integrated to create a complete 3D representation of the terminal.

[0063] Subsequently, data extracted from the processed images is converted into a 3D point cloud, and this point cloud represents the physical shape of the terminal with thousands or millions of points, and each point accurately represents the surface location of the terminal.

[0064] Subsequently, the point cloud is finally synthesized into a single continuous 3D model. This model can be manipulated within a digital environment for subsequent analysis and is reviewed by an algorithm capable of automatically identifying damaged or defective areas. The completed 3D model is visually displayed on a screen, and further analysis is performed by inspection personnel or automated systems. During this process, the 3D model can be rotated, enlarged, and reduced, allowing inspectors to examine the device from all angles. This 3D modeling process enables more accurate identification of damage or defects during the purchase of used devices, provides a means to effectively demonstrate the device's condition to buyers before sale, and can significantly contribute to enhancing transparency and reliability, particularly in the transaction of high-value electronic devices.

[0065] A control unit according to one embodiment of the present invention receives a plurality of images captured while stationary at each rotation interval and rotation angle information corresponding to each image for the first user terminal placed on an automated rotating imaging tool, and extracts reference feature points by identifying the outline, corner line, display boundary line, camera placement area boundary line and button placement area boundary line of the first user terminal in each of the plurality of images.

[0066] In one embodiment of the present invention, an automated rotary imaging tool may include a flat rotating plate so that the first user terminal is seated without shaking, a driving motor that rotates the rotating plate in units of a predetermined angle, an encoder or motor step count detection unit that detects the current rotation position, and a fixing member that supports the back or side of the first user terminal. The upper surface of the rotating plate may be formed as a solid color surface of a matte material so that the outer edge of the first user terminal is clearly separated from the shooting background, and the fixing member may be positioned to support only a part of the bottom corner or rear edge of the terminal so as not to obscure the display area, camera placement area, and button placement area. For example, the rotating plate may rotate by dividing 360 degrees into a plurality of rotation sections at intervals of 10 or 15 degrees, and may be controlled to perform shooting after maintaining a state where the drive is stopped at the end of each rotation section and the vibration subsides for 0.1 to 0.5 seconds. The camera used for shooting may be a high-precision camera including an image sensor of 12 million pixels or more so that shape differences between the front, side, and rear of the first user terminal can be distinguished, and a diffuse light illumination unit may be placed on the top or perimeter of the rotating plate to reduce boundary distortion caused by surface reflection during shooting. When shooting is performed while stationary at each rotational section in this manner, an external image of the same first user terminal at different angles can be secured in a state where shaking is suppressed, and a premise is established to compare the boundary positions of each image in a subsequent step.

[0067] In one embodiment of the present invention, the control unit may receive a plurality of images and rotation angle information corresponding to each image from the automated rotating imaging tool. Here, the rotation angle information may be the absolute angle value of the rotating plate at the time each image is captured or the accumulated amount of rotation from a reference starting angle, and may be generated as any one of the following: a value obtained by converting the number of encoder pulses into an angle unit, a value obtained by converting the number of motor steps into a reference angle, or a value obtained by recording the accumulated amount of rotation after the optical sensor recognizes the reference marker of the rotating plate. For example, when the reference position of the rotating plate is set to 0 degrees and rotated 10 degrees clockwise, the control unit may receive rotation angle information corresponding to each image, such as 0 degrees, 10 degrees, 20 degrees, and 30 degrees, together and record it in the storage unit. At this time, the control unit may insert the rotation angle information into the image file name itself or create a separate index table to map the shooting order, rotation angle, shooting time, and image storage location to each other. When image and rotation angle information are recorded together in this way, the control unit can clearly distinguish which image is an image looking at the front right corner of the terminal and which image is an image looking at the rear left corner, so that the shooting direction of each photo is managed under a set standard, going beyond simply collecting multiple photos.

[0068] In one embodiment of the present invention, the control unit may identify the outline, corner line, display boundary line, camera placement area boundary line, and button placement area boundary line of the first user terminal in each of the plurality of images to extract reference feature points. To this end, the control unit may first correct lens distortion for each image and generate a corrected image based on the average brightness and color temperature of the entire image to reduce brightness deviations due to shooting lighting. Lens distortion correction may be performed from a reference image captured by placing a correction plate having a grid pattern on the rotating imaging tool before shooting begins, and the control unit may correct edge curvature using the difference between the actual arrangement of grid intersection points and the arrangement on the captured image, and then apply the same correction value to the plurality of images. Brightness correction may be performed by storing a reference image in which only the background surface is captured while the lighting unit is turned on, calculating the difference between the average brightness of the background surface in each image and the average brightness of the reference image, and shifting the total pixel value within a certain range. By first reducing errors caused by lenses and lighting in this way, it becomes possible to distinguish between brightness changes caused by the boundaries of the actual terminal and brightness changes caused by the shooting environment.

[0069] In one embodiment of the present invention, the identification of the outline can be performed using the difference in brightness and color between the background and the terminal body. For example, if the top surface of the turntable is configured as a black or gray matte surface, the control unit can sequentially scan from the upper row to the lower row of each image and extract the first point where pixels that are lower or higher than the average brightness of the background area appear consecutively as a candidate point for the terminal outline, and can additionally secure candidate points for the outline by scanning from the left column to the right column in the same manner. Among the candidate points for the outline thus secured, the control unit can form a closed curve by connecting points where the distance between adjacent points is less than or equal to a preset distance, and can confirm the closed curve with the largest area as the outline of the first user terminal. Here, the preset difference value is not determined arbitrarily, but can be determined by collecting the actual difference in brightness between the background and the body boundary from reference images taken multiple times of normal terminals of the same or similar size, comparing the actual outline marked manually by a person with the automatic extraction result, and determining a value that simultaneously minimizes outline omission and background misdetection. For example, after securing angle-based captured images for 20 to 50 normal terminals, the brightness difference value and minimum connection length can be adopted such that the average boundary error is 3 pixels or less by comparing the correct data, in which a person manually marks the outline in each image, with the automatically extracted outline. In this way, since the process of determining the reference value itself is performed using actual captured images and manual marking results, it is possible to prevent the outline extraction conditions from being set arbitrarily.

[0070] In one embodiment of the present invention, a corner line can be identified by utilizing a section where the direction of travel changes abruptly among the outer points continuously arranged along the outline. The control unit can divide the points constituting the outline into regular intervals, calculate the amount of change in two direction vectors connecting the preceding point and the following point relative to the current point, and identify a section where the amount of change is greater than or equal to a preset angle as a candidate for the starting or ending point of the corner line. For example, in straight sections, the direction of travel remains almost constant, so the change in direction between adjacent sections appears small; whereas in unrounded corners or frame bends, the change in direction occurs intensively within a short distance, so this can be used to separate the corner line. If a continuous change in direction is detected within the same section, the control unit can group the entire section into a single corner line and store both ends as reference feature points. In the case of a terminal with rounded corners, the center point where the largest change in direction occurs can be stored as a corner feature point, and the points where the change in direction begins before and after it can be stored as the two end feature points of the corner line. In this way, if the edge line is not simply treated as a single corner position on the screen, but is subdivided and extracted into a start point, a center point, and an end point, it is possible to determine more precisely whether the external shape of the first user terminal is distorted in a subsequent step.

[0071] In one embodiment of the present invention, the display boundary line can be identified as a linear boundary where the illuminated display area and the bezel or frame area meet in a front image. For an image rotated at a front angle, the control unit removes the background while leaving only the area inside the outline, and then sets a plurality of search lines from the center outward to inspect changes in brightness and color distribution along each search line. For example, when the display is photographed while displaying a predetermined monochrome screen, the difference in brightness between the illuminated screen area and the bezel area appears constant; therefore, the control unit can identify the first location where the amount of brightness change continuously exceeds a preset standard while moving from the center outward as a candidate point for the display boundary. Even when the display is photographed in a turned-off state, a boundary exists between the front glass and the frame where the reflection pattern and surface color differ; therefore, the control unit can extract candidate points for the boundary by using not only color changes but also the difference in variance values ​​of adjacent pixel groups. The extracted candidate points for the boundary can be organized into mutually orthogonal or parallel line segments, and the control unit can confirm a line as the display boundary line only if the number of candidate points placed on the same straight line exceeds a standard. Next, the control unit can store the intersection points of the four corners of the display boundary line as display reference feature points.

[0072] In one embodiment of the present invention, the boundary line of the camera placement area can be identified as a closed curve boundary surrounding the lens window, camera bump, flash window, or the outer edge of the camera housing in a rear image or a front top image. The control unit can first select a camera candidate area from within the area inside the outer edge where a circular or elliptical dark center and a ring-shaped reflective area around it are repeated. Subsequently, the control unit tracks a closed curve where brightness changes continuously occur around the candidate area, and if two or more lens apertures are included within the closed curve or if a shadow pattern is formed by changes in height around the closed curve, this can be identified as the boundary line of the camera placement area. For example, if the camera housing protrudes in a rectangular or circular island shape, a boundary line is formed between the top surface of the protrusion and the rear case; therefore, the control unit can determine the camera placement area by considering the length of the closed curve, changes in curvature, the number of internal apertures, and the surrounding shadow distribution together. The control unit can extract the center point, each corner point, or each lens center point of the identified camera placement area boundary line as reference feature points.

[0073] In one embodiment of the present invention, the boundary line of a button placement area can be identified by utilizing sections where the outline is locally protruded or cut off in an image at a side-facing rotation angle. For example, in the case where a power button or a volume button is of the protruding type, the side outline continues in a straight line, moves slightly outward from the button start point, and returns to the original outline at the button end point; therefore, the control unit can identify a section where the local deviation of the outline is maintained for a certain length or longer as a button candidate area. In the case of a recessed or flat button, it may not be distinguishable solely by the protrusion of the outline; therefore, the control unit can identify the boundary line of the button placement area by considering fine shadow lines in the horizontal or vertical direction in the side image, double boundary lines formed by button gaps, or reflection differences between the button material and the frame material. Additionally, by accumulating changes in the side boundary that are repeatedly detected near the same location for each rotation angle, a button boundary that appears weakly in a single image can also be determined. The control unit can extract and store the two end points, the center point, and the longitudinal end point of the button placement area as reference feature points.

[0074] In one embodiment of the present invention, a reference feature point is a point that can be used as a fixed reference for matching angle-specific images of the same terminal in subsequent processing, and may include corner points of an outline, start and end points of a corner line, intersection points of a display boundary line, center points or boundary corner points of a camera placement area, and both ends of a button placement area. For each reference feature point, the control unit may record the x-coordinate and y-coordinate within the image, the rotation angle at the time of shooting, the type of feature point, and a pixel patch of a predetermined size around the feature point. The pixel patch around the feature point is intended to re-verify points of similar shape that are difficult to distinguish by simple coordinates alone; for example, a partial image of 21×21 or 31×31 pixels centered on the reference feature point may be cropped and stored. Additionally, the control unit may record the brightness change amount in 8 or 16 surrounding directions for each feature point so that feature points of the same shape can be found again in a subsequent step. By recording the reference feature point including its coordinates, type, and surrounding image pattern in this manner, identifiability that distinguishes it from other points on the same outline can be secured.

[0075] In one embodiment of the present invention, the procedure for identifying the outline, corner line, display boundary line, camera placement area boundary line, and button placement area boundary line can be performed using verified conditions based on actual shooting data. For example, the control unit may obtain images under the same rotational shooting conditions for a plurality of sample terminals, including a terminal of the same model in a normal state and a terminal with scratches, dents, discoloration, or frame warping, and store verification correct answer data in which an operator directly marks the outline, display boundary line, camera placement area boundary line, and button placement area boundary line on each image. Subsequently, the control unit may calculate the positional error, the number of omissions, and the number of false detections between the automatic extraction result and the correct answer data by applying different criteria for brightness difference, minimum line segment length, direction change, closed curve area, and button boundary repeated detection count. The control unit may store the condition combination in which the average value of the positional error is smallest and the sum of the number of omissions and false detections is smallest as a boundary identification standard for the corresponding model group or shooting environment. Therefore, the above reference feature point extraction procedure does not simply apply arbitrary image processing conditions, but can be performed according to the selected conditions after selecting the conditions under which the outline and each boundary line are most stably reproduced using actual captured images and boundary data verified by a human. This corresponds to an embodiment that is concretized to the extent that a person skilled in the art can directly implement the operation in which the control unit receives a plurality of images captured while stationary at each rotation interval and rotation angle information corresponding to each image, and identifies the outline, corner line, display boundary line, camera placement area boundary line, and button placement area boundary line of the first user terminal in each of the plurality of images to extract reference feature points.

[0076] A control unit according to one embodiment of the present invention identifies a connection relationship between images by matching reference feature points representing the same position among images corresponding to adjacent rotation angles, unifies the coordinate standards of each image based on the identified connection relationship, calculates the depth of a surface point using the time difference between feature points identified as the same surface position in a plurality of images with unified coordinate standards, and generates a point cloud representing the surface shape of the front, side, and rear surfaces of the first user terminal by continuously connecting the calculated surface points.

[0077] In one embodiment of the present invention, the control unit may first check whether each image was acquired under the same shooting conditions before matching reference feature points representing the same location between images corresponding to adjacent rotation angles. To this end, the control unit reads together the rotation angle information, shooting time, focal length, exposure value, and shooting resolution assigned to each image, and may set only images with the same shooting resolution and focal length, where the difference in rotation angle is within a preset adjacent angle range, as a pair of comparison targets. For example, when an automated rotational imaging tool divides 360 degrees into 24 sections and performs still shots at 15-degree intervals, the control unit may set images of immediately adjacent rotation angles as priority matching targets, such as a 0-degree image and a 15-degree image, a 15-degree image and a 30-degree image, or a 30-degree image and a 45-degree image, and may include the next adjacent angle image as an additional matching target only if the number of feature points in one image is insufficient. In this way, by first restricting image pairs based on adjacent rotation angles, it is possible to reduce mismatches between images where reference feature points at the same location have moved excessively far apart.

[0078] In one embodiment of the present invention, the correspondence of reference feature points representing the same location is not simply achieved by connecting points with close coordinates, but can be performed by using the location, type, and surrounding shape pattern of each reference feature point together. Specifically, for the previously extracted corner points of the outline, the start and end points of the edge lines, the intersection points of the display boundary lines, the center points or boundary corner points of the camera placement area, and the two end points of the button placement area, the control unit can cut out a surrounding image patch of a certain size based on the center of each feature point, and record together the direction of brightness change, the direction of progression of the boundary line, the distribution of straight lines or curves, and the order of distances to adjacent feature points within the surrounding image patch. Subsequently, for a specific reference feature point in the first image, the control unit can extract only feature points of the same type in the second image as a candidate group, and for each candidate group, sequentially compare the similarity of the surrounding image patch, the difference in distance from adjacent feature points, and whether the arrangement order on the outline or boundary line matches. For example, regarding a feature point identified as the top-left display boundary intersection in the first image, the candidates are limited to only those points identified as display boundary intersections in the second image as well, and among them, a point that has the same direction of brightness change of surrounding patches and whose relative positional relationship with the end point of the top edge line and the corner point of the left outline matches can be confirmed as a corresponding feature point representing the same location. In this case, if a single feature point matches multiple candidates simultaneously, the control unit does not make a decision based solely on the similarity of surrounding patches, but groups the feature point together with two or more other reference feature points adjacent to it to compare the local arrangement of triangles or rectangles, and then prioritizes the group with the smallest change in area.

[0079] In one embodiment of the present invention, the connection relationship between images can be identified using pairs of corresponding reference feature points as described above. The control unit may group a plurality of reference feature points with a confirmed correspondence relationship in a first image and a plurality of reference feature points corresponding to them in a second image into a group, and generate a correspondence table for each group indicating which feature point is connected to which feature point. This correspondence table may record the feature point identification number of the first image, the feature point identification number of the second image, the types of the two feature points, the areas to which the two feature points belong, such as the front outline, rear camera area, or side button area, and the amount of coordinate movement between the two points. Subsequently, the control unit may review the entire correspondence table obtained for the same pair of images and determine the connection relationship of the corresponding pair of images as valid only if corresponding feature points are secured across at least two different areas among the outline, display boundary, camera placement area, and button placement area in a single image. Conversely, if the corresponding feature points are concentrated only in a specific area or if only feature points of the same type are repeatedly corresponded, it may be considered that there is a high probability of mismatch and a re-search may be performed. For example, if only the center point of the rear camera placement area matches and corresponding points are not secured in other areas, the connection between images is withheld; the connection can be confirmed only when the outline corner point, the center point of the camera placement area, and both ends of the button placement area correspond together. By identifying the connection as a combination of multiple feature points in multiple areas rather than a single point in this manner, stable connections between the entire images are established even when specific areas are obscured or some feature points are weakly detected due to reflection.

[0080] In one embodiment of the present invention, the procedure for unifying the coordinate standards of each image may be a procedure for converting the first image and the second image so that they are interpreted in the same spatial standard based on a determined connection relationship. Specifically, the control unit may select one reference image and correct the position and orientation of the remaining images according to the coordinate system of the reference image. For example, if a 0-degree image is selected as the reference image, the control unit may sequentially calculate the x-axis translation amount, the y-axis translation amount, and the rotation correction amount by using the location in the 0-degree image where the corresponding feature points in the 15-degree image correspond. Here, the translation amount may be determined as the average value of the coordinate difference of the same pair of feature points, and the rotation correction amount may be determined using the difference in direction of the reference line connecting the outline corner point and the display boundary intersection point. If some of the translation and rotation amounts calculated from each of the multiple corresponding feature points show a significantly large deviation, the control unit may exclude them as mismatch points and determine the final correction value using only the remaining feature points. Subsequently, the control unit may translate and rotate the entire second image according to the correction value to convert it into the same coordinate system as the reference image. This procedure can be repeatedly applied not only to 15-degree images but also to 30-degree, 45-degree, and all subsequent images, and the entire set of images can be unified into a single common coordinate system by sequentially utilizing the connections between adjacent images. For example, the coordinate system of a 30-degree image can be corrected by utilizing both the connection with the 15-degree image and the indirect connection with the 0-degree image, and if necessary, the correction error can be reduced by directly comparing it with the previously unified image.

[0081] In one embodiment of the present invention, depth calculation using the parallax between feature points identified at the same surface location in a plurality of images with unified coordinate references can be performed with the relative arrangement of a rotating imaging tool and a high-precision camera corrected in advance. To this end, the control unit may perform a correction procedure using a reference shape body before actual terminal shooting. The reference shape body may be a jig in which a flat portion, a protrusion, and a depression are formed at different heights, and a reference marker with known coordinates may be attached to each face and the protrusion edge. The control unit may place the reference shape body on the rotating imaging tool, acquire a plurality of images at the same rotation interval and camera position as the actual terminal shooting, and measure how much each marker moves in adjacent angle images to generate a correction table that stores the correspondence between the difference in rotation angle, the amount of pixel movement, and the actual depth. For example, by using a reference shape body that includes markers at positions 20 mm, 40 mm, and 60 mm away from the axis of rotation, and surface markers formed at heights of 0 mm, 1 mm, 2 mm, and 3 mm, the control unit can obtain actual measurements of how a feature point moves on an image according to each depth and angle. Therefore, for a surface point of an actual terminal, the depth can be calculated by inputting the number of pixels a specific feature point has moved between adjacent images and the distance information from the axis of rotation of the area to which the feature point belongs into a correction table. In this case, the depth calculation is not determined by a single pair of images alone; rather, if the same surface position is repeatedly detected in two or more adjacent left-right angle images, the depth value obtained from each pair of images can be determined as the median value or the value with the smallest deviation. By converting the depth based on the captured values ​​using the actual reference shape body in this way, a depth calculation standard that reflects the actual device layout and lens characteristics can be used, rather than simply assuming theoretical parallax relationships.

[0082] In one embodiment of the present invention, a feature point at the same surface location does not necessarily refer only to a point on an outline or boundary line, but may also include surface pattern points that are repeatedly observed in the area between the reference feature points formed by the outline and boundary line after alignment is completed. Specifically, after the coordinate reference is unified, the control unit can additionally extract detailed surface patterns for the frame, rear case, camera periphery, button periphery, etc., excluding the uniform area inside the display boundary line. For example, if subtle surface texture changes, the starting point of a scratch, a fixed fine dot pattern that is not a reflection pattern, or processing traces around a camera bump are reproduced at the same location in multiple adjacent images, the control unit can adopt these as additional surface point candidates. Subsequently, for each of these surface point candidates, the control unit measures the pixel displacement between adjacent images and calculates the depth using a previously generated correction table, thereby enabling the construction of a denser set of surface points that does not rely solely on the outline or boundary line. Therefore, the point cloud can be formed not as a sparse set of points representing only the outer contour, but as a dense set of points capable of expressing the flatness of the front, the straightness of the side frame, and even the difference in protrusion height around the rear camera.

[0083] In one embodiment of the present invention, the procedure for generating a point cloud by continuously connecting calculated surface points can be performed by first classifying points belonging to the same surface area by area, and then determining neighbor point relationships within each area. The control unit can distinguish the front display surrounding frame area, side frame area, rear case area, camera placement area, and button placement area using outline and boundary information, and can designate points with close 3D positions as neighbor points for surface points within each area. For example, points whose differences in the x-direction, y-direction, and depth direction from a specific surface point are all within a standard can be set as adjacent points, and if there are two or more adjacent points, the three to six closest points can be selected as priority connection targets. The control unit secures the continuity of the point group by connecting the neighbor points designated in this way with line segments, and at the area boundaries, directly connects with the outline or boundary feature points so that the front, side, and rear can be represented as a single surface shape that is connected to each other. In addition, if some of the connected points protrude or sink excessively compared to surrounding points, the control unit may determine them as error points and remove them, or replace them with positions interpolated from two or more adjacent points. The resulting point cloud is data representing the entire physical shape of the front, side, and rear of the first user terminal as thousands to millions of points, and each point may include an x-position, a y-position, and a depth value in a common coordinate system.

[0084] In one embodiment of the present invention, the reference values ​​and processing order used in the feature point correspondence, coordinate reference unification, depth calculation, and point cloud generation procedures can be determined using actual shooting samples. The control unit may secure verification samples for each model, including a normal terminal, a terminal with corner dents, a terminal with frame warping, a terminal with camera bump damage, and a terminal with a slightly lifted rear cover, and may perform repeated shooting for each sample under the same rotation interval and lighting conditions. Subsequently, an operator may manually designate a reference feature point at the same location in each image and measure the actual amount of frame warping, camera protrusion height, and rear cover lift height using a separate 3D scanner or vernier calipers and a depth gauge. The control unit may determine the feature point correspondence conditions by comparing the manually designated feature point correspondence relationship with the automatic correspondence results, and may adopt a correction table in which the difference between the actual depth and the automatically calculated depth is the smallest. For example, shooting conditions and correction conditions in which the automatic correspondence success rate is 95% or higher and the average difference between the actual depth and the automatically calculated depth is 0.2 mm or less may be determined as actual operating conditions. In addition, the accuracy of the point cloud can be verified by comparing the values ​​of the front width, side thickness, and camera protrusion height of the sample terminal re-measured in the point cloud with the actual measurements, and the control unit can readjust the reference feature point extraction conditions or coordinate correction conditions for the corresponding model group if the difference exceeds the allowable range. As such, feature point correspondence, coordinate unification, depth calculation, and point cloud generation are performed through a shooting and measurement procedure using both the actual terminal sample and the reference shape body, so each operation included in the above sentence can be implemented specifically enough for a person skilled in the art to reproduce it exactly.

[0085] A control unit according to an embodiment of the present invention compares changes in color values, changes in brightness values, and changes in boundary line positions by angle for areas identified at the same surface location in a plurality of images with unified coordinate standards; excludes areas where only the brightness value changes with angle change and there is no change in boundary line position as noise areas caused by reflection; identifies areas where linear color discontinuities are continuous regardless of angle change as scratches; identifies areas where local color degradation forms in the shape of a closed curve as dents; and identifies areas where the boundary line position deviates beyond a reference point and remains in a continuous section as warping; generates an appearance inspection model in which the location, length, area, and type of defects are displayed together by superimposing the locations of the identified scratches, dents, and warping onto a generated point cloud; displays the appearance inspection model on the display unit so that rotation, zooming, and selection by area are possible, and controls the defects to be displayed separately by the display area, side frame area, rear case area, and camera placement area; and when a purchase request is input, superimposes the appearance inspection model onto a real-time captured video so that if the user selects a specific area, the type of defect in that area, Controls the display so that the defect location, rotation angle information at the time of inspection, and the original image are displayed together.

[0086] In one embodiment of the present invention, the control unit may divide the surface area to be compared in each image into local areas of equal size before comparing areas identified as having the same surface location for a plurality of images with unified coordinate standards. For example, the control unit may divide the surface of the terminal on each image into small comparison units of 8×8 pixels to 32×32 pixels using image coordinates corresponding to each point of the point cloud formed on the front, side, and rear, and may store information such as the center coordinate, whether the outer boundary is included, the 3D position on the point cloud, and the rotation angle at the time of shooting together for each comparison unit. By organizing the same surface location into small comparison units on the image in this way, it is possible to determine, by location, where changes in color, brightness, and shape have occurred on the terminal, rather than simply comparing the entire image. Since the original specification already describes a flow for extracting and integrating detailed information such as surface curvature, color, and reflection from each image, generating a point cloud, and automatically identifying damaged or defective areas, the following comparison procedure corresponds to a specific implementation method.

[0087] In one embodiment of the present invention, the control unit may first convert each of a plurality of images with unified coordinate standards into the same color space to compare color value changes by angle. For example, when a camera generates an original image in RGB format, the control unit may calculate not only the average values ​​of red, green, and blue but also the brightness component and the color difference component for each comparison unit to make it easier to distinguish between the influence of lighting intensity and changes in the color itself. More specifically, the control unit may calculate the average red value, average green value, and average blue value of the pixels within each comparison unit, and simultaneously calculate the average brightness value representing the overall brightness and the average color difference value representing the color direction. Subsequently, the control unit may sequentially calculate the change in average red value, the change in average green value, the change in average blue value, the change in average brightness value, and the change in average color difference value between comparison units corresponding to the same surface position in adjacent rotation angle images. Here, the standard for the amount of color change is not determined arbitrarily, but can be derived from a verification image group in which a normal area free from reflection, scratches, dents, and warping is manually designated by an operator after photographing multiple normal terminals of the same model across their entire rotation angle under identical lighting conditions. The control unit accumulates and stores the average color change and average brightness change occurring according to angle changes within the normal area, and can establish a standard value to treat a defect candidate only when the degree of deviation from the distribution range of the normal area appears repeatedly. For example, after confirming that the average color change at the same surface location in the shooting results of 30 or more normal terminals mostly falls within a certain range, only the sections where color discontinuities larger than the upper limit of that range are continuously detected can be adopted as abnormal color sections. Since the process of setting the standard value is based on actual shooting samples and manual designation results in this way, it is possible to prevent arbitrary comparison of color values ​​by angle.

[0088] In one embodiment of the present invention, the control unit can calculate not only the average brightness but also the brightness variance and the brightness gradient around the boundary for each comparison unit identified at the same surface location in order to compare changes in brightness values ​​by angle. For example, an area that becomes temporarily brighter due to reflection tends to show a strong increase in brightness at one rotation angle but decreases again at an adjacent rotation angle, while the actual position of the boundary line is often maintained. Accordingly, the control unit can arrange the average brightness values ​​for comparison units corresponding to the same surface location in order of rotation angles and simultaneously examine whether only brightness changes abruptly between adjacent rotation angles, or whether color discontinuity or boundary shifting is accompanied by changes in brightness. For example, if there are images at 0 degrees, 15 degrees, 30 degrees, and 45 degrees, the control unit sequentially compares the brightness change between 0 and 15 degrees, between 15 and 30 degrees, and between 30 and 45 degrees for the same surface location, and if the direction of the brightness change reverses as the angle increases while the boundary line position is substantially maintained, it can be classified as a candidate for reflection. Conversely, if a color discontinuity is fixed at the same location or the boundary line itself moves simultaneously with the occurrence of a brightness change, it can be determined that it is highly likely to be an actual surface defect rather than a simple reflection. For this brightness comparison as well, a reference value can be set based on the difference in distribution after accumulating and recording the brightness change range of the reflection pattern and the brightness change range of the defect pattern, respectively, using a verification image group in which the operator distinguishes and marks the area where only reflected light exists and the area where actual defects exist.

[0089] In one embodiment of the present invention, the control unit can calculate the actual coordinate displacement of a local boundary passing through the same surface location to compare changes in the boundary line position. Specifically, the control unit can obtain the center position of the boundary line passing through each comparison unit by connecting the outline, the display boundary line, the boundary line of the camera placement area, the boundary line of the button placement area, and local boundary points detected in the vicinity. Subsequently, for the comparison unit corresponding to the same surface location in adjacent rotation angle images, the control unit can calculate, in pixel units, how much the center position of the boundary line has moved in the left-right or up-down direction. If the brightness value changes significantly in a specific comparison unit but the displacement of the center position of the boundary line remains within the allowable error range obtained from the verification image group of a normal terminal, this can be treated as if only the brightness has changed due to surface reflection and there is no actual change in shape. Conversely, if the center position of the boundary line deviates continuously in the same direction in a specific comparison unit and this deviation continues in a plurality of adjacent comparison units, it can be treated as an actual change in shape, such as frame bending or corner deformation. Here, the allowable error range for boundary line position change can also be determined based on the maximum value of position shake detected for the same boundary in the repeated shooting results of a normal terminal. For example, the range of variation of the boundary line center position can be measured by repeatedly shooting the same terminal at the same angle 10 or more times, and it can be considered as actual deformation only when movement exceeding the upper limit of the range of variation is repeated at adjacent angles.

[0090] In one embodiment of the present invention, a procedure for excluding an area as a noise area caused by reflection, in which only the brightness value changes with angle change and there is no change in the boundary line position, can be performed as follows. The control unit organizes the average brightness value and the boundary line center position in order of rotation angle for comparison units at the same surface position, and can identify a reflection noise condition when the brightness change range is greater than or equal to a standard, the color change is within the normal reflection pattern range, and the amount of movement of the boundary line center position is within the allowable error range. Subsequently, the control unit can determine a comparison unit satisfying the reflection noise condition as the final reflection noise area if it appears at only one rotation angle or appears briefly at only two adjacent rotation angles and then disappears. At this time, the control unit can remove the reflection noise area from the defect candidate group and ensure that the color and brightness changes of the excluded area are no longer reflected in subsequent scratch, dent, and warping judgments. Therefore, even when a temporary brightness pattern appears due to lighting reflection around a metal frame, a glossy rear case, or a camera lens, the possibility of mistaking it for an actual defect can be reduced.

[0091] In one embodiment of the present invention, the procedure for identifying an area where linear color breaks are continuous regardless of angle change as a scratch can be performed using the directionality and continuous length of the color breaks. The control unit can detect boundary candidates where the color value changes abruptly at each comparison unit, and then verify whether the boundary candidates are connected to adjacent comparison units to form a single linear pattern. For example, a long, continuous fine scratch on a rear case always appears at similar locations at adjacent rotation angles and is often detected as a color break line in a constant direction regardless of brightness changes. Accordingly, for comparison units identified at the same location in adjacent rotation angle images, the control unit records the start point, end point, direction of progression, and continuous length of the color break line, and can confirm it as a scratch if the 3D positions of the start point and end point are close to each other on a point cloud, the direction of the break line is maintained within a reference direction error range, and the continuous length is greater than or equal to a minimum length determined in advance by experiment. Here, the minimum length criterion can be adopted by comparing the scratch correct answer line directly marked by an operator with the automatically extracted result for a group of verified terminals containing actual scratches, thereby removing short noise lines while not missing actual scratches. Additionally, the control unit determines a line as a scratch only when it is connected across at least three adjacent comparison units, rather than existing only within a single comparison unit, thereby preventing simple pixel noise or lighting boundaries from being mistaken for scratches.

[0092] In one embodiment of the present invention, the procedure for identifying an area where local color degradation forms a closed curve as a dent can be performed by considering both the color degradation within the local area and whether the perimeter boundary is closed. The control unit can extract parts where the average color value is lower or higher than the surrounding normal area in each comparison unit and its surrounding units, and then check whether the boundary is connected by a closed curve. For example, a dent occurring on a corner or rear case appears as local discoloration or peeling of the paint, and a generally closed boundary is formed around it. Accordingly, the control unit can track the outer boundary of the color degradation area and confirm it as a dent if the starting point and the ending point are connected to form a closed curve, the area inside the closed curve is greater than or equal to the minimum area criterion, and the same location is repeatedly detected at adjacent rotation angles. Additionally, if the average depth inside the closed curve area in the point cloud appears lower than that of the surrounding area, the reliability of the dent determination can be further increased. Here, the minimum area criterion and the depth degradation criterion can also be set by comparing the results of automatic extraction with the correct answer data marked by the operator on a terminal sample with an actual dent. For example, since very small color stains may be simple contamination even if they form a closed curve, they can be excluded from the imprint candidates if the area inside the closed curve is less than the standard or if they are not reproduced at adjacent angles.

[0093] In one embodiment of the present invention, the procedure for identifying an area as bending in which the boundary line position is maintained in a continuous section with a deviation from a reference can be performed by using a point cloud and the amount of boundary line position displacement together. The control unit selects boundary reference points arranged at regular intervals along the outline, side frame line, display boundary line, or the joint line between the rear case and the frame, and can calculate how much the 3D position of each reference point deviates from an ideal straight line or plane. For example, in a terminal sample where the side frame is normal, the top frame line is maintained as an almost straight line, so the control unit can store the average frame line obtained from multiple normal terminals of the same model as the reference line. Subsequently, if the boundary reference points constituting the top frame line in the actual terminal subject to inspection deviate from the reference line in the same direction and the deviation is maintained over multiple continuous sections, this can be identified as frame bending. In the case of display boundary lines, if the boundary line continuously deviates inward or outward along a specific side, it may be actual bending or lifting rather than a simple shooting error; therefore, the control unit can determine the bending by using the direction of deviation, the length of deviation, and the duration of the deviation amount together. Here, the continuous duration criterion may be determined as being satisfied when deviation in the same direction is confirmed in at least three adjacent comparison units in order to exclude accidental boundary errors that appear in only one comparison unit.

[0094] In one embodiment of the present invention, the control unit may generate an appearance inspection model by superimposing the locations of scratches, dents, and warping identified as above onto a generated point cloud. To this end, for each defect, the control unit may connect the image coordinates of the comparison unit where the defect was detected with the 3D coordinates on the point cloud, and define a set of points constituting the defect as a single defect object. For example, in the case of a scratch, the length may be calculated using the 3D coordinates of the start and end points, and points arranged along the scratch may be connected to be stored as a linear defect object. In the case of a dent, the area may be calculated using a set of points included within the closed curve boundary, and the internal average depth and the difference between the maximum depth may be recorded together. In the case of warping, a set of points deviating from a reference line or reference plane may be extracted, and the start position, end position, maximum deviation amount, and deviation direction of the deviation section may be recorded together. Subsequently, the control unit can superimpose scratches as linear highlights, dents as closed-curve area highlights, and warping as continuous segment highlights, respectively, onto a 3D model displaying the entire point cloud. Additionally, each defect object may be associated with defect type, detection location, length, area, maximum depth or maximum deviation, detection rotation angle range, and original image reference information. As a result, the appearance inspection model is not merely a simple 3D shape display, but a 3D inspection model configured to reveal both numerical and visual information regarding the location and form in which defects exist. Since the original specification already describes synthesizing the point cloud into a single continuous 3D model, examining damaged or defective areas using an algorithm capable of automatically identifying them, and displaying the model in a manner that allows for rotation, enlargement, and reduction, the above appearance inspection model corresponds to a specific embodiment thereof.

[0095] In one embodiment of the present invention, the control unit may display the appearance inspection model on the display unit so that rotation, zooming, and selection by area are possible. Specifically, the control unit transmits a point cloud or a 3D surface model converted therefrom into a triangular mesh to a graphic rendering engine and can perform rotation in the x-axis, y-axis, and z-axis directions according to the user's drag input or touch input. Zooming in and out can be performed according to mouse wheel input or two-finger zoom input, and the control unit can automatically correct so that the selected defect object moves to the center of the screen. In addition, the control unit may pre-classify which area each point of the appearance inspection model belongs to among the display area, side frame area, rear case area, and camera placement area. This classification can be performed using the previously extracted display boundary line, outline line, and camera placement area boundary line; for example, the area inside the display boundary line may be designated as the display area, the front perimeter between the outline line and the display boundary line as the front frame area, the set of points in the side direction as the side frame area, the part inside the rear outline excluding the camera placement area as the rear case area, and the area inside the camera placement area boundary as the camera placement area. Subsequently, when a user wishes to select and view only a specific area, the control unit can switch points outside the selected area to a translucent or non-display state and highlight only the defects in that area. By performing area-specific display in this way, it becomes possible for a buyer or inspector to check only for scratches on the rear case or to check only for dents in the camera placement area.

[0096] In one embodiment of the present invention, when a purchase request is input, the control unit may superimpose the appearance inspection model onto the real-time captured image. To this end, the control unit receives real-time image from a camera currently connected to the display unit or a separate capturing camera, and may detect reference feature points of the terminal or terminal placement reference plate in the real-time image. Subsequently, the control unit determines the correspondence between the reference feature points of the real-time image and the reference feature points used when generating the appearance inspection model, and may correct the position and direction so that the current rotation state of the appearance inspection model matches the viewpoint of the real-time image. Once the correction is completed, the control unit superimposes the outline and defect markings of the appearance inspection model onto the terminal position in the real-time image, and when a user touches or clicks a specific defect marking on the screen, it may also display detailed information connected to the defect object in the form of a pop-up window or a side information window. This detailed information may include the type of defect, such as whether it is a scratch, dent, or warping; the location of the defect, such as whether it is on the upper left rear or under the right side button; the rotation angle information at the time of inspection; and a thumbnail or full image of the original image where the defect was most clearly detected. Additionally, if there are multiple original images, the control unit may provide a list of original images sorted by rotation angle, allowing the user to directly compare how the defect appeared from different angles. By providing real-time video footage, an exterior inspection model, and original images in conjunction in this manner, the buyer can directly examine the actual location and shape of defects that are difficult to verify through simple text descriptions alone.

[0097] In one embodiment of the present invention, each criterion used in the procedures for exclusion of reflection noise, identification of scratches, identification of dents, identification of warping, generation of appearance inspection models, and overlay display can be determined through actual terminal shooting and manual correct answer marking procedures. Specifically, the control unit may secure normal terminals, highly reflective glossy terminals, terminals with actual scratches, terminals with actual dents, terminals with confirmed frame warping, and terminals with complex defects by model group, and may store image groups repeatedly shot under the same automated rotating imaging tool and lighting conditions. For each image group, the operator directly marks the reflection area, scratch segment, dent boundary, and boundary line where warping occurred, and the control unit may calculate the positional difference, number of omissions, number of false detections, and number of type misclassifications between the automatic judgment result and the operator's mark result, respectively. Subsequently, the control unit may perform iterative calculations while applying different criteria for reflection contrast change, minimum scratch length, minimum dent area, amount of warping deviation, and continuous section, and may store the condition combination closest to the actual display result as a criterion per model or per shooting environment. In addition, the accuracy of the defect location superimposed on the appearance inspection model can be verified by comparing the defect length, area, or deviation amount calculated from the point cloud with the actual measured value obtained through magnified microscopic photography or manual measurement. Since each standard is determined based on actual shooting data and verification data according to this procedure, the entire process of excluding reflection noise, identifying scratches, dents, and warping, generating the appearance inspection model, marking by area, and displaying linked to the original image included in the above sentence can be implemented in such detail that a person of ordinary skill can perform it as is.In addition, since the cited invention describes a classification of appearance defects and a classification of defect types such as bending, scratches, and dents, the present embodiment specifies a specific determination procedure using angle-specific color values, brightness values, boundary line position changes, and point cloud superposition in addition to the defect types themselves, thereby allowing the configuration to reveal not only a simple defect classification level but also the actual 3D appearance inspection model generation and display procedure.

[0098] A control unit according to one embodiment of the present invention performs a virtual cross-section reconstruction procedure to which the principle of cross-section reconstruction of a medical image is substituted, wherein a plurality of virtual cross-sections are set based on the length direction, width direction, and thickness direction of the first user terminal, and surface points intersecting each virtual cross-section are rearranged according to rotation angles to reconstruct the outline of each cross-section.

[0099] In one embodiment of the present invention, the virtual cross-section restoration procedure performed by the control unit may apply the principle of restoring a cross-sectional shape by rearranging information obtained from multiple shooting directions in a medical image into reference plane units to the restoration of the external shape of the first user terminal. Here, the control unit may first determine the length direction, width direction, and thickness direction of the first user terminal using the three-dimensional coordinates of each surface point included in a point cloud already generated. The length direction may be set as the direction connecting the top center and the bottom center of the first user terminal, the width direction may be set as the direction connecting the left side and the right side, and the thickness direction may be set as the direction connecting the front display direction and the rear case direction. The control unit may determine the length axis using the center of the point group farthest in the vertical direction among the outermost points of the outline, determine the width axis using the center of the point group farthest in the horizontal direction, and determine the thickness axis using the difference in average depth between the front point group and the rear point group. At this time, by first determining one axis and then correcting the remaining axis to be orthogonal to that axis, the three directions can be set so that they do not twist relative to each other.

[0100] In one embodiment of the present invention, the control unit may set a plurality of virtual cross-sections based on the length direction, width direction, and thickness direction determined as above. Specifically, the control unit may create a plurality of length direction cross-sections from the top to the bottom of the first user terminal by arranging cross-sections orthogonal to the length direction at regular intervals, create a plurality of width direction cross-sections from the left side to the right side by arranging cross-sections orthogonal to the width direction at regular intervals, and create a plurality of thickness direction cross-sections from the front to the back by arranging cross-sections orthogonal to the thickness direction at regular intervals. For example, the length direction cross-sections may be set by dividing the entire length of the first user terminal into intervals of 0.5 mm to 2.0 mm, the width direction cross-sections may be set by dividing the entire width into intervals of 0.5 mm to 2.0 mm, and the thickness direction cross-sections may be set by dividing the entire thickness into intervals of 0.1 mm to 0.5 mm. However, this interval is not determined arbitrarily, but can be determined by comparing the cross-sectional outline restored from a point cloud with the actual cross-sectional dimensions measured by a separate reference measuring device for multiple sample terminals, including normal terminals and terminals with frame bending or dents, of the same model group. For example, after setting cross-sections at intervals of 0.5 mm, 1.0 mm, 1.5 mm, and 2.0 mm, the width, thickness, and corner curvature of the restored cross-section outline for each interval can be compared with actual measurements, and the interval with the smallest average error and which does not excessively interrupt the shape change between adjacent cross-sections can be adopted as the cross-section setting interval for the corresponding model group. Accordingly, multiple virtual cross-sections can be set according to the interval verified by comparison with the shape data of the actual sample terminals.

[0101] In one embodiment of the present invention, the procedure for extracting surface points intersecting each virtual cutting plane can be performed by determining which cutting plane each surface point of the point cloud belongs to. Specifically, for a single virtual cutting plane, the control unit may first determine the center coordinates and normal direction of the cutting plane and calculate the vertical distance from each surface point to the cutting plane. This vertical distance can be calculated using the difference between the normal vector of the cutting plane and the coordinates of the surface point, and only surface points whose absolute value is within a predetermined allowable thickness range can be adopted as surface points intersecting the cutting plane. For example, if the allowable thickness range for the longitudinal section is set to ±0.15 mm, the allowable thickness range for the width section to ±0.15 mm, and the allowable thickness range for the thickness section to ±0.05 mm, only points included within these ranges can be stored as candidate surface points for the corresponding section. The above allowable thickness range can also be determined by measuring the degree of dispersion of surface points at the same cross-sectional location during repeated shooting of a normal terminal, and can be confirmed by checking whether the cross-sectional outline blurs when using an excessively wide range and whether the cross-section breaks when using an excessively narrow range.

[0102] In one embodiment of the present invention, the control unit can rearrange surface points intersecting each virtual cutting plane according to rotation angle. To this end, the control unit can refer to historical information regarding which rotation angle image each surface point was originally extracted from. As previously mentioned, since each surface point is connected to an image of a specific rotation angle during the reference feature point correspondence, coordinate reference unification, and depth calculation processes, the control unit can store rotation angle information of the corresponding image for each surface point. Subsequently, the control unit aligns surface points belonging to a single virtual cutting plane in order of rotation angle, such as 0 degrees, 15 degrees, 30 degrees, and 45 degrees, and can rearrange surface points extracted from the same rotation angle according to polar coordinate order based on the center point of the cutting plane. Here, the polar coordinate order can be determined by calculating the atan2(y, x) value for the position of each surface point projected onto the cutting plane plane. For example, by setting the center of the cross-section plane as the origin, calculating the relative coordinates of each surface point on the plane, and then sequentially arranging them from the point with the smallest atan2(y, x) value to the point with the largest, the circumferential arrangement order of the points surrounding the cross-section plane can be obtained. Rearrangement by rotation angle is intended to clarify which section of the cross-section the point group observed at the same angle represents, while polar coordinate rearrangement is intended to continuously reconstruct the outline of the cross-section plane.

[0103] In one embodiment of the present invention, the procedure for restoring the outline for each cutting plane can be performed by connecting points forming the outline among surface points rearranged according to rotation angles. The control unit first calculates the radius value from the center point for all surface points projected onto a single cutting plane, and if multiple points belonging to the same angle interval exist, the point farthest from the center point can be adopted as the outline candidate point. For example, by dividing the cutting plane into small angle intervals of 1 or 2 degrees and selecting the point with the largest radius value for each interval, a set of points located at the outermost edge along the cutting plane can be obtained. The control unit can generate a primary outline by connecting these selected outline candidate points in polar coordinate order. However, since there may be sections where outline candidate points are missing because some surfaces are obscured by a support member or some points are deleted during the reflection noise removal process at a specific rotation angle, the control unit can generate interpolation points for the missing sections by referencing outline candidate points in adjacent front and back angle intervals. For example, if the length of the missing section is less than or equal to a predetermined ratio of the total circumference length, interpolation can be performed by connecting the front and rear outer candidate points with straight lines or curves; if the missing section is long, interpolation can be performed by additionally referencing surface points corresponding to the same cross-section from other images with adjacent rotation angles. Subsequently, the control unit can generate a secondary outline by comparing the distance and direction changes between adjacent points to remove sections where abrupt jagged phenomena are repeated on the generated primary outline, and by removing intermittent protrusion points that exceed the allowable range by comparing it with the outline of the cross-section of a normal sample terminal. The secondary outline thus generated becomes the restored outline of the corresponding cross-section.

[0104] In one embodiment of the present invention, to ensure the accuracy of contour restoration for each cross-section, the control unit may perform a verification procedure using a sample terminal group. Specifically, the control unit may prepare a plurality of sample terminals including a normal terminal of the same model, a terminal with a bent frame, a terminal with a damaged corner, and a terminal with a deformed area around the rear camera, and may perform imaging using the automated rotary imaging tool and separate reference shape measurement on these terminals. The reference shape measurement may be performed by obtaining the actual dimensions of the representative cross-section outline in the length, width, and thickness directions of each terminal using a 3D scanner capable of cross-sectional projection, a laser profiler, or high-precision vernier calipers. Subsequently, the control unit may restore a plurality of cross-section contours by applying different cross-section spacing, allowable thickness ranges, angle intervals for selecting outline candidate points, and interpolation conditions to the same sample terminal, and compare the width, thickness, corner curvature, protrusion height, and depression depth of the restored contours with the reference shape measurement results. Based on these comparison results, the control unit can store the condition combination that minimizes the average dimensional error, maximum outer deviation, and the number of outline breakage occurrences as the cross-section restoration condition for the corresponding model group. Accordingly, the above-mentioned multiple virtual cutting plane settings and the outline restoration procedure for each cutting plane can be performed according to conditions determined using the measured values ​​and repeated shooting values ​​of the actual sample terminal.

[0105] In one embodiment of the present invention, the outlines for each cutting plane may be used as a basic shape for determining the degree of curvature, thickness deviation, or local depression of the cross-section in a subsequent step; however, at the present step, the core is to restore the outlines for each cutting plane by rearranging the surface points intersecting each cutting plane according to the rotation angle. Accordingly, after restoring the outlines for each cutting plane as described above, the control unit can store the cutting plane identification number, the direction of the cutting plane, the position of the cutting plane, the coordinates of the outline constituent points, and the outline length together. Once virtual cutting planes in multiple directions and outlines for each cutting plane are secured in this manner, the front, side, and rear shapes of the first user terminal become organized into cross-sectional structural units beyond simple outer surface point groups, and a person skilled in the art can actually implement the virtual cross-section restoration procedure according to the above procedure.

[0106] A control unit according to one embodiment of the present invention calculates, for each outline of a restored cut plane, an amount of curvature deviating from a reference straight line, a spacing deviation between two opposing outlines, a reduction in corner angle, and a separation deviation between a display boundary line and a frame boundary line; if the calculated amount of curvature accumulates in the same direction in three or more consecutive virtual cut planes, it is identified as frame bending; if the separation deviation is repeated in a section adjacent to the display boundary line, it is identified as panel lifting; if the reduction in corner angle is concentrated in a corner section, it is identified as corner deformation; and generates a cross-sectional distortion map in which the sections where the identified frame bending, panel lifting, and corner deformation occur are arranged in the longitudinal order of the first user terminal, and displays it on the display unit together with the appearance inspection model.

[0107] In one embodiment of the present invention, the control unit may not use the outlines of each previously restored virtual cross-section as they are, but may first align them in a state comparable to the reference cross-section outlines obtained from a group of normal terminals of the same model. To this end, the control unit may photograph multiple normal terminals of the same manufacturer and model name as the first user terminal using an automated rotating imaging tool to restore the outlines of virtual cross-sections in the same manner for the length, width, and thickness directions, and may store the average outline shape for each cross-section position as the reference cross-section outline in the storage unit. For example, in the case of a lengthwise cross-section, the entire length from the upper reference point to the lower reference point may be divided in equal proportions to store the reference cross-section outlines corresponding to each position, and in the case of a widthwise cross-section, the entire width from the left frame reference point to the right frame reference point may be divided in equal proportions to store the reference cross-section outlines. In this way, by averaging the outlines of the cut surfaces obtained from 20 to 50 or more normal terminals of the same model, a reference shape can be prepared that reflects shape changes allowed within the manufacturing tolerance range, and subsequently, the extent to which the cut surface of a specific inspection target terminal deviates from this reference shape can be numerically calculated.

[0108] In one embodiment of the present invention, the amount of curvature from a reference straight line calculated by the control unit can be calculated for an outer section defined as a straight section of a frame or rear case among a single cut surface outline. Specifically, the control unit may first distinguish the upper frame section, lower frame section, left frame section, and right frame section among the restored cut surface outlines, and set a straight line connecting the two endpoints of each section as the reference straight line for that section. Subsequently, the control unit may calculate the vertical distance from each outline point constituting the section to the reference straight line and record the sign of that distance. For example, a point protruding convexly outward from the reference straight line may be recorded as a positive directional deviation value, and a point recessed inward from the reference straight line may be recorded as a negative directional deviation value. Afterward, the control unit may calculate the maximum value of the vertical distance, the average value, the length of the continuous section, and the consistency of the directional sign for each outline section. Here, the amount of curvature is not determined simply by the maximum vertical distance, but may be determined by reflecting the average and maximum values ​​of the vertical distances accumulated in the same direction throughout the entire outline section. For example, if outward deviation continues in more than 80% of the upper frame section, with a maximum vertical distance of 0.35 mm and an average vertical distance of 0.21 mm, the control unit may record this as an outwardly bent section. The reference straight line and allowable deviation range can be determined using the vertical distance distribution collected for the same section from a normal terminal group, and can be stored to be processed as an actual deformation only when it exceeds the upper range of the maximum deviation value appearing in the normal terminal group.

[0109] In one embodiment of the present invention, the spacing deviation between two opposing outlines can be calculated using the change in distance between opposing sections of the outlines of the cutting plane. The control unit can select two opposing outline sections, such as an upper frame and a lower frame, or a left frame and a right frame, from a single cutting plane, and then calculate the minimum distance from each point of one section to the opposite section. For example, in a thickness-direction cutting plane, the front outline and the rear outline may be opposing sections, and in a length-direction cutting plane, the left frame and the right frame may be opposing sections. The control unit can organize the opposing distances calculated in this way into fixed positional intervals within the cutting plane and calculate the deviation by comparing them with the distance between reference cross-section outlines at the same position. If the average opposing distance at the same cross-sectional location of a normal terminal group is 7.20 mm with an allowable dispersion range of ±0.08 mm, whereas the terminal under inspection repeatedly measures larger values ​​such as 7.58 mm, 7.60 mm, and 7.55 mm at the same location, the control unit can identify that the gap deviation between outer sections at that location has expanded. Conversely, if a smaller value is repeatedly measured at a specific location, it can be recorded as a reduction deviation caused by compression or pressing. By calculating the gap deviation by location in this way, it is possible to precisely verify which sections have locally widened or narrowed, rather than simply checking whether the overall width or thickness has changed.

[0110] In one embodiment of the present invention, the reduction amount of the corner angle can be calculated directly from the outline of the corner section. The control unit identifies the upper-left corner where the upper frame and the left frame meet, the upper-right corner where the upper frame and the right frame meet, the lower-left corner where the lower frame and the left frame meet, and the lower-right corner where the lower frame and the right frame meet, respectively, among the outlines of the cut surface, and can extract tangent vectors in two directions where the frame straight section begins in each corner section. At this time, the tangent vectors can be calculated using outline points located a certain distance from the corner center point, and in the case of a terminal with rounded corners, the direction of each straight section can be determined based on the starting and ending points of the round. Subsequently, the control unit can calculate the angle between the two tangent vectors using atan2() or cos^-1() and obtain the reduction amount of the angle by comparing it with the reference corner angle of a normal terminal group of the same model. For example, if the average angle of the top-left corner of a normal terminal group is 89.6 degrees and the allowable range is 88.8 to 90.4 degrees, while 84.1 degrees is calculated at the same corner of the terminal under inspection, the control unit can record it as a corner deformation reduced by 5.5 degrees. In addition, the control unit can check whether the angle reduction at the same corner is repeated across adjacent cross-sections corresponding to the vicinity of the corner, rather than simply checking the angle only once on a single cross-section. Therefore, it is possible to identify structural deformation formed by actual corner compression or denting, rather than angle changes caused by temporary image noise or outline reconstruction errors.

[0111] In one embodiment of the present invention, the deviation in separation between the display boundary line and the frame boundary line can be calculated as a value for determining whether the panel is lifted. To this end, the control unit can identify the location where the display boundary line passes and the location where the inner boundary line of the frame passes through, respectively, in a single cut plane. For example, in a cut plane facing the front, the end line of the glass panel and the inner edge line of the frame may appear parallel to each other; therefore, the control unit can calculate the shortest distance between the two boundary lines on the plane of the cut plane and store it as a separation value. Subsequently, the control unit can calculate the deviation in separation by comparing this separation value with a reference separation value at the same cross-sectional location of a normal terminal group. If, in a normal terminal group, the average separation value between the display boundary line and the frame boundary line in the upper left 10 mm section is 0.42 mm with an allowable range of ±0.05 mm, but in the terminal under inspection, repeated measurements such as 0.61 mm, 0.64 mm, and 0.63 mm are taken in the same section, the control unit can determine that there is a high possibility that the panel is lifted away from the frame in the said section. Conversely, if the offset value becomes excessively small, the panel may be recorded with the panel shifted to one side or the frame pressed inward. By calculating the offset deviation for each cross-sectional location in this way, it is possible to distinguish whether the panel lifting occurs across the entire top section, only in the section adjacent to the right corner, or locally only near the camera placement area.

[0112] In one embodiment of the present invention, the procedure for identifying frame bending when the calculated amount of curvature accumulates in the same direction across three or more consecutive virtual cut planes may be a procedure for distinguishing between actual deformation and temporary errors. For each of the cut planes arranged in longitudinal or transverse order, the control unit records the direction of curvature and the amount of curvature of the upper frame, lower frame, left frame, or right frame section of the corresponding cut plane, and then checks whether the direction of curvature of the same frame section matches between adjacent cut planes. For example, if longitudinal cross-sections are arranged sequentially from top to bottom, and outward curvature is detected in the right frame section of the 6th, 7th, and 8th cross-sections, and the average vertical distance values ​​of each cross-section increase continuously to 0.18 mm, 0.24 mm, and 0.27 mm, the control unit may identify this as outward bending of the right frame. Conversely, if a large deviation appears only in a single cut plane while maintaining a normal range in the front and back cut planes, it may be treated as a temporary outline reconstruction error or a local dent. Accordingly, when determining frame bending, the control unit may use as determination conditions whether bending in the same direction is confirmed in at least three consecutive cut surfaces and whether the amount of bending is accumulated or maintained. Here, the number of consecutive cut surfaces and the criteria for accumulating the amount of bending can also be determined by comparing a sample of a terminal with actual bending and a sample of a normal terminal, and can be determined as a combination of conditions that rarely appear in a normal terminal but repeatedly appear in a terminal with actual bending.

[0113] In one embodiment of the present invention, the procedure for identifying panel lifting when the separation deviation is repeated in a section adjacent to the display boundary line may be a procedure for verifying whether local separation is detected by chance in only one cross-section or whether the actual panel separation phenomenon continues in the longitudinal direction. The control unit may select a section where the display boundary line exists among the cut surfaces arranged along the longitudinal direction of the first user terminal, and record the separation deviation measured at the upper, lower, left, or right frame adjacent portions for each cut surface. Subsequently, the control unit may inspect whether the separation deviation repeatedly exceeds the normal range in adjacent cut surfaces along a specific side. For example, if a separation deviation expanded by 0.15 mm or more is detected across four consecutive cut surfaces in the upper left corner adjacent portion, the control unit may identify this as an upper left panel lifting section. In addition, the control unit may increase the reliability of the panel lifting determination by also checking whether the height of the front glass boundary on the point cloud in the same section rises slightly from the rear reference plane. Conversely, if the deviation appears in only one cross-section or changes irregularly in different directions, it can be treated as a shooting error or a boundary line detection error.

[0114] In one embodiment of the present invention, the procedure for identifying corner deformation when the amount of reduction in the corner angle is concentrated in the corner section can be performed using the angle change distribution of the cutting surface section corresponding to each corner. The control unit organizes the amount of reduction in the corner angle for each of the top-left, top-right, bottom-left, and bottom-right corners in the order of the cutting surfaces and can check whether the reduction appears repeatedly only in the same corner. For example, if a value reduced by 4 degrees or more from the reference corner angle is continuously detected in three or more cutting surfaces among the cutting surfaces corresponding to the bottom-right corner, while maintaining a normal range in other corner sections, the control unit can identify this as bottom-right corner deformation. In addition, the control unit can verify whether the deformation is formed by actual corner indentation or denting rather than simple outline extraction noise by additionally checking whether a spacing deviation between opposing outlines and local color degradation exist together in the cutting surfaces where the angle reduction occurred. Therefore, the identification of corner deformation is not determined solely by the angle reduction value calculated from a single cross-section, but can be confirmed through a combination with the cutting surface distribution across the entire corner section and other shape indicators.

[0115] In one embodiment of the present invention, a cross-sectional distortion map in which the occurrence sections of identified frame bending, panel lifting, and corner deformation are arranged according to the longitudinal order of the first user terminal can be generated by aligning the deformation type and deformation amount detected for each cut plane on the terminal's longitudinal axis. Specifically, the control unit can set the longitudinal coordinates from the upper reference point to the lower reference point as the reference axis and map which position each cut plane corresponds to on this reference axis. Subsequently, for each cut plane, the control unit records values ​​such as the amount of left frame bending, the amount of right frame bending, the deviation from the upper or lower separation, and the amount of angle reduction of each corner together, and if the corresponding value exceeds the reference range, it can generate a distortion indication corresponding to the position of the cut plane. For example, if the outer bending of the right frame is continuously detected in the cut planes from the 12 mm point to the 28 mm point in the longitudinal direction, the control unit can display the right frame distortion section as a single continuous block on the cross-sectional distortion map, and if the upper panel lifting is repeatedly detected in the upper 0 mm to 6 mm section, the upper panel lifting section can be displayed as a separate pattern. In addition, if the reduction in corner angles is concentrated in the three cross-sections at the bottom right, a corner deformation indicator can be superimposed near the bottom end in the longitudinal direction. As such, the cross-section distortion map is not a simple text warning list, but rather a map-shaped display information that allows one to check at a glance what type of deformation has occurred at which location along the longitudinal direction of the first user terminal.

[0116] In one embodiment of the present invention, the control unit may display the cross-sectional distortion map together with the appearance inspection model on the display unit. For example, on one side of the display unit, a 3D appearance inspection model in which a point cloud and defect locations are superimposed may be rotatably displayed, and on the other side, a cross-sectional distortion map arranged along a longitudinal axis may be displayed together. When a user selects a specific frame section or corner section on the appearance inspection model, the control unit may highlight the cross-sectional distortion map corresponding to the selected location and display the actual values ​​and reference ranges for the curvature amount, gap deviation, corner angle reduction amount, and separation deviation for each corresponding cross-sectional distortion map. Conversely, when a user selects a specific distortion section on the cross-sectional distortion map, the control unit may highlight the 3D location corresponding to that section on the appearance inspection model and call up the original image that served as the basis for detection or the result of restoring the cross-sectional outline. Through such linked indicators, the inspector can not only see results indicating that the frame is bent or the panel is lifted, but also immediately verify which measurements exceeded the standards at which cut surface, how those measurements extend along the length, and which location on the actual exterior model corresponds to those results.

[0117] In one embodiment of the present invention, the reference values ​​for the curvature amount, gap deviation, corner angle reduction amount, and separation deviation, as well as the identification conditions for frame bending, panel lifting, and corner deformation, can be determined through a verification procedure using actual terminal samples. Specifically, the control unit may secure a normal terminal group, a terminal group with a frame intentionally bent slightly, a terminal group with a part of the panel lifted slightly, and a terminal group with corner dents, respectively, and repeatedly perform rotational imaging and reference measurements for each terminal. Reference measurements may be performed by acquiring the actual frame bending amount, actual panel lifting height, and actual corner angle deformation amount using a laser profiler, digital micrometer, gap gauge, angle measuring jig, etc. Subsequently, the control unit may calculate the average error, maximum error, omission judgment rate, and false detection rate by comparing the curvature amount, gap deviation, corner angle reduction amount, and separation deviation calculated from the restored cut surface outline with the actual measured values. Using these comparison results, the control unit may store a reference range for each model in which the normal terminal group is maintained normally while the deformed terminal group is accurately separated. Accordingly, the procedure for calculating shape indicators by cross-section, identifying frame bending, panel lifting, and corner deformation, generating cross-sectional distortion maps, and displaying linked with the appearance inspection model included in the above sentence can be implemented based on actual sample measurement data and repeated shooting results, and can be specified to the extent that a person skilled in the art can perform it as is.

[0118] A control unit according to one embodiment of the present invention lowers the appearance grade and identifies it as a target for re-inspection if the section where frame bending occurs in the cross-sectional distortion map is continuous up to the section adjacent to the display boundary line, if panel lifting is repeated in three or more virtual cutting planes continuous along one side of the first user terminal, or if corner deformation is simultaneously identified in two or more of four corner sections.

[0119] In one embodiment of the present invention, after a cross-sectional distortion map is generated, the control unit can determine whether to maintain the appearance grade again using the distortion type, distortion location, cutting plane order, and distortion amount recorded in the cross-sectional distortion map. To this end, the control unit may first record which item among frame bending, panel lifting, and corner deformation has been identified for each virtual cutting plane, which side or corner of the first user terminal corresponds to the location where the identified item occurred, and how much the distortion amount of the item exceeds the standard range obtained from the normal terminal group. For example, for each of the first to nth cutting planes arranged in longitudinal order, the presence of distortion and the amount of distortion for the left frame, right frame, upper frame, lower frame, upper-left corner, upper-right corner, lower-left corner, and lower-right corner can be organized in a table form, and this table can be stored in the cross-sectional distortion map in the order of the cutting planes. When the distortion results for each cross-section are organized together with location information in this way, the control unit can distinguish whether the distortion is temporary in a single cross-section or structurally continuous across one side or multiple corners of the terminal.

[0120] In one embodiment of the present invention, the control unit may first determine the positional continuity between the section of the cut surface identified as frame bending and the section of the cut surface identified as the section adjacent to the display boundary line in order to determine whether the section where frame bending occurs continues to the section adjacent to the display boundary line. Here, the section adjacent to the display boundary line may be an area within a predetermined distance from the display boundary line in the direction of the frame from each cut surface. For example, after measuring the average separation value and dispersion range between the display boundary line and the inner boundary line of the frame in a normal terminal group, the section adjacent to the display boundary line may be set to a certain distance outward from the average separation value. The control unit may record the start and end points of the outer section where frame bending is identified for each cut surface, and verify whether the recorded section overlaps with the section adjacent to the display boundary line within the same cut surface or continues without interruption from adjacent cut surfaces. More specifically, if frame bending is identified as an outward curvature of the right frame on a certain cut surface, and the inner boundary of the right frame on the same cut surface is included in the section adjacent to the display boundary line, and this condition appears continuously on adjacent cut surfaces, the control unit may determine that it is not merely a local bending of the outer frame but a continuous bending that can affect the display mounting structure. In this case, continuity can be determined by whether the distortion is maintained as a single connected section between adjacent cut surfaces in the longitudinal or transverse direction on the cross-sectional distortion map, and even if there is a missing cut surface in the middle, if the curvature in the same direction is maintained on the cut surfaces before and after it and the amount of distortion of the missing cut surface is near a boundary value below the standard, it can be treated as a single continuous section. Conversely, if frame bending appears briefly only on the outer surface opposite the display boundary line and does not extend to the section adjacent to the display boundary line, it may be maintained as a general defect rather than being immediately reflected as a reason for downgrading the appearance grade.

[0121] In one embodiment of the present invention, whether panel lifting is repeated in three or more consecutive virtual cutting planes along one side of the first user terminal can be determined by using whether the separation deviation exceeds the same direction for the cutting plane positions corresponding to the one side. To this end, the control unit can arrange the separation deviations between the display boundary line and the frame boundary line in the order of cutting planes for each of the left side, right side, upper side, and lower side. Subsequently, the control unit can check whether the separation deviation repeatedly exceeds the normal range in consecutive cutting planes based on a specific side. For example, if a separation deviation greater than the reference value is detected in the 11th, 12th, and 13th cutting planes among the cutting planes arranged along the right side, and the values ​​exceed the upper limit of the normal terminal group as 0.14 mm, 0.18 mm, and 0.16 mm, respectively, the control unit can determine that panel lifting is repeated across three consecutive cutting planes on the right side. Here, the condition of having three or more consecutive cut surfaces is not set arbitrarily; rather, it can be confirmed by reflecting verification results after photographing a group of actual normal terminals and a group of sample terminals with actual panel lifting, which indicate that while deviations appearing in only one or two cut surfaces may occur temporarily due to boundary detection errors or changes in lighting, deviations that are repeated along the same side in three or more consecutive cut surfaces appear stably only in actual lifted samples. Furthermore, the control unit may not simply look at the number of cut surfaces, but may also examine whether the direction of the deviation in the three or more cut surfaces corresponds to spreading outward toward the frame, or to what extent the deviation value of each cut surface exceeds the normal upper limit. Therefore, values ​​that are temporarily measured significantly in only one cut surface, or irregular deviations with different directions, are not confirmed as panel lifting, and only the actual state of the panel spreading along the same side can be reflected as a reason for re-inspection.

[0122] In one embodiment of the present invention, whether corner deformation is simultaneously identified in two or more of the four corner sections can be determined by using the angle reduction amount and the shape change amount of the cutting surface group corresponding to each corner together. For each of the top-left corner, top-right corner, bottom-left corner, and bottom-right corner, the control unit can organize the reduction amount of the corner angle, whether the outline of the corner-adjacent section is sunken, the deviation in the spacing of opposing outlines near the corner, and the local depth reduction on the point cloud into a single corner-specific determination information. Subsequently, among the corner-specific determination information, the control unit can determine a corner as a deformed corner only when the reduction amount of the corner angle in a plurality of adjacent cutting surfaces including the corner center all exceeds a reference value and the reduction amount appears concentrated in a specific corner. For example, if the angle reduction amounts in the three adjacent cut surfaces at the top-left corner are 4.8 degrees, 5.1 degrees, and 4.6 degrees, respectively, and in the three adjacent cut surfaces at the bottom-right corner are 3.9 degrees, 4.3 degrees, and 4.0 degrees, while the remaining two corners maintain a normal range, the control unit may determine that corner deformation has been identified simultaneously in the two corners, the top-left corner and the bottom-right corner. Here, "identified simultaneously" does not mean only that they appear simultaneously at the same cut surface number, but may also mean a state in which at least two of the four corners each satisfy the deformation determination criteria at the same inspection point for a single terminal subject to inspection. Furthermore, if corner deformation is identified simultaneously in two or more places, the control unit may downgrade the appearance grade by considering that it is highly likely that the entire appearance of the terminal was affected by dropping, crushing, or twisting, rather than a local dent caused by a simple single impact.

[0123] In one embodiment of the present invention, the downward adjustment of the appearance grade is not performed merely because a single type of defect is detected, but only when the structural distortion conditions reflected in the cross-sectional distortion map satisfy certain criteria. To this end, the control unit may store an existing appearance grade system in advance and store the acceptable defect range for each grade in a reference table. For example, in the upper grade representing a normal level of appearance, general defects such as surface scratches or fine dents are allowed, but the upper grade may be set so that it cannot be maintained if frame bending continues to the section adjacent to the display boundary line, if panel lifting is repeated in three or more cut planes along one side, or if corner deformation is simultaneously identified in two or more of four corners. Accordingly, if any one of the three conditions is satisfied, the control unit may lower the currently calculated appearance grade by at least one step, and if two or more conditions are satisfied simultaneously, it may lower it by two or more steps or immediately switch to a sales hold grade. The extent of the downgrade in appearance grade is not determined arbitrarily, but can be finalized through cumulative analysis that correlates the history of terminals returned or subject to customer dissatisfaction after the actual transaction is completed with the cross-sectional distortion map from the time of inspection. For instance, terminals can be divided into groups with no structural distortion conditions, a group with only one structural distortion condition, and a group with two or more structural distortion conditions. By comparing the return rate and appearance-related claim rate after shipment among these groups, the extent of grade downgrade for each combination of conditions can be determined based on the comparison results. Therefore, the downgrade of appearance grade can be performed based on criteria that link actual inspection data with post-transaction data, rather than through simple subjective judgment.

[0124] In one embodiment of the present invention, the procedure for identifying a target for re-inspection may be a procedure that initiates a subsequent inspection procedure immediately after the appearance grade is downgraded. If any of the structural distortion conditions are satisfied and the appearance grade is downgraded, the control unit may separate the corresponding first user terminal from the general sales waiting list and register it in the re-inspection list. At this time, the control unit may record the model name, serial number, inspection time, distortion type, range of the distortion-occurring cut surface, the highlighted section on the cross-sectional distortion map, and original image reference information of the terminal together in the re-inspection target identification information. Subsequently, while displaying a message requiring re-inspection on the display unit, the control unit may specifically indicate which condition corresponds to the reason for re-inspection among a continuous frame bending section, a repeated panel lifting section, or two or more corner deformations. For example, the reason for re-inspection can be directly indicated, such as “right frame bending continues to the section adjacent to the display boundary,” “top panel lifting is repeated in four cut planes,” or “simultaneous deformation of the top-left and bottom-right corners,” and the cut plane outline corresponding to the reason, the section distortion map highlighting section, and the display location of the appearance inspection model can be called together. Once the target for re-inspection is identified in this way, the inspector can immediately determine which section needs to be magnified and checked first, and if necessary, can perform additional re-photography, manual measurement, or component disassembly inspection on the same terminal.

[0125] In one embodiment of the present invention, the continuous frame bending condition, the repeated panel lifting condition, and the simultaneous multiple corner deformation condition can be determined through a verification procedure using actual terminal samples. The control unit can generate a cross-sectional distortion map by preparing a normal terminal group, a terminal group with frame bending, a terminal group with panel lifting, a terminal group with two or more corner deformations, and a terminal group with complex structural deformation, respectively, and then applying the same shooting and cross-sectional restoration procedure. Subsequently, by comparing the results of manual measurement of the actual external condition of each sample terminal by an operator with the automatic judgment results of the cross-sectional distortion map, the control unit can accumulate how the continuous condition appears only in actual cases where frame bending extends to the section adjacent to the display boundary line, how the separation deviation is repeated in the minimum number of continuous cut planes in actual cases with panel lifting, and how multiple corner deformations are distributed in terminals with an actual history of dropping or crushing. Based on these accumulated results, the control unit can select a combination of conditions where structural distortion conditions are rarely established in the normal terminal group and are stably repeated in the actual structurally deformed terminal group, and save it as a re-inspection standard. Accordingly, the procedure for identifying targets for appearance grade downgrade and re-inspection included in the above sentence can be performed according to standards established based on actual sample terminal shooting data, manual measurement data, and subsequent transaction history, and can be implemented specifically enough that a person of ordinary skill can carry it out as is.

[0126] A control unit according to one embodiment of the present invention performs an identity verification procedure in which a feature point correspondence principle used in fingerprint identification is applied by substitution, wherein at the time of purchase, the time of registration for sale, and the time immediately before shipment, it obtains enlarged exterior images including the perimeter of the camera placement area, the arrangement of speaker holes, the perimeter of the charging terminal, the button boundary line, and the distribution of fine scratches on the side frame, and extracts the start and end points of linear scratches, the inflection points of curved scratches, the center points of the hole arrangement, the corner bending points, and the discontinuity points of the paint boundary line as feature points from each enlarged exterior image.

[0127] In one embodiment of the present invention, the identity verification procedure performed by the control unit may be a principle applied to the identification of the appearance of the first user terminal, wherein stable reference points, such as break points, branching points, and endpoints, are extracted from repeated images of the same fingerprint in fingerprint identification, and then the maintenance of such reference points is verified between each shooting time. However, in this embodiment, the control unit does not use the ridge pattern of the fingerprint itself, but rather secures an enlarged appearance image based on the distribution of fine shapes or fine scratches that maintain the same location even during repeated shooting of the appearance of the first user terminal, and can extract reference feature points to be used for identity verification from the enlarged appearance image. When the control unit is configured to secure fine features of the appearance in stages in this manner, basic data can be provided for identifying the unique external traces of a specific object among multiple terminals of the same model group that are difficult to distinguish using only general information such as model name, color, and serial number.

[0128] In one embodiment of the present invention, the control unit may obtain enlarged exterior images at the time of purchase, the time of registration for sale, and the time immediately before shipment. Here, the time of purchase may be the time when the first user terminal is first received and the exterior inspection begins; the time of registration for sale may be the time when the exterior inspection and physical inspection are completed and the first user terminal is converted into inventory available for sale; and the time immediately before shipment may be the time when the final state is checked before packaging or delivery after the purchase is confirmed. The control unit may control the same area to be photographed at each time point with the same magnification ratio, the same shooting angle, and the same lighting conditions, and to this end, a magnification shooting device including a camera for magnification shooting, a macro lens, a diffused lighting unit, and a position fixing jig may be used. For example, after setting the perimeter of the camera placement area, the arrangement of speaker holes, the perimeter of the charging terminal, the button boundary line, and the side frame as shooting targets, the first user terminal may be fixed to a mounting jig, and while positioned so that each shooting area is at the center of the lens, it may be photographed at a magnification ratio of 5 to 20 times. In addition, the control unit may provide a fixed distance maintaining spacer in the lens housing so that the distance between the camera and the first user terminal does not change each time the same area is repeatedly photographed, or control the system so that shooting occurs only after autofocus is completed.

[0129] In one embodiment of the present invention, an enlarged exterior image of the perimeter of the camera placement area may be captured so that the outer boundary of the rear camera module, the starting point of the protrusion of the camera housing, the metal or resin border around the lens window, and fine scratches or processing marks formed along the perimeter are sufficiently distinguishable. For example, the control unit may sequentially illuminate from at least some of the up, down, left, right, and diagonal directions relative to the center of the camera placement area to obtain a plurality of original enlarged images so that the directionality and reflection pattern of the fine scratches are revealed, and among them, the image in which the boundary line and the fine scratches are most clearly revealed may be adopted as the enlarged exterior image. An enlarged exterior image of the speaker hole arrangement may be captured so that the centers of a plurality of holes, the spacing between holes, whether the edges of the holes are dented, and whether there is fine paint damage around the perimeter of the holes are revealed. An enlarged exterior image of the perimeter of the charging terminal may be captured so that the upper and lower boundaries of the terminal opening, the left and right corners, and fine dents and insertion wear marks around the terminal are distinguishable. An enlarged external image of the button boundary line can be captured so that the starting point, ending point, linear boundary of the button gap of the power button or volume button, and the distribution of fine scratches formed on the frame adjacent to the button can be identified. An enlarged external image of the side frame can be captured so that the straight edge of the frame, the corner bend section, the paint boundary line, and the distribution of fine scratches formed along the longitudinal direction are revealed.

[0130] In one embodiment of the present invention, the control unit may generate reference shooting coordinates before obtaining each enlarged exterior image. For example, the control unit may first record reference coordinates of the camera placement area, speaker hole arrangement, charging terminal perimeter, button boundary line, and side frame in the overall exterior inspection model or point cloud of the first user terminal, and during enlarged shooting, move a jig or stage so that the actual position corresponding to the reference coordinates is at the center of the lens. At this time, to reduce the error from the reference coordinates at each shooting point, the control unit may receive a low-magnification preview image immediately before enlarged shooting, identify the outline of the camera placement area, the two end holes of the speaker hole arrangement, the left and right end points of the charging terminal opening, the start and end points of the button boundary line, or the corner bending point of the side frame as temporary reference points in the preview image, and then finely move the shooting stage until the temporary reference points match the pre-stored target position. Accordingly, enlarged exterior images taken at the time of purchase, the time of registration for sale, and the time immediately before shipment, respectively, can be obtained to reproduce the same exterior parts within almost the same field of view.

[0131] In one embodiment of the present invention, the control unit can extract feature points from each enlarged exterior image, including the start and end points of linear scratches, the inflection points of curved scratches, the center points of hole arrangements, corner bending points, and the discontinuity points of paint boundary lines. To this end, the control unit can first perform lens distortion correction, brightness leveling, and reflection suppression processing on the enlarged exterior image. Lens distortion correction can be performed using the positional difference of grid intersection points in an image captured of a reference pattern plate corresponding to the enlargement imaging device, and brightness leveling can be performed using the average brightness difference between the current enlarged exterior image and a reference background image obtained with only the diffuse lighting unit turned on. Reflection suppression processing can be performed by removing patterns in which only the brightness changes rapidly while the actual boundary position is maintained among multiple original images of the same area captured from different lighting directions. Once such preprocessing is completed, the control unit becomes able to distinguish between fine boundaries fixed to the actual exterior and reflection patterns that vary depending on the lighting.

[0132] In one embodiment of the present invention, the start and end points of a linear defect may be extracted from the two end points of a color or brightness discontinuity section that extends in the form of a thin line with a width of less than a certain width within an enlarged exterior image. The control unit first detects linear candidates in which the color difference and brightness difference between adjacent pixels in the enlarged exterior image are greater than or equal to a certain standard, and can verify whether the linear candidates are continuous along adjacent pixels. Subsequently, the control unit aligns the points constituting the linear candidates in the length direction and records the foremost point and the last point as the start and end points, respectively. At this time, the control unit may adopt a linear defect only if the total length of the linear candidate is greater than or equal to a pre-verified minimum length and the width of the linear candidate is significantly smaller than the minimum length. For example, after securing multiple enlarged exterior images of a normal terminal of the same model and an enlarged exterior image of a terminal with actual defects, the length, width, and continuity conditions of the automatically extracted linear candidates are adjusted based on verification correct data in which an operator manually marks the start and end points of the actual linear defects, thereby determining a combination of conditions that minimizes false detection and minimizes the omission of actual defects. When the starting and ending points of a linear scratch are extracted according to the verified conditions as described above, a reference point is established that allows for the repeated comparison of the positions of both ends of the linear scratch even when the same area is re-photographed.

[0133] In one embodiment of the present invention, the inflection point of a curved scratch can be extracted from a point where the direction of travel changes within a scratch path that continues in a curved shape. After detecting a candidate for a curved scratch, the control unit can select sample points at regular intervals along the candidate curve and calculate the change in the tangent direction connecting adjacent sample points. Subsequently, the control unit records a point where the tangent direction changes beyond a certain range as a candidate for an inflection point, and among them, the point where the amount of change in direction is greater than the average of adjacent sections can be adopted as the final inflection point. For example, if a curved micro-scratch formed on a side frame continues in a straight line and then bends downward at a specific point, the control unit can identify the point of direction change as the inflection point of the curved scratch. The inflection point of the curved scratch can also be determined by comparing verification data manually marked by an operator with the results of automatic extraction, determining how much change in direction to consider as an inflection point and when the minimum number of consecutive sample points must exist to allow the determination of an inflection point. Therefore, the inflection point is not arbitrarily selected by simple noise, but can be identified according to conditions where the actual change in the direction of travel of the curved scratch is verified.

[0134] In one embodiment of the present invention, the center point of a hole array can be extracted by calculating the center of each hole in an area where a plurality of micro-openings are repeatedly arranged, such as a speaker hole array or a microphone hole array. The control unit may first detect dark opening candidates surrounded by a circle or an ellipse in an enlarged external image and extract a plurality of boundary points along the outer boundary of each candidate. Subsequently, the control unit may calculate the center point of each hole using the average coordinates of the boundary points or the center coordinates of the least squares fit circle. In addition, the control unit may check whether the plurality of hole center points are arranged at regular intervals and may adopt them as final feature points only if the center points of the holes at both ends and the middle of the hole array both satisfy the normal arrangement conditions. For example, if the speaker hole array is a straight line arrangement, it can be checked whether the plurality of hole center points exist within an allowable deviation range from a single reference straight line, and if it is a curved arrangement, the normal arrangement can be checked using the deviation from the reference curve. Since the hole center point is a reference point whose position hardly changes during the exterior manufacturing process, if the same hole center is repeatedly extracted at the time of purchase, the time of registration for sale, and the time immediately before shipment, it can be used to re-identify the same exterior part.

[0135] In one embodiment of the present invention, a corner bend point can be extracted from a point where the outer direction changes at a corner section where the side frame and the upper or lower frame are connected, or at a corner section of a camera placement area protrusion. The control unit tracks points along the frame outline or protrusion boundary line in an enlarged exterior image and sets a point where the amount of change in direction of an adjacent section exceeds a certain standard as a bend point candidate. Subsequently, the length of the line segment before and after the bend point candidate, the duration of the change in direction, and the change in curvature near the bend point are examined together to select only the point where the actual bend occurs as the final bend point. For example, if the side frame bends gently at the top corner, the point with the greatest curvature becomes the bend point center, and the start and end positions of the section where the curvature increases above a certain standard around it can be stored together. The reason bend points are useful for identity verification is that their relative position is stably maintained even within the manufacturing tolerance range, and if deformation occurs due to impact, the change appears relatively clearly in the enlarged image. Therefore, the control unit can record the bending direction, the length of the preceding and succeeding line segments, and the curvature value near the bend, along with the bending point coordinates.

[0136] In one embodiment of the present invention, the discontinuity point of the paint boundary line can be extracted from a point where the paint color or surface finish is interrupted and does not continue continuously around the frame or camera placement area of ​​the first user terminal. The control unit first calculates the average color and average reflection values ​​of the parts where the same material and the same paint area are connected in the enlarged exterior image, and can detect boundary candidates that deviate abruptly from the average values. Subsequently, if the boundary candidate continues for a certain length or longer and then suddenly breaks at a specific point, or if the direction of the paint boundary line ends at that point, the control unit can adopt that point as the discontinuity point of the paint boundary line. For example, if fine dents or wear occur on the paint portion of the side frame, the paint boundary may not continue linearly but may appear as a form that is cut off in the middle or bypassed; the control unit can record that very point of interruption as a discontinuity point feature. Furthermore, if the discontinuity point is repeatedly detected at adjacent rotation angles or adjacent shooting times, it can be determined as an actual trace of paint discontinuity rather than an accidental change in lighting.

[0137] In one embodiment of the present invention, the control unit may store not only simple coordinates for each feature point extracted as described above, but also the type of feature point, the external area to which the feature point belongs, the magnification ratio, the time of shooting, rotation angle or camera attitude information, and partial image patches around the feature point. For example, for the starting point of a linear scratch, additional information such as “linear scratch starting point,” “side frame,” “time of registration for sale,” “right side,” and “surrounding 31×31 pixel patch” may be recorded together, and for the center point of a hole array, additional information such as “speaker hole center point,” “lower frame,” “time of purchase,” “front bottom,” and “surrounding 21×21 pixel patch” may be recorded together. By storing the feature point surrounding patches and additional information together in this manner, preparations are made to reliably determine whether the location and surrounding pattern of the same type of feature point are maintained not only in comparisons between the same time points but also in comparisons between different time points.

[0138] In one embodiment of the present invention, the conditions for securing enlarged exterior images and extracting feature points can be determined through a verification procedure using actual sample terminals. The control unit prepares multiple groups of terminals of the same model with normal exteriors, terminals with fine linear scratches, terminals with curved scratches, terminals with a partially deformed hole arrangement, terminals with corner deformation, and terminals with paint discontinuity, respectively, and can perform repeated shooting assuming the time of purchase, the time of registration for sale, and the time immediately before shipment. Subsequently, an operator can generate correct data by manually marking the start and end points of actual linear scratches, the inflection points of actual curved scratches, the center points of each hole, the actual bending points, and the discontinuity points of the paint boundary lines in each enlarged exterior image. The control unit calculates the positional difference, the number of omissions, and the number of false detections between the automatically extracted feature points and the correct data, respectively, and can perform repeated comparisons while applying different magnification ratios, lighting intensity, minimum linear length, criteria for change of direction, hole center tolerance, criteria for bending point curvature, and criteria for paint discontinuity length. Subsequently, the control unit can store the condition combination closest to the actual correct answer data as the feature point extraction condition for the corresponding model group. Accordingly, the operation of the control unit securing enlarged exterior images including the perimeter of the camera placement area, the arrangement of speaker holes, the perimeter of the charging terminal, the button boundary line, and the distribution of fine scratches on the side frame at the time of purchase, the time of registration for sale, and the time immediately before shipment, and extracting the start and end points of linear scratches, the inflection points of curved scratches, the center points of the hole arrangement, the corner bending points, and the discontinuity points of the paint boundary line as feature points from each enlarged exterior image, can be configured specifically enough so that a person skilled in the art can implement it exactly as is based on actual shooting samples and manual verification data.

[0139] A control unit according to one embodiment of the present invention generates a unique appearance pattern for each terminal by combining the relative distance between extracted feature points, the pinch angle with adjacent feature points, the arrangement order of feature points, and the location of repeated appearance within the same area; identifies whether the exterior housing has been replaced or parts have been replaced by comparing the unique appearance pattern at the time of purchase with the unique appearance pattern at the time of registration for sale; and identifies whether new damage has occurred during the storage or packaging process by re-comparing the unique appearance pattern at the time of registration for sale with the unique appearance pattern immediately before shipment.

[0140] In one embodiment of the present invention, the control unit may classify the previously extracted linear scratches, inflection points of curved scratches, center points of hole arrangements, corner bending points, and discontinuity points of paint boundary lines into the same shooting area, and then generate a unique appearance pattern for the first user terminal using the relative positional relationship between feature points within each shooting area. To this end, the control unit may first divide the perimeter of the camera placement area, the speaker hole arrangement, the charging terminal perimeter, the button boundary line, and the side frame into different reference areas, place the extracted feature points for each reference area onto an image coordinate system, and then select one or more reference feature points that serve as centers within the same reference area. For example, in the speaker hole arrangement, the center point of the two end holes or the center point of the central hole may be set as the reference feature point; in the button boundary line, either the button start point or the end point may be set as the reference feature point; and in the perimeter of the camera placement area, the lens center point or the corner point of the camera housing may be set as the reference feature point. The control unit can record together the relative distance from the reference feature point to the remaining feature points, the angle between the line segments connecting the reference feature point and the adjacent feature points, the arrangement order in which feature points appear along the outer edge, and the positional relationship in which the same type of feature points appear repeatedly within the same area.

[0141] In one embodiment of the present invention, the relative distance may be calculated not as a simple absolute length value, but as the pixel distance between a reference feature point and a comparison target feature point in an enlarged exterior image obtained at the same magnification ratio, or as a corrected actual distance value. The control unit obtains the x-coordinates and y-coordinates of each feature point after completing lens distortion correction and magnification ratio correction, and can calculate the distance value between two feature points as sqrt((x2-x1)^2+(y2-y1)^2). However, even if the control unit assumes that a fixed jig is used to ensure that the shooting magnification and distance conditions are constant for each enlarged exterior image, in order to correct minute shooting errors, it may generate a ratio value for the reference length within the same area in addition to using only the absolute distance value. For example, in the case of a speaker hole array, the total length between the left end hole and the right end hole may be used as the reference length, and the distance between each hole center point and the left end hole may be converted into a ratio value. In the button boundary area, the length between the button start point and the end point is used as a reference length, and the position of the button adjacent scratch's start point and end point relative to the reference length can be converted into a ratio. By recording the relative distance together with the absolute distance and ratio value in this way, it becomes possible to compare identity despite minute differences in magnification that may occur when repeatedly photographing the same area.

[0142] In one embodiment of the present invention, the incline angle with adjacent feature points can be calculated as the difference in direction between two line segments connecting two feature points adjacent to a single feature point. The control unit may arrange the feature points for each reference area in a specific order and then designate the previous feature point and the next feature point as adjacent feature points for each feature point. Subsequently, the control unit may obtain the direction vector from the current feature point to the previous feature point and the direction vector from the current feature point to the next feature point, respectively, and calculate the incline angle between the two direction vectors using atan2() or cos^-1(). For example, since the inflection point of a curved scratch is a point where the direction of travel changes, the incline angle with adjacent feature points may appear significantly smaller or larger than that of a straight section of a linear scratch. Additionally, since the center points of a hole array are generally arranged in a nearly straight line, the incline angle with adjacent hole center points is maintained within a certain range, and if the edge of a specific hole is deformed due to impact or wear, the relative angle with surrounding points may change slightly. By storing these interlocking angle values ​​together with relative distance values, the control unit can determine whether the same set of feature points is not only similar in location but also has the same shape relationship with surrounding feature points.

[0143] In one embodiment of the present invention, the arrangement order of feature points may be information indicating the order in which feature points appear within each reference area. The control unit may align feature points along the boundary in a clockwise or counterclockwise direction starting from the top-left corner point of the camera housing around the camera placement area, arrange them in a rightward direction starting from the center point of the left end hole in the speaker hole arrangement, and arrange them from the top or left starting point to the bottom or right end point direction in the button boundary line. In the distribution of fine scratches on the side frame, they may be arranged from the top reference point to the bottom direction, or from the front reference point to the rear reference point direction. Since the arrangement order is information for comparing the order of appearance regardless of the absolute coordinates of the feature points, when verifying identity, it is possible to first check whether the same type of feature points are repeated in the same order in the unique pattern of the exterior at each shooting point. For example, in the speaker hole arrangement, the order of the hole center points basically does not change, but if the exterior housing is replaced and the hole arrangement shape or spacing changes, the order relationship between a specific hole center point and surrounding fine scratches may change. In addition, if the order of the paint break point and the scratch start point adjacent to the button boundary is reversed, it is more likely that the exterior part has been replaced or new damage has been added rather than a simple shooting error.

[0144] In one embodiment of the present invention, the repeated appearance location within the same area may refer to a relative position pattern in which feature points of the same or different types appear repeatedly within a specific reference area. For example, in the side frame area, a plurality of fine scratches extending in the longitudinal direction may appear repeatedly in mutually parallel positions, near the button boundary line, paint discontinuity points adjacent to both ends of the button gap may always appear at a constant relative position, and in the speaker hole array, hole center points may appear repeatedly at regular intervals. The control unit may set a reference axis in the longitudinal or perimeter direction along each reference area, and then convert and record where each feature point appears repeatedly on the reference axis into a ratio value or an index value. For example, after normalizing the total length of the speaker hole array to 100, the position of each hole center point may be recorded as a position value such as 0, 12.5, 25.0, and 37.5, and the distribution of fine scratches on the side frame may be recorded as 18.2, 21.4, and 46.8 according to the distance ratio from the top reference point. By storing the locations of repeated occurrences as normalized position values ​​in this way, the same pattern can be compared even if the shooting range shifts slightly when re-shooting the same area.

[0145] In one embodiment of the present invention, the control unit can generate a unique appearance pattern for each terminal by combining the relative distance, pinch angle, arrangement order, and repeated appearance location. To this end, the control unit can generate a feature point combination unit for each reference area, wherein the feature point combination unit may be a local structure comprising one reference feature point and two or more adjacent feature points around it. For each local structure, the control unit can generate a pattern block comprising the type of reference feature point, a list of relative distances from the reference feature point to adjacent feature points, a list of pinch angles formed with each adjacent feature point, an appearance order based on the reference feature point, a repeated appearance location value, and a partial image patch around the feature point. Subsequently, the control unit stores the pattern blocks generated for the perimeter of the camera placement area, the arrangement of speaker holes, the perimeter of the charging terminal, the button boundary line, and the side frame, respectively, as a set of patterns by area, and can integrate the set of patterns by area into a unique appearance pattern for the entire terminal. In this case, the unique external pattern may not be a simple string or a simple list of coordinates, but rather structured data in which pattern blocks are sequentially arranged by region, and each block may record at what point in time the image was taken, from which region it was extracted, how many feature points it contains, and what priority it will be used for mutual comparison. For example, the arrangement of speaker holes and the perimeter of the charging terminal may be assigned a high priority as relatively stable reference patterns due to manufacturing tolerances, while the distribution of fine scratches may be assigned as an auxiliary pattern because, although it is advantageous for object identification, it can change due to new damage.

[0146] In one embodiment of the present invention, the procedure of comparing the unique exterior pattern at the time of purchase with the unique exterior pattern at the time of registration for sale may be a procedure for identifying whether the terminal has undergone replacement of the exterior housing or some parts after purchase. To this end, the control unit may first match identical reference areas and then calculate the degree of match in the form of pattern blocks within each reference area. For example, regarding the perimeter of the camera placement area, the control unit may match the pattern block composed of the camera housing corner point, lens center point, and surrounding fine scratch distribution at the time of purchase with the pattern block of the same area at the time of registration for sale to calculate the relative distance difference, the pinch angle difference, whether the arrangement order matches, and the amount of movement of the repeated appearance location, respectively. Subsequently, if the difference values ​​all exist within the allowable range obtained from the normal terminal re-shooting verification data, the control unit may determine that the exterior is in the same state, and if multiple pattern blocks in a specific area simultaneously deviate from the allowable range, it may identify that there is a possibility of replacement of the exterior housing or parts. For example, if the relative distance between the center points of the speaker hole arrangement and the charging terminal perimeter is maintained, but the distribution of fine scratches and the pattern of paint discontinuities around the camera placement area change overall, and the pattern of repeated appearances at the camera housing boundary also does not match, it can be determined that the camera housing or rear exterior parts have been replaced. Conversely, if only the patterns of the button boundary area and the side frame area change significantly while other areas remain the same, it can be identified that there is a possibility that the button or side frame parts have been replaced.

[0147] In one embodiment of the present invention, the identification of whether an external housing or a part has been replaced is not determined solely by the discrepancy of a single feature point, but can be performed based on whether a combination of multiple feature points within the same area changes simultaneously. The control unit can aggregate the number of matching pattern blocks and the number of discrepancy pattern blocks for each reference area from the comparison results between the purchase time and the time of registration for sale. For example, if more than 70% of the total pattern blocks around the camera placement area are discrepant, and the discrepancy appears across the lens center point, housing corner point, and the entire distribution of surrounding scratches, it can be determined that this is closer to a part replacement than a simple addition of new scratches. Furthermore, if the relative distance and arrangement order of structural feature points formed during manufacturing, such as the charging terminal perimeter and speaker hole arrangement, change simultaneously, it can be determined that there is a high probability of replacing the entire external housing or the lower assembly. The control unit can standardize the combination of these multiple conditions by comparing it with actual replacement history data of a sample terminal group. For example, the same shooting and feature point extraction can be performed on sample terminals in which only the rear housing has been replaced, only the button module has been replaced, and the bottom frame, including the charging port housing, and after cumulatively analyzing which pattern block mismatches are repeated in which areas when replacements occur, replacement determination criteria for each area can be stored based on the results. Therefore, identification of whether the exterior housing or parts have been replaced can be performed based on the comparison results between the actual replaced sample and the normal sample.

[0148] In one embodiment of the present invention, the procedure of re-contrasting the unique exterior pattern at the time of registration for sale with the unique exterior pattern immediately before shipment may be a procedure for identifying whether new damage has occurred during storage or packaging. In this case, the control unit may focus on verifying whether new scratches, new dents, new paint breaks, or local deformations have been added, rather than structural changes such as the replacement of the entire exterior housing. To this end, the control unit may use the unique exterior pattern at the time of registration for sale as a reference pattern and correlate it with pattern blocks at the time immediately before shipment to calculate the change in the position of existing feature points, the number of newly appearing feature points, the degree to which existing feature points have disappeared, and whether the location of repeated appearances has shifted. For example, if the start and end points of a linear scratch that did not exist at the time of registration for sale in the side frame area are newly detected at the time immediately before shipment, and the location corresponds to the contact position of a packaging jig or the corner contact section of a storage tray, the control unit may determine that new damage has occurred during storage or packaging. Furthermore, if new paint discontinuities appear adjacent to button boundaries or around charging terminals while other structural reference points remain intact, it can be identified as simple new damage rather than part replacement. As such, re-comparison with the pattern immediately prior to shipment serves as a procedure to isolate only exterior damage added after a specific point in time.

[0149] In one embodiment of the present invention, identification of whether new damage has occurred can be performed by considering the appearance of new feature points and the maintenance status of existing feature points together. The control unit may record as a new damage pattern when structural reference feature points, such as the center point of the speaker hole, structural corner points around the charging terminal, and the start and end points of the button, remain the same between the time of registration for sale and the time immediately before shipment, while the start and end points of linear scratches, the inflection points of curved scratches, or paint discontinuity points are added to the surrounding area. Conversely, if the structural reference feature points themselves appear to have moved or been replaced overall, the possibility of part replacement may be re-examined rather than considering it as new damage. The control unit may confirm new damage by using the location where the new feature point appeared, the type of new feature point, the relative distance from existing feature points, and whether repeated detection occurred at multiple time points together. For example, if a discontinuity point detected only once at the time immediately before shipment disappears in a subsequent re-photograph, it may not be confirmed as new damage due to the possibility of false detection; conversely, if it is repeatedly detected at the same location in two or more enlarged exterior images at the time immediately before shipment, it may be confirmed as actual new damage. In this way, the identification of new damage can also be performed by reflecting repeated shooting and whether structural reference points are maintained, rather than relying on a single feature point of a single image.

[0150] In one embodiment of the present invention, the conditions for generating the unique exterior pattern and comparing it between time points can be determined through a verification procedure using actual sample terminals. The control unit secures multiple groups of terminals of the same model, including a normal terminal group, a terminal group with a rear housing replacement, a terminal group with a button replacement, a terminal group with a bottom frame or charging terminal component replacement, a terminal group with new scratches occurring during storage, and a terminal group with corner dents occurring during packaging, and can perform repeated shooting for each sample at the time of purchase, the time of registration for sale, and the time immediately before shipment. Subsequently, an operator generates correct answer data indicating the actual replacement history and the actual history of new damage occurrence, and the control unit compares the results of the comparison between time points of the automatically generated unique exterior pattern with the correct answer data to adjust the relative distance tolerance, the pinch angle tolerance, the number of allowed discrepancies in arrangement order, the allowable range of movement of the repeated appearance location, the minimum number of new feature points, and the replacement judgment conditions for each area. For example, in a normal terminal group, it is confirmed that the relative distance error due to repeated shooting is ±0.03 mm or less, the pinch angle error is ±2 degrees or less, and the amount of movement of the repeated appearance location is 1% or less of the reference length, and in replacement or newly damaged samples, it is possible to determine when the pattern change exceeding the above range accumulates to a certain extent to best match the actual variation. Based on such verified criteria, the control unit generates a unique exterior pattern for each terminal by combining the relative distance between extracted feature points, the pinch angle with adjacent feature points, the arrangement order of feature points, and the repeated appearance location within the same area; identifies whether the exterior housing has been replaced or parts have been replaced by comparing the unique exterior pattern at the time of purchase with the unique exterior pattern at the time of registration for sale; and identifies whether new damage has occurred during the storage or packaging process by re-comparing the unique exterior pattern at the time of registration for sale with the unique exterior pattern immediately before shipment. The procedure can be implemented in such detail that a person skilled in the art can perform it as is.

[0151] A control unit according to one embodiment of the present invention determines that the terminal is the same when pairs of corresponding feature points are simultaneously identified in at least two areas among the camera placement area, the charging terminal perimeter, and the side frame while maintaining the same arrangement order of the feature points, and identifies that the terminal is the same when non-corresponding feature points are concentrated in a specific area, and identifies that the terminal is the same when parts are replaced or new damage occurs in that area, and controls that sales are withheld and re-photographing and re-inspection are requested when the terminal is not the same or when parts replacement or new damage occurs, and performs final shipment approval only when the terminal is the same and no new damage occurs.

[0152] In one embodiment of the present invention, when the control unit performs a time-series comparison of feature points constituting a previously generated unique exterior pattern, it may first separate and sort the sets of feature points of two time points to be compared by region. Here, the region may include a camera placement area, a charging terminal perimeter, and a side frame, and the feature points within each region may include the start and end points of already extracted linear scratches, the inflection points of curved scratches, the center points of holes, the corner bend points, and the discontinuity points of paint boundary lines. The control unit may not immediately compare the types of feature points in each region while they are mixed, but may first determine a reference starting feature point, and then reassign an arrangement order of the feature points along the perimeter or length direction from the starting feature point. For example, in the camera placement area, the top-left corner point or the center point of the first lens can be set as the starting feature point and an arrangement order can be assigned in a clockwise direction; around the charging terminal, the left end point of the terminal opening can be set as the starting feature point and an arrangement order can be assigned from left to right; and in the side frame, an arrangement order can be assigned from the top reference point downwards or from the front reference point to the rear reference point. Once the arrangement order is re-determined in this manner, the control unit can first check whether the feature points at two points in time that are the subject of comparison in the same area appear in the same order along the same direction of progression, and if the arrangement order is misaligned from the beginning, it may not immediately confirm them as the same area even if the individual feature point coordinates are partially similar.

[0153] In one embodiment of the present invention, the control unit can identify pairs of corresponding feature points after verifying whether the arrangement order of the feature points is maintained identically. To this end, the control unit can extract a priority candidate group consisting only of feature points of the same type from the same area at another time point, for a reference feature point selected in a specific area at a given time point. Subsequently, for each candidate group, the control unit can sequentially compare the difference in relative distance from the reference feature point, the difference in the angle of inclusion with adjacent feature points, the difference in the arrangement order index, and the difference in the location of repeated occurrence within the same area. For example, in the camera placement area, the relative distance between the lens center point and the housing corner point, the pinch angle formed by the two side corner points relative to the lens center point, the clockwise arrangement order of the corner points, and the location of repeated occurrences around the camera housing can be compared together; around the charging terminal, the relative distance between the left and right end points and the micro-scratch start point, the angle formed by the terminal opening center point and both end points, the left-right arrangement order of feature points, and the location of repeated discontinuities along the terminal perimeter can be compared together; and in the side frame, the distance between the bend point and the scratch start point, the pinch angle of adjacent feature points centered on the bend point, the arrangement order in the up-down or front-back direction, and the location of repeated occurrences relative to the frame length can be compared together. The control unit can determine the allowable range for each comparison item by comparing the variation range obtained from a normal terminal repeated shooting sample with actual re-shooting samples of the same terminal, samples mixed with other terminals, parts replacement samples, and newly damaged samples. For example, in repeated shooting of the same terminal, relative distance error, pinch angle error, array order retention rate, and repeated appearance position displacement amount are accumulated and stored, and a range separated from the error distribution appearing in other terminals or replacement parts terminals can be adopted as a reference value.Based on the criteria established in this way, the control unit can determine as corresponding feature point pairs only those feature points among the candidate group for which the difference in relative distance and the difference in the included angle are within the allowable range, the array order index matches, and the difference in the location of repeated occurrence is also within the allowable range.

[0154] In one embodiment of the present invention, the control unit may determine that the device is the same terminal if corresponding pairs of feature points are simultaneously identified in at least two of the camera placement area, the charging terminal perimeter, and the side frame while maintaining the same arrangement order of the feature points. Here, "simultaneously identified" may mean a state in which a certain number of corresponding pairs of feature points are established for each of the two areas during the comparison process at the same two points in time, and the arrangement order of those corresponding pairs of feature points is maintained within the area. For example, if the lens center point, the housing corner point, and the paint cut point correspond in the camera placement area, and the terminal left and right end points, part of the speaker hole center point, and the fine scratch start point correspond in the charging terminal perimeter, the control unit may determine that the device is the same terminal by considering that simultaneous correspondence has been secured in the two areas of the camera placement area and the charging terminal perimeter, even if some feature points are missing in the side frame. Conversely, if a sufficient number of corresponding pairs of feature points are secured only in the camera placement area, and the arrangement order is misaligned or the number of corresponding pairs of feature points is significantly insufficient in the remaining areas, the device may not be immediately confirmed as the same terminal but may maintain a deferred judgment. The condition for the simultaneous establishment of the two aforementioned regions can also be set by comparing a sample of the same terminal being re-captured with a sample of a different terminal being mis-inserted. For example, judgment criteria can be stored using verification results showing that while pairs of corresponding feature points are repeatedly established stably in two or more regions within a group of the same terminal, in a group of different terminals of the same model, cases of simultaneous matching in two or more regions rarely occur when including fine scratch distributions or paint break patterns, even if one structural region is similar.

[0155] In one embodiment of the present invention, the control unit may identify the replacement of parts or the occurrence of new damage in a specific area when non-corresponding feature points are concentrated in that area. To this end, the control unit may first calculate the total number of feature points, the number of corresponding feature points, the number of non-corresponding feature points, and the number of newly appearing feature points for each area. Subsequently, the control unit may determine whether the non-corresponding feature points and newly appearing feature points within a single area account for more than a preset ratio relative to the total number of feature points. For example, if structural reference feature points such as corner points and lens center points are maintained in a camera placement area, but a large number of new start and end points of perimeter micro-scratches and paint discontinuity points appear, it may be determined that there is a high probability of new damage occurring. Furthermore, if the relative distance and arrangement order of the structural reference feature points change overall, and a large number of existing feature points are simultaneously lost while a new set of reference points is formed, it may be determined that there is a high probability of replacing the camera housing or rear exterior parts. Even around the charging terminal, if the overall positional relationship of structural reference feature points, such as left and right end points and perimeter corner points, changes, or if the arrangement order or spacing pattern between hole center points differs, it can be identified as replacement of the bottom housing or charging terminal components; however, if the structural reference feature points are maintained while new discontinuity points, fine dents, or linear scratch end points are added only around the terminal, it can be identified as new damage during storage or packaging. Similarly, regarding the side frame, if the basic position of the bending point is maintained while only the distribution of fine scratches is added, it can be classified as new damage; and if the relative position between the bending point and the button boundary line itself changes, it can be classified as replacement of the side frame or button components. The control unit can confirm these classification criteria by using a sample terminal with a history of actual component replacement and a sample terminal in which actual new damage has occurred.For example, the same magnified imaging and feature point extraction are performed on a rear housing replacement sample, a button module replacement sample, a bottom frame replacement sample, a sample with scratches added during storage, and a sample with corner dents added during packaging. Then, judgment criteria for each area can be stored by cumulatively comparing which areas correspond to parts replacement when structural reference feature points also change, and which areas correspond to new damage when only fine feature points are added while structural reference points are maintained.

[0156] In one embodiment of the present invention, the control unit may suspend sales and request re-photography and re-inspection if it is determined that the terminal is not the same, or if parts replacement or the occurrence of new damage is identified. To this end, if the comparison point is between the time of registration for sale waiting and the time immediately before shipment, the control unit may immediately switch the sales status of the first user terminal from a general saleable state to a suspended state, and the storage unit may record together the suspension reason code, suspension time, area that became the reason for suspension, list of mismatched feature points, list of new feature points, and the path of the original enlarged exterior image. In addition, the control unit may display a re-photography request message on the display unit and specifically indicate which area has an arrangement order mismatch, which area has a lack of corresponding feature points, or which area has a concentration of new feature points. For example, it may display phrases such as “lack of corresponding feature points in camera placement area,” “concentration of new discontinuity points around charging terminal,” or “side frame arrangement order mismatch,” and call up an enlarged exterior image thumbnail corresponding to the phrase. Upon a request for re-shooting, the control unit automatically retrieves the magnification, lighting direction, shooting position, and shooting sequence to ensure that the same conditions as the original shooting are applied again, and can repeat the identity verification procedure once the re-shooting result is saved. The request for re-inspection is not limited to simply re-shooting; if necessary, it can also display precision inspection items to the operator, including manual measurement or component disassembly verification. For example, if structural reference feature points in the camera placement area have been generally altered, a precision inspection can be requested to verify whether the rear housing has been replaced; and if new dents are concentrated around the charging terminal, a precision inspection can be requested to check for contact marks from packaging or storage.

[0157] In one embodiment of the present invention, control can be performed to execute final shipment approval only when the terminal is determined to be the same and no new damage is identified. To this end, the control unit may store as a shipment approval condition a state in which corresponding feature point pairs are established in at least two areas among the camera placement area, the charging terminal perimeter, and the side frame, the arrangement order is maintained, feature points not corresponding to a specific area are not concentrated, and the amount of new feature point appearance is below an allowable range. Subsequently, the control unit may switch the state of the first user terminal to a shipment-ready state only when the result of comparing the unique appearance pattern at the time of sales-waiting registration with the unique appearance pattern immediately before shipment satisfies all of the above shipment approval conditions, and may store the shipment approval time, the number of corresponding feature points per area used for approval, the number of mismatched feature points, the number of new feature points, and the enlarged appearance image at the time of approval as the final history. In addition, the control unit may display a summary of the number of corresponding areas, the number of corresponding feature points, and the result of no new damage being detected, along with the shipment approval message, on the display unit. Conversely, if any of the release approval conditions are not met, the release approval button or the release slip issuance function of the terminal can be controlled to be disabled. Therefore, final release approval is not simply made by the operator's judgment, but can be systematically performed only when all of the multiple specific conditions are satisfied, such as maintaining the arrangement order of the feature points, simultaneous correspondence of two or more regions, absence of region concentration discrepancies, and no new damage occurs.

[0158] In one embodiment of the present invention, the conditions for determining identical terminals, the conditions for area concentration mismatch, the conditions for identifying parts replacement, the conditions for identifying new damage, and the conditions for shipment approval can be determined through a verification procedure using an actual sample terminal group. The control unit prepares, respectively, a group of repeated shots of normal terminals of the same model, a group of mixed terminals, a group of rear housing replacements, a group of charging terminal bottom parts replacements, a group of button or side frame parts replacements, a group of newly scratched terminals occurring during storage, and a group of newly dented terminals occurring during packaging, and can then repeatedly perform shots corresponding to the time of purchase, the time of registration for sale, and the time immediately before shipment. Subsequently, when an operator records whether the terminals are actually identical, whether parts have actually been replaced, and whether new damage has actually occurred as correct answer data, the control unit can compare the automatic determination results with the correct answer data to adjust the minimum number of corresponding feature points per area, the relative distance tolerance, the pinch angle tolerance, the number of allowed arrangement order mismatches, the specific area concentration mismatch ratio, the minimum number of newly occurring feature points, and the allowable range for shipment approval. For example, in actual identical terminal samples, a certain number of corresponding feature point pairs are always maintained in two areas, such as the camera placement area and the perimeter of the charging terminal, and the arrangement order is preserved, whereas in samples mixed with other terminals, there may be a tendency for only one area to partially match or for only some structural reference points to match, so the control unit can store minimum conditions for separating them. In addition, in parts replacement samples, structural reference feature points in specific areas are also changed, whereas in newly damaged samples, structural reference points are maintained and only fine scratch feature points are newly added, a tendency may repeatedly appear, so the control unit can store conditions for distinguishing between replacement and damage by utilizing these differences.As the criteria are determined based on actual samples and correct answer data in this manner, the procedures for the control unit to determine whether it is the same terminal, identify component replacement or the occurrence of new damage, and perform actions such as sales suspension, re-photography, re-inspection requests, and final shipment approval can be implemented in such detail that a person of ordinary skill can perform them as is.

[0159] A control unit according to an embodiment of the present invention performs a procedure for determining a droplet stagnation vulnerable area to which the water collection principle of a rain gauge is substituted, wherein the front, side, and rear of the first user terminal in the external inspection model are divided into a plurality of unit areas, and for each unit area, the depth of indentation of the surface, the height of the protrusion of the periphery edge, the slope of connection to adjacent unit areas, and the number of open discharge directions are calculated, and for each unit area, virtual droplets falling in a reference direction are set to flow in, and then if the depth of indentation is greater than or equal to a reference, the height of the protrusion of the periphery edge is greater than or equal to a reference, and the slope of connection is not formed in an outward discharge direction or the number of open discharge directions is less than a reference, it is identified as a water collection area where droplets remain, and if the water collection area overlaps with the boundary of a speaker hole arrangement area, a charging terminal perimeter area, a button gap area, a camera placement area, or a frame joint line area, it is identified as a droplet inflow vulnerable area, and the number of droplet inflow vulnerable areas, the area of ​​each droplet inflow vulnerable area, the continuous length between adjacent droplet inflow vulnerable areas, and the droplet inflow vulnerable area including an opening The risk of moisture stagnation is identified based on the degree of overlap with the area, and if the risk of moisture stagnation is greater than or equal to a preset standard, the first user terminal is switched to a sales hold state or a precision inspection state, and the display unit is controlled to display the location, area, connection slope direction, and type of overlapping opening of the vulnerable area, along with an external inspection model showing the corresponding droplet inflow vulnerable area.

[0160] In one embodiment of the present invention, the control unit can perform a procedure to determine a droplet stagnation vulnerable area by applying the principle of water collection, in which rainwater falls on a rain gauge and collects at a sunken point, and remains for a certain period of time or longer when the perimeter is formed relatively high or there is insufficient discharge path leading to the outside, to the surface shape analysis of the external inspection model. To this end, the control unit first separates the front, side, and rear of the first user terminal from the external inspection model generated in the form of a point cloud or mesh, and can divide each surface into unit areas of a predetermined size. Relatively flat sections, such as the front and rear, can be divided into square or polygonal patches of the same area, and sections with large changes in curvature, such as the side frame, the perimeter of the camera placement area, the button gap adjacent area, and the frame joint line, can be subdivided into smaller polygonal patches based on point density and the amount of change in the surface normal direction. The size of a single unit area can be set to a level sufficient to distinguish micro-indentations where actual water droplets may reside. By repeatedly photographing a normal terminal and a terminal with a deformed exterior of the same model and comparing the restored exterior inspection model, the segmentation size that shows the highest degree of correspondence with the location where the droplet resided in the actual drop test can be adopted. For example, the flat rear surface can be set to a patch with a side length of several hundred micrometers to several millimeters, while structurally complex areas, such as the area adjacent to button gaps or the perimeter of the charging terminal, can be set to smaller patches.

[0161] In one embodiment of the present invention, the control unit can calculate the depth of indentation on the surface for each unit area. To this end, the control unit may set the lowest point or the average height of points corresponding to a certain lower proportion among a plurality of surface points within a specific unit area as the floor height, and select areas among adjacent unit areas surrounding the specific unit area that belong to the same exterior surface and are relatively less affected by openings, boundaries, or defects as reference areas. Subsequently, the control unit may set the average height of the reference area or the average of the upper height group as the reference height, and calculate the difference between the reference height and the floor height as the depth of indentation of the unit area. At this time, the reference area is not arbitrarily fixed, but may be configured to pre-store the distribution of the average surface height at the corresponding location in an exterior inspection model for multiple normal terminals of the same model, and then select only the areas that do not contain abnormal distortion by comparing them with the height of the adjacent area of ​​the terminal to be inspected. For example, when calculating the depression adjacent to the boundary of the camera placement area, the protrusion of the camera housing or the lens window itself can be excluded from the reference area, and a relatively stable rear case section outside of it can be selected as the reference area.

[0162] In one embodiment of the present invention, the control unit can calculate the protruding height of the surrounding rim for each unit area. To this end, the control unit sets a first surrounding area that encircles the outer edge of a specific unit area and can extract the highest surface point among adjacent unit areas included in the first surrounding area, or the average height of surface points corresponding to a certain upper ratio. Subsequently, the control unit can record the difference between the high height and the bottom height of the specific unit area as the protruding height of the surrounding rim. If unit areas of similar height and unit areas of very high height are mixed around a specific unit area, the control unit can calculate a separate rim height for each outer edge of the specific unit area and then adopt only the rim height in the direction that substantially blocks outward discharge as a valid value. For example, in the area adjacent to the button gap, the height of the edge on the side of the button body and the outer edge of the frame may differ from each other; therefore, the control unit can record the rim height for each direction separately to specifically identify in which direction the liquid droplet is likely to remain.

[0163] In one embodiment of the present invention, the control unit can calculate the connection slope to an adjacent unit area for each unit area. To this end, the control unit connects the center point of a specific unit area with the center points of adjacent unit areas that abut the specific unit area, and for each connection line, calculates the slope direction and slope magnitude using the height difference between the two center points and the horizontal displacement distance. The slope direction can be distinguished based on whether the height decreases or increases from the specific unit area toward the adjacent unit area, and the slope magnitude can be calculated by dividing the height difference by the displacement distance. In this way, the control unit records the connection slopes for the up, down, left, and right directions or all adjacent directions in the mesh structure, and can classify which direction is the outward discharge direction, which direction leads to a deeper depression, and which direction is blocked by a rim protrusion. For example, if a specific unit area around a charging terminal has a downward slope toward the charging terminal opening direction but an upward slope toward the outer frame direction, the control unit can classify the opening direction as the direction of droplet inflow or stagnation expansion, and the outer frame direction as the direction where discharge is blocked.

[0164] In one embodiment of the present invention, the control unit can calculate the number of open discharge directions for each unit area. To this end, the control unit tracks consecutive adjacent unit areas in each direction starting from a specific unit area, and if the height gradually decreases along the tracking path, and there are no protruding edges or blocking sections exceeding a standard in the middle, and the direction finally leads to the outer edge of the first user terminal or connects to a relatively wider flat discharge surface, the direction can be counted as an open discharge direction. Conversely, even if it decreases initially, if it rises again in the second or third adjacent unit area, or if it connects to a stagnant section around a deeper depression or opening, it may not be recognized as a discharge direction. In this way, the control unit calculates the total number of open discharge directions for each unit area, and if there are 0, it can be classified as a completely stagnant candidate area with no discharge path; if there is 1, it can be classified as a restricted discharge area where only unilateral discharge is possible; and if there are 2 or more, it can be classified as an area with a relatively high possibility of discharge.

[0165] In one embodiment of the present invention, the control unit may set the inflow of virtual droplets falling in a reference direction for each unit area. Here, the reference direction may be a gravity direction reflecting the storage posture or inspection posture of the first user terminal, and at least one of a front upward posture, a rear upward posture, a left side upward posture, and a right side upward posture may be selected. The control unit may store the average posture in which the device is placed on a storage tray during an actual transaction, the posture in which it is placed during inspection immediately before packaging, or the store display posture as the reference posture, and assign a gravity direction vector corresponding to the reference posture to the appearance inspection model. Subsequently, the control unit may set the virtual droplets to move from an upper unit area to a lower unit area according to the gravity direction and determine whether a specific unit area is included in the inflow path of the virtual droplets. For example, in a rear upward posture, a virtual droplet that falls on the upper rear flat area may move around the camera placement area or in the direction of the frame connection line, and in a side upward posture, it may move along the side frame in the direction adjacent to the button gap or around the charging terminal. By setting up virtual droplet inflow in this way, it is possible to analyze, on a path-by-path basis, where droplets can gather during actual falling or inflow situations, rather than simply viewing the shape of the current unit area.

[0166] In one embodiment of the present invention, the control unit may identify a collection area where droplets remain if the indentation depth is greater than or equal to a reference, the protrusion height of the surrounding rim is greater than or equal to a reference, the connection slope is not formed in an outward discharge direction, or the number of open discharge directions is less than a reference. Here, the indentation depth reference, the rim protrusion height reference, and the open discharge direction reference are not determined arbitrarily, but can be determined using actual droplet test results for a normal terminal and a terminal with a deformed exterior of the same model. For example, the control unit may repeatedly drop a certain amount of test droplets onto the perimeter of the camera placement area, the area adjacent to the button gap, the perimeter of the charging terminal, the area adjacent to the speaker hole arrangement, and the frame joint line area, and record the remaining droplet locations by performing magnified photography after 10, 30, and 60 seconds have elapsed. Subsequently, by backtracking the unit area shape values ​​of the exterior inspection model corresponding to each remaining location, the indentation depth range, the rim protrusion height range, and the number of open discharge directions commonly observed in the area where droplets actually remained can be accumulated and stored. Based on the above accumulated results, the control unit may adopt a combination of shape conditions that best distinguishes between a normal area where no droplets remain and a stagnant area where droplets repeatedly stay as a criterion for determining the collection area. Therefore, the identification of the collection area can be performed not by simple shape assumptions, but by a data-based procedure that cross-references actual droplet tests with the results of external shape restoration.

[0167] In one embodiment of the present invention, the control unit may identify the water collection area as a droplet ingress vulnerable area if it overlaps with the speaker hole arrangement area, the charging terminal perimeter area, the button gap area, the boundary of the camera placement area, or the frame joint line area. To this end, the control unit may first separately label the structural area corresponding to each opening or gap on the external inspection model. The speaker hole arrangement area may be defined as an area within a certain distance from the center of each hole and the outer boundary of the hole; the charging terminal perimeter area may be defined as a lower frame section within a certain distance from the outer edge of the terminal opening; and the button gap area may be defined as the gap line between the button start point and the end point and the surrounding frame section. The boundary of the camera placement area may be defined as a band-shaped area of ​​a certain width inside or outside the outer boundary line of the camera housing, and the frame joint line area may be defined as a linear area around the joint line where the front, side, and rear parts meet. The control unit calculates the extent to which the area or point cloud set of the water collection area overlaps with the structural area, and if the overlap ratio is greater than or equal to a standard or if the center of the water collection area is included within the structural area, the water collection area can be determined as a droplet inflow vulnerable area. In this way, by separately classifying only the areas within the water collection area that overlap with openings or gaps as vulnerable areas, it is possible to clearly distinguish between flat areas where water simply remains and dangerous areas where actual internal inflow is possible.

[0168] In one embodiment of the present invention, the control unit can identify the risk of moisture stagnation based on the number of droplet inflow vulnerable areas, the area of ​​each droplet inflow vulnerable area, the continuous length between adjacent droplet inflow vulnerable areas, and the degree to which the droplet inflow vulnerable area overlaps with an area including an opening. First, the control unit can calculate the total number of vulnerable areas and calculate the number of unit areas and the total area included for each vulnerable area. Subsequently, the control unit can determine the boundary point distance between adjacent vulnerable areas and merge them into a single continuous vulnerable section if they are arranged continuously within a distance of less than a certain distance. Afterward, the control unit calculates the continuous length using the start and end positions of the merged vulnerable section and can separate and record, for example, a continuous longitudinal section extending from the perimeter of the charging terminal to the arrangement of speaker holes, or a linear section extending along the button gap. The degree of overlap can be represented by the ratio of the area of ​​each vulnerable area that overlaps with the opening or gap area, or by the minimum distance from the center of the vulnerable area to the center of the opening or gap. The control unit can determine the risk level in stages by combining such multiple elements. For example, if there is one vulnerable area with a small surface area and a low opening overlap ratio, it can be identified as low risk; if there are multiple vulnerable areas, two or more of which overlap with the perimeter of the charging terminal or the gap between buttons, and the continuous length exceeds a standard, it can be identified as medium risk; and if there is a continuous vulnerable section extending across the boundary of the camera placement area and the frame joint line area, and the opening overlap ratio is also high, it can be identified as high risk. The above risk level can also be determined by conducting a fine drop or spray test on an actual test terminal group and then comparing the results of confirming internal moisture traces, residual moisture traces, or whether water has entered the opening with the vulnerable area determination value.

[0169] In one embodiment of the present invention, the control unit may switch the first user terminal to a sales hold state or a precision inspection state if the risk of moisture stagnation exceeds a preset standard. To this end, the control unit may pre-store state transition rules corresponding to the risk level and the distribution pattern of vulnerable areas in the storage unit. For example, if the risk level is at a medium risk level and the vulnerable area is limited to the vicinity of a single opening, the device may be switched to a precision inspection state to request additional inspection; and if the risk level is at a high risk level and the vulnerable area is simultaneously distributed around the charging terminal, the button gap area, and the frame joint line area, or extends into a long continuous section, the device may be switched to a sales hold state. When switching states, the control unit may record the number of vulnerable areas, total area, maximum continuous length, opening overlap ratio, and risk level together and store them as an inspection history. Furthermore, in the precision inspection state, subsequent inspection items such as re-dip testing, enlarged exterior re-photography, enlarged gap inspection, or verification of internal moisture traces may be automatically called, and in the sales hold state, the device may be controlled to be automatically excluded from the list of available for sale until a separate release procedure is completed.

[0170] In one embodiment of the present invention, the control unit can control the display unit to display the location, area, direction of connection slope, and type of overlapping opening of the vulnerable area along with an external inspection model in which the corresponding droplet inflow vulnerable area is displayed. To this end, the control unit can distinguish and display each vulnerable area on the external inspection model by color or pattern, and can also indicate which side (front, side, or rear) each vulnerable area is located on, and which section (top, center, bottom, left, or right) it corresponds to. In addition, the control unit can display information such as the calculated area value and the most dominant direction of connection slope for each vulnerable area, such as the direction of the charging terminal, the direction of the button gap, the direction of the camera boundary, or the direction of the frame joint line, and can also display the type of opening overlapping the vulnerable area, such as the speaker hole arrangement, the perimeter of the charging terminal, the button gap, the boundary of the camera placement area, or the frame joint line. When a user selects a specific vulnerable area, the control unit can call up the depression depth, rim protrusion height, number of discharge directions, and continuous length of the unit area surrounding that area to an additional information window, and can also display under which reference position the risk level was calculated to be high for the same area. The visual inspection model displayed in this way can be utilized as an inspection tool that goes beyond a simple 3D visual representation and presents structural vulnerable areas with a high probability of droplet stagnation and inflow in a form combining location and numerical values.

[0171] In one embodiment of the present invention, the unit area division condition, sinkhole depth criterion, rim protrusion height criterion, connection slope judgment criterion, open discharge direction criterion, water collection area judgment criterion, vulnerable area overlap criterion, and risk level criterion can be determined through repeated testing using actual sample terminals. The control unit secures a group of normal terminals of the same model, a group of terminals with frame bending, a group of terminals with button gap widening, a group of terminals with charging terminal perimeter deformation, and a group of terminals with camera boundary dents, respectively, and after restoring the appearance inspection model under the same posture conditions, can perform a fine drop test or a spray test. Subsequently, the location where the test droplet actually stayed, the location where it moved toward the opening, and the location where it remained after a certain period of time are recorded by high-magnification photography or dye tracing, and the results can be correlated with the unit area on the appearance inspection model of each terminal. Using this accumulated data, the control unit can store condition combinations in which the normal terminal group is classified as low risk, and samples where actual retention or inflow is confirmed are classified as medium or high risk. Therefore, the above procedure for determining the droplet stagnation vulnerable area is not a simple estimation but can be implemented by cross-referencing the external shape restoration data of an actual terminal sample with the results of repeated droplet tests, and can be configured specifically enough so that a person skilled in the art can perform it as is. Explanation of the symbols

[0173] 100: Computing device 110: User Input Section 120: Display section 200: Server 210: Communications Department 220: Storage section 230: Control unit

Claims

Claim 1 A used product management system comprises a display unit for displaying a screen of a used user terminal purchase and shipment management program; and a computing device provided with a user input unit for user input, and a server connected to the computing device via a wired or wireless network, wherein the server comprises a storage unit; a communication unit; and a control unit; wherein the control unit, upon receiving a purchase request for a first user terminal subject to purchase from the user input unit, requests first product information including the customer's estimated price and first image information for the first user terminal; upon receiving the first product information from the user input unit, the control unit identifies the first image information to identify first defect information regarding the display screen and appearance of the first user terminal, generates inspection peculiarity information, and displays it on the display unit; upon receiving first physical inspection information and first physical inspection price including an inspection result of at least one of vibration, speaker, face recognition, fingerprint sensor, communication connection, camera, touch display sensitivity, and button for the first user terminal from the user input unit, generates a first inspection result list for the first user terminal and displays it on the display unit; the control unit switches the first user terminal for which the first inspection result list has been generated to a sales waiting state and registers information as a used user terminal available for sale on the program; and the control unit When user input regarding manufacturer-based first terminal source information stored in the first user terminal is received from the user input unit, the first terminal source information is input through a pre-configured loss / theft solution lookup API for blocking the distribution of illegal terminals to perform a loss / theft lookup for the first user terminal, and when the control unit identifies the first user terminal as being in a normal release state as a result of the loss / theft lookup, it displays a popup window to the display unit for deleting existing data and connecting a function diagnostic program for the first user terminal.When the control unit receives user input for a purchase request for the first user terminal from the user input unit, it identifies a first physical sales price in which an amount of a preset ratio is added to the first physical inspection price and displays the first physical sales price on the display unit; the control unit 3D models the external state of the first user terminal and visualizes and provides the external appearance and functions of the first user terminal in a virtual reality environment; the control unit generates a 3D model of the first user terminal by combining image capture technology and point cloud data with multiple images of the first user terminal placed on an automated rotating imaging tool and captured from multiple angles; the control unit receives multiple images captured while stationary at each rotation interval and rotation angle information corresponding to each image for the first user terminal placed on the automated rotating imaging tool, identifies the outline, corner line, display boundary line, camera placement area boundary line, and button placement area boundary line of the first user terminal in each of the multiple images to extract reference feature points, and adjacent to each other Identifying the connection between images by matching reference feature points representing the same location among images corresponding to the rotation angle, unifying the coordinate standards of each image based on the identified connection, calculating the depth of a surface point using the parallax between feature points identified as the same surface location in multiple images with unified coordinate standards, and generating a point cloud representing the surface shape of the front, side, and rear of the first user terminal by continuously connecting the calculated surface points, comparing the color value change, brightness value change, and boundary line position change by angle for areas identified as the same surface location in multiple images with unified coordinate standards, excluding areas where only the brightness value changes with angle change and there is no change in boundary line position as noise areas due to reflection, and identifying areas where linear color discontinuity is continuous regardless of angle change as scratches.Areas where localized color degradation forms in the shape of a closed curve are identified as dents, and areas where the boundary line position deviates beyond a standard and remains in a continuous section are identified as warping; an appearance inspection model is generated by superimposing the locations of the identified scratches, dents, and warping onto the generated point cloud to display the location, length, area, and type of defects together; the appearance inspection model is displayed on the display unit to enable rotation, zooming, and selection by area, and control is provided so that defects are displayed separately by the display area, side frame area, rear case area, and camera placement area; when a purchase request is entered, the appearance inspection model is superimposed on the real-time captured video, and control is provided so that when the user selects a specific area, the defect type, defect location, rotation angle information at the time of inspection, and the original image of that area are displayed together; the control unit performs a virtual cross-section reconstruction procedure in which the principle of tomography reconstruction of medical images is applied by substitution, by setting multiple virtual cross-sections based on the length, width, and thickness directions of the first user terminal, and rearranging surface points intersecting each virtual cross-section according to the rotation angle The outlines for each cut plane are restored, and for each restored outline for each cut plane, the amount of curvature deviating from a reference straight line, the spacing deviation between two opposing outlines, the reduction in corner angle, and the separation deviation between the display boundary line and the frame boundary line are calculated; if the calculated amount of curvature accumulates in the same direction across three or more consecutive virtual cut planes, it is identified as frame bending; if the separation deviation is repeated in the section adjacent to the display boundary line, it is identified as panel lifting; and if the reduction in corner angle is concentrated in the corner section, it is identified as corner deformation; a cross-sectional distortion map is generated by arranging the occurrence sections of the identified frame bending, panel lifting, and corner deformation according to the longitudinal order of the first user terminal, and is displayed on the display unit together with the appearance inspection model; and if the occurrence section of frame bending in the cross-sectional distortion map continues to the section adjacent to the display boundary line, orA used product management system characterized by lowering the appearance grade and identifying it as a target for re-inspection when panel lifting is repeated in three or more consecutive virtual cutting planes along one side of the first user terminal, or when corner deformation is simultaneously identified in two or more of four corner sections. Claim 2 delete Claim 3 delete Claim 4 delete Claim 5 delete Claim 6 In claim 1, the control unit performs an identity verification procedure to which the feature point correspondence principle used in fingerprint identification is substituted, wherein at the time of purchase, the time of registration for sale, and the time immediately before shipment, it secures enlarged exterior images including the perimeter of the camera placement area, the arrangement of speaker holes, the perimeter of the charging terminal, the button boundary line, and the distribution of fine scratches on the side frame, respectively, and extracts the start and end points of linear scratches, the inflection points of curved scratches, the center points of the hole arrangement, the corner bending points, and the discontinuity points of the paint boundary line as feature points from each enlarged exterior image, generates a unique exterior pattern for each terminal by combining the relative distance between the extracted feature points, the pinch angle with adjacent feature points, the arrangement order of the feature points, and the location of repeated appearance within the same area, identifies whether the exterior housing has been replaced or parts have been replaced by comparing the unique exterior pattern at the time of purchase with the unique exterior pattern at the time of registration for sale, identifies whether new damage has occurred during the storage or packaging process by re-comparing the unique exterior pattern at the time of registration for sale with the unique exterior pattern at the time immediately before shipment, and while the arrangement order of the feature points is maintained identically A used product management system characterized by controlling that if a pair of corresponding feature points is simultaneously identified in at least two areas among the camera placement area, the charging terminal perimeter, and the side frame, it is determined to be the same terminal; if non-corresponding feature points are concentrated in a specific area, it is identified as a replacement of parts or the occurrence of new damage in that area; if it is not determined to be the same terminal or if a replacement of parts or the occurrence of new damage is identified, the sale is suspended and a re-photograph and re-inspection are requested; and only if it is determined to be the same terminal and the occurrence of new damage is not identified, final shipment approval is performed.

Images