AN ARTIFICIAL INTELLIGENCE-BASED AUGMENTED OR VIRTUAL REALITY DESIGN ADAPTATION SYSTEM

Abstract

This invention relates to a system (1) that enables the creation of 3D designs in AR / VR and holographic environments, their positioning on real-world objects or user body models, and their optimization by artificial intelligence by ensuring the fit of dimensions, color, proportions and materials.

Description

1 TARIFF AI-BASED AUGMENTED OR VIRTUAL REALITY DESIGN ADAPTATION SYSTEM Technical Area This invention enables the creation of 3D designs in AR / VR and holographic environments, bringing them to life. positioning of world objects or user body models on an artificial one. optimization by intelligence through the coordination of size, color, proportion, and material. It is related to a system that provides. Previous Technique In the current state of the technology, virtual design is applied to a real-world image. by placing "elements" on a product (especially clothing items like shoes) 15 its design focuses on visualization in an AR environment, thus allowing the camera to... The process involves overlaying virtual graphic layers onto the captured image, allowing the user to create different options. various design variations that allow one to see them on the actual product Solutions are encountered. In another known case of the technique, a physical object (e.g., furniture, product) fitting into a specific three-dimensional volume (room, shelf, space, etc.) 20 enabling the assessment of whether the virtual object will fit, with the help of AR. placement within the physical environment seen through the user's camera, boundaries and various Solutions are being found. However, in the current state of the technology, user gestures or UI controls conversion of real-time 3D deformation commands, optimization The engine assesses the complexity of the scene and determines which object is which. 2 Automatic determination of the resolution to be maintained, via the adaptation module. the designed object's connection to a real-world target surface (table, wall, body, vehicle) (parts, etc.) are fitted with physical accuracy thanks to the AI ​​interaction engine. the created model is fixed to the scene as an AR hologram in the real world Various solutions providing this have not been encountered. 5 International patent number WO2025100911, which falls under the prior art. The document states that the video was created by a transparent augmented reality device. A method is described. The method in the invention involves at least one imaging sensor. The image frames of a scene captured using and the depth associated with the scene 10 obtaining the data, pixels corresponding to human skin to the rest of the scene using an artificial intelligence model trained to separate the moving parts Creating masks associated with an object, image frames, depth data and the reconstruction of images of the moving object based on masks, 15 images of static scene content based on image frames and depth data. reconstruction of moving object images and static scene content combining its images and at least one virtual feature, combined This may involve the processing of images on at least one screen. The sensors in the invention, also one or more buttons for touch input, one or more microphone, depth sensor, motion sensor, gyroscope or gyroscope sensor, air 20 pressure sensor, magnetic sensor or magnetometer, accelerometer or It may include an accelerometer, grip sensor, proximity sensor, and color sensor. Image and Depth data capture involves acquiring depth data related to the captured image frames. It can also be used to obtain data. For example, on a VST XR device or At least one depth sensor used with it, using transparent camera(s) 25 It can capture depth data within the displayed scene. Light detection. and suitable sensors such as rangefinding (LIDAR) or time-of-flight (ToF) depth sensors. Any depth sensor can be used. However, in this invention, a movable one is required. a purely optical processing such as masking objects through pixels (optical (by going beyond the processing) approach, the 3D geometry within the scene is directly processed. 3 pixels that allow focus on the construction and manipulation of missing data not through the geometry, but through the structural integrity of the designed 3D model. Features that enable resolution in the finishing and surface matching layer. It is not included. Brief Description of the Invention The aim of this invention is to render user gestures or UI controls in real-time in 3D. converting the deformation commands, optimizing the scene with the optimization engine 10 automatic determination of the actual object designed with the adaptation module with physical accuracy to the target surface in the world (table, wall, body, vehicle parts, etc.) the placement of the model, created thanks to the AI ​​interaction engine, in the real world an AI-based system that enables the AR hologram to be fixed onto the stage. The goal is to implement an augmented or virtual reality design adaptation system. 15 Detailed Description of the Invention To achieve the goal of this invention, a "20 Artificial Intelligence-Based Project" was carried out. The "Augmented or Virtual Reality Design Adaptation System" is shown in the figure. and in this form; Figure 1: The subject of the invention is an artificial intelligence-based augmented or virtual reality design. This is a schematic view of the adaptation system. 25 The parts shown in the figure are individually numbered, and the corresponding numbers correspond to these numbers. It is given below. 1. System 30 4 2. Electronic device 3. Server Creating 3D designs in AR / VR and holographic environments, bringing them into the real world. positioning objects or the user's body model and artificial intelligence 5 optimization by ensuring size, color, proportion and material harmony the subject of the invention is an artificial intelligence-based augmented or virtual reality system. design adaptation system (1); - used by the user, with multiple depth sensors and cameras. possessing, or at least communicating through a virtual reality headset, 10 enabling the user to experience a virtual or augmented reality environment with the hardware it contains, the user can access services on global networks, at least one configured to enable it to connect with platforms electronic device (2) and - connecting with the electronic device (2) and the complex 15 GPU hardware on it. capable of performing image calculations, with at least a mapping module, Production module, optimization engine, geometry completion module, interaction to operate the engine, material engineering analysis module, electronic device (2) by establishing a connection between the real objects in the user's environment or the person's body Extraction of the 3D geometric representation using the (2) camera or 20 in the electronic device SLAM-based local mapping of raw image streams from the depth scanner combining multi-frame depth data in a volumetric field (TSDF – Truncated Signed Distance Fields (TRDs) are collected to create a high-accuracy mesh model. creation, extraction of body measurements using an anthropometric model, surface topological representation of curves, edge junction regions, and orientation information 25 enabled by initiating NURBS / SDF / voxel primitives within the scene translating the design process into a consistent mesh, user gestures or UI converting controls into real-time 3D deformation commands enabling the complexity of the scene with its optimization engine. the evaluation and automatic determination of which object will be stored at what resolution. determination of the virtual reality glasses connected with the electronic device (2) user's line of sight, eye-tracking sensor data, ray-distance values, scene Continuous monitoring of polygon density and GPU usage, optimization. select LOD tiers (L0-L4) according to the importance level of the model Elimination of distant / unimportant objects, ultra-mesh processing of near / focused objects. to enable rendering, to transfer the designed object to the target surface in the real world. To ensure that (table, wall, body, vehicle parts, etc.) is positioned with physical accuracy. The AR hologram of the model created thanks to the interaction engine is a real-world model. fixing the object to the scene, the user's virtual (2) electronic device with the object EN 10 is structured to enable physical interaction through the glasses. It includes at least one server (3). The equipment on the electronic device (2) in the system that is the subject of the invention (1) enabling the user to use virtual reality glasses, with the hardware on them allowing the user to... enabling it to navigate the virtual world, thanks to its depth sensor and camera 15 The mobile phone is configured to transmit the data stream it obtains to the server (3). It is a smart device in the form of a phone, tablet, or computer. Mapping module on server (3) in the system (1) that is the subject of the invention thanks to the 20 captured by the camera and depth sensor of the electronic device (2). 3D models of real objects in the user's environment or the user's body It is configured to enable the extraction of a geometric representation. Server (3) thanks to the mapping module on the electronic device (2) camera or SLAM-based positioning of raw image streams from a depth scanner combining mapping with multi-frame depth data in a volumetric area 25 (TSDF – Truncated Signed Distance Field) is collected to create a high-accuracy mesh. creating a model, thereby enabling the user's body measurements to be anthropometrically determined. extraction of surface curves, edge junction regions and orientation using the model It is structured to enable the topological marking of information. 6 Within the mapping module on the server (3), LiDAR, ToF, Structured Light, The use of 3D scanning techniques in RGB-D Fusion format, SLAM, ORB- Mapping in the form of SLAM2, KinectFusion, ARKit LiDAR Mesh API. the use of techniques such as TSDF Volume Integration, Poisson Surface The use of mesh reconstruction techniques in the form of reconstruction, 5 SMPL human body model, Anthropometric mapping of the body in ML shape. to enable the use of modeling techniques, thereby realizing the truth the object / environment's 3D mesh (RealObjectMesh), normal, slope, and surface classes, segments (SurfaceTopologyData), optional, user body 10 to enable the generation of parameters (UserBodyMetrics) It is being structured. To run the production module on the server (3), mapping in the production module To enable the module's output data to be used as a reference, production After the module connects with the electronic device (2), the user's hand 15 movements, controller, UI inputs, predefined template models, Size, style, purpose: design in the form of furniture, accessories, medical devices. enabling the use of parameters as input, production as output. obtaining a high-resolution 3D representation of the object designed in the module It is configured to provide 20 in the production module on the server (3). Combining voxel / mesh / NURBS / SDF primitives with user gestures and commands Generating a draft 3D geometry; if the 3D geometry is voxel-based, then a volumetric grid. or using a sparse voxel octree, followed by surface extraction and meshing. If the extraction is 3D geometry mesh-based, then extrusion, boolean, subdiv, smooth Running classic modeling operations as shown in Figure 25, if desired, with AI. to enable the use of techniques that allow for "making suggestions" It is configured. In the production module on the server (3), OpenVDB, SDF Unity uses geometry kernels in the form of Signed Distance Fields. Processing in the form of HDRP / Unreal Nanite / Vulkan compute shaders the use of 3D primitives in the form of Extrude, Boolean, Smooth, Inflate, and Retopo. 7 the use of ops, form fluency prediction with Graph Neural Networks It is structured to enable its use. The output of the production module with the optimization engine on the server (3) Interacting with a 3D object ready for processing, 5 of the 3D scene in question assessment of complexity and at what resolution each object should be stored to enable automatic identification, using the 3D object model in question as input, scene information (other objects, polygon counts, light sources), electronic devices (2) telemetry (GPU / CPU load, FPS, VRAM usage), user's view 10 Enable the use of direction / eye tracking data (if VR / AR eye tracking is available) It is structured accordingly. The output of the production module with the optimization engine on the server (3) Transformer + Policy Gradient RL is applied to a 3D object ready for processing. Using the LOD decision network, Quadric Edge Collapse / Nanite-style clusters 15 use of mesh reduction in this form, Frame Timing, VRAM Pressure Monitor the use of scene telemetry in this form, Dynamic Render Scaling and Foveated Enabling the use of performance optimization in the form of rendering, this Thanks to this, the 3D object model becomes very light and fluid inside the XR engine. obtaining an output of the delivered version and / or LOD 20 for each part. Obtaining the aforementioned 3D scene / object model with defined and scaled-down levels It is structured to ensure that this is done. Instantaneous 3D Scene within the optimization engine on the server (3) Complexity: total number of vertices / triangles, number of draw calls, shader density 25 Measurement in this way, the area the user is looking at, the distance to the object, interaction importance map of signals such as probability (e.g., proximity) translation, AI-based LOD decision model (e.g., a small Transformer + RL) (policy) to select the LOD level as L0-L3 for each 3D object / sub-part, Apply mesh / voxel destruction for the necessary objects (quadric edge collapse, 30 8 cluster-based simplification), and, if necessary, proxy meshes with reduced resolution, The system is designed to facilitate the production of billboards and imposters. 3D designed thanks to the geometry completion module on the server (3) placing the object onto a real-world target surface with physical accuracy, word 5 The subject is the surface normals, slope, material color, and light vectors of a 3D object. and is configured to enable the analysis of aspect ratios. Server (3) The 3D model optimized in the geometry completion module on the target surface / object information, 3D model's slope, normal, and segment information, user 10 to enable the use of body metrics as input It is being structured. In the geometry completion module on the server (3), the design 3D model To align with the actual surface / body, use ICP (Iterative Closest Point) or feature- To ensure the implementation of based alignment, 3D 15 according to the target use case. To ensure the model is positioned correctly on reference points, and for color and material consistency; Analyzing the texture / color distribution on RealObjectMesh, the design To enable color matching using AI, 3D completion networks (PointNet++, NeRF-based volume) for geometry deficiencies (completion) to enable automatic filling, simple 20 for stability and ergonomics To enable physics calculations, thereby determining size, color, position, and geometry. and to enable the production of a printout of the physically adapted 3D model. It is being structured. Output of the interaction engine and geometry completion module on the server (3) is 25 The 3D model in question appears electronically as an AR hologram in the real world. to ensure that the virtual reality glasses (2) on the device are fixed to the scene, so that the user's 3D object through the virtual glasses (2) on the electronic device to enable it to touch, turn, and pull It is being structured. 30 9 Geometry completion module within the interaction engine on the server (3) The 3D model, which is the output of the electronic device (2), is the instantaneous camera sensor. data of the world coordinate system / anchor points Enabling the use of (ARKit / ARCore / HoloLens) information as input, these 5 This allows for plane detection (of floors / walls / tables, etc.) in the right places in a real-world environment. the creation of spatial anchors, and the ContextReadyModel connecting to these anchors. making it appear as if it's fixed in the real world, depth of field for closure by comparing the depth information of the model with the map and identifying intervening objects (hand, etc.) (another object) should be displayed in front of the model, ambient light for lighting harmony 10 making predictions (extracting HDR environment map from camera feed), model adjusting the shaders accordingly, via the user's electronic device (2) With hand / gesture inputs, the model can be rotated, repositioned, scaled, and clicked. It is structured to enable interactions to take place. Created thanks to the materials engineering analysis module on the server (3) Calculating the suitability of the 3D model for actual physical production, materials. strength, heat tolerance, elasticity, stretch allowance, load-bearing capacity, and production. Cost estimation and, if necessary, using AI to find more suitable alternatives. to ensure that materials are recommended to the user via electronic device (2), word 20 The processes in question involve FEA (Finite Element Analysis), GPU Simulation, and Material DB. PLA / ABS / CFRP / Aluminium / Stainless Composite, Structural Stress Solver and Buckling detection, Cost estimator: Monte Carlo Manufacturing Cost AI-shaped It is structured to enable the use of these techniques. Within the materials engineering analysis module on the server (3), the 3D running Finite Element Analysis (FEA) on the model, as specified Stress, deformation, and fracture risk are simulated according to load / force scenarios. to ensure that the part thickness and infill ratio in the 3D model are determined, Geometric parameters such as curvature / radius vary according to manufacturing techniques. evaluation, at least from a materials database, of suitable materials selection, calculation of the manufacturability score of the 3D model between 0-100, “this The relevant 3D statements include phrases like "the area should be thickened" and "this junction radius should be increased". It is structured to allow feedback to be transmitted to the design. Server (3) on the outputs of the material engineering analysis module To ensure that the simulation and technical reporting module is accessed, simulation and output selected by the user from the electronic device (2) with the technical reporting module using the format information, thereby enabling automatic report generation for the user, production technical drawings, step-by-step construction documentation, material list and 10 the preparation of prototype variants, as well as simulation animations, SRD (System Research Report), which includes fracture test results and improvement recommendations. It is configured to enable the creation of the (Document) document. Server (3) 15 on the outputs of the material engineering analysis module processing in the simulation and technical reporting module, 3D during processing Suitable models for production include STEP, IGES, STL, FBX, and GLB / GLT. converting to formats, creating technical drawings and dimensioning plans, with the results obtained from simulations (stress maps, deformation animations, heat distribution, etc.) report files and video (if desired) 20 the production of animations, all of these are CAD files, bill of materials (BOM), production guidelines, simulation visuals, and development suggestions Actions such as packaging under the name "FinalDesignPack" It is structured to ensure its implementation. Industrial application of the invention In the system subject to the invention, (1) the user has an electronic device with virtual reality glasses (2) It uses the depth sensor and camera sensor on the electronic device (2) 30 11 The data is sent to the server (3). The server (3) receives the data from the electronic device (2). 3D environmental information is obtained through various AI-powered modules located on it. It is being processed. As a result of these processes, 3D designs are converted into AR / VR and holographic versions. created in the environment, based on real-world objects or a user body model Positioning and AI-controlled measurement, color, proportion, and material matching. 5 This is achieved by optimizing the process. When a user designs an object... system (1) scans the real environment or body measurements to create a digital model of the physical automatic resizing according to context and surface adaptation. It enables correction and intelligent completion of missing geometries. Thus The design goes beyond being just a displayed hologram; it becomes a real-world phenomenon. It is transforming into a prototype that can be experienced and implemented. The AI-based LOD (Level of Detail) mechanism assesses the density and visibility of the scene. By analyzing the area, computational load, and user interaction potential, it creates a dynamic system. It determines the resolution levels; this allows even complex holographic scenes to be rendered at 15 It can operate with high performance and low latency. The system (1) is only designed for It not only optimizes but also evaluates manufacturability. Material selection, Strength estimation, production technique recommendation (3D printing, molding, CNC, etc.), cost Technical reports are generated that include simulations and development pathways. The invention system (1) enables the realization of user gestures or UI controls. Converting real-time 3D deformation commands into optimization engine commands. Assessing the complexity of the scene and determining which object should be used at what resolution. Automatic determination of which object will be retained, designed with the adaptation module. Physical 25 on a real-world target surface (table, wall, body, vehicle parts, etc.) The accuracy of the model is ensured by the AI ​​interaction engine, which creates a realistic representation of the model. In the world, AR holograms are fixed to the stage. The system in question (1) is within the scope of the Personal Data Protection Law (KVKK). It operates. 30 12 The invention concerns an artificial intelligence-based augmented or virtual reality design. It is possible to develop a wide variety of applications of the adaptation system (1), The invention cannot be limited to the examples described here, but mainly to the claims. as stated. 5

Claims

13 REQUESTS 1. Creating 3D designs in AR / VR and holographic environments, integrating them into the real world. the positioning of objects or the user's body model and artificial intelligence 5. Optimizing the size, color, proportion, and material by ensuring harmony. providing; - used by the user, with multiple depth sensors and cameras. possessing, or at least communicating through, a virtual reality headset enabling the user to experience a virtual or augmented reality environment with the hardware located on it, the user can access services on global networks, 10 at least one configured to enable it to connect with platforms electronic device (2) containing and - connecting with the electronic device (2) and its complex GPU hardware. capable of performing image calculations, with at least a mapping module, Production module, optimization engine, geometry completion module, interaction 15 to operate the engine, material engineering analysis module, electronic device (2) by establishing a connection between the real objects in the user's environment or the person's body Extraction of the 3D geometric representation using the camera (2) in the electronic device or SLAM-based local mapping of raw image streams from the depth scanner combining multi-frame depth data in a volumetric field (TSDF – 20 Truncated Signed Distance Fields (TRDs) are collected to create a high-accuracy mesh model. creation, extraction of body measurements using an anthropometric model, surface topological representation of curves, edge junction regions, and orientation information enabled by initiating NURBS / SDF / voxel primitives within the scene translating the design process into a consistent mesh, user gestures or UI 25 converting controls into real-time 3D deformation commands enabling the complexity of the scene with its optimization engine. automatic evaluation and determination of which object should be stored at what resolution determination of the virtual reality glasses connected with the electronic device (2) User's line of sight, eye-tracking sensor data, ray-distance values, scene 30 14 Continuous monitoring of polygon density and GPU usage, optimization. select LOD tiers (L0-L4) according to the importance level of the model Eliminating distant / unimportant objects, ultra-mesh processing of near / focused objects. to enable rendering, to transfer the designed object to the target surface in the real world. (Ensuring that the table, wall, body, vehicle parts, etc.) are positioned with physical accuracy, 5 The AR hologram of the model created thanks to the interaction engine is a real-world model. fixing the object to the scene, the user's virtual (2) electronic device with the object designed to enable physical interaction through the glasses an artificial intelligence based augmented or characterized by at least one server (3) virtual reality design adaptation system (1). 10 2. The hardware on it allows the user to use virtual reality glasses, hardware that allows the user to navigate the virtual world, with depth transmitting the data stream obtained by the sensor and the camera to the server (3) a smart device in the form of a mobile phone, tablet, or computer configured to provide artificial intelligence as in Claim 1 characterized by the electronic device (2) which is a device an augmented or virtual reality design adaptation system based on (1).

3. Thanks to its mapping module, the camera (2) of the electronic device and The depth sensor captures real objects around the user or 20 to enable the creation of a 3D geometric representation of the user's body artificial as in Claim 1 or 2 characterized by the configured server (3). an intelligence-based augmented or virtual reality design adaptation system (1).

4. Thanks to its mapping module, the electronic device can receive 25 from the (2) cameras. or SLAM-based positioning of raw image streams from a depth scanner combining mapping with multi-frame depth data in a volumetric field (TSDF – Truncated Signed Distance Field) is collected to create a high-accuracy mesh. creating a model, thereby enabling the user's body measurements to be anthropometrically determined. Extraction of surface curves, edge junction regions, and orientation using the model 30 Server configured to enable topological marking of information (3) artificial intelligence as in any of the above-mentioned requests characterized by an augmented or virtual reality design adaptation system based on (1).

5. Its mapping module includes LiDAR, ToF, Structured Light, RGB-D 5 The use of fusion-based 3D scanning techniques, SLAM, ORB-SLAM2, KinectFusion utilizes mapping techniques in the form of the ARKit LiDAR Mesh API. using TSDF Volume Integration, Poisson Surface Reconstruction the use of mesh reconstruction techniques in the form of SMPL human body The model, 10 of the body modeling techniques in the form of Anthropometric Mapping (ML). to enable the use of the 3D mesh of the real object / environment (RealObjectMesh), normal, slope, surface classes, segments. (SurfaceTopologyData), optional, user body parameters Server (3) configured to enable the generation of (UserBodyMetrics) Artificial intelligence, as in any of the above characterized requests, 15 an augmented or virtual reality design adaptation system based on (1).

6. Running the production module on it, mapping in the production module. To enable the module's output data to be used as a reference, production After the module connects with the electronic device (2), the user's hand 20 movements, controller, UI inputs, predefined template models, Size, style, purpose: design in the form of furniture, accessories, medical devices. enabling the use of parameters as input, production as output. obtaining a high-resolution 3D representation of the object designed in the module The above 25 is characterized by the server (3) configured to provide an AI-based augmented or virtual system, as in any of the requests. reality design adaptation system (1).

7. On the production module, with the user's gestures and commands Combining voxel / mesh / NURBS / SDF primitives to generate a prototype 3D geometry, 30 16 If the 3D geometry is voxel-based, then volumetric grid or sparse voxel octree can be used. using, then surface extraction to create a mesh, 3D geometry mesh If it's based on; classic modeling options like extrusion, boolean, subdiv, and smooth. The operation can be run, and if desired, AI can make "shape suggestions". Server (3) and 5 configured to enable the use of techniques such as artificial intelligence as in any of the above characterized requests an augmented or virtual reality design adaptation system based on (1).

8. In the production module on it, OpenVDB, SDF Signed Distance Fields The use of the following geometry kernels in Unity HDRP / Unreal Nanite / 10 Vulkan allows the use of compute shaders, Extrude, Boolean, The use of 3D primitive ops such as Smooth, Inflate, and Retopo, Graph To enable the use of Neural Networks for form fluency prediction. any of the above requests characterized by the configured server (3) an AI-based augmented or virtual reality design like the one in 15 adaptation system (1).

9. The processing output of the production module, with its optimization engine. Interacting with a ready-made 3D object, the complexity of that 3D scene the evaluation and automatic determination of which object will be stored at what resolution. to enable its determination, the 3D object model in question as input, the scene information (other objects, polygon numbers, light sources), electronic device (2) telemetry (GPU / CPU load, FPS, VRAM usage), user's view Enable the use of direction / eye-tracking data (if VR / AR eye tracking is available) 25 of the above requests characterized by the server (3) configured for. an AI-based augmented or virtual reality design, like any other adaptation system (1).

10. The processing output of the production module, with its optimization engine. LOD decision 30 in Transformer + Policy Gradient RL format on a ready-made 3D object 17 the use of the network, in the form of Quadric Edge Collapse / Nanite-style clusters using mesh reduction, Frame Timing, and VRAM Pressure Monitoring. the use of scene telemetry, Dynamic Render Scaling and Foveated Rendering to enable the use of performance optimization in this way, thereby... The subject is a 3D object model that has been made very light and fluid within the XR engine. obtaining the output of the version and / or LOD levels for each part obtaining the specified and scaled-down 3D scene / object model the above characterized by the server (3) configured to provide an AI-based augmented or virtual system, as in any of the requests. Reality Design Adaptation System (1). 10 11. The instantaneous complexity of the 3D scene within its optimization engine. in the form of total vertex / triangle count, draw call count, shader density. Measurement of the area the user is looking at, distance to the object, and likelihood of interaction. 15 importance map of signals such as (e.g., proximity of hand) translation, AI-based LOD decision model (e.g., a small Transformer + RL) (policy) to select the LOD level as L0-L3 for each 3D object / sub-part, Applying mesh / voxel destruction for the necessary objects (quadric edge collapse, cluster-based simplification), and, if necessary, proxy meshes with reduced resolution, Server configured to enable the production of billboards and imposters (3) 20 artificial intelligence as in any of the above-mentioned requests characterized by an augmented or virtual reality design adaptation system based on (1).

12. Thanks to its geometry completion module, the designed 3D object the 3D 25 in question ensures that the target is placed on the real-world surface with physical accuracy. object surface normals, slope, material color, light vectors, and dimensions It is configured to enable the analysis of the ratios. Server (3) The 3D model, optimized in the geometry completion module on it, is the target. surface / object information, 3D model's slope, normal, and segment information, user 30 structured to allow the use of body metrics as input. 18 as in any of the above requests characterized by the server (3) an artificial intelligence-based augmented or virtual reality design adaptation system (1).

13. In the geometry completion module, the design 3D model is rendered as real. To align with the surface / body, use ICP (Iterative Closest Point) or feature-based 5. to ensure alignment is applied, 3D model according to the target use case. To ensure proper alignment with reference points, and for color and material harmony; Analyzing the texture / color distribution on RealObjectMesh, the design To enable color matching using AI, 3D completion networks for geometry deficiencies (PointNet++, NeRF-based volume 10) Enabling automatic filling with (completion) ensures simplicity for stability and ergonomics. To enable physics calculations, thereby determining size, color, position, and geometry. and to enable the production of a printout of the physically adapted 3D model. any of the above requests characterized by the configured server (3) an AI-based augmented or virtual reality design like the one in 15 adaptation system (1).

14. The output of the geometry completion module with its interaction engine. the 3D model in question appears as an AR hologram on an electronic device in the real world. (2) to ensure that the virtual reality glasses are fixed to the scene, so that the user can 20 (2) on the electronic device, to be able to touch the relevant 3D object through the virtual glasses, with the server (3) configured to enable translation and extraction. artificial intelligence as in any of the above characterized requests an augmented or virtual reality design adaptation system based on (1).

15. Output of the geometry completion module within its interactive engine. The 3D model in question, the instantaneous camera sensor data of the electronic device (2), Earth coordinate system / anchor points (ARKit / ARCore / HoloLens) to enable the use of this information as input, thus allowing for real-world, plane-based simulations. Spatial anchor 30 with detection (on floor / wall / table etc.) in appropriate places. 19 the creation of a ContextReadyModel by connecting to anchors in the real world to appear fixed, with a depth map of the real environment for closing. the depth information of the model is compared and intervening objects (hand, another object) the model being displayed in front of it, ambient light prediction for light matching (camera (Extracting HDR environment map from feed) and model shaders accordingly. adjustment according to the user's hand / gesture inputs via electronic device (2). model interactions such as rotation, displacement, scaling, and clicking characterized by the server (3) configured to enable it to perform an AI-based augmented reality like in any of the above requests or virtual reality design adaptation system (1). 10 16. 3D created thanks to its materials engineering analysis module. calculating the model's suitability for actual physical production, material strength, heat tolerance, elasticity, stretch allowance, load-bearing capacity, and production cost. deriving the prediction, and if necessary, using AI to identify more suitable alternative materials. to ensure that the electronic device (2) is recommended to the user via the said device (2) FEA (Finite Element Analysis) GPU Simulation, Material DB in operations: PLA / ABS / CFRP / Aluminium / Stainless Composite, Structural Stress Solver and Buckling detection, Cost estimator: Monte Carlo Manufacturing Cost AI-shaped 20 characterized by the server (3) configured to enable the use of techniques an AI-based system, as in any of the above requests Augmented or virtual reality design adaptation system (1).

17. The 3D model in question within the materials engineering analysis module. Run Finite Element Analysis (FEA) on it, at the specified load / force 25 To enable the simulation of stress, deformation, and fracture risk according to the scenarios, such as part thickness, infill ratio, and curvature / radius in the 3D model in question. at least the evaluation of geometric parameters according to production techniques. Selecting the appropriate material from a material database, 3D rendering from 0–100 The model's manufacturability score is calculated by specifying, "this area should be thickened," "this 30..." feedback should be provided on the relevant 3D design, such as "the junction radius should be increased". the above characterized by the server (3) configured to provide an AI-based augmented or virtual system, as in any of the requests. reality design adaptation system (1).

18. Simulations on the outputs of the materials engineering analysis module and To ensure that the simulation and technical reporting are forwarded to the technical reporting module. The module provides the output format information selected by the user from the electronic device (2). to use, thereby enabling automatic report generation for the user, production techniques drawings, step-by-step construction documentation, bill of materials, and prototype 10 the preparation of variants, as well as simulation animations, breakpoints System Research Document (SRD) containing test results and development recommendations the server (3) configured to enable the creation of the document artificial intelligence as in any of the above characterized requests an augmented or virtual reality design adaptation system based on (1). 15 19. Simulation on the outputs of the materials engineering analysis module and processing in the technical reporting module, 3D model for production during processing Technical requirements include converting the data to the appropriate formats such as STEP, IGES, STL, FBX, and GLB / GLT. Creating drawings and dimensioning plans, 20 derived from simulations Report with results (stress maps, deformation animations, heat distribution, etc.) Production of files and, if desired, video animations, all of this in CAD files, bill of materials (BOM), production instructions, simulation visuals, and development proposals are packaged under the name "FinalDesignPack" 25 with server (3) configured to enable the execution of actions artificial intelligence as in any of the above characterized requests an augmented or virtual reality design adaptation system based on (1).