Virtual 3D Model Creation via 2D Scan Vector Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for generating virtual 3D models of real objects using imaging techniques, such as CT scans, cannot be processed to automatically produce tailor-made products, as they lack real-time three-dimensional viewing and editing capabilities, and voxel-based models are limited in resolution and flexibility.
Innovation Solution
A computer-implemented method that uses 3D scanners to create a virtual 3D model by converting slice-by-slice 2D scans into vector format, then into voxel format, allowing for precise manipulation and conversion of the model into a solid body, enabling real-time editing and processing for product creation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If voxel-based 3D models are used to represent three-dimensional space, then the model can represent equidistantly sampled space effectively, but the resolution becomes too coarse for many purposes and real-time processing is not possible
Solution Approach 1:
The patent segments the 3D space into two distinct data structures: voxel data for volumetric representation and grid data for surface representation. This segmentation allows each data structure to serve its optimal purpose - voxels for internal volume accuracy and grids for high-resolution surface detail, thereby resolving the contradiction between volumetric representation and surface resolution.
Solution Approach 2:
The patent transitions from purely volumetric voxel data to a hybrid representation that includes surface grid data. By adding this dimensional aspect (surface geometry), the system achieves both accurate volumetric sampling and high-resolution surface representation, overcoming the resolution limitation of voxel-only models.
2Ease of operation
If grid models consisting of triangles and quads are used to display virtual objects, then the number of nodes is predetermined, but excessively pulling out a node results in very long bars that can no longer be divided and deformed
Solution Approach 1:
The patent introduces an intermediary conversion process that transforms rigid grid models into flexible voxel models. This intermediary step allows the model to be rerepresented in a data structure that supports flexible deformation and editing without the topological constraints of predetermined grid nodes, thereby enabling node pulling and deformation operations.
3Quantity of substance
If 2D scans are recorded layer by layer to create virtual 3D models, then a large number of scans can be captured, but the complex calculation of three-dimensional view cannot take place in real time
Solution Approach 1:
The patent performs preliminary conversion of 2D scan layers into both voxel and grid data structures during the scanning process itself. By pre-processing the data into multiple representations, the system eliminates the need for complex real-time calculations when generating 3D views, allowing instantaneous rendering and interaction with the model.
4Measurement precision
If CT scans are used to generate virtual 3D models, then the imaging method can capture internal structures, but the model cannot be processed to automatically produce tailor-made products
Solution Approach 1:
The patent transforms the static CT scan data into dynamic, editable voxel and grid models that can be manipulated in real-time. This dynamic representation allows users to modify the model geometry, extract specific features, and automatically generate customized products based on the scanned internal structures, thereby enabling product adaptability while preserving measurement precision.
Data Source
Figure 1~4
AI summary
The method involves generating several layered two-dimensional scans in a pixel format by scanner, while identifying two-dimensional scans structures. The two-dimensional scans of pixel format are converted in vector format. The vectors are extended along the respective edges of the labeled structures. A grid is placed in respective edges of laminated vectorized two-dimensional scans, to connect layered two-dimensional scans in vertical direction. The grid is converted to voxel format, to create virtual solid with edges of selected structure as three-dimensional virtual model. An independent claim is included for medical, therapeutic or cosmetic product.