3D Apparel Scanning with Mannequin Removal and Fold Line Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computer imaging systems are inadequate for generating accurate multiple 3D representations of apparel items for use in 3D and mixed reality products, failing to effectively depict items in various formats such as hanging, folded, and staged on a mannequin.
Innovation Solution
A computer imaging system and method that uses a single depth sensing scan and various apparel images to generate 3D representations with folding line indications, involving point cloud determination, high polygon model creation, mannequin pose estimation and removal, and folding line detection, enabling multiple illustrative formats like hanging, mannequin display, and folding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional computer imaging systems are used to scan apparel items, then the scanning process is simple, but the system cannot generate accurate multiple 3D representations in various formats (hanging, folded, mannequin-staged)
Solution Approach 1:
The system segments the apparel item detection process into distinct components: point cloud generation from depth sensing, high-polygon model creation, mannequin detection and removal, and fold line detection. Each component processes specific data aspects independently, enabling the system to generate multiple 3D representations (hanging, folded, mannequin-staged) from a single scan while managing complexity through modular processing stages.
2Manufacturing precision
If multiple separate scans are taken to create different 3D representations of apparel, then accurate representations can be achieved, but the scanning time and process complexity increase significantly
Solution Approach 1:
The system performs preliminary processing of the single depth scan by generating point clouds and high-polygon models that contain all necessary geometric information. Fold line detection and mannequin pose estimation are performed in advance on this comprehensive model, enabling the system to quickly generate multiple format representations (hanging, folded, mannequin-staged) without requiring additional scanning time, thus maintaining accuracy while reducing total processing time.
3Productivity
If a single depth sensing scan is used, then the scanning process is fast and simple, but the system lacks sufficient data to generate accurate 3D representations in multiple formats
Solution Approach 1:
The system transforms the single depth scan data into multiple dimensional representations by generating point clouds (3D spatial data), high-polygon models (detailed mesh structures), and extracting fold line information (2D pattern data). This multi-dimensional data extraction from a single scan source enables accurate generation of multiple 3D representations (hanging, folded, mannequin-staged) while maintaining fast scanning speed, as all necessary information is derived from the original single scan through sophisticated processing algorithms.
Data Source
AI summary
A computer system and process for scanning apparel items for generation of three-dimensional (3D) reproduced representations of scanned apparel items having folding line indications. An electronic scan is created of at least one apparel item by an imaging device. Point clouds are determined from the at least one scanned apparel item and at least one high polygon 3D model is determined from the point clouds. 3D translation and rotational camera reference images are generated which are associated with the at least one scanned apparel item. Next, a mannequin pose estimation and correction is determined utilizing the generated 3D translation and rotational camera reference images and mannequin detection and removal is then determined relative to the at least one scanned apparel item. Information (data) for indication of fold lines for folding the at least one scanned apparel item is the determined.