3D Clothing Model Generation via Point Cloud Difference Tables
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for creating three-dimensional models of clothing are labor-intensive, time-consuming, and require skilled technicians, making it inefficient to generate 3D models for a large number of stock keeping units (SKUs) with variations in sizes and materials, especially in online retail where customers cannot physically try on clothing before purchasing.
Innovation Solution
A system and method that captures a 3D point cloud of a reference mannequin and the mannequin wearing clothing using depth or image sensors, processing the data to create a difference table for the wearable, allowing for simplified and parallel processing to generate 3D clothing models quickly and efficiently without the need for complex physical simulations or skilled operators.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional methods are used to create 3D clothing models, then model accuracy and detail can be achieved, but the process becomes labor-intensive and time-consuming requiring skilled technicians
Solution Approach 1:
The patent uses photogrammetry to capture multiple 2D images of a physical mannequin wearing clothing and automatically generates a 3D model by processing these images. This copying approach replaces manual 3D scanning and modeling, eliminating the need for skilled technicians while maintaining model accuracy through automated image processing algorithms.
Solution Approach 2:
The patent replaces manual mechanical operations (skilled technicians physically creating 3D models) with an automated optical and computational system. The system uses cameras to capture images and software to automatically generate 3D models, substituting human skill with automated image processing and computer vision algorithms.
2Reliability
If physical stores are maintained to allow customers to try on clothing, then customer satisfaction and fit verification improve, but retail costs and complexity increase
Solution Approach 1:
The patent creates virtual try-on experiences by generating accurate 3D models of clothing that can be digitally overlaid on customer images or avatars. This virtual copying of the physical try-on experience allows customers to verify fit and appearance online without visiting physical stores, eliminating the need for extensive inventory in multiple locations.
Solution Approach 2:
The patent enables dynamic adjustment of clothing parameters (size, color, style) in the virtual try-on system, allowing customers to explore different variations without physical handling. This parameter-based approach replaces physical inventory management with digital configuration options.
3Adaptability or versatility
If centralized warehouses are used for online fulfillment, then inventory consolidation and SKU variety improve, but the ability to provide pre-sales fitting is lost
Solution Approach 1:
The patent creates virtual copies of the fitting experience through 3D modeling and augmented reality technologies. Customers can visualize clothing on their own images or digital avatars, receiving the benefit of pre-sales fitting guidance without needing physical proximity to inventory. This virtual copying bridges the gap between centralized warehousing and customer fitting needs.
Data Source
AI summary
A system for creating a model of an article of clothing or other wearable, the system includes a mannequin or other model of at least a portion of a human form; a sensing device configured to scan the mannequin without the wearable to generate a first scan information and configured to scan the surface of the wearable on the mannequin to generate a second scan information; a processor communicatively coupled to the sensing device to receive the first and second scan information, the processor configured to: generate point clouds using the scan information; aligning the point clouds; generating a plurality of slices along at least one longitudinal axis through the point clouds, each slice having a centroid along a corresponding longitudinal axis; and generating a table having a plurality of entries each representing a distance between corresponding vertices for the pair of point clouds; the table representing the wearable.