Virtual Apparel Fitting Using 3D Body Mesh and Fit Overlay
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge of selecting clothing items online that fit well is exacerbated by sizing discrepancies between brands and styles, leading to inconvenient returns and dissatisfaction due to the inability to physically try on clothes, assess color, and predict fit accurately.
Innovation Solution
A machine-learning model processes a subject's image to determine dimensions, generates a mesh model, and overlays a selected clothing item to create a composite image showing fit, using heat maps to indicate tight or loose areas, thereby enhancing online shopping confidence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If customers purchase clothing online without physically trying on items, then shopping convenience is improved, but fit accuracy deteriorates
Solution Approach 1:
The patent creates a digital copy (3D mesh model) of the customer's body from 2D images, enabling virtual try-on experiences that replicate the physical fitting process. This digital twin allows customers to visualize clothing fit online without visiting physical stores, resolving the contradiction between shopping convenience and fit accuracy.
Solution Approach 2:
The patent introduces an intermediary system (machine learning model with mesh generation and clothing simulation) that bridges the gap between online shopping and physical fitting. This intermediary processes customer images, generates body models, simulates clothing drape and fit, and provides feedback, enabling accurate fit assessment remotely.
2Loss of information
If sizing charts are provided for online shopping, then information availability is improved, but reliability deteriorates due to brand variations
Solution Approach 1:
The patent transforms sizing from standardized brand charts to personalized body measurements derived from customer images. By extracting actual body dimensions (chest, waist, hips, inseam) and matching them to garment specifications, the system adapts sizing parameters to individual customers, eliminating brand variation issues.
Solution Approach 2:
The patent performs preliminary body measurement and analysis before the customer makes a purchase decision. By pre-generating the 3D mesh model and pre-simulating clothing fit, the system provides fit information in advance, allowing customers to make informed decisions without relying on unreliable sizing charts.
3Measurement precision
If machine learning models process customer images to determine fit, then fit prediction accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent segments the complex fit prediction task into distinct modules: image processing and body measurement extraction, 3D mesh model generation, clothing item analysis, virtual try-on simulation, and fit assessment. This segmentation allows each component to be optimized independently and processed efficiently.
Solution Approach 2:
The patent replaces complex physical fitting mechanics with computational models. Instead of requiring physical mannequins or multiple measurement devices, the system uses machine learning algorithms to infer body dimensions from 2D images and simulate clothing drape through computational physics, reducing hardware complexity while maintaining accuracy.
Data Source
AI summary
In one implementation of remote apparel fitting, a processing device receives an input image that depicts a subject person (e.g., an online consumer). A selection of a clothing item is also received. For example, the subject person is browsing an online catalog of clothing items to find clothing items (e.g., shirts) that fit well. A machine-learning model uses the image to determine measurements of the subject person that correlate to one or more dimensions of the clothing item. In some implementations, the machine-learning model determines the measurements after generating a mesh model of the subject person. The machine-learning model is then used to determine the fit of the clothing item on the subject person. The processing device then presents a composite image that represents the fit of the clothing item on the subject person overlayed on a reproduced image of the subject person wearing the clothing item.