Virtual Dressing System Using Segmented Garment Layers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems fail to provide users with a realistic visualization of how clothing and accessories would appear on their own body, leading to difficulties in online purchasing due to the lack of personalized and accurate representations, resulting in user hesitation and potential returns.
Innovation Solution
A system utilizing image processing and computer vision to capture and analyze user images, allowing for virtual dressing of clothing and accessories, which are then superimposed onto the user's image, considering gravity, shading, and fit, enabling realistic simulations of how items would appear on the user's body.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If simple image overlay is used to display garments on user images, then implementation complexity is reduced, but realism and accuracy of the virtual dressing representation deteriorates
Solution Approach 1:
The system segments the garment image into multiple adjustable components (garment layer, shadow layer, highlight layer) that can be independently processed and positioned. This allows precise control over how the garment appears on the user image without requiring complex monolithic processing, resolving the contradiction between simplicity and accuracy.
Solution Approach 2:
The system applies different processing techniques to different regions of the garment-image composite. Shadowing is applied in specific areas to simulate depth, highlights are added to certain surfaces to mimic light interaction, and the garment is positioned and scaled locally to match the user's body proportions. This localized enhancement achieves realism without uniformly complicating the entire system.
2Manufacturing precision
If detailed image processing and machine learning are applied to achieve realistic virtual dressing, then visualization accuracy is improved, but computational complexity and processing time increases
Solution Approach 1:
The system performs preliminary actions by pre-processing garment images to extract key features, pre-calculating shadow patterns, and pre-identifying critical body points from user images. This preparation work is done before the actual virtual dressing composition, reducing the complexity of real-time processing while maintaining high realism in the final output.
Solution Approach 2:
The system introduces intermediary elements such as shadow layers and highlight layers that mediate between the simple garment overlay and the complex realistic rendering. These intermediary layers handle the computationally intensive aspects of lighting and depth simulation, allowing the core system to remain relatively simple while achieving photorealistic results.
3Measurement precision
If critical points are adjusted to match user body configuration, then garment fit accuracy is improved, but processing time and computational resources increase
Solution Approach 1:
The system applies partial action by focusing computational effort on identifying and adjusting only the critical points that most significantly affect garment fit (shoulders, waist, hips, etc.) rather than processing every point on the user's body. This selective approach achieves sufficient measurement precision for accurate garment fitting while minimizing processing time and computational resource requirements.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
System, device, and method of image processing, particularly utilizing machine learning and computer vision. A user selects a particular shirt from an online catalog; and instead of seeing an image of the shirt on the shelf or being worn by a model, the system generates and shows to the user an image of that particular shirt being virtually dressed by that particular user. The user requests the system to show search results for blue pants; and instead of producing generic images of blue pants, on the shelf or worn by various models, the system generates user-tailored search results that shown various types of blue pants being virtually worn by the user himself, or virtually worn by a particular other person that is selected by the user. The system enhances the product image by correcting it and by adding realistically-looking shading and wrinkles.