User-Selected Clothing Image Compositing for Virtual Dressing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing online shopping platforms lack the ability for customers to dynamically create properly-proportioned images with customer-selected clothing combinations, and the one-click paradigm only allows for the purchase of one item at a time, making it difficult to visualize how items will look when worn and to facilitate multiple purchases simultaneously.
Innovation Solution
A system and method for processing digital images to identify and interpolate hidden areas, fill in regions with color or texture, and generate composite images that depict combined subject matters, allowing users to virtually dress themselves with selected products and enable a single-click purchase mechanism for multiple items.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If fixed images of human models wearing outfits are displayed on clothing vendor websites, then customers can see how outfits look when worn, but customers cannot dynamically create properly-proportioned images with customer-selected clothing combinations or visualize themselves wearing different combinations
Solution Approach 1:
The patent uses image copying and compositing techniques to transfer clothing items from source images onto model images. The system extracts clothing items from vendor images and overlays them onto standardized model templates, creating customized try-on images without requiring complex 3D modeling or physical garments.
Solution Approach 2:
The system extracts specific elements (clothing items) from complex source images by identifying and separating the garment from the background and model. This extraction allows the clothing to be reused and recombined with different models and configurations to create customized combinations.
2Productivity
If the shopping cart paradigm is used for online purchasing, then multiple items can be selected for purchase, but the process requires multiple steps including adding items to cart, reviewing, and checking out
Solution Approach 1:
The system performs preliminary actions by pre-configuring the purchase process. When a customer completes a try-on session, the system has already gathered all necessary purchase information (item details, pricing, customer data) and can immediately execute the transaction without requiring separate cart management steps.
Solution Approach 2:
The patent merges the try-on visualization function with the purchasing function into a single integrated workflow. The same interface used to visualize clothing combinations also serves as the purchase interface, combining what were previously separate functions (browsing/visualizing and purchasing) into one unified process.
3Ease of operation
If the one-click paradigm is used for purchasing, then a single mouse-click operation is required, but only one item can be purchased per mouse-click and multiple items cannot be purchased simultaneously
Solution Approach 1:
The purchase button serves multiple functions simultaneously: it acts as both a visualization confirmation (agreeing to the try-on result) and a multi-item purchase trigger. This universal button handles both single and multiple item purchases, as well as coordinating purchases across different vendors, without requiring separate operations for each function.
4Loss of information
If clothing items are displayed separately or in random combinations on vendor websites, then all clothing items can be showcased, but it is difficult to visualize how items will look when actually worn in combination
Solution Approach 1:
The system introduces standardized model images as intermediaries between the clothing items and the customer's self-image. These model templates serve as a neutral canvas that receives clothing overlays, allowing customers to visualize combinations without requiring the system to directly manipulate or edit the customer's actual photograph in complex ways.
Data Source
AI summary
A system including one or more processors and one or more non-transitory computer-readable media storing computing instructions that, when executed on the one or more processors, cause the one or more processors to perform operations: identifying a first subject matter in a first digital image; defining shapes for one or more regions in a second digital image of a second subject matter, wherein the one or more regions comprise at least one hidden area of the second subject matter that is obstructed by one or more objects; overlaying at least a first region of the one or more regions in the second digital image with another portion of the second subject matter; generating a composite image that depicts the first subject matter combined with the second subject matter having the at least the first region overlaid; and causing the composite image to be displayed. Other embodiments are described.