Apparel Fit Calibration Using Reference Object
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Online apparel purchases often result in ill-fitting items due to inaccurate size grading rules and limited generic size charts, leading to consumer frustration, costly returns, and a lack of confidence in sizing accuracy.
Innovation Solution
A system that captures user body measurements using a reference object to calibrate camera images, calculates a total weighted variance to determine the correct apparel size, and evaluates fit quality by comparing user body shape to a fit model, providing personalized size recommendations and fit anomalies before purchase.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If generic size charts are used for online apparel purchases, then the purchasing process is simple and fast, but the fit accuracy deteriorates because they do not account for different body shapes and sizes
Solution Approach 1:
The system performs preliminary body measurement capture and analysis before the purchase decision. Users capture images of their body with a mobile device, and the system processes these images to generate personalized measurement data and size recommendations in advance, enabling informed purchasing decisions without requiring multiple trial purchases
Solution Approach 2:
The system transitions from using fixed, generic size chart parameters to dynamic, personalized measurement parameters. By capturing actual body measurements and comparing them against apparel size specifications, the system adapts the sizing process to individual body shapes and proportions, significantly improving fit accuracy
2Measurement precision
If apparel is purchased in multiple sizes to evaluate fit, then the fit accuracy improves, but the time consumption and cost increase significantly
Solution Approach 1:
The system performs the fit evaluation process in advance by capturing body measurements and comparing them against multiple apparel size options before purchase. This preliminary analysis identifies the most suitable size, eliminating the need for consumers to purchase and try multiple sizes physically
Solution Approach 2:
The system creates a digital copy of the consumer's body measurements and uses this virtual model to simulate fit evaluation against different apparel sizes. This digital twin approach allows accurate fit prediction without requiring physical trial of multiple garments
3Ease of operation
If body measurements are captured using uncalibrated camera images, then the measurement process is simple, but the measurement precision deteriorates due to camera rotation and positioning errors
Solution Approach 1:
The system introduces a reference object as an intermediary element in the measurement process. This reference object with known dimensions serves as a calibration standard, allowing the system to correct for camera rotation, distance, and positioning variations. The reference object acts as a mediator between the camera and the body being measured, enabling accurate measurements from uncalibrated camera images
Solution Approach 2:
The system dynamically adjusts measurement parameters based on the detected position and orientation of the reference object. By calculating transformation matrices from the reference object's known geometry and its captured image position, the system compensates for camera parameters and corrects measurement data to achieve accurate body measurements
4Productivity
If standard apparel grading rules are used to create sizes, then the manufacturing process is efficient, but the fit quality deteriorates because they do not account for reality of different body shapes
Solution Approach 1:
The system applies local quality by providing customized size recommendations for different body regions rather than using uniform sizing. By analyzing specific body measurements (waist, hips, chest, etc.) and comparing them against apparel size specifications, the system identifies which sizes will fit particular body areas well, accounting for regional variations in body shape
Data Source
AI summary
A reference object is used in measurement capturing features of the invention to calibrate a camera taking pictures of a user. The position of the reference object in the pictures of the user helps correct for rotations of the reference object that is held by the user. Once user measurements are captured, a total weighted variance is defined to identify correct size of apparel to the user. In addition to the variance, overall fit quality is determined by adding user body shape and fit anomalies of the apparel to the variance.