Closed Loop Object Fitting via Neural Network Image Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for manufacturing and adjusting closed loop objects, such as rings and bangles, do not allow for precise fitting over joints and can lead to increased manufacturing costs and delivery times.

Innovation Solution

A computer-implemented training method using a learning machine, such as a neural network, to analyze images of users' members and associate opening dimensions of closed loop objects with joint and wearing portion sizes, allowing for precise fitting and reduced manufacturing operations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If traditional methods are used to determine ring size, then manufacturing process is simple, but fitting precision is poor leading to loss or discomfort

Engineering Contradiction:
Improvefitting precisionVSAvoidsystem complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent replaces traditional mechanical measurement methods (physical ring sizers, manual measurement) with an optical imaging system and machine learning algorithm. The system captures images of the user's finger, processes them through a trained neural network, and predicts the optimal ring size automatically, substituting mechanical measurement with optical-digital processing to achieve higher precision.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent creates a digital copy (image) of the user's finger and processes this copy through the machine learning model to predict ring size. Instead of directly measuring the physical finger with mechanical tools, the system works with an optical replica, allowing for non-contact, high-precision measurement and analysis.

Inventive Principle:
Principle #26Copying

2Manufacturing precision

If physical tests and adjustments are performed, then fitting accuracy is improved, but delivery time and manufacturing cost increase

Engineering Contradiction:
Improvefitting accuracyVSAvoiddelivery time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent performs the fitting determination action in advance during the online ordering process itself. The machine learning model predicts the optimal ring size before the product is manufactured or shipped, eliminating the need for post-delivery adjustments or returns. This preliminary sizing action prevents future fitting issues and reduces delivery time.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system enables users to determine their own ring size accurately through the automated image capture and machine learning prediction process during online ordering. This self-service approach eliminates the need for manual intervention by sellers or customers for size adjustment, streamlining the entire process from ordering to delivery.

Inventive Principle:
Principle #25Self-service

3Productivity

If manual measurement and adjustment methods are used, then manufacturing cost is low, but productivity is reduced due to rework operations

Engineering Contradiction:
Improvemanufacturing efficiencyVSAvoidmanufacturing simplicity
Core Design Contradiction:
ProductivityVSEase of manufacture

Solution Approach 1:

The patent implements a feedback loop where the machine learning model is trained on actual fitting data and continuously improves its prediction accuracy. The system learns from real-world outcomes (whether rings fit properly or need adjustment) and uses this feedback to refine its sizing predictions, thereby reducing rework and improving manufacturing efficiency over time.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent changes the key parameter from manual measurement values to machine learning prediction values derived from image analysis. By transforming the sizing determination from a manual process to an automated algorithmic process, the system dramatically improves productivity while maintaining manufacturing simplicity through software-based solutions.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3764307B1Closed loop object fitting to user member
Publication Date: 2025.05.28 RICHEMONT INTERNATIONAL SA
  • EP3764307B1 patent drawingFigure 1~2
  • EP3764307B1 patent drawingFigure 3~4
  • EP3764307B1 patent drawingFigure 5~6

AI summary

The present invention provides a computer implemented method with a training phase to train a learning machine, and comprises a predicting phase for predicting a predicted fitting opening dimension of a closed loop object to be worn by a user. The training phase comprises the step of supplying to the learning machine pictures of individuals and opening dimensions of a closed loop object worn by the individuals. The predicting phase comprises the step a supplying a picture of a member of a user, and is followed by a step of manufacturing or adjusting or selecting a closed loop object having an opening sized according to the predicted fitting opening dimension.