Freeform Gradient Reconstruction via Iterative Color Handle Refinement

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing technologies struggle to accurately extract and reproduce freeform gradients from images, leading to inaccurate representations and difficulties in modifying color gradients, especially for inexperienced users.

Innovation Solution

A digital design system that reconstructs freeform gradients by determining an outline of the input image, computing candidate color handles, and iteratively refining the gradient until the reconstruction error meets a threshold, allowing for accurate representation and modification of gradients.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If linear gradient extraction is used, then the process is simple, but only linear gradients can be captured and other gradient types cannot be reproduced

Engineering Contradiction:
Improvesimplicity of gradient extractionVSAvoidability to capture different gradient types
Core Design Contradiction:
Ease of manufactureVSAdaptability or versatility

Solution Approach 1:

The patent changes the parameters of gradient representation from linear constraints to freeform constraints by allowing color handles to be positioned at arbitrary locations and orientations, enabling the system to capture radial, angular, and other non-linear gradient patterns while maintaining an automated extraction process

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent introduces dynamic color handles that can be freely positioned and oriented along the gradient path, allowing the system to adapt to different gradient types (linear, radial, angular, freeform) rather than being constrained to a fixed linear structure

Inventive Principle:
Principle #15Dynamics

2Adaptability or versatility

If freeform gradients are manually created by establishing color handles, then gradient flexibility is achieved, but the process becomes difficult and time-consuming for inexperienced users

Engineering Contradiction:
Improvegradient flexibilityVSAvoidtime to create gradient
Core Design Contradiction:
Adaptability or versatilityVSLoss of time

Solution Approach 1:

The system performs automatic gradient extraction and color handle placement by analyzing the input image itself, allowing the gradient to reconstruct itself without requiring user intervention for handle positioning or color selection, thus saving time while maintaining flexibility

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system pre-computes the optimal positions and colors of color handles by analyzing the input image before the user needs to create the gradient, performing the complex task of gradient analysis in advance and presenting only the final result to the user

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If more color handles are added to refine gradient reconstruction, then reconstruction accuracy improves, but the complexity of the process increases

Engineering Contradiction:
Improvegradient reconstruction accuracyVSAvoidnumber of color handles
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system uses an iterative refinement process where each added color handle provides feedback on the reconstruction error, allowing the system to automatically determine when sufficient accuracy has been achieved and stop adding handles, thus balancing accuracy with complexity

Inventive Principle:
Principle #23Feedback

Data Source

PatentUS11699252B2Reconstructing freeform gradients from an input image
Publication Date: 2023.07.11 ADOBE INC
  • US11699252B2 patent drawing
  • US11699252B2 patent drawing
  • US11699252B2 patent drawing

AI summary

Embodiments are disclosed for reconstructing freeform gradients from an input image. In particular, in one or more embodiments, the disclosed systems and methods comprise receiving an input image, computing an outline of the input image, identifying a set of candidate color handles for the input image, each candidate color handle of the set of candidate color handles representing an extremum point for a color in the input image, generating a reconstructed image using a subset of the set of candidate color handles, determining a reconstruction error by computing a difference between the input image and the reconstructed image, and providing the reconstructed image when the reconstruction error is below a threshold value.