Interactive Image Recoloring With Scribble-Guided Region Selection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional image recoloring methods require substantial user intervention and computational resources, making them tedious and computationally expensive, especially on mobile devices.

Innovation Solution

A system that allows users to interactively recolor image regions using a single scribble and specified color, employing graphical model inference and preprocessing techniques to optimize computational efficiency and user interface flow.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional image recoloring methods are used, then recoloring accuracy can be maintained, but user intervention and computational resources increase significantly

Engineering Contradiction:

Improverecoloring accuracyVSAvoiduser intervention requirements

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs preprocessing on the input image before recoloring, including converting to appropriate color spaces and preparing data structures. This preliminary preparation enables the subsequent recoloring to proceed with minimal user intervention while maintaining accuracy.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The recoloring system automatically processes the image through multiple stages including segmentation, color transfer, and post-processing without requiring continuous user input. The system serves itself by making autonomous decisions about color mapping and region identification.

Inventive Principle:

Principle #25Self-service

2Manufacturing precision

If conventional image recoloring methods are used, then recoloring quality can be maintained, but computational processing time increases

Engineering Contradiction:

Improverecoloring qualityVSAvoidcomputational processing time

Core Design Contradiction:

Manufacturing precisionVSLoss of time

Solution Approach 1:

The image processing is divided into distinct segments: preprocessing, recoloring, and post-processing. Each segment handles specific tasks efficiently, allowing parallel processing and optimization of individual stages without compromising overall quality.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system transforms the image between different color spaces (RGB, LAB, HSV) and adjusts processing parameters dynamically based on image characteristics. This enables efficient computation while preserving color accuracy and visual quality.

Inventive Principle:

Principle #35Parameter changes

3Adaptability or versatility

If conventional image recoloring methods are used, then thorough color adjustment can be achieved, but computational resources and energy consumption increase

Engineering Contradiction:

Improvecolor adjustment capabilityVSAvoidcomputational energy consumption

Core Design Contradiction:

Adaptability or versatilityVSUse of energy by moving object

Solution Approach 1:

The system applies recoloring selectively to specific regions or portions of the image rather than processing the entire image uniformly. This partial action approach maintains color adjustment versatility while reducing overall computational energy consumption.

Inventive Principle:

Principle #16Partial or excessive action

Solution Approach 2:

The system replaces complex mechanical or manual color adjustment processes with automated computational algorithms. This substitution enables versatile color adjustment capabilities while significantly reducing the energy resources required compared to conventional methods.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS12517637B2Interactive image recoloring

Publication Date: 2026.01.06 SNAP INC

AI summary

Disclosed are systems, methods, and computer-readable storage media to perform an interactive image recolorization process. The method includes receiving user input including a stroke drawn on an image presented on a client device. The stroke comprises a user-specified color. The method further includes determining a region of interest in the image. The method further includes recolorizing the region of interest on the image based on the user-specified color and causing presentation of a result of the recolorization on the client device.