AI Image Keying With Chroma-Based Transparency Blending

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing image keying technologies often result in the loss of edge details, poor highlight and shadow effects, inaccurate processing of semitransparent objects, and require extensive user operations, leading to low image processing accuracy and quality.

Innovation Solution

An artificial-intelligence-based method that determines a background color and blending transparency using chroma differences between pixels and the background color, performing a smoother image blending process to retain original edge details and improve composite image quality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If image keying is performed using tolerance range-based methods, then the keying process is simple to implement, but edge details are lost and image processing accuracy is low

Engineering Contradiction:

Improveease of keying implementationVSAvoidimage processing accuracy

Core Design Contradiction:

Ease of manufactureVSManufacturing precision

Solution Approach 1:

The patent changes the parameter from fixed tolerance range to dynamic chroma difference calculation. By computing chroma difference between each pixel and background color, and using this to determine blending transparency, the system achieves smooth transitions and retains edge details while maintaining ease of use through automated processing

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces the mechanical tolerance-range-based keying system with an AI-driven chroma difference calculation system. This substitution enables more precise and intelligent image keying that automatically adapts to different images without requiring manual tolerance adjustment

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If traditional image keying is used, then processing speed is fast, but highlight and shadow effects are poor

Engineering Contradiction:

Improveprocessing speedVSAvoidhighlight and shadow quality

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent applies local quality by calculating chroma difference and blending transparency for each pixel individually. This pixel-level analysis enables different regions of the image to be processed with appropriate attention, preserving highlight and shadow effects while maintaining overall processing efficiency through parallel computation

Inventive Principle:

Principle #3Local quality

3Ease of operation

If tolerance range keying is applied, then the method is simple and quick, but semitransparent objects are processed inaccurately

Engineering Contradiction:

Improveoperation simplicityVSAvoidsemitransparent object accuracy

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The patent changes from binary tolerance-based classification to continuous chroma difference-based transparency calculation. This enables accurate processing of semitransparent objects by calculating blending transparency as a continuous value rather than a binary mask, while maintaining operational simplicity through automated processing

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS12536618B2Artificial-intelligence-based image processing method and apparatus, electronic device, computer-readable storage medium, and computer program product

Publication Date: 2026.01.27 TENCENT TECHNOLOGY (SHENZHEN) CO LTD

AI summary

This application provides an artificial-intelligence-based image processing method and apparatus, an electronic device, a computer-readable storage medium, and a computer program product. The method includes determining a background color in an image to be processed; determining a blending transparency of the pixel according to a chroma difference between a pixel and the background color in the image to be processed; obtaining a background image corresponding to the image to be processed; and performing in pixels an image blending process on the image to be processed and the background image according to the blending transparency of the pixel to obtain a composite image.