Image Matting With Variable-Width Trimaps for Precise Edges

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image matting methods often result in rough edges of the foreground region due to assigning a fixed category to each pixel, leading to a poor image matting effect.

Innovation Solution

Perform segmentation on a target object to obtain multiple segmented images, generate a target trimap with different line widths corresponding to various sub-regions of the foreground, and perform image matting processing based on this trimap to improve precision.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If fixed category segmentation is applied to each pixel, then segmentation speed is improved, but edge precision deteriorates resulting in rough foreground boundaries

Engineering Contradiction:
Improvesegmentation speedVSAvoidedge precision
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The patent divides the foreground region into multiple sub-regions (first target region and second target region) with different segmentation requirements. The first target region contains pixels requiring high precision edge processing, while the second target region contains pixels that can use faster processing. This hierarchical segmentation allows different processing strategies to be applied to different areas, resolving the contradiction between speed and precision.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different processing qualities to different local regions: high-precision processing is applied to the first target region (edge areas) while standard processing is applied to the second target region (internal areas). This local differentiation ensures that edge precision is maintained where needed without sacrificing overall processing efficiency.

Inventive Principle:
Principle #3Local quality

2Manufacturing precision

If multiple segmented images with different line widths are generated, then image matting precision is improved, but device complexity increases

Engineering Contradiction:
Improveimage matting precisionVSAvoidprocessing complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent segments the foreground region into multiple sub-regions and generates different trimmed images for each sub-region with appropriately differentiated line widths. This segmentation approach allows precise control over edge rendering in different areas while maintaining an organized processing framework that manages complexity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent varies the line width parameter across different trimmed images corresponding to different sub-regions. By changing this single parameter (line width) rather than the entire processing pipeline, the patent achieves improved matting precision while minimizing the increase in system complexity.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4276754B1Image processing method and apparatus, device, storage medium, and computer program product
Publication Date: 2026.04.08 TENCENT TECHNOLOGY (SHENZHEN) CO LTD
  • EP4276754B1 patent drawingFigure 1
  • EP4276754B1 patent drawingFigure 2
  • EP4276754B1 patent drawingFigure 3~4

AI summary

An image processing method is provided. During image matting on an original image, a plurality of segmented images including different regions are first obtained through semantic segmentation. Further, according to the segmented images, lines of different widths are drawn on a contour line of a foreground region to obtain a target trimap. Finally, a target image is generated based on the target trimap. For the target trimap, because lines of different widths are drawn on the contour line of the foreground region, pertinent image matting can be implemented for different regions, so that image matting precision can be improved for a region requiring fine image matting, and image matting precision of another region can also be ensured. In this way, a final image obtained through image matting is fine and natural.