3D Face Model Rotation for Portrait Image Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current portrait editing techniques lack the ability to adjust the angle of a face in captured images, limiting face enhancement options.
Innovation Solution
A method and device for image processing that involves receiving an image with a face, detecting and establishing a three-dimensional model, rotating it by a specified angle, and projecting it back into the image coordinate system, while also processing the background to maintain a seamless integration with the face adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional portrait editing techniques are used, then face texture editing (removal of spots, wrinkles, skin whitening, smoothing) can be performed, but the face angle cannot be adjusted
Solution Approach 1:
The patent transforms the two-dimensional image processing into three-dimensional space by establishing a three-dimensional face model from the image, rotating this 3D model to change the face angle, and then projecting it back to the image coordinate system. This dimensional transition enables face angle adjustment while maintaining image consistency.
Solution Approach 2:
The patent segments the image processing into distinct modules: face detection, three-dimensional model establishment, rotation, and projection back to image coordinates. By dividing the complex task into separate processing stages, the system manages complexity through structured organization while achieving the new functionality.
2Manufacturing precision
If the three-dimensional model is rotated to adjust face angle, then the face appearance is enhanced, but the background region must also be processed to maintain seamless integration
Solution Approach 1:
The background processing also utilizes three-dimensional transformation by establishing a three-dimensional background model, rotating it by the same angle as the face model, and projecting it back to the image coordinate system. This ensures the background moves consistently with the face angle adjustment, maintaining spatial relationships and visual seamless integration.
Solution Approach 2:
The patent merges the face processing and background processing into a unified workflow where both the face model and background model are transformed together using the same rotation parameters. This coordinated transformation ensures consistency between the face and background, reducing the complexity of maintaining seamless integration.
Data Source
AI summary
A method and a device for image processing are disclosed. The method includes: receiving a first image, wherein the first image includes a face; detecting the face and a background region in the first image, establishing a three-dimensional model of the face according to the first image; rotating the three-dimensional model of the face by a first angle; and projecting the three-dimensional model of the face rotated by the first angle to an image coordinate system of the first image and fusing a face region with a processed background region to obtain a second image.