Facial Model Deformation for Marker-Free Eyelid Motion Capture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for facial animation in computer graphics either require manual artist intervention or fail to accurately reproduce face motions due to marker attachment limitations, particularly in low reproduction areas like the eyelid.
Innovation Solution
An information processing apparatus that deforms a face model using marker positions and contour detection to generate high-quality facial animations, incorporating contour position detection and deformation units to enhance motion reproduction in low reproduction areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If motion capture with markers is used to deform polygon mesh directly, then cost is reduced and high-quality animation is achieved, but portions without markers cannot be reproduced accurately
Solution Approach 1:
The patent segments the face model into multiple regions based on marker attachment positions. The deformation process is divided into two stages: first deforming the entire face model based on marker positions, then selectively refining low-reproduction portions (areas without markers) in a second deformation process. This segmentation allows different deformation strategies for different face regions.
Solution Approach 2:
The patent applies local quality by identifying low-reproduction portions (regions without markers) and applying specialized deformation processing only to those specific areas. The second deformation execution unit targets only the low-reproduction portions, adjusting their shapes based on contour information from the original face image, while leaving marker-based regions unchanged.
2Manufacturing precision
If markers are attached to all face portions for accurate motion capture, then motion reproduction accuracy is improved, but the complexity of marker attachment and processing increases
Solution Approach 1:
The patent introduces contour information from the original face image as an intermediary for regions without markers. Instead of requiring physical markers on every face portion, the system uses the contour data as a mediator to guide the deformation of low-reproduction portions, enabling accurate representation of unmarked areas through computational processing.
Solution Approach 2:
The patent creates a copy of the contour information from the original face image and uses this copied data to guide the deformation of low-reproduction portions. The contour copy serves as a reference pattern that is applied to the corresponding regions in the deformed face model, ensuring accuracy without requiring physical markers.
3Manufacturing precision
If manual rigging is used to construct facial animation mechanisms, then realistic reproduction is achieved, but extensive manual work and trial-and-error are required
Solution Approach 1:
The patent enables the system to self-adjust by automatically identifying low-reproduction portions and performing secondary deformation on them. The system serves itself by detecting which areas need improvement and autonomously applying corrective deformation based on contour information, eliminating the need for manual artist intervention in refining specific face regions.
Solution Approach 2:
The patent performs preliminary deformation of the entire face model based on marker positions before conducting the second deformation on low-reproduction portions. This preliminary action establishes a baseline deformation that captures most motions, allowing the subsequent refinement process to focus only on problematic areas and reduce overall processing time.
Data Source
AI summary
The information processing apparatus includes a first deformation execution unit and a second deformation execution unit. The first deformation execution unit deforms a face model of an actor on the basis of positions of a plurality of markers in a face image of the actor to which the plurality of markers are attached, and generates a first deformed face model. The second deformation execution unit deforms a shape of a low reproduction portion of the first deformed face model based on the face image of the actor so that a position of a contour of the low reproduction portion having relatively low reproducibility in the first deformed face model matches a position of a contour of the low reproduction portion in the face image of the actor, and generates a second deformed face model.