Projection-based disparity control for toed-in multiview images.

Area of Science:

• Computer Vision
• 3D Imaging
• Stereoscopic Displays

Background:

• Multiview images from toed-in camera arrays offer vivid 3D scene reproduction.
• Adjusting scene depth in these images for specific visual effects is challenging.
• Existing methods lack sufficient control over disparity for desired depth manipulation.

Purpose of the Study:

• To propose a novel disparity control method for toed-in multiview images.
• To enable adjustable scene depth for enhanced visual effects.
• To improve the stereoscopic visual quality of processed multiview images.

Main Methods:

• A projection-based disparity control method is introduced.
• The zero disparity point from a reference view is projected to other views.
• Views are shifted based on calculated disparities derived from matched corresponding points.

Main Results:

• The proposed method effectively controls the depth of toed-in multiview images.
• Experiment results demonstrate superior performance compared to conventional methods.
• Processed images exhibit desirable stereoscopic visual quality.

Conclusions:

• The novel disparity control method is effective and easy to implement.
• Scene depth adjustment in toed-in multiview images is achievable as required.
• The technique enhances stereoscopic visual quality for various applications.