Iterative calibration of a multiview 3D display with linear extrinsic crosstalk using camera feedback

Area of Science:

• Computer Vision
• Display Technology
• Image Processing

Background:

• Multiview 3D displays require precise layer alignment for optimal image quality.
• Misalignment between display layers causes significant visual distortion, degrading the 3D viewing experience.

Purpose of the Study:

• To develop and validate an automated calibration algorithm for multiview 3D displays.
• To compensate for layer misalignment and minimize 3D image distortion.

Main Methods:

• An iterative algorithm utilizing camera feedback to detect and correct layer misalignment.
• Development of a mathematical model to predict distortion and define alignment-indicating features.
• Implementation of a compensation algorithm using cascaded one-dimensional root-finding blocks for 2D optimization.

Main Results:

• Experimental verification of the image distortion model and algorithm convergence.
• Demonstrated improvement in 3D image quality post-calibration.
• Achieved accuracy exceeding requirements for linear extrinsic crosstalk compensation.

Conclusions:

• The proposed algorithm effectively calibrates multiview 3D displays by compensating for layer misalignment.
• The iterative approach ensures accurate alignment and enhances overall 3D image fidelity.
• This method offers a robust solution for improving 3D display performance.