Refocusing distance of a standard plenoptic camera

Area of Science:

• Computational Photography
• Optical Systems

Background:

• Plenoptic cameras allow post-capture refocusing.
• Existing ray models inaccurately predict refocused distances.

Purpose of the Study:

• Develop a precise method for estimating object distances in refocused plenoptic images.
• Enhance depth map recovery and synthetic focus applications.

Main Methods:

• Model light rays as linear functions to find their intersection point.
• Experimental validation using various lenses and focus settings.
• Devise a blur metric to compare estimated distances with refocused images.

Main Results:

• The proposed method accurately predicts refocused object plane distances.
• Quantitative assessments show prediction deviations under 0.35% over a 24m range.
• Results are comparable to optical design software predictions.

Conclusions:

• The refocusing estimator enhances object distance prediction for plenoptic cameras.
• Essential for high-precision synthetic focus and depth map recovery.
• Aids in the prototyping stage of plenoptic camera development.