Unfocused plenoptic metric modeling and calibration.

Area of Science:

• Optics and Photonics
• Computer Vision
• Metrology

Background:

• Unfocused plenoptic imaging systems require metric calibration for high-quality imaging and metrology.
• Existing models may not fully account for imaging distortions inherent in these systems.

Purpose of the Study:

• To derive an explicit unfocused plenoptic metric model.
• To investigate and model the impact of imaging distortions on depth computation.
• To establish a complete metric model combining explicit modeling and distortion analysis.

Main Methods:

• Derivation of an explicit unfocused plenoptic metric model.
• Experimental exploration of imaging distortion effects on depth computation.
• Modeling radial distortion parameters as depth distortions.
• Development of a three-step calibration strategy using the Levenberg-Marquardt algorithm.

Main Results:

• A complete unfocused plenoptic metric model was established by integrating explicit and distortion models.
• Radial distortion was identified to contain depth-dependent factors, modeled as depth distortions.
• A calibration strategy was proposed to determine 12 internal parameters per microlens unit.
• Depth measurement precision improved to 0.25 mm within a 300 mm depth range.

Conclusions:

• The proposed metric model and calibration strategy enhance the accuracy of unfocused plenoptic imaging.
• The findings demonstrate the potential of consumer unfocused plenoptic cameras for high-accuracy 3D measurements.
• Accurate depth computation is achievable through comprehensive modeling of imaging properties and distortions.