Reconstruction of partially occluded objects encoded in three-dimensional scenes by using digital holograms

Area of Science:

• Optics and Photonics
• 3D Imaging Technologies
• Computational Imaging

Background:

• Digital holography enables 3D reconstruction but faces challenges with data volume and occlusions.
• Existing methods often require prior knowledge of object distance or occlusion patterns.

Purpose of the Study:

• To develop a task-specific digital holographic capture system for efficient 3D scene reconstruction.
• To address challenges of data reduction and reconstruction of partially occluded objects.
• To enable reconstruction without a priori knowledge of occlusion or object distance.

Main Methods:

• A task-specific digital holographic capture system is proposed.
• Subwindows of the camera-plane Fresnel field are digitally propagated.
• Reconstructions from subwindows are combined to overcome unknown foreground occlusions.
• The Wigner distribution function is utilized for analyzing occlusions and subwindow effects.

Main Results:

• The system effectively reconstructs partially occluded objects.
• Data transmission from the camera system to the receiver is reduced.
• Combining reconstructions from subwindows reveals features not apparent in whole-hologram reconstructions.
• Successful demonstrations were achieved with optically captured holograms of real-world objects and simulated occlusions.

Conclusions:

• The proposed system offers an effective solution for 3D scene reconstruction with reduced data and improved handling of occlusions.
• Task-specific digital holography provides a pathway to overcome limitations of traditional holographic imaging.
• The method demonstrates the potential for reconstructing complex, partially occluded scenes.