Numerical twin image suppression by nonlinear segmentation mask in digital holography

Area of Science:

• Optics and Photonics
• Digital Imaging
• Holographic Microscopy

Background:

• In-line holography offers wider spatial bandwidth compared to off-axis methods.
• High-quality reconstruction in in-line holography requires separation of direct current (DC)-noise and twin images.
• Existing twin image suppression methods have limitations, particularly in overlapped regions.

Purpose of the Study:

• To propose a novel approach for suppressing twin images in digital in-line holography.
• To enhance the quality of reconstructed images by effectively separating real and twin image components.
• To provide a robust and efficient method for digital holography applications.

Main Methods:

• Utilizing a blind separation matrix to obtain initial regions of real and twin images.
• Calculating a segmentation mask via nonlinear quantization from the segmented image to suppress the twin image.
• Comparing the proposed method with existing overlapping block variance and manual-based schemes.

Main Results:

• The proposed method demonstrates superior performance in suppressing twin images, especially in overlapped regions.
• The new approach results in less loss of the real image compared to the overlapping block variance-based scheme.
• Experimental validation confirms the effectiveness of the proposed twin image suppression technique.

Conclusions:

• The developed scheme is an effective tool for high-quality reconstruction in digital in-line holography.
• The method offers improved performance and reduced real image loss compared to existing techniques.
• This advancement contributes to the practical application of in-line holography in various scientific fields.