MVMFCam: an all-in-focus optical synthetic aperture imaging system based on multi-view multi-focus computational...

Area of Science:

• Computational Imaging
• Optical Systems
• Computer Vision

Background:

• Traditional imaging systems have limited depth of field, blurring objects outside the focal plane.
• Existing multi-focus imaging methods capture images sequentially, unsuitable for dynamic scenes.

Purpose of the Study:

• To develop a novel all-in-focus imaging system capable of capturing dynamic scenes in a single exposure.
• To introduce an effective neural network for fusing multi-view multi-focus images.

Main Methods:

• Proposed MVMFCam system: a nine-sub-camera array capturing multi-view multi-focus (MVMF) images simultaneously.
• Developed MVMF-Net: an end-to-end neural network for image alignment and adaptive fusion.
• Image alignment uses feature transfer matching; fusion employs a densely connected network with adaptive weights.

Main Results:

• MVMFCam successfully captured high-quality all-in-focus images in dynamic scenes.
• The MVMF-Net achieved accurate alignment and adaptive fusion, producing superior all-in-focus results.
• Experimental datasets validated the system's effectiveness.

Conclusions:

• MVMFCam enables real-time all-in-focus imaging for dynamic scenes.
• The proposed MVMF fusion network significantly enhances image quality.
• This technology has broad applications in microscopic imaging, close-range photography, and non-destructive testing.