Related Experiment Video
Updated: Aug 15, 2026
07:56
A Photonic System for Generating Unconditional Polarization-Entangled Photons Based on Multiple Quantum Interference
Published on: September 5, 2019
Structural design of an improved SPIDER optical system based on multimode interference coupler: an update
Optics Express
|August 14, 2026
Summary
This study introduces a compact synthetic aperture imaging system using photonic integrated circuits. A reduced-output design reveals a null space, addressed by a novel reconstruction framework for improved image quality.
Area of Science:
- Photonics
- Optical Engineering
- Signal Processing
Background:
- Segmented Planar Imaging Detector for Electro-Optical Reconnaissance (SPIDER) utilizes photonic integrated circuits (PICs) for compact synthetic aperture imaging.
- Previous work established an 8x8 multimode interference (MMI) coupler as a three-input primitive for simultaneous mutual intensity extraction across three baselines.
- This approach replaces conventional two-aperture, one-baseline measurement units, enhancing imaging capabilities.
Purpose of the Study:
- To investigate a reduced-output 3x6 MMI-SPIDER architecture.
- To characterize the information structure of this reduced system using a unified design matrix framework.
- To develop and evaluate a reconstruction strategy for estimating missing spectral components and improving image quality.
Main Methods:
- Formulation of any three-input MMI system with N output ports as a real-valued linear model M^(N)v = y.
- Analysis of system solvability based on the rank of the design matrix M^(N).
- Development of a two-stage reconstruction framework: iterative spectral projection with wavelet sparsity (ISP-WS) followed by deep prior refinement (DPR).
Main Results:
- The 8x8 system with full-rank port combinations allows for closed-form solutions.
- The reduced 3x6 system exhibits a one-dimensional null space related to the real part of a mutual-intensity component, governed by Type-I port geometry.
- A port randomization strategy transforms spectral blind zones into a dispersed half-sampling pattern, enabling effective spectral inpainting.
Conclusions:
- The reduced MMI-SPIDER architecture's information structure is well-defined and predictable.
- The proposed ISP-WS and DPR reconstruction framework successfully estimates missing spectral data and suppresses artifacts.
- Numerical simulations demonstrate target-dependent image quality improvements compared to direct-inverse methods across various image categories.
