Guided Lensless Polarization Imaging

2026-03-28

AI Summary

This work addresses the limitations of conventional lensless polarimetric imaging systems, which often suffer from poor reconstruction quality and rely on expensive, bulky hardware. The authors propose an RGB-guided lensless polarimetric imaging method that, for the first time, leverages an external RGB image as a structural prior. By integrating a compact polarization-RGB sensor with an off-the-shelf RGB camera, the approach combines physics-based initial reconstruction with a Transformer-based fusion network for refinement, enabling high-quality recovery of multi-angle polarimetric images. Notably, the method requires no fine-tuning for real-world hardware and consistently outperforms purely lensless baselines across multiple datasets and real-world scenes. The resulting prototype system achieves high-fidelity, high-resolution polarimetric imaging while demonstrating strong generalization and practical applicability.

Abstract

Polarization imaging captures the polarization state of light, revealing information invisible to the human eye yet valuable in domains such as biomedical diagnostics, autonomous driving, and remote sensing. However, conventional polarization cameras are often expensive, bulky, or both, limiting their practical use. Lensless imaging offers a compact, low-cost alternative by replacing the lens with a simple optical element like a diffuser and performing computational reconstruction, but existing lensless polarization systems suffer from limited reconstruction quality. To overcome these limitations, we introduce a RGB-guided lensless polarization imaging system that combines a compact polarization-RGB sensor with an auxiliary, widely available conventional RGB camera providing structural guidance. We reconstruct multi-angle polarization images for each RGB color channel through a two-stage pipeline: a physics-based inversion recovers an initial polarization image, followed by a Transformer-based fusion network that refines this reconstruction using the RGB guidance image from the conventional RGB camera. Our two-stage method significantly improves reconstruction quality and fidelity over lensless-only baselines, generalizes across datasets and imaging conditions, and achieves high-quality real-world results on our physical prototype lensless camera without any fine-tuning.

Problem

• lensless imaging
• polarization imaging
• reconstruction quality
• computational imaging
• RGB guidance

Innovation

• lensless imaging
• polarization imaging
• RGB guidance
• Transformer-based fusion
• computational reconstruction