Scattered light field separation and high-quality imaging based on low-rank sparse matrix decomposition

• Published in: Proceedings of SPIE, the international society for optical engineering Vol. 13418; pp. 134180S - 134180S-6
• Main Authors: Liu, Kang, Wang, Pinghe
• Format: Conference Proceeding
• Language: English
• Published: SPIE 05.12.2024
• ISBN: 9781510686168; 1510686169
• ISSN: 0277-786X
• DOI: 10.1117/12.3046629

The absorption and scattering properties of light in biological tissues reflect their key characteristics. However, strong background light often obscures these properties. Currently, contrast enhancement in imaging primarily focuses on noise reduction from an image processing perspective rather than considering the properties of the light field itself. This approach has limitations in preserving imaging details and information, particularly under strong background light conditions, leading to information loss. This paper introduces an imaging technique based on the Low-rank Sparse Matrix Decomposition (LSMD) to achieve high-quality imaging of biological tissues. Experiments in Gaussian and ring-shaped light fields validate that the low-rank component of the light field represents background light, while the sparse component indicates the scattering and absorption properties of the target. This technique effectively highlights microscopic details and scattering/absorption characteristics. Compared to traditional image-based filtering methods, this light field-based technique significantly improves the extraction of absorption and scattering information. It has the potential to enhance tissue visualization in medical diagnostics and research.