Physical origin and boundary of scalable imaging through scattering media: a deep learning-based exploration

• DOI: 10.1364/prj.490125
• Publication Date: 2023-04-14T15:00:40Z
• Authors: Xuyu Zhang; Shengfu Cheng; Jingjing Gao; Yu Gan; Chunyuan Song; Dawei Zhang; Songlin Zhuang; Shensheng Han; Puxiang Lai; Honglin Liu
• Keywords: 02 engineering and technology; 0210 nano-technology

Imaging through scattering media is valuable for many areas, such as biomedicine and communication. Recent progress enabled by deep learning (DL) has shown superiority especially in the model generalization. However, there a lack of research to physically reveal origin or define boundary scalability, which important utilizing DL approaches scalable imaging despite with high confidence. In this paper, we find amount ballistic light component output field prerequisite endowing generalization capability using “one-to-all” training strategy, offers physical meaning invariance among multisource data. The findings are supported both experimental simulated tests roles scattered components revealed contributing scalability. Experimentally, performance network enhanced increasing portion photons detection. mechanism understanding practical guidance our beneficial developing methods descattering adaptivity.