The extreme performance of video snapshot compressive imaging under system noise constraints

Open Webpage

Yanxin Cai, Xing Liu, Zhibin Wang, Xun Liu, Wei Li, Guoqing Wang, Tianyu Li, Xin Yuan

Published: 01 Dec 2025, Last Modified: 21 Nov 2025Optics & Laser TechnologyEveryoneRevisionsCC BY-SA 4.0
Abstract: Highlights•A noise-integrated reconstruction framework is proposed for video snapshot compressive imaging (SCI) in real imaging systems.•Intrinsic detector noise and light intensity are modeled and evaluated throughout the full SCI reconstruction pipeline.•The reconstruction performance under various signal-to-noise ratios and compression ratios is systematically analyzed. An extreme performance bound is revealed under noisy conditions, which is lower than the noise-free benchmark.•A real video SCI system is built with intrinsic noise calibration to validate the proposed method experimentally.