Hierarchical calibration based on multi-source noise dynamics unraveling for ultra-precision fluorescence lifetime measurement via a streak camera

Fluorescence lifetime measurements using time-resolved optical streak camera imaging are highly susceptible to internal noise, particularly under the high-gain, low-signal conditions common in ultrafast experiments. This study systematically analyzes noise-induced errors in lifetime extraction across a wide signal-to-noise ratio (SNR) range and introduces a robust multi-stage processing framework incorporating grayscale filtering, Otsu adaptive thresholding, morphological operations, and area-based connected-component analysis. The proposed method significantly improves accuracy and robustness, as reflected by reduced deviation, coefficient of variation (CV), and fixed-reference normalized ratio bias. For Rhodamine B (1.74±0.02<i>n</i><i>s</i>), uncorrected lifetimes range from 1.6072 to 1.9947 ns (up to ±14.64<i>%</i> deviation) while corrected results converge to 1.6023-1.7869 ns, with normalized ratios of 0.92-1.03 and a 48.2% CV reduction to 3.82%. For Rhodamine 6G (4.08 ns), uncorrected estimates span 2.6882-5.5670 ns (deviations >±30<i>%</i>), whereas corrected lifetimes stabilize at 2.3821-4.0945 ns, with over 80% of values within ±15<i>%</i> of the theoretical lifetime and a 28.6% CV reduction. Overall, the method lowers the mean deviation from 18.7% to 4.2% and reduces CV from 27.5% to 5.8% under moderate-to-low SNR conditions. These results demonstrate that the proposed framework provides a practical and effective solution for improving the accuracy and reliability of streak camera-based fluorescence lifetime measurements in ultrafast optical applications.