Correction of CMOS-related noise in fluorescence microscopy

Abstract

CMOS cameras are very appealing for fluorescence microscopy but they suffer from high readout noise and fixed-pattern noise. We propose a non-iterative, fast, and unsupervised algorithm that erases camera-related noise without losing the quantitative information of fluorescence signal. Finally, we demonstrate the performance of the algorithm using different sensors, from low-cost CMOS chips to high-end sCMOS cameras.