Search

This paper introduces a novel lightweight computational framework for enhancing images under low-light conditions, utilizing advanced machine learning and convolutional neural networks (CNNs). Traditional enhancement techniques often fail to adequately address issues like noise, color distortion, and detail loss in challenging lighting environments. Our approach leverages insights from the Retinex theory and recent advances in image restoration networks to develop a streamlined model that efficiently processes illumination components and integrates context-sensitive enhancements through optimized convolutional blocks. This results in significantly improved image clarity and color fidelity, while avoiding over-enhancement and unnatural color shifts. Crucially, our model is designed to be lightweight, ensuring low computational demand and suitability for real-time applications on standard consumer hardware. Performance evaluations confirm that our model not only surpasses existing methods in enhancing low-light images but also maintains a minimal computational footprint.

We demonstrate the trajectory measurement of the single neutral atoms deterministically using a high-finesse optical micro-cavity. Single atom strongly couples to the high-order transverse vacuum TEM_{10} mode, instead of the usual TEM_{00} mode, and the parameter of the system is (g_{10},\kappa ,\gamma )=2\pi \times (20.5,2.6,2.6)MHz. The atoms simply fall down freely from the magneto-optic trap into the cavity modes and the trajectories of the single atoms are linear. The transmission spectrums of atoms passing through the TEM10 mode are detected by a single photon counting modules and well fitted. Thanks to the tilted cavity transverse TEM10 mode, which is inclined to the vertical direction about 45 degrees and it helps us, for the first time, to eliminate the degenerate trajectory of the single atom falling through the cavity and get the unique atom trajectory. Atom position with high precision of 0.1{\mu}m in the off-axis direction (axis y) is obtained, and the spatial resolution of 5.6{\mu}m is achieved in time of 10{\mu}s along the vertical direction (axis x). The average velocity of the atoms is also measured from the atom transits, which determines the temperature of the atoms in magneto-optic trap, 186{\mu}K {\pm} 19{\mu}K., Comment: 13 pages, 5figures