Image enhancement algorithm for absorption imaging

Pengcheng Zheng,Songqian Zhang,Zhu Ma,Haipo Niu,Jiatao Wu,Zerui Huang,Chengyin Han,Bo Lu,Peiliang Liu,Chaohong Lee
2024-03-07
Abstract:The noise in absorption imaging of cold atoms significantly impacts measurement accuracy across a range of applications with ultracold atoms. It is crucial to adopt an approach that offers effective denoising capabilities without compromising the unique structure of the atoms. Here we introduce a novel image enhancement algorithm for cold atomic absorption imaging. The algorithm successfully suppresses background noise, enhancing image contrast significantly. Experimental results showcase that this approach can enhance the accuracy of cold atom particle number measurements by approximately tenfold, all while preserving essential information. Moreover, the method exhibits exceptional performance and robustness when confronted with fringe noise and multi-component imaging scenarios, offering high stability. Importantly, the optimization process is entirely automated, eliminating the need for manual parameter selection. The method is both compatible and practical, making it applicable across various absorption imaging fields.
Quantum Physics
What problem does this paper attempt to address?
This paper focuses on the noise problem in absorption imaging of cold atoms, which is a key challenge for precise measurements in applications of ultracold atoms. The author proposes a new image enhancement algorithm designed to effectively reduce background noise and significantly improve image contrast. The algorithm consists of adaptive Gaussian filtering and nonlinear grayscale stretching. Experimental results show that this method can improve the accuracy of cold atom particle number measurements by approximately ten times while preserving the unique structural information of the atoms. In addition, the algorithm performs well and exhibits high stability in handling stripe noise and multi-component imaging scenarios. The optimization process is fully automated and does not require manual parameter selection. This method is applicable to various absorption imaging fields and has potential application value in quantum control, precision measurement, and many-body quantum physics research.