https://www.selleckchem.com/products/ABT-263.html Single-pixel imaging allows for high-speed imaging, miniaturization of optical systems, and imaging over a broad wavelength range, which is difficult by conventional imaging sensors, such as pixel arrays. However, a challenge in single-pixel imaging is low image quality in the presence of undersampling. Deep learning is an effective method for solving this challenge; however, a large amount of memory is required for the internal parameters. In this study, we propose single-pixel imaging based on a recurrent neural network. The proposed approach succeeds in reducing the internal parameters, reconstructing images with higher quality, and showing robustness to noise.Ultra-smooth surfaces with low contamination and little damage are a great challenge for aluminum optical fabrication. Ion beam sputtering (IBS) has obvious advantages of low contamination and non-contact that make it a perfect method for processing aluminum optics. However, the evolution laws of aluminum surface morphology are quite different from conventional amorphous materials, which affects the roughness change and needs systematic research. Thus, in this paper, the roughness evolution of an aluminum optical surface (i.e., aluminum mirror) subjected to IBS has been studied with experimental and theoretical methods. The surface morphology evolution mechanisms of turning marks and second phase during IBS are revealed. The newly emerging relief morphology and its evolution mechanism are studied in depth. The experimental results find that IBS causes the coarsening of optical surfaces and the appearance of microstructures, leading to the surface quality deterioration. Turning marks have been through the process of deepening and vanish, while second phase generates microstructures on the original surface. The corresponding mechanism is discussed exhaustively. Preferential sputtering, curvature-dependent sputtering and material properties play important roles