A5016886877- Random lasers and scattering media
- Advanced X-ray Imaging Techniques
- Advanced Optical Sensing Technologies
- Image Enhancement Techniques
- Digital Holography and Microscopy
- X-ray Diffraction in Crystallography
- Seismic Imaging and Inversion Techniques
- Cold Atom Physics and Bose-Einstein Condensates
- Optical measurement and interference techniques
- Radio Wave Propagation Studies
- Computer Graphics and Visualization Techniques
- Advanced Optical Imaging Technologies
- Neural Networks and Applications
- Terahertz technology and applications
- Geochemistry and Geologic Mapping
- Quantum optics and atomic interactions
Shanghai Institute of Optics and Fine Mechanics
2020-2024
Chinese Academy of Sciences
2020-2024
University of Chinese Academy of Sciences
2021-2024
Abstract Most of the neural networks proposed so far for computational imaging (CI) in optics employ a supervised training strategy, and thus need large set to optimize their weights biases. Setting aside requirements environmental system stability during many hours data acquisition, practical applications, it is unlikely be possible obtain sufficient numbers ground-truth images training. Here, we propose overcome this limitation by incorporating into conventional deep network complete...
Abstract Imaging through dynamic scattering media is one of the most challenging yet fascinating problems in optics, with applications spanning from biological detection to remote sensing. In this study, we propose a comprehensive learning-based technique that facilitates real-time, non-invasive, incoherent imaging real-world objects dense and media. We conduct extensive experiments, demonstrating capability our see turbid water natural fog. The experimental results indicate proposed...
Imaging through non-static and optically thick scattering media such as dense fog, heavy smoke, turbid water is crucial in various applications. However, most existing methods rely on either active coherent light illumination, or image priors, preventing their application situations where only passive illumination possible. In this study we present a universal method for imaging that does not depend any prior information. Combining the selection of small-angle components out incoming...
Imaging through scattering media is a long-standing challenge in optical imaging, holding substantial importance fields like biology, transportation, and remote sensing. Recent advancements learning-based methods allow accurate rapid imaging optically thick media. However, the practical application of data-driven deep learning faces hurdles due to its inherent limitations generalization, especially scenarios such as highly non-static Here we utilize concept transfer toward adaptive dense...
According to the atmospheric scattering model (ASM), object signal's attenuation diminishes exponentially as imaging distance increases. This imposes limitations on ASM-based methods in situations where medium one wish look through is inhomogeneous. Here, we extend ASM by taking into account spatial variation of density, and propose a two-step method for inhomogeneous media. In first step, proposed eliminates direct current component scattered pattern subscribing estimated global...
In this paper, we propose a single-shot three-dimensional imaging technique. This is achieved by simply placing normal thin scattering layer in front of two-dimensional image sensor, making it light-field-like camera. The working principle the proposed technique based on statistical independence and spatial ergodicity speckle produced layer. Thus, local point responses should be measured advance are used for reconstruction. We demonstrate method with proof-of-concept experiments analyze...
Abstract According to the atmospheric scattering model (ASM), object signal's attenuation diminishes exponentially as imaging distance increases. This imposes limitations on ASM-based methods in situations where medium one wish look through is inhomogeneous. Here, we extend ASM by taking into account spatial variation of density, and propose a two-step method for inhomogeneous media. In first step, proposed eliminates direct current component scattered pattern subscribing estimated global...
It is well known that neural networks including deep learning have been widely employed to solve the problems in recognition and classification. was not until recently people started use them imaging problems. In this talk, we focus on how phase retrieval