Traditional cameras are limited to capturing only the intensity distribution of light, lacking other important information. We present a novel approach to overcome this limitation by incorporating metasurfaces - thin layers of nanostructures - into standard cameras. These metasurfaces can encode light information of different dimensions, including polarization, frequency, and time, through the unique optical responses of their constituent meta-atoms. As a result, a single snapshot of the light field can capture all its properties. We also use deep-learning-based recovery algorithms that enable real-time retrieval of multi-dimensional images. This breakthrough has enormous potential to revolutionize the way we collect and analyze light information, leading to new applications and insights in various fields.
|