The system relies on a technology called a ‘metasurface’. Metasurfaces are thin two-dimensional metamaterial layers that allow or inhibit the propagation of electromagnetic waves.
In this application, the metasurface comprises 1.6 million cylindrical posts and can be produced much like a computer chip, on a space 0.5 millimetres wide. These posts are carefully designed to bend light in just the right manner so that the array, as a whole, shapes the entire optical wave front. A key advancement was the development of a model to efficiently approximate the metasurfaces’ image production capabilities.
Using machine learning-based algorithms, the posts’ interactions with light combine to produce crisp, high-definition colour images that are comparable with a traditional lens set-up 500,000 times larger in volume. The camera could be applied to medical imaging within the human body, or providing small robots with better vision. It has even been suggested that thousands of these sensors could turn a whole surface (like the back of a mobile phone) into one giant camera.
Source: Princeton University