Realization of tunable edge-enhanced images based on computing metasurfaces.

Bright-field imaging and edge imaging can extract different characteristics from objects, and therefore play important roles in image processing and pattern recognition. Here, we propose a fast, convenient, and electrically driven adjustable scheme to achieve tunable edge-enhanced images based on computing metasurfaces. The computing metasurface can perform spatial differential operation as optical waves propagate through it. This optical differential operation is polarization-dependent, thus any desirable contrast can be realized by the interplay between two orthogonal polarization components. By regulating the external voltages applied on the liquid-crystal phase plate, different phase retardances between two orthogonal polarization components are introduced; this allows us to quickly switch between the bright-field image and the edge image.