Metalens for light field imaging

An image depth estimation method, comprising the following steps:

• obtaining an imaging signal which comprises a plurality of sub-images of an object and is generated from a metalens array by an imaging sensing unit, wherein the metalens array comprises a plurality of metalenses, and each metalens comprises a plurality of nanostructures and a dielectric layer, wherein the nanostructures are disposed on the dielectric layer, and the nanostructures comprises an array of first phase compensation structures and an array of second phase compensation structures, wherein the array of the first phase compensation structures are disposed to surround the array of the second phase compensation structures so as to define a single metalens, the first and second phase compensation structures are complementary to each other and substantially satisfy the Babinet's principle;
• selecting one of the sub-image;
• deciding a square patch in the selected sub-image, wherein the square patch has a size around a quarter of a diameter of the metalens which generates the selected sub-image;
• calculating radiance differences between the selected sub-image and an adjacent sub-image of the selected sub-image so as to obtain a disparity between the selected sub-image and the adjacent sub-image;
• and
• calculating the obtained disparities to obtain a depth information of the imaging signal.