Wide field of view and full Stokes polarization imaging using metasurfaces inspired by the stomatopod eye

Abstract

Wide field of view and polarization imaging capabilities are crucial for implementation of advanced imaging devices. However, there are still great challenges in the integration of such optical systems. Here, we report a bionic compound eye metasurface that can realize full Stokes polarization imaging in a wide field of view. The bionic compound eye metasurface consists of a bifocal metalens array in which every three bifocal metalenses form a subeye. The phase of the bifocal metalens is composed of gradient phase and hyperbolic phase. Numerical simulations show that the bifocal metalens can not only improve the focusing efficiency in the oblique light but also correct the aberration caused by the oblique incident light. And the field of view of the bionic compound eye metasurface can reach 120° × 120°. We fabricated a bionic compound eye metasurface which consists of three subeyes. Experiments show that the bionic compound eye metasurface can perform near diffraction-limited polarization focusing and imaging in a large field of view. The design method is generic and can be used to design metasurfaces with different materials and wavelengths. It has great potential in the field of robot polarization vision and polarization detection.

Keywords: bionic compound eye; full Stokes polarization imaging; metasurfaces; wide field of view.

Figure 1:

Schemes of BCEM. (a) Schematic diagram of mantis shrimp’s ommatidium distribution. (b) Anatomical schematic of ommatidium in the compound eye of mantis shrimp (adapted from [36]). (c) Schematic diagram of a bionic compound eye with polarized vision; there is a mismatch between the curved image surface and the polarizer and planar detector. (d) Schematic diagram of the BCEM. (e) Schematic diagram of the bifocal metalens in a sub eye, whereα and β are the azimuth and elevation angles of the sub eye visual axis, respectively.

Figure 2:

Design of the metasurface. (a) Schematic diagram of the elliptical silicon pillar, the substrate is silica. (b) and (c) Calculated transmittance and phase shift of a single period for X-LP normal incident light with a wavelength of 780 nm.

Figure 3:

Theoretical analysis and numerical simulation of the metalens. (a) Schematic diagram of the optical path of the bifocal metalens in the XOZ plane, where θ₁ and θ₂ are the off-axis angles of the X-LP and Y-LP components, respectively. The line cd represents the optical path difference between the oblique light vector and the vertical light vector on the bifocal metalens. (b) When α = 0, the off-axis angles of the X-LP and Y-LP components in (a), where the solid and dashed lines represent the traditional bifocal metalens and the bifocal metalens with gradient phase, respectively, and the dotted line represents the initial value. (c) and (d) When α = 0, the focused spots of X-LP and Y-LP components of the two bifocal metalenses at different elevation angles, respectively. Scale bar: 3 μm. (e) and (f) When α = 0, the transmittance and focusing efficiency of the two bifocal metalenses, the solid and dashed lines represent the traditional bifocal metalens and the bifocal metalens with gradient phase, respectively. (g) and (h) MTFs of the X-LP components at different elevation angles for the traditional bifocal metalens and the bifocal metalens with gradient phase, respectively. (i) and (j) MTFs of the Y-LP components at different elevation angles for the traditional bifocal metalens and the bifocal metalens with gradient phase, respectively.

Figure 4:

Manufactured BCEM sample and its test optical path. (a) Photograph of the BCEM sample. Scale bar: 1 mm. (b) Microscopic image of the sub eye, which consists of three bifocal metalenses. Three bifocal metalenses decompose the light into X-LP and Y-LP, 45°-LP and 135°-LP, RCP, and LCP, respectively. Scale bar: 20 μm. (c) SEM image of the bionic compound eye metasurface. Scale bar: 600 nm. (d) Schematic diagram of the optical path for testing the focusing performance of the BCEM. Where LP and λ/4 denote line polarizer and quarter-wave plate.

Figure 5:

Experiment of polarization focusing by the metalens. (a) and (b) The focused spot of the X-LP and Y-LP components of the bifocal metalens located at (0,0), respectively. (c) and (d) The focused spot of the X-LP and Y-LP components of the bifocal metalens located at (0,30), respectively. The angle in (a)–(d) is the value of β in the experiment. Scale bar 10 μm. (e) and (f) Transmittance and focusing efficiency of the X/Y bifocal metalenses. The solid line and dots represent the X/Y bifocal metalens located at (0,0) and (0,30), respectively. (g) and (h) MTFs of X-LP and Y-LP components of the X/Y bifocal metalenses, where the solid and dashed lines represent the X/Y bifocal metalens located at (0,0) and (0,30), respectively, and the dashed line represents the diffraction limit.

Figure 6:

Experiment of polarization imaging by the BCEM. (a) Schematic diagram of the optical path used to test the imaging performance of the BCEM. (b)–(d) Full Stokes polarization imaging tests of sub eyes located at (0,0), (0,30), and (90,30), respectively. The results in (d) were tested with the BCEM rotated 90° counterclockwise.

Figure 7:

Characterization results of polarization imaging by the BCEM. (a) Stokes parameter reconstruction results of Figure 6(b)–(d). (b) Stokes parameter reconstruction results after normalizing the image intensity of Figure 6(b)–(d). Where gray, blue, red, and cyan bars represent the expected values, and experimental results of sub eyes located at (0,0), (0,30), and (90,30), respectively.