Complementary nonlinear optics for polarimetric computing in tellurium

Polarization serves as a high-dimensional information carrier, yet current polarization-sensitive optical systems typically separate sensing from processing, generating massive redundancy and latency. Here, we demonstrate an intelligent paradigm based on the complementary integration of low-dimensional tellurium nanowires and nanosheets, leveraging their intrinsic optical anisotropy and opposite bias-modulated second-harmonic generation behaviors. We systematically exploit these mechanisms across two key domains. In information security, optical anisotropy serves as an entropy source for physical unclonable functions, while integrating tellurium nanowires and nanosheets constructs robust encryption units. In polarized image processing, anisotropy facilitates optical filtering of interference, and, crucially, the complementary bias modulation empowers the system to synergistically perform background denoising and edge extraction. This strategy effectively executes complex information encoding and image pre-processing tasks directly at the detection stage, offering a compact solution for next-generation intelligent optical systems. Polarization-sensitive systems often move computation out of the optical domain, separating detection from processing. Here, authors use opposite bias-tuned nonlinear optics in tellurium nanowires and nanosheets for complementary polarimetric computing directly at the sensing front end.