Incoherent Mid‐Infrared Edge Detection via Configurable Nonlinear Filtering

ABSTRACT Incoherent optical edge detection enables contour‐feature extraction under passive illumination, making it attractive for practical machine vision and intelligent sensing. However, existing frameworks based on constructing bipolar point spread functions have mainly been developed in the visible and near‐infrared regimes, while extending this capability to mid‐infrared (MIR) imaging remains challenging because of the limited sensitivity of MIR detectors and the lack of efficient platforms for high‐fidelity spatial‐frequency modulation at long wavelengths. Here, we demonstrate a high‐sensitivity incoherent MIR upconversion edge‐detection framework based on configurable nonlinear filtering. By exploiting the intrinsic phase‐matching selectivity of a nonlinear crystal, the angular‐spectrum transmission of the MIR signal and the system optical transfer function (OTF) are directly engineered during the frequency upconversion process. Through crystal‐temperature‐controlled OTF reconfiguration and dual‐channel image subtraction, an effective hollow OTF is synthesized for high‐contrast incoherent edge enhancement within a fixed optical layout. The proposed approach features high detection sensitivity and flexible wavelength tunability, providing a versatile route toward incoherent MIR edge‐enhanced imaging and nonlinear computational photonics.