Loop differential ghost imaging based on line pattern

The pursuit of ghost imaging (GI) in high speed and miniaturization has led researchers to explore photonic integrated circuits (PICs). This study delves into the pattern characteristics of optical phased arrays (OPAs) based on PICs and identifies a limitation in traditional GI denoising algorithms when employing line pattern illumination field generated by one dimensional OPA (1D OPA) with grating waveguide (GW), attributed to noise symmetry disruption. To overcome this challenge, we introduce the symmetry of noise metric ${\kappa }_s$, tailored to evaluate noise symmetry across different field matrix. Subsequently, we propose the loop differential ghost imaging (LDGI) algorithm, demonstrating orders of magnitude improvement compared to traditional algorithms in GI. This research advances GI technology towards integrated miniaturization and achieving high speed imaging capabilities, with implications for fields such as virtual reality, LiDAR, and photomicrography.