Enabling Real-Time Imaging and Onboard RFI Localization for Three-Level Quantized Microwave Interferometric Radiometers

Real-time imaging processing of microwave interferometric radiometer (MIR) has great potential in various application field, such as onboard data processing, onboard information fusion and alternative visual application. The primary challenge lies in the computational complexity of the entire processing chain, including both visibility function pre-processing and brightness temperature (TB) reconstruction. In this study, the real-time estimation of the normalized threshold level is identified as the key step for enabling re-al-time imaging of three-level quantized MIR system. Three algorithms - Acklam's algorithm (AKA), polynomial fitting algorithm (PFA), and Taylor expansion algorithm (TEA) - are proposed and evaluated. The PFA shows balanced performance in terms of estimation accuracy and computation efficiency. Leveraging the proposed algorithms, this paper further establishes an onboard real-time processing framework for three-level quantization MIRs, enabling real-time TB imaging and RFI localization. A real-time imaging experiment has been carried out with a 15-element, 50GHz one-dimensional MIR system, which successfully realizes real-time imaging of fast-moving vehicles on the expressway with greatly reduced computational latency (an imaging time of 570.9μs for 159 baselines). A further flight experiment employing an L-band system verifies the feasibility of onboard RFI localization and achieves a localization accuracy within 2.1% of the spatial resolution.