Low-Complexity Algorithm for People Detection Using FMCW Radar

The increasing demand for energy-efficient buildings and optimized space utilization in smart and healthcare environments such as hospitals and clinics to improve patient flow and room management highlights the importance of advanced occupancy detection and people-counting systems. Traditional methods often face high computational costs and privacy issues, mainly from relying on AI and extensive datasets, leading to challenges like substantial storage, complex processing, and data biases. This paper presents a simple algorithm for occupancy detection and people counting using post-processed radar-angle images, suitable for various radar systems. Our method employs a straightforward Maximum Likelihood estimator to align reference images with radar data, identifying target areas efficiently with low computational load, making it viable for low-capacity systems.