Sparse sampling‐based microwave 3D imaging using interferometry and frequency‐domain principal component analysis

Abstract

Microwave radar 3D imaging with high resolution generally requires a great number of samples. The authors aim at accurate reconstruction of microwave radar images while significantly reducing the required number of samples. A novel algorithm is proposed which realises sparse sampling with nearly 50% data reduction and high‐quality restoration, based on interferometry and principal component analysis (PCA) in frequency domain. Interferometric processing is utilised to concentrate the frequency spectrum into low‐frequency stage, thereby reaching an effective sparse representation of radar image. Furthermore, PCA is introduced to reform radar image according to its principal characteristics in frequency spectrum, without side‐lobe artefacts and receiver noise. Experimental data in anechoic chamber demonstrates the great potential of the proposed approach.

Keywords

• sparse sampling‐based microwave 3D imaging
• frequency‐domain principal component analysis
• PCA
• data reduction
• high‐quality restoration
• interferometric processing