Moving targets detection and tracking in reverberation environments

Abstract

The moving target detection and tracking is always an important but challenging task with applications in radar and sonar, video surveillance, web search, bioinformatics, etc. An effective method for the moving target detection and tracking in reverberation environments is presented by utilizing spare and low-rank matrix decomposition. Due to the strong coherence between the beamspace data with the almost same background (reverberation structure), the matrix with beamspace data has potentially low-rank structure. Accordingly, the moving targets can be modeled as the sparse component, and separated from the reverberation.