This paper presents a new approach, called progressive band subset fusion (PBSF) for hyperspectral anomaly detection. Unlike band selection (BS) which selects bands according to band prioritization or band search… Click to show full abstract
This paper presents a new approach, called progressive band subset fusion (PBSF) for hyperspectral anomaly detection. Unlike band selection (BS) which selects bands according to band prioritization or band search strategies, PBSF fuses band subsets progressively during data collection processing. It is completely opposite to BS which must be done after data is acquired and then selects bands by removing spectral redundancy as a post data processing. To accomplish PBSF, two versions of PBSF are derived, PBSF of multiple band subsets (PBSF-MBS) and PBSF of uniform band selection (PBSF-UBS). In particular, the fusion process takes place in an anomaly detector from a real time processing perspective. Three approaches are developed to realize PBSF of two band subsets simultaneously, PBSF-band sequential (PBSF-BSQ), PBSF-real time (PBSF-RT) and PBSF-zigzag. Extensive experiments demonstrate that PBSF has advantages over BS in many ways.
               
Click one of the above tabs to view related content.