Improved filter for point target detection in multidimensional imagery Conference Paper uri icon

abstract

  • We consider a method to improve the standard RX algorithm for point target detection. In this algorithm, we weight the results from an anti-mean or anti-median filter by dividing by the standard deviation of the local environment of each suspect pixel. In this way, we lower the false alarms caused by edge points. The results of each band are then combined. Results will be shown for visible and SWIR hyperspectral imagery.

publication date

  • January 1, 2004