Priebe, Carey E.; Chen, Dalei (2001, ASQ and American Statistical Association) The Johns Hopkins University, Baltimore, MD
[This abstract is based on the authors' abstract.] A common concern in automatic target recognition is the detection and identification of regions of interest in spatial and temporal data. One approach to region-of-interest identification involves the use of spatial scan statistics, but a difficulty arises due to competing concerns. Small scan windows are necessary to detect potentially small targets, but larger scan windows are needed to improve the accuracy of the detector. When the scan statistics are mixture-model density estimates, a borrowed strength profile likelihood approach is shown to be superior to conventional likelihood estimators. These spatial scan density estimates are investigated on example imagery from an unmanned spacecraft.
Kernel density estimates,Poisson distribution,Likelihood methods