Unsupervised target detection in hyperspectral images using projection pursuit

In this paper, we present a projection pursuit (PP) approach to target dete ction. Unlike most of developed target detection algorithms that require st atistical models such as linear mixture, the proposed PP is to project a hi gh dimensional data set into a low dimensional data space while retaining d esired information of interest. It utilizes a projection index to explore p rojections of interestingness. For target detection applications in hypersp ectral imagery, an interesting structure of an image scene is the one cause d by man-made targets in a large unknown background. Such targets can be vi ewed as anomalies in an image scene due to the fact that their size is rela tively small compared to their background surroundings. As a result, detect ing small targets in an unknown image scene is reduced to finding the outli ers of background distributions. It is known that "skewness," is defined by normalized third moment of the sample distribution, measures the asymmetry of the distribution and "kurtosis" is defined by normalized fourth moment of the sample distribution measures the flatness of the distribution. They both are susceptible to outliers. So, using skewness and kurtosis as a base to design a projection index may be effective for target detection. In ord er to find an optimal projection index, an evolutionary algorithm is also d eveloped to avoid trapping local optima. The hyperspectral image experiment s show that the proposed PP method provides an effective means for target d etection.