Projection Pursuit and Lowpass Filtering for Preprocessing of Hypespectral Images

Hyperspectral data potentially contain more information than multispectral data because of their higher spectral resolution. However, the stochastic data analysis approaches, successfully applied to classification of multispectral data, are not as effective as those for hyperspectral data. Various investigations indicate that the key problem causing poor performance in the stochastic approaches to hyperspectral data classification is inaccurate class parameters estimation. It has been found that the conventional approaches can be retained if a preprocessing stage is established before feature extraction stage in the classification process. This paper, presents a combined preprocessing algorithm which includes dimensionality reduction followed by class separability improvement. For the dimensionality reduction, the Sequential Parametric Projection Pursuit was used because of its special characteristics. For class separability improvement, a Lowpass filter was used. Experimental results showed that applying such a combination, improves the classification accuracy as compared with the case where either a dimensionally reduction or a class separability improvement algorithm is used individually.