Independent component analysis for remotely sensed image classification with limited data dimensionality

Abstract

The application of independent component analysis (ICA) to remotely sensed image classification has been studied recently. It is particularly useful for classifying objects with unknown spectral signatures in an unknown image scene, i.e., unsupervised classification. Since the weight matrix in ICA is a square matrix for the purpose of mathematical tractability, the number of objects that can be classified is equal to the data dimensionality, i.e., the number of spectral bands. When the number of spectral bands is very small (e.g., 3-band CIR photograph and 6-band Landsat image), it is impossible to classify all the different objects present in an image scene with the original data. In order to solve this problem, we present a data dimensionality expansion technique to generate artificial bands. Its basic idea is to use nonlinear functions to capture the second and high order correlations between original bands, which can provide additional information for detecting and classifying more objects. The results from such nonlinear band generation approach are compared with a linear band generation method using cubic spline interpolation of pixel spectral signatures. The experiments demonstrate that nonlinear band generation approach can significantly improve unsupervised classification accuracy, while linear band generation method cannot since no new information can be provided.