Radar target recognition using wavelet transform and information fusion

A new target recognition approach with wavelet transform and information fusion is detailed, and tested on backscatter returns of high range resolution (HRR) radar. Features of targets are extracted by computing the logarithms of the ratios of energies in adjacent scales defined by wavelet transform. With multiresolution analysis, the features are decomposed into two parts: the highpass features reflecting detailed structure of targets, and the lowpass features showing the outline of targets. A radial basis function (RBF) neural network with information fusion technique is applied for classification purposes.