The performance of power transform in high resolution radar target identification

Power transform is a nonlinear pretreatment method, which can improve the recognition rate of high resolution range profiles obviously. This paper mainly analyses how we should decide the power transform coefficient in order to get good recognition results. We process the range profiles in different noise condition using different power transform coefficients, and then we classify them with radial basis function classifiers. According to the real data processing results, we know that we should decide power transform coefficients in accordance with the signal-to-noise ratio.