AN ADAPTIVE CONSTANT FALSE ALARM RATE DETECTION SYSTEM

摘要

A method and apparatus of Constant False Alarm Rate (CFAR) detection results in optimal detection in the Neyman Pearson sense independent of the type of the input data. The whole system consists three major blocks: pre-processing block, data fusion block and adaptive multi-CFAR detection block. For any kind of input data, various appropriate data processing techniques could be chosen and applied in parallel in the pro-processing unit for the purpose of enhancing Signal to Noise Ratio (SNR) and fusion. The SNRs of the output data from the pre- processing block will be compared and the top couples will be transmitted to the data fusion block fo r further SNR enhancement by performing appropriate data fusion technique. The adaptive multi-CFAR detector will then be applied at the last stage to get the final decision. In the application of this system on hyperspectral data, a new Transform Bas ed Matched Filter (TBMF) detection algorithm is proposed in the pre-processing unit. The TBMF algorithm employs a transformation before the traditional matched filter to make the background subspace and target subspace orthogonal, hence enhance the detection performance of the matched filter algorithm.