Classification and neural networks to detect buried landmines

— The use of neural networks will not be able to detect buried objects because the high error rate of resulting from using this method. The analysis of the signal after the change in the explosives must progress. Conventional methods cannot detect new plastic mines. In recent methods, classification and neural network are used. In a neural network, a feature is extracted from the data and is trained to a network that can be used to identify land mines. In the classification method, all data in a class has a common property which are different from other. New data can be categorized according to this classification. In this paper, an algorithm is designed based on classification, data reduction and neural networks. In fact, this algorithm simultaneously uses the neural network and classification. Simple methods that only use the neural network or classification have a high error rate. In this method, the data are classified according to similarity. The similarity between all the signals in a class is more than 90%. Scattering parameter, which has amplitude and phase parts, is used to create a parallel process algorithm