Signal Processing and Pattern Recognition for Object Detection by Combination of Neural Network and Wavelet Using a Magnetic Sensor

The practical utili:ation ofobject detection and classiJication, in high-petformance struclut'al ntine detection or proxinti!1t fuses is somewhal intpeded due to some contplicated phenomena such as: existence of multiple u,ave ntodes, jamnting, high susceptibility to dit erse inletferences, bullty sampled data, clutters and dfficulty in signal interpretation An intelligent signal processing approach using the wayelet transform and artificial neural nehuork algorithnts y,as det,eloped; this vas actuali:ed in a signal processing package. The intelligent signal processing lechnique comprehensit,ely functions as signal f ltration, data compression and pattern recognition, capable of extracting essential features front acquired rau, yi,at,e signals andfurther assisting in structural mine detection or proximity fuses evaluation For ttalidation, the algorithm v,as applied to the detection and classification of l0 dffirent objects.

Keywords:

Wavelet , Classifi cation , Mine detection , Intelligent signal processing