Optimized GPR signal deconvolution using an adaptative conjugate gradient method

Presents a signal processing method to improve the identification of interface between different layered media, using a deconvolution process. Our methodological approach is based on an inverse problem algorithm, called conjugate gradient method. The emitted pulse is assumed to be known, and the convolution is considered as a linear operator that must be inverted. We take into account the distortion of the signal and the noise of the recorded signal.