Detail Cantuman

Compressive Sensing (CS) framework is
mathematical framework to recover the signal by having less
measurement data compared to Shannon-Nyquist theorem. It
indicates the underdetermined linear system where the
dimension of measurement data is much lower compared to
dimension of the projected data. The basic idea of CS is to shift
the sensing load into image reconstruction load. Thus, even
though the sensing process produces less measurement data
subject to the recovery data dimension, the CS theoretically is
able to perform good signal recovery. Theoretically, CS should
be working for natural sparse signal or sparse in transform
domain. Electrical Capacitance Volume Tomography (ECVT)
imaging forms naturally underdetermined linear system since
the dimension of capacitance as the measurement data is much
lower compared to dimension of predicted permittivity
distribution. In addition, the ECVT signal is naturally sparse.
Thus, the compressive sensing framework is theoretically
promising for ECVT imaging. This paper will introduce ECVT
static imaging based on compressive sensing framework. The
early simulations show that compressive sensing with l1
optimization on the sparse recovery succeed to eliminate the
elongation error on ECVT imaging by ILBP (Iterative
Learning Back Propagation).