In this paper, the results of the simulation study to reconstruct the size of the defects from the data obtained using the active thermography technique based on transient induction heating, will be presented. The forward problem of electro-magnetic induction was solved with an axi-symmetric model using finite element method and from the temperature history profiles, an inverse analysis was performed using Genetic Algorithm (GA) to size the defect. Simulations were performed using the finite element model to obtain the temperature data which are then used to reconstruct the radius (r d ) and depth (d d ) of the wall thinning defects in aluminum plate using inversion method. Two cases, coil inner radius less than the defect radius (rc<r d ) and coil inner radius greater than the defect radius (r c >r d ), were considered. The analysis of the sensitivity of coil dimensions to the calculated peak temperature at the observation point was carried out. © 2013 American Institute of Physics.