Abstract
Advantages and limitations associated with each NDE modality raises a tradeoff in which no single modality can be identified for a particular application. The necessary elements for building an intelligent NDE system based on image fusion are introduced. An application is presented considering the fusion of optical and eddy current images. Fusion algorithms are implemented using intensity hue saturation (IHS) transform, discrete wavelet transform and IHS with shift invariant wavelet decomposition. Developed image evaluation measures (quality metrics) are adopted to cross the gap between subjective and objective evaluation, which is essential to automate NDE systems in industrial environments. Introduced quality metric depends on mean square error (RMSE), correlation (CORR), peak signal to noise ratio (PSNR) and mutual information (MI). Analysis of the developed system is presented based on experimental data.