New Advances in Multidimensional Processing for Thermal Image Quality Enhancement

Abstract

This chapter presents three recent methodologies about multidimensional processing for signal to noise ratio (SNR) and thermal contrast enhancement on sequences of thermographic images, acquired from active pulsed thermography experiments over composite slabs, mainly, carbon fiber reinforced plastic (CFRP). The first technique corresponds to noise pre-processing by means of iterative 3D filtering to take advantage of the high space-time correlation in thermal sequences; the other two techniques correspond to thermal contrast enhancement processing: one of them using an atypical median filtering scheme, and the other based on heat propagation discrete models. Beginning with their heuristic and mathematical foundations, and following with the algorithmic procedures development, advantages and limitations will be shown through suitable indexes for evaluation, and some comparisons with other similar techniques.