Optical fringe pattern processing and analysis [1] plays crucial role in metrological applications (e.g., interferometry, moiré and structured illumination methods). It might be often a troublesome task because of fringe pattern defects such as noise, uneven background, low modulation and generally complex fringe shapes in a wide spatial frequency range. In this paper we present adaptive optical fringe pattern processing (filtering and normalization) techniques, robust to mentioned pattern imperfections, based on the empirical mode decomposition (EMD).