The potential of a new procedure of image processing for the characterization of a given combustion state through flame visualization is here presented and discussed. Experimental tests were carried out in a swirl-stabilized, semi-industrial scale burner of 500 kW(th). Using an advanced vision based system, flame images have been recorded and subsequently processed, obtaining both luminous and spectral parameters from the grey values registered by each individual pixel. The acquisition system is based on a CCD (charge-coupled device) camera of high-speed frame rate. The innovative nature of the analysis lies in the 2D distribution of statistical and oscillatory parameters which can be interpreted as a "fingerprint" of the flame condition. By this method, flame spatial characterization was achieved allowing the identification of areas with different luminous and oscillating patterns. Their evolution regarding primary air-to-fuel ratio was also studied. First results suggest changes on flame symmetry and oscillation regimen. Additionally, quantitative flame analysis through global values of selected parameters and regression studies were conducted in order to analyse their usefulness for the development of monitoring and control algorithms in the combustion facility. (C) 2012 Elsevier Ltd. All rights reserved.