Temporally and spatially resolved two-dimensional imaging of turbulent reacting flows promises to enhance our knowledge of flame chemistry during ignition, propagation, and quenching. There are a number of active research programs in the combustion imaging diagnostics field, for example in high-speed laser shadowgraph recording, and two-dimensional laser-induced fluorescence imaging. The image data from these experiments is typically used to provide quantitative visualization of species concentrations in propaga-ting flames. Up until now, image sequences have always been interpreted with the aid of fairly simple image display or processing techniques, such as pseudocolor enhancement. Relatively little work has been done in the areas of computer recognition, classification, and interpretation of fluid flow. This paper takes an initial step towards this goal. We address the probem of tracking and displaying the flow of local features which occur at fluid interfaces in laser shadowgraph imagery. Image sequences are computer processed to produce two-dimensional maps showing the motion of the interfaces between hot combustion products and cold reactants gases. This is analogous to a simple form of optical flow map, typically used for describing solid object motion in video sequences. The computer processing is applied in the following stages: (1) unwanted background noise is subtracted from each frame in the sequence , (2) frames are thresholded and thinned to produce a sequence of skeletons representing the essential fluid structure, (3) localized correlation is performed between adjacent frames to produce an optical flow map.