A new method has been developed to apply image processing tech-niques to a more conventional stereoscopic measurement method. This new technique will provide both qualitative information and quantitative data, which are needed to achieve a better characterization of complex flow fields such as those in agitated vessels. The technique of flow visualization has been widely used to acquire qualitative realizations of flow structures and mixing mecha-nisms. However, the quantity of data obtained from such flow visualization studies has always been limiting. Sheu et al. [Chem. Eng. Comm. 17, 67 (1982)] used a stereoscopic cine camera to record the flow field in an agitated tank containing small, neutrally buoyant tracer particles and showed that three-dimensional velocities can be obtained from discrete sets of particle locations in stereo scenes. In this paper, the application of digital image processing to automate data acquisition procedures is presented. Instead of tracking particles manually, films (visual images) were digitized and were stored in digital form on magnetic tapes. A software package, utilizing THRESHOLDING and PATTERN MATCHING, has been developed to perform the following tasks: (1) particle image identification, (2) particle image tracking in consecutive frames, (3) stereo pair matching of particle trajectories, and (4) velocities evaluation in three-dimensional space evolving in time. The important attributes for flow characterization, such as velocity, acceleration, turbulent intensity, and possi-bly temperature, can be collected for analysis in much larger quantities at a much faster rate than ever before. Furthermore, considerable improvement in accuracy over that of a manually tracked method is also achieved. Most attrac-tive of all, this new combination of image processing and stereoscopic motion pictures is capable of providing sufficient data with high resolution to character-ize the flow field from a Lagrangian/Eulerian point of view.