Authors: Mai El-Shehaly, Denis Gracanin, Mohamed Gad, Hicham Elmongui, Kresimir Matkovic
Abstract: Flow visualization aims primarily at the occlusion-free and controllable visual exploration of dense flow fields. Flow fields are usually visualized using streamlines (steady-state fields) or pathlines (time-dependent flows). The challenge is how to highlight crucial features while reducing clutter to convey relevant trends to the expert user in a timely manner. We present a novel querying and selection technique for the ad-hoc exploration of dense flow fields. Our goal is to allow the user to create a geometric shape that acts as a "probe" for the field, displaying only the information that flows through a region covered by the probe. The main contribution of our technique is that it does not require the construction of expensive hierarchical data structures for pathlines representing the flow. Unlike other techniques, ours operates directly on pathlines' vertices and requires no knowledge of line features. We implemented our technique using GPU-parallelism and new features in OpenGL 4.4 shaders to achieve interactive frame rates.