Authors: Jun Tao, Chaoli Wang, Ching-Kuang Shene, Seung Hyun Kim
Abstract: Striking a careful balance among coverage, occlusion, and complexity is a resounding theme in the visual understanding of large and complex three-dimensional flow fields. In this paper, we present a novel deformation framework for focus+context streamline visualization that reduces occlusion and clutter around the focal regions while compacting the context region in a full view. Unlike existing techniques that vary streamline densities, we advocate a different approach that manipulates streamline positions. This is achieved by partitioning the flow field's volume space into blocks and deforming the blocks to guide streamline repositioning. We formulate block expansion and block smoothing into energy terms and solve for a deformed grid that minimizes the objective function under the volume boundary and edge flipping constraints. Leveraging a GPU linear system solver, we demonstrate interactive focus+context visualization with 3D flow field data of various characteristics. Compared to the fisheye focus+context technique, our method can magnify multiple streamlines of focus in different regions simultaneously while minimizing the distortion through optimized deformation. Both automatic and manual feature specifications are provided for flexible focus selection and effective visualization.