| dc.description.abstract | Flow visualisation has already been a very attractive part of visualisation research for a long time. Usually very large data sets need to be processed, which often consist of multivariate data with a large number of sample locations, often arranged in multiple time steps. Recently, the steadily increasing performance of computers again has become a driving factor for a new boom in flow visualisation, especially in techniques based on feature extraction, vector field clustering, and topology extraction. In this state-of-the-art report, an attempt was made to (1) provide a useful categorisation of FlowVis solutions, (2) give an overview of existing solutions, and (3) focus on recent work, especially in the field of feature extraction. In separate sections we describe (a) direct visualisation techniques such as hedgehog plots, (b) visualisation using integral objects, such as streamlines, (c) texture-based techniques, including spot noise and line integral convolution, and (d) techniques based on extraction of features or flow topology. | en_US |
