Visualizing the evolution of image features in time-series: supporting the exploration of sensor data
PhD was about helping users of time-series images (geoscientists) to explore their data. Sensor image repositories are becoming the fastest growing archives of spatio-temporal information. This data flow leads to large time-series and, geoscientists are often confronted with the amount of data that need to be explored. Because geoscientists are primarily interested in image features, my solution was to focus on those features - to extract, track and visualize them.
I have applied and evaluated this approach to two water-related case studies: a) Short term forecasting about potentially dangerous weather phenomena: convective clouds and studying the conditions in which the convections are formed; b) Coastal morphodynamics: main task was to explore the evolution of rip channels (dangerous swimming sites) in highly dynamic coastal zones...
Research prototype: relationships between multiple views are based on the Space-Time Cube metaphor.
3D event viewer showing precipitating clouds. Images and contours on the bottom of the cube mark the position of image objects in geographic space. Glyphs above the base represent the evolution of the precipitating clouds. The radius and colour of the glyphs (spheres) is in this example proportional to the size of the cloud object.
Visual query for the clouds with the negative temperature gradients in the Space Time Cube and Attribute viewer. The upper graphs demonstrate the overview and the lower graphs display spatial, temporal and attribute distribution of the gradient query
The image viewer displaying the same frame of animation twice: the full view of the cloud objects (left) and the view of the brushed cloud objects with superimposed movement trajectory (right).