Novel Visualisation and Image Understanding for Environmental Remote-Sensed Data, in collaboration with Plymouth Marine Laboratory (Remote Sensing Group)

Applications are invited for an Engineering Doctorate (EngD) in the Centre for Digital Entertainment (CDE), a collaboration between the University of Bath and Bournemouth University, with PML http://www.pml.ac.uk.

This 4 year studentship, fully funded by the EPSRC, offers the opportunity to conduct research in an industrial context. Please see http://www.digital-entertainment.org/ (click current opportunities) for more information and some data examples to play with.

The project aims to explore and develop efficient and effective scientific visualisations of these datasets. This involves handling large datasets, intelligent combination of complementary data and possibly image processing and interpretation to summarise complex data into understandable visualisations. Beyond this starting point, there is freedom to research into related and applicable areas. The ultimate goal is to provide simple combination and compelling exploration of scientific datasets, to improve the accessibility and use of essential data backing climate research.

You will have the following datasets available to work with:

- Environmental satellite data: Full global scope over >10 years, two dimensional with ~15 colour bands and numerous derived products (mainly
relating to chlorophyll estimates, a proxy for biological activity in the oceans)
- Aircraft hyperspectral and camera data: local acquisitions over ~20 years, two dimensional with ~200-500 bands (hyperspectral) or RGB.
- Aircraft LIDAR data: local acquisitions over ~10 years, 2.5D / 3D range data, with multiple “depths” per coordinate. World-leading full waveform data in the last 3 years, giving ultra-high depth density (~10cm resolution from 2km up) .

The starting points already in place are simple visualisations using images on a website through to more advanced OpenGL / 3D modelling of LIDAR data, and processing techniques including simple slicing operations, basic overlaying, and data reduction/dynamic downsampling (quad/oct trees). Taking things forward, we propose to explore some of the following areas, however they best align with your research interests:

- Full and dynamic combination of multimodal aircraft data (e.g. map a 3D surface with user-selected bands, allowing one to visually search for the best bands to, say, identify a particular tree species)
- Exploration of high resolution LIDAR data; like the videos linked above, but with more data and a better visualisation interface (user-controlled); voxelisation and translucency may be a starting point.
- Animations / slicing through time on satellite and model data; ways to visualise how ocean conditions change over time and space, with particular interest in highlighting repeated/seasonal, unusual events or exploring biological/climatic structures in the oceans.
- Web-oriented 3D visualisation, using WebGL
- Database-oriented methods for accessing and slicing large datasets, in a networked environment; dynamic and interactive exploration of very large datasets.
- Interpretation of full-waveform LIDAR data to determine canopy thickness/structure or recognise the flora type.
- Scalable representations for event recognition and visualisation in very large datasets


REQUIREMENTS: Computer Science background, interest in large data handling or data structures/algorithms, visualisation/3d experience

Please read http://www.digital-entertainment.org/info-applicants.html carefully.
For more information, please contact [Email Address Removed]

To apply, please complete the application form on http://www.digital-entertainment.org/.