Awarded VEIV EngD
This project is concerned with GIS, GPS and Vision data fusion. There is no way at present to indicate the relationship between roads, buildings and other features which a traveler sees with those described by GNSS position and GIS data. The research will consider: the trade off between monocular of stereo/multi camera systems, what kinds of features to identify from camera data, how to identify those features and how to align the features to GIS data in real time. The research may also include Image Parsing, unifying segmentation, detection and recognition, using GIS data as an image prior.
Alastair P. Moore, Simon J. D. Prince and Jonathan Warrell; 2010; “Lattice Cut” - Constructing Superpixels Using Layer Constraints; IEEE Conference on Computer Vision and Pattern Recognition. 2010
Jonathan Warrell, Simon J.D. Prince and Alastair P. Moore; 2009; Epitomized Priors for Multi-labeling Problems; CVPR 2009 Proceedings
Alistair P. Moore, Simon J.D. Prince, Jonathan Warrell, Umar Mohammed and Graham Jones; 2009; Scene Shape Priors for Superpixel Segmentation; Computer Vision, IEEE 12th International Conference 2009
Jonathan Warrell, Alistair P. Moore and Simon J.D. Prince; Vistas: 2009; Hierarchical Boundary Priors Using Multiscale Conditional Random Fields; Proceedings of British Machine Vision Conference 2009