Handwritten Text Recognition
As of first January 2017 I have left the position as director of Swedish eScience Education to work 40% with research in handwritten text recognition. A call for PhD students is now out and a call for postdoc will follow (see Upcoming Events on this page).
Feature point descriptors for image stitching
When microscopy images are to be put together to form a larger image than one field of view, images are stitched together based on key point features in the images. Several methods for matching these images exist, but are often general in the sense that they can handle scale and rotation, which are not present in this particular case. Therefore, these methods are like cracking a nut with a sledge hammer, and we have investigated how simpler and therefore more efficient and also faster methods can be developed and applied for solving this task. Several key point descriptors have been investigated that are based on new sampling strategies and also new ways of combining these samples, using for instance elements of the Fourier transform, instead of histograms of gradients etc. A paper describing two versions of fast and simple feature point descriptor with or without rotation invariance was presented at the WSCG conference.
The whole pipeline of matching has been investigated and several improvements have been suggested. We have shown that for instance RANSAC can be substituted by a fast clustering method, which makes computation of the transformation between images and false positives removal not only faster, but also deterministic, which otherwise is a problem with RANSAC as it is based on a random sampling approach. This alternative to RANSAC was presented at the second workshop on Features and Structures (FEAST), co-located with the International Conference on Machine Learning, Lille, France, in July.
Stereo Visualisation of Historical Aerial Photographies
Stereo images are important as they give a much better understanding of what is actually seen on the ground than single photos ever can. The important factor is the depth cue that helps understanding the content and adds the ability to distinguish between bushes and trees, stones and pillars, hills and valleys etc. During the operation Crossbow, the English meticulously photographed Europe during WWII to obtain stereo photos, which helped them in their search for military objects. This rich source of information that is now provided by the AFN in Rome, as well as other archives can be utilised by archeologists in a similar way. There are however still challenges to be faced in order to create useful stereo images that we have addressed and there are many possibilities and advantages. Together with IIT at CNR in Pisa, Italy, we have published papers in two international conferences discussing these things, namely Computer Applications and Quantitative Methods in Archaeology (CAA) in Siena, Italy and Digital Heritage in Granada, Spain. Moreover, in a collaboration with the University of Udine in Italy, we have published a new method for synthesising new views from photos at the International Conference on 3D Imaging in Liege, Belgium, making it possible to look at the stereo pairs from different views, which is otherwise not possible. A collaboration with several universities in Spain lead to a publication in which it was investigated how tumulus can be searched in what is now submerged megalithic landscape. All this, thanks to the fact that historical aerial photos are available and can be converted into stereo representation.