ECCV '98 Proceedings of the 5th European Conference on Computer Vision-Volume II - Volume II
Object Recognition from Local Scale-Invariant Features
ICCV '99 Proceedings of the International Conference on Computer Vision-Volume 2 - Volume 2
Representation and Detection of Deformable Shapes
IEEE Transactions on Pattern Analysis and Machine Intelligence
Weakly Supervised Group-Wise Model Learning Based on Discrete Optimization
MICCAI '09 Proceedings of the 12th International Conference on Medical Image Computing and Computer-Assisted Intervention: Part II
Robust autonomous model learning from 2D and 3D data sets
MICCAI'07 Proceedings of the 10th international conference on Medical image computing and computer-assisted intervention - Volume Part I
Constructing part-based models for groupwise registration
ISBI'10 Proceedings of the 2010 IEEE international conference on Biomedical imaging: from nano to Macro
Computing Accurate Correspondences across Groups of Images
IEEE Transactions on Pattern Analysis and Machine Intelligence
Automatic learning sparse correspondences for initialising groupwise registration
MICCAI'10 Proceedings of the 13th international conference on Medical image computing and computer-assisted intervention: Part II
Weakly supervised learning of part-based spatial models for visual object recognition
ECCV'06 Proceedings of the 9th European conference on Computer Vision - Volume Part I
A unified information-theoretic approach to groupwise non-rigid registration and model building
IPMI'05 Proceedings of the 19th international conference on Information Processing in Medical Imaging
| Hi-index | 0.00 |
Groupwise non-rigid image registration plays an important role in medical image analysis. As local optimisation is largely used in such techniques, a good initialisation is required to avoid local minima. Although the traditional approach to initialisation--affine transformation--generally works well, recent studies have shown that it is inadequate when registering images of complex structures. In this paper we present a more sophisticated method that uses the sparse matches of a parts+geometry model as the initialisation. The choice of parts is made by a voting scheme. We generate a large number of candidate parts, randomly construct many different parts+geometry models and then use the models to select the parts with good localisability. We show that the algorithm can achieve better results than the state of the art on three different datasets of increasing difficulty. We also show that dense mesh models constructed during the groupwise registration process can be used to accurately annotate new images.