Authors:
Shaomin Zhang
;
Lijia zhi
;
Dazhe Zhao
and
Hong Zhao
Affiliation:
Northeastern University, China
Keyword(s):
Medical Image Registration, Entropic Spanning Graph Estimator, Minimum Spanning Tree (Mst), Rényi Entropy, Salient Point.
Related
Ontology
Subjects/Areas/Topics:
Computer Vision, Visualization and Computer Graphics
;
Image and Video Analysis
;
Image Registration
Abstract:
In this paper, we propose a novel registration algorithm based on minimal spanning tree. There are two novel aspects of the new method. First, instead of a single feature points, we extracted corner-like as well as edge-like points from image, and also added a few random points to cover the low contrast regions; Second, the hierarchical mechanism which fusing multi-salient points is used to drive the registration during the registration procedure. The new algorithm has solved the low robustness brought by the instability of extraction of feature points and the speed bottleneck problem when using MST to estimate the Rényi entropy. Experiment results show that on the simulated and real brain datasets, the algorithm achieves better robustness while maintaining good registration accuracy.