Robust and efficient object segmentation using pseudo-elastica

  • Authors:

  • Matthias Krueger;Patrice Delmas;Georgy Gimel'Farb

  • Affiliations:

  • Department of Computer Science, The University of Auckland, Auckland 1142, New Zealand;Department of Computer Science, The University of Auckland, Auckland 1142, New Zealand;Department of Computer Science, The University of Auckland, Auckland 1142, New Zealand

  • Venue:
  • Pattern Recognition Letters
  • Year:
  • 2013

Quantified Score

Hi-index 0.10

Visualization

Abstract

A new object segmentation method based on second-order energy minimisation is proposed. It is called pseudo-elastica as it relates to the classic Euler's elastica but resulting contours cannot be expected to converge towards continuous elastica if the resolution is increased. Comparing to prior works, our segmentation technique can be easily applied to both closed contours and open contours with fixed endpoints, and its computational complexity, O(NlogN), is significantly lower. The efficiency is increased by extending the idea of bidirectional Dijkstra-type search to second-order energies and incorporating heuristics with some sacrifice in exact energy minimisation. Our pseudo-elastica generalises the classic first-order path-based schemes to second-order energies while maintaining the same low complexity. Experiments suggest that it scores similar or better results and usually requires considerably less user input than the state-of-the-art approaches. The algorithm can be made anisotropic in order to allow corners in the contour.