Open AccessActive contours driven by novel LGIF energies for image segmentation
Author(s) -
Han Bin,
Wu Yiquan
Publication year - 2017
Publication title -
electronics letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.375
H-Index - 146
ISSN - 1350-911X
DOI - 10.1049/el.2017.1987
Subject(s) - active contour model , segmentation , artificial intelligence , image segmentation , scale space segmentation , energy (signal processing) , intensity (physics) , computer vision , segmentation based object categorization , computer science , noise (video) , pattern recognition (psychology) , image (mathematics) , mathematics , optics , physics , statistics
An active contour model driven by novel local and global intensity fitting (LGIF) energies for image segmentation is presented. First, the GIF energy is defined by the squared‐chord distance, which is more robust to noise. Second, the LIF energy is defined by the Lorentzian distance to calculate the local intensity information, which improves the generality of the presented model. Experiments are carried out on synthetic and real images and the results illustrate that the presented model can obtain higher segmentation accuracy and better segmentation efficiency; moreover, it is not sensitive to initial contour.