Due to its intrinsic advantages such as the ability to automatically handle complex shapes and topological changes, the level set method has been widely used in image segmentation. Nevertheless, in addition to be computational expensive, it has the limitation to very often lead to a local minimum because of the energy functional to be minimized is non-convex. In this work, we use the geometric active contours and the image thresholding frameworks to design a novel method for global image segmentation. The local lattice Boltzmann method is used to solve the level set equation. The proposed algorithm is therefore effective and highly parallelizable. Experimental results on satellite, natural and medical images demonstrate the effectiveness and the efficiency of the proposed method when implemented using an NVIDIA graphics processing units. © 2013 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Balla-Arabé, S., Gao, X., & Xu, L. (2013). Texture-aware fast global level set evolution. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8261 LNCS, pp. 529–537). Springer Verlag. https://doi.org/10.1007/978-3-642-42057-3_67
Mendeley helps you to discover research relevant for your work.