A Hybrid Active Contour without Re-initialization
Azizi Abdallah, Kaouther Elkourd · 2015
In this paper, we propose a novel hybrid edge and region based active contour in a variational level set formulation without reinitialization for image segmentation. The proposed model associate a penalizing term that penalizes the deviation of the level set function from a signed distance function, into a variational level set formulation which consists of both region and gradient information. Experimental results indicate the robustness and efficiency of the proposed model compared with edge-based active contour without re-initialization model, and the Chan and Vese model. The proposed model has the property of local / global segmentation according to the approximated Dirac function chosen.