Automatic object segmentation using mean shift and growcut
Elyor Kodirov, Guee-Sang Lee · 2013
In this paper, we propose an efficient unsupervised object segmentation algorithm that provides effective and robust segmentation of color images by incorporating the advantages of Mean Shift (MS) and GrowCut (GC) methods. In the first stage, the image is divided into different segments using MS algorithm and the generated segments are labeled using Mahalanobis distance. Then, the labeled segments are given to the GC method for grouping the clustered segments. The proposed method requires low computation complexity and is therefore very feasible for real time image segmentation processing. The superiority of the proposed method is examined and demonstrated through a number of experiments using color natural scene images. Experimental result shows that the proposed method gives better performance than other methods.