Feature space analysis is the main module in many computer vision tasks. The most popular technique, k-means clustering, however, has two inherent limitations: the clusters are co...
In this paper, a mean shift-based clustering algorithm is proposed. The mean shift is a kernel-type weighted mean procedure. Herein, we first discuss three classes of Gaussian, C...
Object tracking based on Mean Shift (MS) algorithm has been very successful and thus receives significant research interests. Unfortunately, traditional MS based tracking only ut...
Abstract. The mean-shift algorithm is an efficient technique for tracking 2D blobs through an image. Although it is important to adapt the mean-shift kernel to handle changes in il...