Our goal is to turn an intensity image into its perceived luminance without parsing it into depths, surfaces, or scene illuminations. We start with jarring intensity differences at...
An accurate and reproducible segmentation of the mitral valve annulus from 3D ultrasound is useful to clinicians and researchers in applications such as pathology diagnosis and mi...
Robert Schneider, Douglas Perrin, Nikolay V. Vasil...
Abstract. This paper presents a novel approach for motion segmentation from feature trajectories with missing data. It consists of two stages. In the first stage, missing data are...
In this paper we introduce a novel edge-based, implicit approach for single object segmentation in 3D images. From a couple of curves, traced by the user on the object to be segme...
Today's Content-Based Image Retrieval (CBIR) techniques are based on the "k-nearest neighbors" (kNN) model. They retrieve images from a single neighborhood using lo...