Many images--especially those used for page design on web pages--as well as videos contain visible text. If these text occurrences could be detected, segmented, and recognized auto...
Abstract. In this paper we describe a flexible approach for determining the relative orientation of the camera with respect to the scene. The main premise of the approach is the fa...
Let n ≥ 4 be even. It is shown that every set S of n points in the plane can be connected by a (possibly self-intersecting) spanning tour (Hamiltonian cycle) consisting of n str...
We o er a simple paradigm for tting models, parametric and non-parametric, to noisy data, which resolves some of the problems associated with classic MSE algorithms. This is done ...
In this paper, we rely on the theory of marked point processes to perform an unsupervised road network extraction from optical and radar images. A road network is modeled by a Mar...