El objetivo de este proyecto es el estudio, desarrollo y evaluación de un sistema robusto de rep-resentación de trayectorias de objetos para vídeo vigilancia. El sistema debe permitir trabajar de forma flexible con las trayectorias, generando una representación útil para aplicaciones de más alto nivel. La solución propuesta comprende cuatro fases, cada una con un propósito determinado. La primera fase consiste en un algoritmo que se encarga de realizar las tareas necesarias para ex-traer las trayectorias de los objetos en movimiento de los vídeos. La segunda fase o comprende una serie de técnicas destinadas a mejorar y depurar la información obtenida en la fase ante-rior. En la tercera etapa, las trayectorias se dividen para facilitar su posterior uso y análisis. Por último, se desarrolla una representación precisa de las trayectorias que también permite la reconstrucción de las trayectorias originales. Para la evaluación del sistema se utilizan dos bases de datos distintas, tratando de incluir los escenarios y situaciones más frecuentes en vídeo vigilancia. Cada técnica utilizada en el sistema se evalúa de manera individual, determinando su eficacia para la consecución del objetivo final. Para la correcta evaluación de algunas partes del proyecto ha sido necesario construir una base

The copyright law of the United States (title 17, U.S. Code) governs the making of photocopies or other reproductions of copyrighted material. Any copying of this document without permission of its author may be prohibited by law.

.m a.= as 1111 (.2 11. *16

Scientists can capture images of objects from the microscopic scale to the plane-tary scale but manual time-consuming work is often required to process these images. Meanwhile, the public’s interaction with image processing has exploded, e.g. Face-book users annotate pictures an incredible 80 million times each day. People also

TOPICS: Differential GPS/MEMS-IMU integration performance, trajectory analysis, lateral slipping of tires, determination of characteristics of tires Abstract: Inertially aided satellite positioning can bring its benefits to all disciplines in which detailed knowledge of the trajectory is a

this paper we consider the problem of generating velocity proles. As for the more general problem of control of non-holonomic systems, this non-trivial problem has been widely studied in the past (see, e.g. Astol 1999, Aicardi et al. 1994, Feng et al. 1993)

trajectory information using singular value decomposition (SVD), taking advantage of the linear algebraic properties of orthographic projection. However, an orthographic formulation limits the range of motions the method can accommodate. Paraperspective projection, first introduced by Ohta, is a projection

The task of planning trajectories for a mobile robot has received considerable attention in the research literature. Most of the work assumes the robot has a complete and accurate model of its environment before it begins to move; less attention has been paid to the problem of partially known

are presented and their error sources are described. The assumptions often made to account for the vertical wind velocity are explained. Most studies agree now that fully three-dimensional trajectories are the most accurate trajectory type. Methods to assess trajectory errors are outlined and a summary

-trivial because we need to process streams of highly inaccurate GSM localization samples (average error of over 175 meters) and produce an accurate trajectory. CTrack meets this challenge using a novel two-pass Hidden Markov Model that sequences cellular GSM fingerprints directly without converting them