Robot Localization by Particle Filter using Visual Database

Abstract

One of the major problems in robotics is to recognize the robots position with respect to a given environment. More recently researchers have begun to exploit the structural properties of robotic domains that have led to great success. A general solution for such problem is the implementation of particle filters. The particle filter is more efficient than any other tracking algorithm because this mechanism follows Bayesian estimation rule of conditional probability propagation. In this paper we would like to present an approach to improvise the particle filter algorithm using SIFT pattern recognition technique and image database processing to obtain unimodal uncertainty for effective position tracking.

Performance Analysis of Particle Filter Localization Algorithm for Mobile Robot

“…To overcome the problem of localization for mobile robots, several types of localization algorithms can be applied. There are Kalman Filter algorithm, Extended Kalman Filter algorithm (EKF) and Particle Filter algorithm [5][6][7][8].…”

Multiarquitectura distribuida para el desarrollo de misiones multi-robot

“…Las medidas externas que necesitan los algoritmos de localización se obtienen normalmente utilizando sensores láser [31] [32], transceptores inalámbricos [33] [34] o cámaras utilizando algoritmos de detección de patrones [35]. En la mayoría de los casos, la implementación de estas soluciones requiere un estudio previo del entorno del robot (tomando imágenes, creando mapas basados en las lecturas láser) o la inserción de balizas.…”