Qualitative vision for the guidance of legged robots in unstructured environments

Elisa Martínez, Carme Torras

Research output: Contribution to journalArticle

17 Citations (Scopus)

Abstract

Visual procedures especially tailored to the constraints and requirements of a legged robot are presented. They work for an uncalibrated camera, with pan and zoom, freely moving towards a stationary target in an unstructured environment that may contain independently moving objects. The goal is to dynamically analyse the image sequence in order to extract information about the robot motion, the target position and the environment structure. The development is based on the deformations of an active contour fitted to the target. Experimental results confirm that the proposed approach constitutes a promising alternative to the prevailing trend based on the costly computation of displacement or velocity fields.

Original languageEnglish
Pages (from-to)1585-1599
Number of pages15
JournalPattern Recognition
Volume34
Issue number8
DOIs
Publication statusPublished - Aug 2001
Externally publishedYes

Fingerprint

Robots
Cameras

Cite this

@article{7694c53895cb4eda9a8c198714dfefd0,
title = "Qualitative vision for the guidance of legged robots in unstructured environments",
abstract = "Visual procedures especially tailored to the constraints and requirements of a legged robot are presented. They work for an uncalibrated camera, with pan and zoom, freely moving towards a stationary target in an unstructured environment that may contain independently moving objects. The goal is to dynamically analyse the image sequence in order to extract information about the robot motion, the target position and the environment structure. The development is based on the deformations of an active contour fitted to the target. Experimental results confirm that the proposed approach constitutes a promising alternative to the prevailing trend based on the costly computation of displacement or velocity fields.",
keywords = "3D reconstruction, Contour tracking, Egomotion recovery, Image sequences, Uncalibrated camera, Unstructured environment, Visual robot navigation",
author = "Elisa Mart{\'i}nez and Carme Torras",
year = "2001",
month = "8",
doi = "10.1016/S0031-3203(00)00093-5",
language = "English",
volume = "34",
pages = "1585--1599",
journal = "Pattern Recognition",
issn = "0031-3203",
publisher = "Elsevier Limited",
number = "8",

}

Qualitative vision for the guidance of legged robots in unstructured environments. / Martínez, Elisa; Torras, Carme.

In: Pattern Recognition, Vol. 34, No. 8, 08.2001, p. 1585-1599.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Qualitative vision for the guidance of legged robots in unstructured environments

AU - Martínez, Elisa

AU - Torras, Carme

PY - 2001/8

Y1 - 2001/8

N2 - Visual procedures especially tailored to the constraints and requirements of a legged robot are presented. They work for an uncalibrated camera, with pan and zoom, freely moving towards a stationary target in an unstructured environment that may contain independently moving objects. The goal is to dynamically analyse the image sequence in order to extract information about the robot motion, the target position and the environment structure. The development is based on the deformations of an active contour fitted to the target. Experimental results confirm that the proposed approach constitutes a promising alternative to the prevailing trend based on the costly computation of displacement or velocity fields.

AB - Visual procedures especially tailored to the constraints and requirements of a legged robot are presented. They work for an uncalibrated camera, with pan and zoom, freely moving towards a stationary target in an unstructured environment that may contain independently moving objects. The goal is to dynamically analyse the image sequence in order to extract information about the robot motion, the target position and the environment structure. The development is based on the deformations of an active contour fitted to the target. Experimental results confirm that the proposed approach constitutes a promising alternative to the prevailing trend based on the costly computation of displacement or velocity fields.

KW - 3D reconstruction

KW - Contour tracking

KW - Egomotion recovery

KW - Image sequences

KW - Uncalibrated camera

KW - Unstructured environment

KW - Visual robot navigation

UR - http://www.scopus.com/inward/record.url?scp=0035427427&partnerID=8YFLogxK

U2 - 10.1016/S0031-3203(00)00093-5

DO - 10.1016/S0031-3203(00)00093-5

M3 - Article

VL - 34

SP - 1585

EP - 1599

JO - Pattern Recognition

JF - Pattern Recognition

SN - 0031-3203

IS - 8

ER -