Human Detection and Identification by Robots

Only available on StudyMode
  • Download(s) : 137
  • Published : December 26, 2012
Open Document
Text Preview
J Intell Robot Syst (2012) 66:223–243
DOI 10.1007/s10846-011-9612-2

Human Detection and Identification by Robots
Using Thermal and Visual Information
in Domestic Environments
Mauricio Correa · Gabriel Hermosilla ·
Rodrigo Verschae · Javier Ruiz-del-Solar

Received: 11 December 2010 / Accepted: 30 May 2011 / Published online: 12 July 2011 © Springer Science+Business Media B.V. 2011

Abstract In this paper a robust system for enabling robots to detect and identify humans in domestic environments is proposed. Robust human detection is achieved through the use of thermal
and visual information sources that are integrated
to detect human-candidate objects, which are further processed in order to verify the presence of humans and their identity using face information
in the thermal and visual spectrums. Face detection is used to verify the presence of humans, and face recognition to identify them. Active vision
mechanisms are employed in order to improve the
relative pose of a candidate object/person in case

M. Correa · G. Hermosilla · R. Verschae ·
J. Ruiz-del-Solar
Department of Electrical Engineering,
Universidad de Chile, Av. Tupper 2007,
Santiago, Chile
M. Correa
e-mail: macorrea@ing.uchile.cl
G. Hermosilla
e-mail: ghermosi@ing.uchile.cl
J. Ruiz-del-Solar
e-mail: jruizd@ing.uchile.cl
M. Correa · G. Hermosilla · R. Verschae (B) ·
J. Ruiz-del-Solar
Advanced Mining Technology Center,
Universidad de Chile, Av. Tupper 2007,
Santiago, Chile
e-mail: rodrigo@verschae.org

direct identification is not possible. The response
of the different modules is characterized, and the
proposed system is validated using image databases of real domestic environments, and human detection and identification benchmarks of the
RoboCup@Home research community.
Keywords Human detection · Human
identification · Service robot · Thermal image

1 Introduction
There is increasing interest in domestic service
robots in the robotics community. A domestic
service robot is a subclass of mobile service robots designed to interact with humans in a homeenvironment, and to provide different kinds of services (cleaning, cooking, entertainment, companionship, and surveillance, to name just a few). The home environment is defined as ’any place

where people live their daily lives’, which can
include, for example, a kitchen, a bedroom, or a
garden. Although some special-purpose domestic
robots are already popular (e.g. vacuum robots
[35]), we are still far from having general-purpose
domestic robots.
Among the basic skills of domestic service robots are the ability to move autonomously in domestic environments, the ability to recognize and manipulate ‘home’ objects (cups, books, glasses,

224

medications, chairs, door handles, etc.), and the
capability of identifying humans and interacting with them using intuitive interfaces such as speech, gestures, and facial information. The following analysis will focus on human’s detection and identification using visual information, which,

from our point of view, is required in domestic
robots that need to be general purpose and used
by non-expert users.
The robust detection of humans in real-home
environments is a challenging task, mainly because of variable illumination conditions, cluttered backgrounds, and variable poses of a human body with respect to the robot’s camera. In fact, a
human body is a complex and deformable object
with several degrees of freedom, whose appearance can change greatly when mapped onto a 2D image. Thus, the problem of the detection of a
human body or a human body-part using standard
CCD and CMOS cameras that work in the visible spectrum is far from being solved! Depending on the specific circumstances, humans can be detected by using information about their faces,
silhouettes, skin, or movement, as well as by using
depth information. None of these methods is allpurpose and any of them can fail depending on the specific...
tracking img