OCTOBER 19-20, 2012 - YMCA University of Science & Technology

Proceedings of the National Conference on
Trends and Advances in Mechanical Engineering,
YMCA University of Science & Technology, Faridabad, Haryana, Oct 19-20, 2012
NAVIGATION CONTROL AND LOCALIZATION OF MOBILE
ROBOT
Meghana S 1 and Dr. D.N Drakshayani 2
1,2 Department of Mechanical Engineering, Sir M Visvesvaraya Institute of Technology, Bengaluru
1 Corresponding author: megha87_s@yahoo.co.in
Abstract
This paper presents an experimental study on localization and navigation control based on vision system. The
system is designed for a robot navigating in an indoor environment with a single web camera. A colour image
segmentation method based on RGB (Red, Green, and Blue) colour information of the colour image is applied
for the recognition of position and orientation of a small mobile robot. First, the lower and higher threshold
values of RGB are determined for the sample colours. The colour image segmentation and processing is
performed using MATLAB tool. The localization and navigation control of the robots are efficiently and
accurately established by knowing the intensity of each extracted colour region. Experimental results of the
colour image segmentation or threshold segmentation method applied to the recognition of an object are
studied. This work proposes the navigation control of a mobile robot with inexpensive hardware.
1. Introduction
Autonomous mobile robots need the capability to explore, navigate and localize themselves in dynamic or
unknown environments in order to suit the wide range of industrial applications. In the past two decades, a
number of different approaches have been proposed to develop flexible and efficient navigation systems for
manufacturing industry, based on different sensor technologies such as odometry, Inertial Navigation, Landmark
Navigation, laser scanners, inertial sensors, sonar and vision [1] etc.
Localization and navigation are the fundamental problems of mobile robots. In the past, a variety of approaches
for mobile robot localization has been developed. They mainly differ in the techniques used to represent the
belief of the robot about its current position and according to the type of sensor information that is used for
localization [2]. The basic requirements for the autonomous navigation of a mobile robot are environmental
recognition, path planning, driving control, and location estimation/correction capabilities [3], [4]. The location
estimation and correction capabilities are practically indispensable for the autonomous mobile robot to execute
the given tasks efficiently. There are two general methods for the estimation of location: 1) using deadreckoning
sensors attached at the wheels and body to add the displacement to the initial position data and 2)
using a camera, ultrasonic, laser, radar and/or infrared sensors to recognize and locate beacons. The former
scheme, though cheap and easy to implement, is subject to the aggregate error that may result in inaccurate
position data. The latter scheme has become very popular recently, as it can provide precise location data
instantaneously. However, there are many factors involved in obtaining accurate location information while the
mobile robot is moving [5]. To get reliable and precise location data, sensor fusion techniques [6], [7] have also
been developed.
When a charge-coupled device (CCD) camera is utilized under good illumination conditions, certain patterns or
shapes of objects are also effective for determining the location [8], [9]. Cameras are excellent sensors in robotic
systems. However, a camera being extreme sensitive to illumination variation restricts its capability. The image
sensor of a camera has extremely high sensitivity to varying light illumination. The image sensor in a camera
usually results in inconsistent colour data due to little illumination variation because of fixed aperture. Lighting
is one of the most important considerations for any vision application. Varying intensity of light source surely
makes captured image data different. Without the right type of lighting, the application will not be successful
[10]. Most researchers [11], [12] focus on the indoor navigation of a mobile robot in a well-structured
environment. In other words, beacons, doors, and corridor edges are utilized to estimate the current location of
the mobile robot.Hence in this work, the localization system based on the Infrared sensors and vision based
system has been developed. It works on relative positioning technique promising with better performance and at
low cost. Even though using a vision sensor or combination of vision sensor and infrared sensors can provide
plenty of information about the environment, extraction of visual features for positioning is not easy.
Localization is done by the identification of colours with the predefined information of their intensity for the
given environment. During its operation, the robot uses its Infrared sensors to scan obstacles present in its
surrounding.
289