Detection and Tracking of Wheeled Mobile Robot by Image Processing

Jacob, Jennifer and Mathew, Robins and Hiremath, Somashekhar S and Bhat, Chandrashekhar (2017) Detection and Tracking of Wheeled Mobile Robot by Image Processing. International Journal of Engineering Research in Mechanical and Civil Engineering, 2 (3). pp. 505-511. ISSN 2456-1290

[img] PDF
2773.pdf - Published Version
Restricted to Registered users only

Download (357kB) | Request a copy


Mobile robotics is a fast developing area due to their wide applications in exploration on-ground, under-ground, on-water, under-water and space. Detection and tracking of the paths followed by these robots are required to analyze their motion and to guide them through optimal path. This paper focuses on the application of image-processing for estimating the position and velocity of mobile robot on an indoor workspace. The motion of the robot on the workspace was captured using monocular vision system. In order to detect the path of the mobile robot, the obtained image was processed using a full featured high-level programming language-MATLAB. This work is fully dependent on how well the feature of the robot in the image plane was extracted and tracked from the initial frame to the subsequent frames. Thus, color-based feature extraction and blob analysis are used to serve the purpose of detecting the path of the mobile robot. The resulting data was further used to obtain path coordinates which in turn gives the position and velocity errors of the robot. A calibration phase was carried out to analyze the relationship between the measured coordinates (pixels) with the world coordinates. Some of the challenges faced during implementation of the system are also mentioned in the paper

Item Type: Article
Uncontrolled Keywords: Blob analysis, Feature-extraction, Image-processing, Position estimation, Velocity estimation, Wheeled mobile robot
Subjects: Engineering > MIT Manipal > Mechatronics
Depositing User: MIT Library
Date Deposited: 24 Jun 2017 08:45
Last Modified: 24 Jun 2017 08:45

Actions (login required)

View Item View Item