Mobile Robot Navigation Based on Circle Recognition
| Autor: | Peter Hubinský, Peter Pásztó |
|---|---|
| Rok vydání: | 2013 |
| Předmět: |
Computer science
business.industry ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Mobile robot Image processing Mobile robot navigation Hough transform law.invention Stereopsis law Digital image processing Trajectory Computer vision Noise (video) Artificial intelligence Electrical and Electronic Engineering business |
| Zdroj: | Journal of Electrical Engineering. 64:84-91 |
| ISSN: | 1335-3632 |
| DOI: | 10.2478/jee-2013-0012 |
| Popis: | This paper presents a navigation method for a mobile robot using a visual system. Circular marks with specific colors are used for marking the significant points of the mobile robot’s trajectory that it needs to pass. The colors of the used marks are signalizing the way of their bypassing with the mobile robot (from the left or right side). The mobile robot uses only one camera for the marks recognition task and it is able to determine its own relative position from the detected marks. The image processing and the mobile robot’s trajectory planning algorithm working in real-time are described in this paper. K e y w o r d s: circular mark detection, mobile robot trajectory planning, image processing, Hough transform The aim of this paper is to propose a navigation method for a mobile robot using a visual system that consists of only one ordinary camera (web camera). The algorithm should be able to determine the relative position between the mobile robot and the detected marks and to plan the mobile robot’s trajectory to bypass the marks from a desired side. Most of visual system applications are using stereo vision techniques for object’s position determination. In this paper the position of the mobile robot relative to the detected mark is computed using only one camera with unknown optical parameters. The navigation consists of two main parts: the image processing part and the trajectory planning part. Image processing part is a combination of more image processing algorithms (steps) and its goal is to detect all the circular marks in an unknown environment while the changing of light conditions is possible. This part also determines the colors of the detected marks. Changing of the light conditions is an important disturbance bringing noise into the image processing. Therefore the image processing application needs to be designed the way to be able to troubleshoot this disturbance. Trajectory planning part is computing the positions of the detected marks relative to the position of the mobile robot using the data obtained with image processing part. After it the mobile robot’s trajectory is planned using ? + |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|