Machine vision use in navigation system for task of route and its landmarks recognition is considered. A special recognition algorithm based on the Kenny detector was applied for image processing, it was decided to use the programming language Python, connected to the library OpenCV. Due to this it became possible to implement the program with a minimal amount of code with microcontroller-based board Raspberry PI.
The developed software for mark recognition uses the method «Homography to find Objects», which scans the image and search matches of preset elements’ quantity of scanned landmark and original landmark’s image.
The developed system for automatic orientation, positioning and navigation of robot can be used as part of control systems for AGV, service robots, robots for safety control and for work in harsh environments.
The work was performed under the grant RFBR № 12-07-97526-r_tsentr_a.
Control, robotics, machine vision, tags, robot, mobile platforms.
Year 2015 Issue number 4 Consecutive issue number 9 Pages 30-33
UDC identifier: 004.93'1:007.52