In this study, Mean-shift, Cam-shift and Color Tracking algorithms which all be used in object tracking are compared in a real-time robotic application. The environment in which the robot and the object followed by robot is recorded by a camera and that data is sent to a computer through cables. To identify this object by using appropriate image process tracking, the location of the robot and the object is could be monitored at the same time. According to the results, the most effective method is shared at the discussion of the outcomes. The robot used in this application has omni-wheel and were designed in our own laboratory.
Keywords — Target tracking, Mean-Shift, Cam-Shift, image processing, remote control, omni wheel
link: https://www.icens.eu/sites/default/files/icens_2016_chapter_07.pdf
Prepared by M. Mustafa Kelek, M. Fatih Aslan and Akif Durdu
RACLAB 