Title :
Target tracking and measuring based on binocular vision
Author_Institution :
Coll. of Autom. Sci. & Technol., South China Univ. of Technol., Guangzhou, China
Abstract :
Vision system which can be used for real-time tracking and measuring of moving target is very important in robot soccer match. Based on DSP embedded vision subsystem, this paper proposes a method to measure the binocular distance based on binocular parallax between two images´ color. First, camera´s intrinsic parameters can be determined by calculating the R component of the pixels´ color of the image in the experiment of camera calibration. Then image segment is used in real-time capture of the target´s images, and the binocular parallax and the world coordinates of target are calculated. Then the target can be located by regional color matching method. The results of the experiments show that this algorithm can detect and locate the target commendably. It sets up a platform for further research which can supply plenteous information for strategy system and tracking target.
Keywords :
cameras; distance measurement; embedded systems; image colour analysis; image matching; image segmentation; object detection; robot vision; target tracking; DSP embedded vision subsystem; binocular distance measurement; binocular parallax; binocular vision; camera calibration; digital signal processing; image color; image segment; moving target tracking; real time tracking; regional color matching method; robot soccer match; target detection; Calibration; Cameras; Image color analysis; Mathematical model; Robot kinematics;
Conference_Titel :
Information Science and Technology (ICIST), 2013 International Conference on
Conference_Location :
Yangzhou
Print_ISBN :
978-1-4673-5137-9
DOI :
10.1109/ICIST.2013.6747797