Open Collections

A computational vision system for joint angle sensing

University of British Columbia

The research described was aimed at developing passive vision techniques to perform kinematic calibration for a multi-link manipulator, in near real time. Working with a single image of the arm of a Caterpillar excavator, edges are extracted first, then the constrained nature of the problem allows filtering of edge elements to select only those potentially arising from the arm. A model based matching process is performed on this refined data to determine the desired joint angles quickly and efficiently. Methods of exploiting the parallel nature of these techniques are also described.

Text

2010-08-31

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.

http://hdl.handle.net/2429/28029

Computer Science

University of British Columbia

Graduate