Dear All,
I have implemented this variable track optimization. I have tested
the functionality for simple and real applications and it works
fine.
From the performance point of view, e.g. for KLT feature tracking
application, the optimized version is 49% execution time efficient
and 30% memory efficient.
Let me know about your comments before I merge the pull request to
celab develop branch.
https://github.com/celabtud/QUAD/pull/10
Best Regards,
--
====================================
Imran Ashraf
PhD Student
Computer Engineering Group
S&C Technology, EEMCS, TU Delft
The Netherlands
E: I.As...@tudelft.nl
T: 0031 15 27 89379
M: 0031 645 388 558
W: http://ce.et.tudelft.nl/~iashraf/
====================================