Created by the PUT Vision Lab.
Successful establishing of point correspondences between consecutive image frames is important in tasks such as visual odometry,structure from motion or simultaneous localization and mapping. We present the architecture of the compact, energy-efficient dedicated hardware processors, enabling fast feature detection and matching.
You are free to use the code for research/hobby purposes. If you plan to use the code or its fragments in a commercial application, please contact the Authors, so that further agreements can be made. If the posted code is useful to you, the Authors would be really grateful if you cite the folowing articles:
- System on chip coprocessors for high speed image feature detection and matching
@inproceedings{ACIVS2011,
author = {Marek Kraft and Michal Fularz and Andrzej Kasinski},
title = {System on Chip Coprocessors for High Speed Image Feature Detection and Matching},
booktitle = {Proceedings of the 13th International Conference on Advanced Concepts for Intelligent Vision Systems},
series = {ACIVS'11},
year = {2011},
isbn = {978-3-642-23686-0},
location = {Ghent, Belgium},
pages = {599--610},
publisher = {Springer-Verlag},
}
- High-speed image feature detection using FPGA implementation of FAST algorithm
@inproceedings{VISAPP2008,
author = {Marek Kraft and Adam Schmidt and Andrzej Kasinski},
booktitle = {VISAPP (1)},
date = {2008-04-07},
editor = {Ranchordas, Alpesh and Araújo, Helder},
isbn = {978-989-8111-21-0},
pages = {174-179},
publisher = {INSTICC - Institute for Systems and Technologies of Information, Control and Communication},
title = {High-Speed Image Feature Detection Using FPGA Implementation of Fast Algorithm.},
year = 2008
}