Title:
|
A COMPARISON OF OPTICAL FLOW TECHNIQUES FOR A BLIND SPOT CAMERA |
Author(s):
|
Lorenz Van den Heuvel, Kristof Van Beeck, Toon Goedemé |
ISBN:
|
978-972-8939-48-9 |
Editors:
|
Yingcai Xiao |
Year:
|
2011 |
Edition:
|
Single |
Keywords:
|
Computer vision, object recognition, optical flow, quantitative comparison. |
Type:
|
Reflection Paper |
First Page:
|
331 |
Last Page:
|
334 |
Language:
|
English |
Cover:
|
|
Full Contents:
|
click to dowload
|
Paper Abstract:
|
In this paper we present an overview and comparison of the most important optical flow calculation techniques. Our comparison is based upon three aspects: accuracy, speed and robustness against noise. Based on the results of our comparison, we determine which of the algorithms is best suited for real-time implementation on dedicated hardware. This decision is based on the trade-off that we found between our evaluation aspects. In the future we will use this result to achieve the automatic detection of vulnerable road users using a blind spot camera. |
|
|
|
|