Faculty of Informatics Vienna University of Technology Institute of Computer Aided Automation PRIP Home PRIP Home
Personal tools
You are here: Home Teaching PRIP-Talks Harald Reingruber: An asynchronous Stereo Vision Concept for Event-based Vision Systems
Navigation
 

Harald Reingruber: An asynchronous Stereo Vision Concept for Event-based Vision Systems

Spezifikationsvortrag Diplomarbeit

What
  • Presentation
When Dec 15, 2009
from 04:35 pm to 04:55 pm
Where Sem 183/2
Add event to calendar vCal
iCal
Computer vision systems usually operate by capturing and processing
sequences of frames, which are then processed frame by frame and in most
cases pixel by pixel. The human brain doesn’t operate on a frame by frame
concept. Each ganglion cell sends spikes (in this case called events) to the
visual cortex, when its activity level reaches a threshold. During the
current decade researchers developed bio-inspired sensors which mimic the
human visual sensing. The output of this event-based vision sensor is a
stream of events, generated by autonomous pixels, which generate events as
soon as they occur and do not wait for an artificial frame time.

Current Stereo Vision concepts based on Address-Event Representation (AER)
data are abandoning the advantages of this asynchronous data representation
by buffering incoming events into artificially introduced pseudo-frames. Aim
of this thesis is to investigate in an asynchronous stereo matching concept
which mostly preserves these advantages and, for example, is applicable on
fast as well as slow movements in the sensors field of view. The key
questions will be: How much data is necessary to be considered for a stereo
matching approach and how can it be measured when this point is reached?
Document Actions