State-of-the-art Solutions for Vision-based Estimation

of 5

Please download to get full document.

View again

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
PDF
5 pages
0 downs
40 views
Share
Description
1. 1/5G. Silveira, “State-of-the-art Solutions for Vision-based Estimation” - State-of-the-art Solutions for Vision-based Estimation Geraldo Silveira, PhD…
Transcript
  • 1. 1/5G. Silveira, “State-of-the-art Solutions for Vision-based Estimation” - State-of-the-art Solutions for Vision-based Estimation Geraldo Silveira, PhD Geraldo.Silveira@cti.gov.br
  • 2. 2/5G. Silveira, “State-of-the-art Solutions for Vision-based Estimation” - Problem Statement • Vision-based estimation of parameters that best describe a phenomenon of interest • Performance requirements: high accuracy, fast response and robustness to lighting changes • Applications in 3D mapping, augmented reality, visual tracking, pose estimation, etc. Example: Estimate motion of a beating heart
  • 3. 3/5G. Silveira, “State-of-the-art Solutions for Vision-based Estimation” - Existing Methods  are not robust to arbitrary illumination changes; or  require particular image features; or  require particular objects; or  require particular cameras; or  require particular camera-object situations; or  require off-line learning steps; or  require relatively slow camera motion
  • 4. 4/5G. Silveira, “State-of-the-art Solutions for Vision-based Estimation” - State-of-the-art Solution  We have proposed a vision-based estimation solution which – is highly accurate: all possible visual information is exploited; no need of particular image features (e.g., points, lines, etc.) – is versatile: works regardless of object shape, of camera motion and of camera type (e.g., standard and omni cameras) – is robust to arbitrary illumination changes and to aberrant measures (e.g., occlusions) – has large domain and rate of convergence; starts immediately after task definition; no off-line learning steps Further information can be obtained upon request
  • 5. 5/5G. Silveira, “State-of-the-art Solutions for Vision-based Estimation” - Application: Image Stabilization  Experimental setup – a fully uncalibrated conventional camera – completely unknown illumination conditions – a freely moving object of unknown dimensions  Typical result Region to be tracked and stabilized Some excerpts of the experiment
  • Related Search
    We Need Your Support
    Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

    Thanks to everyone for your continued support.

    No, Thanks