Arts and Technology. First International Conference, ArtsIT 2009, Yi-Lan, Taiwan, September 24-25, 2009, Revised Selected Papers

Research Article

Improving Optical Flow Using Residual and Sobel Edge Images

Download64 downloads
  • @INPROCEEDINGS{10.1007/978-3-642-11577-6_27,
        author={Tobi Vaudrey and Andreas Wedel and Chia-Yen Chen and Reinhard Klette},
        title={Improving Optical Flow Using Residual and Sobel Edge Images},
        proceedings={Arts and Technology. First International Conference, ArtsIT 2009, Yi-Lan, Taiwan, September 24-25, 2009, Revised Selected Papers},
        proceedings_a={ARTSIT},
        year={2012},
        month={5},
        keywords={},
        doi={10.1007/978-3-642-11577-6_27}
    }
    
  • Tobi Vaudrey
    Andreas Wedel
    Chia-Yen Chen
    Reinhard Klette
    Year: 2012
    Improving Optical Flow Using Residual and Sobel Edge Images
    ARTSIT
    Springer
    DOI: 10.1007/978-3-642-11577-6_27
Tobi Vaudrey1, Andreas Wedel2, Chia-Yen Chen3, Reinhard Klette1
  • 1: The University of Auckland
  • 2: Daimler Research, Daimler AG
  • 3: National University of Kaohsiung

Abstract

Optical flow is a highly researched area in low-level computer vision. It is a complex problem which tries to solve a 2D search in continuous space, while the input data is 2D discrete data. Furthermore, the latest representations of optical flow use Hue-Saturation-Value (HSV) colour circles, to effectively convey direction and magnitude of vectors. The major assumption in most optical flow applications is the intensity consistency assumption, introduced by Horn and Schunck. This constraint is often violated in practice. This paper proposes and generalises one such approach; using residual images (high-frequencies) of images, to remove the illumination differences between corresponding images.