ECCV 2012 - LNCS 7572-7578 and 7583-7585

Fast Parameter Sensitivity Analysis of PDE-Based Image Processing Methods

Torben Pätz^{1, 2} and Tobias Preusser^{1, 2}

¹School of Engineering and Science, Jacobs University Bremen, Germany

²Fraunhofer MEVIS, Bremen, Germany

Abstract. We present a fast parameter sensitivity analysis by combining recent developments from uncertainty quantification with image processing operators. The approach is not based on a sampling strategy, instead we combine the polynomial chaos expansion and stochastic finite elements with PDE-based image processing operators. With our approach and a moderate number of parameters in the models the full sensitivity analysis is obtained at the cost of a few Monte Carlo runs. To demonstrate the efficiency and simplicity of the approach we show a parameter sensitivity analysis for Perona-Malik diffusion, random walker and Ambrosio-Tortorelli segmentation, and discontinuity-preserving optical flow computation.

LNCS 7578, p. 140 ff.

Full article in PDF | BibTeX