Abstract

The increasing importance of outdoor applications such as driver assistance systems or video surveillance tasks has recently triggered the development of optical flow methods that aim at performing robustly under uncontrolled illumination. Most of these methods are based on patch-based features such as the normalized cross correlation, the census transform or the rank transform. They achieve their robustness by locally discarding both absolute brightness and contrast. In this paper, we follow an alternative strategy: Instead of discarding potentially important image information, we propose a novel variational model that jointly estimates both illumination changes and optical flow. The key idea is to parametrize the illumination changes in terms of basis functions that are learned from training data. While such basis functions allow for a meaningful representation of illumination effects, they also help to distinguish real illumination changes from motion-induced brightness variations if supplemented by additional smoothness constraints. Experiments on the KITTI benchmark show the clear benefits of our approach. They do not only demonstrate that it is possible to obtain meaningful basis functions, they also show state-of-the-art results for robust optical flow estimation. Document type: Part of book or chapter of book

Abstract

It is widely acknowledged that many geotechnical properties are correlated over space and/or time. Consequently, crosscorrelated random fields play a pivotal role in geotechnical reliability analysis for properly modeling both the auto- and cross-correlation structures of correlated geotechnical properties. Existing methods for simulating cross-correlated random fields typically require precise knowledge of random field parameters as input. However, in a typical site investigation program, engineering constraints such as limited time, budget, and space often lead to sparse measurements of geotechnical properties. Estimating reliable random field parameters, particularly the auto-correlation and crosscorrelation structures of a two-dimensional (2D) cross-correlated random field, from such sparse data is a notorious challenge. To address this issue, this study introduces a 2D cross-correlated random field generator that can directly simulate 2D multivariate cross-correlated geotechnical random field samples (RFSs) from sparsely measured data points. This generator leverages the method developed by Guan and Wang (2023), which employs a joint sparse representation to simultaneously exploit auto- and cross-correlation structures of various spatial/temporal quantities directly from sparse measurements. The effectiveness of the proposed generator is demonstrated using real geotechnical properties data. The results demonstrate that RFSs generated using this method from sparse measurements accurately capture the spatial auto- and cross-correlation structures of different geotechnical properties.