Fast Guided Global Interpolation for Depth and Motion
We study the problems of upsampling a low-resolution depth map and interpolating an initial set of sparse motion matches, with the guidance from a corresponding high-resolution color image. The common objective for both tasks is to densify a set of sparse
- PDF / 3,891,233 Bytes
- 17 Pages / 439.37 x 666.142 pts Page_size
- 72 Downloads / 240 Views
3
Advanced Digital Sciences Center, Singapore, Singapore [email protected] 2 Chungnam National University, Daejeon, Korea University of Illinois at Urbana-Champaign, Champaign, USA
Abstract. We study the problems of upsampling a low-resolution depth map and interpolating an initial set of sparse motion matches, with the guidance from a corresponding high-resolution color image. The common objective for both tasks is to densify a set of sparse data points, either regularly distributed or scattered, to a full image grid through a 2D guided interpolation process. We propose a unified approach that casts the fundamental guided interpolation problem into a hierarchical, global optimization framework. Built on a weighted least squares (WLS) formulation with its recent fast solver – fast global smoothing (FGS) technique, our method progressively densifies the input data set by efficiently performing the cascaded, global interpolation (or smoothing) with alternating guidances. Our cascaded scheme effectively addresses the potential structure inconsistency between the sparse input data and the guidance image, while preserving depth or motion boundaries. To prevent new data points of low confidence from contaminating the next interpolation process, we also prudently evaluate the consensus of the interpolated intermediate data. Experiments show that our general interpolation approach successfully tackles several notorious challenges. Our method achieves quantitatively competitive results on various benchmark evaluations, while running much faster than other competing methods designed specifically for either depth upsampling or motion interpolation. Keywords: Image-guided interpolation flow
1
·
Depth upsampling
·
Optical
Introduction
Dense depth or optical flow maps often serve as a fundamental building block for many computer vision and computational photography applications, e.g., 3D Jiangbo Lu—This study is supported by the HCCS grant at the ADSC from Singapore’s Agency for Science, Technology and Research (A*STAR). Electronic supplementary material The online version of this chapter (doi:10. 1007/978-3-319-46487-9 44) contains supplementary material, which is available to authorized users. c Springer International Publishing AG 2016 B. Leibe et al. (Eds.): ECCV 2016, Part III, LNCS 9907, pp. 717–733, 2016. DOI: 10.1007/978-3-319-46487-9 44
718
Y. Li et al.
Fig. 1. Using the same guided interpolation pipeline, our technique gives strong results for two tasks: (top) depth upsampling and (bottom) optical flow field interpolation. Local or non-local methods (e.g. GF [21] and Epic-LA [35]) are usually efficient, but suffer from limitations like copying texture from the color guidance (green arrows) and inability to interpolate pixels in a distance (green rectangle). Methods using complicated models and global optimization (e.g. AR [43]) can obtain high quality results, but are often rather slow in computation. Our method, with a unified framework for both problems, is 1000× faster than AR [43] and even faster than local me
Data Loading...
