A Spatio-Angular Binary Descriptor For Fast Light Field Inter View Matching

Abstract

Light fields are able to capture light rays from a scene arriving at different angles, effectively creating multiple perspective views of the same scene. Thus, one of the flagship applications of light fields is to estimate the captured scene geometry, which can notably be achieved by establishing correspondences between the perspective views, usually in the form of a disparity map. Such correspondence estimation has been a long standing research topic in computer vision, with application to stereo vision or optical flow. Research in this area has shown the importance of well designed descriptors to enable fast and accurate matching. We propose in this paper a binary descriptor exploiting the light field gradient over both the spatial and the angular dimensions in order to improve inter view matching. We demonstrate in a disparity estimation application that it can achieve comparable accuracy compared to existing descriptors while being faster to compute.

Light field multi-scale blocked local binary pattern feature in the spatial-angular domain

“…To facilitate feature matching between multiple views, Alain et al [17] propose an improved feature description for binarized features in the light field spatial-angular domain. Liu et al [17] propose a focused all-optical camera feature detection based on a center-projected stereo focus stack.…”

“…To facilitate feature matching between multiple views, Alain et al [17] propose an improved feature description for binarized features in the light field spatial-angular domain. Liu et al [17] propose a focused all-optical camera feature detection based on a center-projected stereo focus stack. The above light field feature extractions and matchings outperform the classic 2D image feature algorithms in terms of precision and robustness, and show a great potential to solve the problem of feature matching ambiguity.…”

A Light Field FDL-HCGH Feature in Scale-Disparity Space

“…Ji et al [26] propose a LF directional gradient histogram (LFHoG) feature to achieve high precision live face detection, while a LF local binary patterns (LFLBP) is introduced in [27], which enhances the LF based face recognition. The authors in [28] propose a solution of accurate and fast disparity estimation by introducing a binary descriptor, which exploits the light field gradient over both the spatial and the angular dimensions.…”