PT Journal
AU Doshi, A
   Starck, J
   Hilton, A
TI An Empirical Study of Non-Rigid Surface Feature Matching of Human from 3D Video
SO JVRB - Journal of Virtual Reality and Broadcasting
PY 2010
VL 7(2010)
IS 3
DI 10.20385/1860-2037/7.2010.3
DE Feature Matching; Qualitative Analysis; Recall-Accuracy; Sift; Video Sequences
AB This paper presents an empirical study of affineinvariant feature detectors to perform matching onvideo sequences of people with non-rigid surfacedeformation. Recent advances in feature detection andwide baseline matching have focused on static scenes.Video frames of human movement capture highlynon-rigid deformation such as loose hair, cloth creases,skin stretching and free flowing clothing. This studyevaluates the performance of six widely used featuredetectors for sparse temporal correspondence on singleview and multiple view video sequences. Quantitativeevaluation is performed of both the number of featuresdetected and their temporal matching against and withoutground truth correspondence. Recall-accuracy analysis offeature matching is reported for temporal correspondenceon single view and multiple view sequences of people   with variation in clothing and movement. This analysis  identifies that existing feature detection and matching  algorithms are unreliable for fast movement with common  clothing.
ER