Abstract
To utilize the geometry structure information and similarity information within the neighborhood surrounding a spatio-temporal interest point for human action recognition task, we employ the axes of a regular polyhedron as a reference locating system, and build a novel local feature named polyhedron neighborhood feature (PNF). Then, to reduce quantization error in the coding stage, locality-constrained linear coding method is used to encode the obtained PNFs. Next, multi-temporal-scale PNFs (MPNFs) are created for handling the problem of various action speeds. In classification, support vector machine (SVM) based on linear kernel is used as classifier taking time consumption into account. The experiments on the KTH and UCF sports datasets show that the recognition system based on PNFs achieves better performance than the competing local spatio-temporal feature-based human action recognition methods.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
More From: International Journal of Multimedia and Ubiquitous Engineering
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.