Multi‐view human action recognition using 2D motion templates based on MHIs and their HOG description

Multi‐view human action recognition using 2D motion templates based on MHIs and their HOG description

In this study, a new multi‐view human action recognition approach is proposed by exploiting low‐dimensional motion information of actions. Before feature extraction, pre‐processing steps are performed to remove noise from silhouettes, incurred due to imperfect, but realistic segmentation. Two‐dimens...

Full description

Bibliographic Details
Main Authors: Fiza Murtaza, Muhammad Haroon Yousaf, Sergio A. Velastin
Format: Article
Language:English
Published: Wiley 2016-10-01
Series:IET Computer Vision
Subjects:
multiview human action recognition
2D motion templates
low-dimensional action motion information
feature extraction
silhouette noise removal
realistic segmentation
Online Access:https://doi.org/10.1049/iet-cvi.2015.0416

Internet

https://doi.org/10.1049/iet-cvi.2015.0416

Similar Items