Open Access Open Access  Restricted Access Subscription Access

Measurement of Active Quantity using Local Measurement of Motion Region and Proposal of Body Model for Active Recognition

Miwa Takai

Abstract


This paper presents Active Quantity, which uses Motion Region occurring from human behavior, to measure how a human body moves activity. And, we propose a human body model for action recognition by the characteristic postures and the outbreak parts of human movement having high Active Quantity on the body. The experiment divides a dynamic image, which photographed a human behavior, into frames every constant time, and finds Motion Region by the temporal gradient between frames. As Active Quantity, we denote the active state of a human behavior, which measures how much the body moves lively, by size and dense of Motion Region. This method for measurement Active Quantity sets a target pixel on frame, and the neighborhood region between neighboring frames.

Full Text:

PDF

References


Cristian Canton-Ferrer, Carlos Segura, Josep R. Casas, Montse Pard‘ as, and Javier Hernando, Audiovisual Head Orientation Estimation with Particle Filtering in Multisensor Scenarios, Hindawi Publishing Corporation, EURASIP Journal on Advances in Signal Processing Volume 2008, Article ID 276846, 12 pages.

S.J.McKenna, H.Nait-Charif, Tracking human motion using auxiliary particle filters and iterated likelihood weighting, Image and Vision Computing, Vol. 25, pp852-862, 2007.

Qiang Zhou, Limin Ma, David Chelberg, Adaptive object detection and recognition based on a feedback strategy, Image and Vision Computing, Vol. 24, pp80-93, 2006.

Paul Brasnett, Lyudmila Mihaylova, David Bull, Nishan Canagarajah, Sequential Monte Carlo tracking by fusing multiple cues in video equences, Image and Vision Computing, Vol. 25, pp1217- 1227, 2007.

Yuan-KaiWang and Kuang-You Cheng, A Two-Stage Bayesian NetworkMethod for 3D Human Pose Estimation fromMonocular Image Sequences, Hindawi Publishing Corporation, EURASIP Journal on Advances in Signal Processing Volume 2010, Article ID 761460, 16 pages

Cristian Canton-Ferrer, Josep R. Casas, andMontse Pard‘as,

Marker-Based HumanMotion Capture inMultiview Sequences,Hindawi Publishing Corporation, EURASIP Journal on Advances in Signal Processing Volume 2010, Article ID 105476, 11 pages.

Feng Guo and Gang Qian, Monocular 3D Tracking of Articulated HumanMotion in Silhouette and Pose Manifolds, Hindawi Publishing Corporation, EURASIP Journal on Image and Video Processing Volume 2008, Article ID 326896, 18 pages.

M.Mozerov, I. Rius, X. Roca, and J. Gonz´alez, Nonlinear Synchronization for Automatic Learning of 3D Pose Variability in Human-Motion Sequences, Hindawi Publishing Corporation, EURASIP Journal on Advances in Signal Processing Volume 2010, Article ID 507247, 10 pages.

Toshiyuki Kirishima, YoshitsuguManabe, Kosuke Sato, and Kunihiro Chihara, Real-TimeMultiview Recognition of Human Gestures by Distributed Image Processing, Hindawi Publishing Corporation, EURASIP Journal on Image and Video Processing Volume 2010, Article ID 517861, 13 pages.

Shaogang Gong, Jeffrey Ng, and Jamie Sherrah, On the semantics of visual behavior, structured events and trajectories of human action, Image and Vision Computing, Vol. 20, pp873-888, 2002.

Richard Souvenir and Kyle Parrigan, ViewpointManifolds for Action Recognition, Hindawi Publishing Corporation, EURASIP Journal on Image and Video Processing Volume 2009, Article ID 738702, 13 pages.




DOI: http://dx.doi.org/10.21535%2FProICIUS.2011.v7.390

Refbacks

  • There are currently no refbacks.