Title :
Silhouette-Based 2D Human Pose Estimation
Author :
Li, Meng ; Yang, Tao ; Xi, Runping ; Lin, Zenggang
Author_Institution :
Shaanxi Key Lab. of Speech & Image Inf. Process., Northwestern Polytech. Univ., Xi´´an, China
Abstract :
In this paper we present a novel silhouette-based method to estimate 2D human pose. It takes a pre-defined human skeleton model as the prior information and a video sequence as the data source, and estimates human pose in each frame by the following steps: Firstly, the Gaussian Mixture Background Model (GMM) is adopted to extract silhouette from an image and this silhouette will be the human body data set after treatment. Then the Distance Transform (DT) and Principal Component Analysis (PCA) are introduced to locate the base point of the human skeleton model, with which the human skeleton model is initialized automatically. Afterwards an iterative process, based on the Expectation Maximization (EM), is constructed out to cluster the human body data set and estimate the parameters of the skeleton iteratively. Finally the pose of human body is figured out in the form of a corresponding skeleton model. Extensive experiments show that this method is robust and precise, and is feasible to apply in the real-time system.
Keywords :
Gaussian processes; expectation-maximisation algorithm; feature extraction; image sequences; pose estimation; principal component analysis; real-time systems; Gaussian mixture background model; PCA; distance transform; expectation maximization; human skeleton model; iterative process; principal component analysis; real time system; silhouette based 2D human pose estimation; video sequence; Biological system modeling; Computer graphics; Computer vision; Data mining; Humans; Parameter estimation; Principal component analysis; Robustness; Skeleton; Video sequences;
Conference_Titel :
Image and Graphics, 2009. ICIG '09. Fifth International Conference on
Conference_Location :
Xi´an, Shanxi
Print_ISBN :
978-1-4244-5237-8
DOI :
10.1109/ICIG.2009.91