Limits...
Articulated Non-Rigid Point Set Registration for Human Pose Estimation from 3D Sensors.

Ge S, Fan G - Sensors (Basel) (2015)

Bottom Line: We introduce a visible point extraction method to initialize a new template for the current frame from the previous frame, which effectively reduces the ambiguity and uncertainty during registration.Third, to support robust and stable pose tracking, we develop a segment volume validation technique to detect tracking failures and to re-initialize pose registration if needed.The experimental results on both benchmark 3D laser scan and depth datasets demonstrate the effectiveness of the proposed framework when compared with state-of-the-art algorithms.

View Article: PubMed Central - PubMed

Affiliation: School of Electrical and Computer Engineering, Oklahoma State University, Stillwater, OK 74078, USA. song.ge@okstate.edu.

ABSTRACT
We propose a generative framework for 3D human pose estimation that is able to operate on both individual point sets and sequential depth data. We formulate human pose estimation as a point set registration problem, where we propose three new approaches to address several major technical challenges in this research. First, we integrate two registration techniques that have a complementary nature to cope with non-rigid and articulated deformations of the human body under a variety of poses. This unique combination allows us to handle point sets of complex body motion and large pose variation without any initial conditions, as required by most existing approaches. Second, we introduce an efficient pose tracking strategy to deal with sequential depth data, where the major challenge is the incomplete data due to self-occlusions and view changes. We introduce a visible point extraction method to initialize a new template for the current frame from the previous frame, which effectively reduces the ambiguity and uncertainty during registration. Third, to support robust and stable pose tracking, we develop a segment volume validation technique to detect tracking failures and to re-initialize pose registration if needed. The experimental results on both benchmark 3D laser scan and depth datasets demonstrate the effectiveness of the proposed framework when compared with state-of-the-art algorithms.

No MeSH data available.


Overview of the proposed human pose tracking framework.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4541828&req=5

f2-sensors-15-15218: Overview of the proposed human pose tracking framework.

Mentions: An overview of the proposed framework is shown in Figure 2, which involves five steps. First, we learn a subject-specific articulated model to initialize the body shape and size for a new subject. Second, visible point extraction is performed from the subject-specific model to create a partial template model, which either involves previous pose estimation or a “T-pose” template. Third, our recently proposed non-rigid registration algorithm (GLTP) is used for correspondence estimation from the observed target model. Fourth, segment volume validation is invoked to detect tracking failures and to trigger pose re-initialization if needed. Last, segment-aware AICP (SAICP) is used for articulated pose estimation by refining correspondence estimation at each segment iteratively. For 3D point sets, only Steps 1, 3 and 5 are needed; while for depth sequences, sequential pose tracking will involve all steps, and Steps 1, 2, 3 and 5 will support frame-by-frame pose estimation.


Articulated Non-Rigid Point Set Registration for Human Pose Estimation from 3D Sensors.

Ge S, Fan G - Sensors (Basel) (2015)

Overview of the proposed human pose tracking framework.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4541828&req=5

f2-sensors-15-15218: Overview of the proposed human pose tracking framework.
Mentions: An overview of the proposed framework is shown in Figure 2, which involves five steps. First, we learn a subject-specific articulated model to initialize the body shape and size for a new subject. Second, visible point extraction is performed from the subject-specific model to create a partial template model, which either involves previous pose estimation or a “T-pose” template. Third, our recently proposed non-rigid registration algorithm (GLTP) is used for correspondence estimation from the observed target model. Fourth, segment volume validation is invoked to detect tracking failures and to trigger pose re-initialization if needed. Last, segment-aware AICP (SAICP) is used for articulated pose estimation by refining correspondence estimation at each segment iteratively. For 3D point sets, only Steps 1, 3 and 5 are needed; while for depth sequences, sequential pose tracking will involve all steps, and Steps 1, 2, 3 and 5 will support frame-by-frame pose estimation.

Bottom Line: We introduce a visible point extraction method to initialize a new template for the current frame from the previous frame, which effectively reduces the ambiguity and uncertainty during registration.Third, to support robust and stable pose tracking, we develop a segment volume validation technique to detect tracking failures and to re-initialize pose registration if needed.The experimental results on both benchmark 3D laser scan and depth datasets demonstrate the effectiveness of the proposed framework when compared with state-of-the-art algorithms.

View Article: PubMed Central - PubMed

Affiliation: School of Electrical and Computer Engineering, Oklahoma State University, Stillwater, OK 74078, USA. song.ge@okstate.edu.

ABSTRACT
We propose a generative framework for 3D human pose estimation that is able to operate on both individual point sets and sequential depth data. We formulate human pose estimation as a point set registration problem, where we propose three new approaches to address several major technical challenges in this research. First, we integrate two registration techniques that have a complementary nature to cope with non-rigid and articulated deformations of the human body under a variety of poses. This unique combination allows us to handle point sets of complex body motion and large pose variation without any initial conditions, as required by most existing approaches. Second, we introduce an efficient pose tracking strategy to deal with sequential depth data, where the major challenge is the incomplete data due to self-occlusions and view changes. We introduce a visible point extraction method to initialize a new template for the current frame from the previous frame, which effectively reduces the ambiguity and uncertainty during registration. Third, to support robust and stable pose tracking, we develop a segment volume validation technique to detect tracking failures and to re-initialize pose registration if needed. The experimental results on both benchmark 3D laser scan and depth datasets demonstrate the effectiveness of the proposed framework when compared with state-of-the-art algorithms.

No MeSH data available.