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.


Result comparison on SCAPE data with average joint position errors (cm).
© Copyright Policy
Related In: Results  -  Collection

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

f14-sensors-15-15218: Result comparison on SCAPE data with average joint position errors (cm).

Mentions: We compare pose estimation results in terms of joint position error (cm) in Figure 14. We can see that directly using the estimated corresponding points to compute joint positions cannot achieve a reasonable pose estimation result. Although compared with CPD, GLTP provides much better results, the correspondence estimation around the connection area between two adjacent segments is not reliable due to the lack of segmental information during the registration, which leads to inaccurate pose estimation. As we mentioned before, without a good initialization, AICP is usually trapped into local minima, which results in large estimation errors. Our framework significantly outperforms other options, including CPD, GLTP, AICP and CPD + SAICP, showing the effectiveness of GLTP for correspondence estimation and the necessity of SAICP for pose estimation, which involves the segmental information to refine the GLTP results. We also present some pose estimation results in Figure 15.


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

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

Result comparison on SCAPE data with average joint position errors (cm).
© Copyright Policy
Related In: Results  -  Collection

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

f14-sensors-15-15218: Result comparison on SCAPE data with average joint position errors (cm).
Mentions: We compare pose estimation results in terms of joint position error (cm) in Figure 14. We can see that directly using the estimated corresponding points to compute joint positions cannot achieve a reasonable pose estimation result. Although compared with CPD, GLTP provides much better results, the correspondence estimation around the connection area between two adjacent segments is not reliable due to the lack of segmental information during the registration, which leads to inaccurate pose estimation. As we mentioned before, without a good initialization, AICP is usually trapped into local minima, which results in large estimation errors. Our framework significantly outperforms other options, including CPD, GLTP, AICP and CPD + SAICP, showing the effectiveness of GLTP for correspondence estimation and the necessity of SAICP for pose estimation, which involves the segmental information to refine the GLTP results. We also present some pose estimation results in Figure 15.

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.