Automatic reconstruction of 3D human motion pose from uncalibrated monocular video sequences based on markerless human motion tracking

School of Information Science and Engineering, Central South University, Changsha 410083, People's Republic of China
Pattern Recognition (Impact Factor: 2.58). 07/2009; 42(7):1559-1571. DOI: 10.1016/j.patcog.2008.12.024
Source: DBLP

ABSTRACT We present a method to reconstruct human motion pose from uncalibrated monocular video sequences based on the morphing appearance model matching. The human pose estimation is made by integrated human joint tracking with pose reconstruction in depth-first order. Firstly, the Euler angles of joint are estimated by inverse kinematics based on human skeleton constrain. Then, the coordinates of pixels in the body segments in the scene are determined by forward kinematics, by projecting these pixels in the scene onto the image plane under the assumption of perspective projection to obtain the region of morphing appearance model in the image. Finally, the human motion pose can be reconstructed by histogram matching. The experimental results show that this method can obtain favorable reconstruction results on a number of complex human motion sequences.

1 Follower
  • Source
    • "Based on the type of footage used, the markerless pose reconstruction (or motion capture) problem can be roughly categorized into two groups [24]: using video sequences from one camera or using footage from multiple calibrated cameras. Pose estimation from monocular video sequences [2], [3], [24], [17], [1], [18] can be more convenient for some applications as it imposes less restrictions to the user, but it has an inherent depth ambiguity. This ambiguity can be solved using structure from motion approaches, a very difficult problem in vision [13], [14]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: We propose a data-driven, multi-view body pose estimation algorithm for video. It can operate in uncontrolled environments with loosely calibrated and low resolution cameras and without restricting assumptions on the family of possible poses or motions. Our algorithm first estimates a rough pose estimation using a spatial and temporal silhouette based search in a database of known poses. The estimated pose is improved in a novel pose consistency step acting locally on single frames and globally over the entire sequence. Finally, the resulting pose estimation is refined in a spatial and temporal pose optimization consisting of novel constraints to obtain an accurate pose. Our method proved to perform well on low resolution video footage from real broadcast of soccer games.
    International Conference on 3D Imaging, Modeling, Processing, Visualization and Transmission, 3DIMPVT 2011, Hangzhou, China, 16-19 May 2011; 01/2011
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In wireless body sensor networks, sensors may be installed on various body limbs to wirelessly collect body information for homecare services. The orientations and accel-erations on each limb are different for various motion states. For example, each limb has different acceleration when walking versus running, and orientation when standing versus lying. According to the above information, the body motion state may be decided. Furthermore, each person has unique body characteristics such as height, foot pitch, and motion habit to effect the body reconstruction. Therefore, it is a challenging issue how to present human motions through 3D skeleton system simulation, and achieve an adaptive reconstruction of human motion according to the different body characteristics of each person. In this study, we proposed a novel scheme to utilize multiple triple axis accelerometer and gyroscopes to measure limb accelerations, then calculated the locations of limbs and try to employ kinematic theory to reconstruct human body skeleton, called 3D Adaptive human Motion Reconstruction (AMR). And we applied Body Correction Algorithm (BCA) to correct human body characteristics and fighted the error of transmission noise. This system was tested and validated with success.
  • [Show abstract] [Hide abstract]
    ABSTRACT: In order to study human motion in biomechanical applications, a critical component is to accurately obtain the 3D joint positions of the user's body. Computer vision and inverse kinematics are used to achieve this objective without markers or special devices attached to the body. The problem of these systems is that the inverse kinematics is “blinded” with respect to the projection of body segments into the images used by the computer vision algorithms. In this paper, we present how to add image constraints to inverse kinematics in order to estimate human motion. Specifically, we explain how to define a criterion to use images in order to guide the posture reconstruction of the articulated chain. Tests with synthetic images show how the scheme performs well in an ideal situation. In order to test its potential in real situations, more experiments with task specific image sequences are also presented. By means of a quantitative study of different sequences, the results obtained show how this approach improves the performance of inverse kinematics in this application.
    EURASIP journal on advances in signal processing 01/2010; 1110-8657. DOI:10.1155/2010/142354 · 0.89 Impact Factor
Show more