3D Human model adaptation by frame selection and shape-texture optimization
Computer Vision and Image Understanding
| Hi-index | 0.00 |
We propose a real-time method for simultaneously refining the reconstructed volume of a human body with loose-fitting clothing and identifying body-parts in it. Time-series volumes, which are acquired by a slow but sophisticated 3D reconstruction algorithm, with body-part labels are obtained offline. The time-series sample volumes are represented by trajectories in the eigenspaces using PCA. An input visual hull reconstructed online is projected into the eigenspace and compared with the trajectories in order to find similar high-precision samples with body-part labels. The hierarchical search taking into account 3D reconstruction errors can achieve robust and fast matching. Experimental results demonstrate that our method can refine the input visual hull including loose-fitting clothing and identify its body-parts in real time.