NRST: Non-rigid Surface Tracking from Monocular Video
This addresses a specific challenge in computer vision for tracking non-rigid surfaces, particularly useful for applications involving fabrics or similar materials, but it appears incremental as it builds on existing registration methods with a tailored texture term.
The paper tackles the problem of non-rigid surface tracking from monocular RGB videos by proposing an efficient method that registers a template mesh to each frame, introducing a novel texture term to handle objects with uniform texture but fine-scale structure like fabrics. The results show effectiveness on both general textured objects and monochromatic fabrics, though no concrete numbers are provided.
We propose an efficient method for non-rigid surface tracking from monocular RGB videos. Given a video and a template mesh, our algorithm sequentially registers the template non-rigidly to each frame. We formulate the per-frame registration as an optimization problem that includes a novel texture term specifically tailored towards tracking objects with uniform texture but fine-scale structure, such as the regular micro-structural patterns of fabric. Our texture term exploits the orientation information in the micro-structures of the objects, e.g., the yarn patterns of fabrics. This enables us to accurately track uniformly colored materials that have these high frequency micro-structures, for which traditional photometric terms are usually less effective. The results demonstrate the effectiveness of our method on both general textured non-rigid objects and monochromatic fabrics.