Automated Tracking and Estimation for Control of Non-rigid Cloth

This report is a summary of research conducted on cloth tracking for automated textile manufacturing during a two semester long research course at Georgia Tech. This work was completed in 2009. Advances in current sensing technology such as the Microsoft Kinect would now allow me to relax certain assumptions and generally improve the tracking performance. This is because a major part of my approach described in this paper was to track features in a 2D image and use these to estimate the cloth deformation. Innovations such as the Kinect would improve estimation due to the automatic depth information obtained when tracking 2D pixel locations. Additionally, higher resolution camera images would probably give better quality feature tracking. However, although I would use different technology now to implement this tracker, the algorithm described and implemented in this paper is still a viable approach which is why I am publishing this as a tech report for reference. In addition, although the related work is a bit exhaustive, it will be useful to a reader who is new to methods for tracking and estimation as well as modeling of cloth. Introduction Recent work on the modeling and tracking of cloth is focused primarily either on realistic recreation of cloth geometry and behavior or being able to simulate cloth in real-time. Research in the area of the fashion or movie industries has focused on the accuracy of cloth data capture in being able to accurately represent clothing and even replace it in post-production of a film [8][26][28]. Often this kind of work requires extensive setup, perhaps numerous cameras and does not run in near real-time. On the other hand, the emphasis in the video game industry is to model or simulate cloth in real-time. However most of these implementations or applications are focused on being aesthetically pleasing rather than physically accurate [30]. This means that although some aspects of cloth parameter estimation and tracking have become highly developed, the ability to control a piece of cloth using these models in real time with current methods is infeasible. In this work I developed an Extended Kalman filter cloth tracker that can run in near real-time (4-20 Hz) and accurately represent the state of the cloth for translations and rotations involving minor deformations. In addition I implement an EKF with a mesh model for the cloth, borrowing from finite element methods and show some promising initial results …