New multicolour illuminated contour-based markers and their use in motion capture

This research investigates how some of the most pertinent problems with current motion capture technology can be addressed by devising a new set of Multicolour Illuminated Contour-based markers, and a prototype optical motion capture system, which is capable of capturing and reconstructing the 3D positions of the new markers. Optical motion capture systems use cameras to extract significant movement information from dynamic systems in 3D space. Currently, optical motion capture is an important tool in a range of areas such as digital entertainment, medical analysis, and sports applications. In the future it has the potential to offer great advantages in advanced control interfaces for robots, which can be used for telemanipulation tasks. In each of these fields, accurately captured movement can be used to support suspension of disbelief, accurate analysis and detailed control in an effective and efficient way. There are a number of problems with current optical motion capture systems. These problems include: i) noise in data signals, ii) missing data caused by occlusions, and iii) marker identification. To explore how these problems can be solved, this research proposes a set of new markers, which are constructed ofintersecting self-illuminating line segments of different colours. To guide the development ofthis new marker system, the systems development research methodology has been adopted and a prototype optical motion capture system capable oftracking the new markers has been constructed. To evaluate the accuracy and robustness ofthe developed technology, a range of experiments have been conducted. Results from these experiments, indicate that the new motion capture technology, support improved performance when compared to current optical motion capture systems. The improved performance demonstrated included making it possible to: i) filter out noise more accurately than when traditional motion capture technology is used, ii) reduce the average number of occlusions, iii) reduce the average length of occlusions, iv) making it possible to automatically identify markers within a single image, without constraining the user’s movements, and v) making it possible to identify markers further away from capturing cameras. Improvements in these areas represent a significant improvement in the performance and accuracy of optical motion capture systems and may greatly increase their range of application areas.