TOP  >  Thesis/Dissertation  >  Analyzing Relationship between Sensor Data of DataGlove and Configuration of Finger Segments for Manipulating 3D Hand Model

Analyzing Relationship between Sensor Data of DataGlove and Configuration of Finger Segments for Manipulating 3D Hand Model

DataGlove is useful to manipulate 3D hand model which is used in the field of movies, computer games, medical care, and so on.Especially, it is effective for immersive virtual reality systems or making movies without a feeling of strangeness to use realistically individualized 3D hand model. We aim to manipulate such 3D hand model reconstructed from 3D measurement of actual human hand, by using DataGlove.

Such 3D hand model has only shape, and does not have articulated structure for manipulation. In order to achieve realistic manipulation, we have to analyze the relationship between the sensor data of DataGlove and the configuration of finger segments. We need to observe optical markers put on the finger segments for analyzing the configuration, and we also need to wear DataGlove to capture the sensor data of it. Therefore, it is difficult to capture them simultaneously.

To solve this problem, we fixed plastic plates with optical markers on the DataGlove. This DataGlove makes it possible to get the sensor data of DataGlove and the configuration of finger segments at the same time. We analyzed the relationship from them, and resulted that second-order multiple regression equation approximates the relationship in the highest accuracy. We manipulated 3D hand model based on it. The error of the configuration resulted in about 5.32mm on average, and verified that we could manipulate the hand model naturally.