A Novel Data Glove for Hand-object Pressure Capture

摘要

We propose a novel data glove for hand-object pressure capture and realize the synchronization of pressure data and video data of hand-object interaction motion. We process the collected motion video data, labeling it automatically with 2D key points of the hand gestures and segmenting the object images. Then, we process the interactive pressure data and generate a pressure data set of hand-object interaction motion with a synchronized video. Three different types of pressure sensors are used and fixed on the gloves and objects respectively to collect the data. We analyze and compare the collected pressure data using the three different sensors.