Hand Gesture Recognition and Tracking based on Distributed Locally Linear Embedding

Due to the variations of posture appearance, the recognition and tracking of human hand gesture remain a difficult problem. In this paper, an unsupervised learning algorithm, Distributed Locally Linear Embedding (DLLE), is proposed to discover the intrinsic structure of the data, such as neighborhood relationships and global distributions. These discovered properties are used to compute their corresponding low-dimensional embedding, and then Probabilistic Neural Network (PNN) is employed and a database is set up for static gesture classification. For dynamic gesture tracking, we make use of the similarity measures among the images, hand gesture motion from a sequence of images can be tracked and dynamically reconstructed according to the image's relative position in the corresponding motion database. Experimental results show that our approach is able to successfully separate different hand postures and track the dynamic gesture.