A Spatiotemporal Robust Approach for Human Activity Recognition

摘要: Nowadays, human activity recognition is considered to be one of the fundamental topics in computer vision research areas, including human-robot interaction. In this work, a novel method proposed utilizing depth and optical flow motion information silhouettes from video for recognition. The utilizes enhanced independent component analysis (EICA) on silhouettes, features, hidden Markov models (HMMs) local features are extracted collection exhibiting various activities. Optical flow- based also silhouette area used an augmented form spatiotemporal features. Next, by generalized discriminant (GDA) better representation. These then fed into HMMs model activities recognize them. experimental results show superiority approach over conventional ones.