In recent years, human action analysis is a focal point in video processing, especially on action recognition and safety surveillance. It always performs as an auxiliary tool to minimize the manpower-resource on special tasks. This paper explores the human action analysis in a specified situation, based on the human posture extraction by pose-estimation algorithm. Deep neural network (DNN) methods was used, composed of residual learning blocks for feature extraction and recurrent neural network for time-series data learning. All these modules can be applied on real-time videos, classifying different security levels of actions between two people, with 91.8% accuracy on test set. Meanwhile, some other classical network structures were compared as baselines. After forward inference process of the neural network model, a logic enhancement algorithm was raised and applied in this paper, due to the prediction error between two classes. Experiments were conducted on real-time videos, achieving satisfying performance.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.