Study on 3D Action Recognition Based on Deep Neural Network

Abstract

Recognition of the detailed human action in a harsh environment such as low light environment is an issue that needs to be constantly resolved in a video surveillance system. Action recognition technology using 3D depth map information is understood as a way to solve this issue. In this paper, we propose efficient action recognition using 3D image based on convolution neural network (CNN). And we apply it to the intelligent video surveillance system and measure the detection accuracy by the action. Experimental results show that the security action recognition accuracy of 61.5%.