Human activity recognition using body pose features and support vector machine

TLDR: This paper addresses the problem of human activity recognition using support vector machine (SVM) classifier by extracting the 3D joint skeleton representation of individual as a compact representation of postures providing adequate accuracy for real-time full body tracking.

Abstract: In this paper, we address the problem of human activity recognition using support vector machine (SVM) classifier. Human action recognition can be viewed as a process of detecting the actions of the individuals by monitoring their actions and environmental conditions. It is an important technology which is widely spread because of its promising applications in surveillance, health care and elderly monitoring. This is achieved by capturing the videos from depth sensor (Microsoft kinect) through which we extract the 3D joint skeleton representation of individual as a compact representation of postures providing adequate accuracy for real-time full body tracking. A complete human activity dataset depicting all the activities including RGBD videos and motion capture data is created. The skeleton data from these videos are used to best recognize the activities. The method is tested on detecting and recognizing 13 different activities performed by 10 individuals with varied views in both indoor and outdoor environments achieving good performance. We show better results for detection of activities even if the individual is not present in the training set before, and achieve an overall detection accuracy of 89%.