A Technique for Human Upper Body Parts Movement Tracking

Abstract

In this digital world, technology mainly revolves around multimedia content, which is massive. The existing techniques are inadequate to capture the movement of the human upper body parts in real-time. Therefore, a system is urgently required to track the human upper body parts movements from image and video content. The Viola–Jones algorithm based on a novel technique is proposed to detect upper body parts like the eyes and face in this work. Moreover, a face tracking algorithm is employed to track the movement of the upper body parts in real-time. The proposed model improves 13%, 18%, and 10% accuracy for face, eye, and body on Grimace dataset, Face94 Dataset, and Milborrow University of Cape Town (MUCT) dataset, respectively. Furthermore, it is also tested directly on the MATLAB software on a new image or Cam video. It takes less than 25 milliseconds per frame; hence, applicable for real-time applications.