Innovative framework for human motion tracking for hci

Abstract

Human Computer Interaction is an emerging field in the computer science context, which tends to discover new systems that use user-centered view instead of machine centered view. This research specifically targets the full body motion interaction with computer. Therefore it develops an application which is more biased towards a user centered view. The process of building a new framework for HCI is challengeable and should be done very carefully. The system needs to indicate higher performance gain for measurements such as usability, efficiency, accuracy, and reliability. This paper describes a passive marker based approach which feeds input from web cameras and tracks blue colored markers. Input brings the marker position details with respect to time. Algorithms which regulate this framework are used for the calculation of derived attributes, clustering markers, and the prediction of positions coordinates of markers and initializing markers. They will give a meaningful interpretation to the marker position details. The next task of motion detection is facilitated using the Artificial Intelligence techniques which take preprocessed marker details as input. According to the prediction made, the user will be able to move in an immersive virtual environment. Although the project components were implemented in C++ java, they were integrated using java Native Interface (JNI).