This master thesis presents an efficient solution for remote collaboration over video. The implemented system is based on providing a generic framework for collaboration over video, which is then extended to several vertical social software solutions. Two major vertical social software scenarios covered in the implemented prototype include a Real-time assistance module and an Offline support module. Real-time assistance module provides a system integrating gesture and live video to support collaboration on physical tasks. This outlines benefits of using gestures and shared visual informa- tion for performing tasks, particularly in a highly mobile environment. A component, which does real-time hand segmentation and overlaying segmented hand’s video on remote user’s problem space, enables the prototype to facilitate communication with remote partner in a natural and intuitive manner. Offline support module ensures the possibility of achieving collaboration over video even if there is no Internet connection available in the task environment. Provision of drawing tools and an effective annota- tion model makes the offline support module of this prototype an upright system for remote assistance.
Results of several evaluation cycles done on the provided prototype suggest that out gesture over video communication system enhances the performance of tasks over traditional audio-video only systems. These results suggest that addition of gestural communication to remote video collaboration can be a very effective solution for han- dling interaction-intense activities. Furthermore, inclusion of offline support module makes this system more robust and applicable in highly remote areas with poor or no Internet connectivity.
