Evaluation of frame estimation techniques in video-lossy environments

Abstract EN

The integration of different types of media such as audio, video, text, and graphics has been a consequence of increasing computational power, research, and need for multimedia.

With the further advancement of telecommunications and networks, the transmission of multimedia data, collaboration, and the sharing of a wide spectrum of different types of information became a feasible reality.

Besides handling the large amount of data involved, a major challenge is the integration of the different types of media in applications.

This challenge is harder in the case of continuous media such as audio and video, which have to abide by stringent time requirements and specifications in most cases.

The delivery of such streams of data over any given network will inevitably introduce delays and disruptions affecting the synchronization between media objects.

The research is mainly divided into two sections.

The first section establishes a classification of human tolerance to synchronization loss.

The second section tries to evaluate frame-estimation based techniques for solving the synchronization problems resulting from the loss in video frames during transmission.

The synchronization restoration techniques: one-way frame duplication, two-way frame duplication, and linear interpolation, are evaluated and compared according to the classification of human tolerance, signal to noise ratio, and mean square error.

The results substantiated that linear interpolation will be the best technique overall.