Real-time, wavelet-based, object-oriented video compression on FPGA

Abstract

Advances in video compression techniques have led to the development of object-oriented video compression systems. However, there have been very few attempts to implement such a system on configurable hardware. This dissertation presents a research initiative leading to the design of a real-time, wavelet-based, object-oriented video compression system, designed for implementation on configurable hardware devices, namely Field Programmable Gate Arrays. The system uses motion detection and segmentation as the basis for identifying moving objects.