Image processing using hardware accelerator

OData support
Supervisor:
Dr. Fehér Béla
Department of Measurement and Information Systems

The use of heterogeneous architecture could mean several orders of magnitude performance increase, compared with conventional processors, in case of well-defined compute-intensive tasks. A new methodology become available thanks to the developments, the High-Level Synthesis. Thanks to the efficient architecture and new design method, we are able to design complex systems with significantly higher performance in a shorter time.

In the thesis, I studied the structure of modern heterogeneous systems, hardware and software components, furthermore I summarized the properties of the latest generation Xilinx’s System on Chip, namely the ZYNQ Z-7020. I reviewed the development technologies used in the realization of heterogeneous environments. Since the video signal processing requires more and more performance, so I chose this application field to introduce the advantage of heterogeneous architecture and High-Level Synthesis. My research studies the most commonly used image processing techniques.

Finally, I’ve created an image processing library that takes advantage of the mentioned architecture. This library contains several image processing algorithms, which operates on the processor or the reconfigurable logic.

Downloads

Please sign in to download the files of this thesis.