Hardware-Software Techniques for Fast Real-Time Image Processing


This project investigates the use of novel hardware and software techniques to speed up image processing to enable real-time manipulation and analysis of very large images.  The challenge of this project is in discovering how to implement the image processing algorithms using very limited hardware resources while at the same time processing image pixels at a very high rate.  Efficient design space exploration of hardware, architectural, and algorithmic variables is key to finding a good solution.  This project is funded by DARPA and a DoD SBIR with additional support from Xilinx, who has donated FPGA boards to implement and test our algorithms.

Current People Involved:

- R. Iris Bahar
- Sherief Reda
- Kumud Nepal
- Onur Ulusel