Tutorial 5: Multiresolution Analysis for Imaging on Reconfigurable Hardware
Presented by
Abbes Amira, Brunel University
Abstract
This tutorial will focus on the design and implementation of multiresolution analysis algorithms used in image processing systems.
A range of techniques including distributed arithmetic and systolic design will be used for the design and implementation of algorithms such as wavelet, ridglet and curvelet. Applications such as medical imaging, biometrics and information retrieval where multiresolution techniques are used will be addressed with their efficient FPGA implementation. The implementation process of these algorithms on Field Programmable Gate Arrays (FPGAs) will be explained in detail and will be carried out using Handel-C programming language- an extension to standard C used for hardware compilation. AccelDSP, a MATLAB to FPGA tool will be also introduced for image processing. The tutorial will review FPGA technology and methods for developing novel architectures for multiresolution techniques used in image processing applications, and will conclude with comprehensive case studies demonstrating the use of these low power efficient architectures to accelerate computationally-intensive image processing applications. The tutorial will introduce the use of Handel-C language based FPGA programming environment, AccelDSP and hardware-software co-design approaches for high performance computing and embedded solutions.
Outline
- Introduction to Reconfigurable Computing
- FPGA Technology and Architectures
- FPGA platforms for Imaging
- Multiresolution Analysis for Imaging
- Theory and Concepts
- Medical Imaging
- Information Retrieval
- Biometrics
- Architectures and FPGA Implementation
- Discrete Wavelet Transform (DWT)
- Finite Ridglet Transform (FRIT)
- FPGA Programming Tools (Handel-C and AccelDSP)
- Examples and Demos
- Conclusions, Q&A
Useful Links
http://people.brunel.ac.uk/~eestaaa2/
http://www.xilinx.com/
http://www.mentor.com/
