Real Time Implementation of MPEG-2 Decoder on NVIDIA GPU
Computer Vision


Project Description:
  A lot of effort has been put into improving the performance of different signal processing algorithms (image processing in particular) implemented on standard computers: from designated command sets like MMX, SSE, SSE2, SSE3 etc; till the development of compilers with advanced optimization methods (Intel). All of the above implement the algorithm on the CPU. Another powerful processor, more suitable for these tasks was being developed at the same time, the GPU, the Graphical Processor Unit. Whilst the CPU is most suitable for control, logic implementation and software with many referrals to routines and subroutines, the GPU is most efficient when stream processing is required.

Project Goal:
  Developing a MPEG2 video Interpreter implemented on a GPU. In association with: Elbit Systems.
note: For Outstanding Students!

Project Details:
  1. Supervisors:
    Yaniv Frishman
    Phone: 04-8295741

    Yair Moshe
    Phone: 04-8294746

  2. Field:
    Computer Vision and Computer Graphics
  3. Requirements:
    Basic computer graphics course with some knowledge of OpenGL concepts, motivation to work closely with HW, Basic signal processing (FFT, filters, sampling), Good programming background , Image Processing course and MATLAB programming skills
  4. Project status:
    Taken by the students Pavel Berman and Anton Kypiatkov
Current Projects Archive
CG&M Lab    Contact Us EE Labs ECE Department Technion